WIP gim: I had started working on some file vertex.go, not sure if I'll end up needing it or not

BUILD

@@ -0,0 +1,5 @@
+RELEASE=RELEASE_381 # this may have to be changed based on llvm version
+svn co https://llvm.org/svn/llvm-project/llvm/tags/$RELEASE/final $GOPATH/src/llvm.org/llvm
+cd $GOPATH/src/llvm.org/llvm/bindings/go
+./build.sh
+go install llvm.org/llvm/bindings/go/llvm

NOTES

@@ -0,0 +1,431 @@
+Been thinking about the stack and heap a lot. It would be possible, though
+possibly painful, to enforce a language with no global heap. The question really
+is: what are the principles which give reason to do so? What are the principles
+of this language, period? The principles are different than the use-cases. They
+don't need to be logically rigorous (at first anyway).
+
+##########
+
+I need to prioritize the future of this project a bit more. I've been thinking
+I'm going to figure this thing out at this level, but I shouldn't even be
+working here without a higher level view.
+
+I can't finish this project without financial help. I don't think I can get a v0
+up without financial help. What this means at minimum, no matter what, I'm going
+to have to:
+
+- Develop a full concept of the language that can get it to where I want to go
+    - Figure out where I want it to go
+- Write the concept into a manifesto of the language
+- Write the concept into a proposal for course of action to take in developing
+  the language further
+
+I'm unsure about what this language actually is, or is actually going to look
+like, but I'm sure of those things. So those are the lowest hanging fruit, and I
+should start working on them pronto. It's likely I'll need to experiment with
+some ideas which will require coding, and maybe even some big ideas, but those
+should all be done under the auspices of developing the concepts of the
+language, and not the compiler of the language itself.
+
+#########
+
+Elemental types:
+
+* Tuples
+* Arrays
+* Integers
+
+#########
+
+Been doing thinking and research on ginger's elemental types and what their
+properties should be. Ran into roadblock where I was asking myself these
+questions:
+
+* Can I do this without atoms?
+* What are different ways atoms can be encoded?
+* Can I define language types (elementals) without defining an encoding for
+  them?
+
+I also came up with two new possible types:
+
+* Stream, effectively an interface which produces discreet packets (each has a
+  length), where the production of one packet indicates the size of the next one
+  at the same time.
+* Tagged, sort of like a stream, effectively a type which says "We don't know
+  what this will be at compile-time, but we know it will be prefixed with some
+  kind of tag indicating its type and size.
+    * Maybe only the size is important
+    * Maybe precludes user defined types that aren't composites of the
+      elementals? Maybe that's ok?
+
+Ran into this:
+https://www.ps.uni-saarland.de/~duchier/python/continuations.htm://www.ps.uni-saarland.de/~duchier/python/continuations.html
+https://en.wikipedia.org/wiki/Continuation#First-class_continuations
+
+which is interesting. A lot of my problems now are derived from stack-based
+systems and their need for knowing the size input and output data, continuations
+seem to be an alternative system?
+
+I found this:
+
+http://lambda-the-ultimate.org/node/4512
+
+I don't understand any of it, I should definitely learn feather
+
+I should finish reading this:
+http://www.blackhat.com/presentations/bh-usa-07/Ferguson/Whitepaper/bh-usa-07-ferguson-WP.pdf
+
+#########
+
+Ok, so I'm back at this for the first time in a while, and I've got a good thing
+going. The vm package is working out well, Using tuples and atoms as the basis
+of a language is pretty effective (thanks erlang!). I've got basic variable
+assignment working as well. No functions yet. Here's the things I still need to
+figure out or implement:
+
+* lang
+    * constant size arrays
+        * using them for a "do" macro
+    * figure out constant, string, int, etc... look at what erlang's actual
+      primitive types are for a hint
+    * figure out all needed macros for creating and working with lang types
+* vm
+    * figure out the differentiation between compiler macros and runtime calls
+        * probably separate the two into two separate call systems
+    * the current use of varCtx is still pretty ugly, the do macro might help
+      clean it up
+* functions
+    * are they a primitive? I guess so....
+    * declaration and type
+    * variable deconstruction
+    * scoping/closures
+* compiler macros, need vm's Run to output a lang.Term
+* need to learn about linking
+    * figure out how to include llvm library in compiled binary and make it
+      callable. runtime macros will come from this
+    * linking in of other ginger code? or how to import in general
+* comiler, a general purpose binary for taking ginger code and turning it
+  into machine code using the vm package
+    * swappable syntax, including syntax-dependent macros
+* close the loop?
+
+############
+
+I really want contexts to work. They _feel_ right, as far as abstractions go.
+And they're clean, if I can work out the details.
+
+Just had a stupid idea, might as well write it down though.
+
+Similar to how the DNA and RNA in our cells work, each Context is created with
+some starting set of data on it. This will be the initial protein block. Based
+on the data there some set of Statements (the RNA) will "latch" on and do
+whatever work they're programmed to do. That work could include making new
+Contexts and "releasing" them into the ether, where they would get latched onto
+(or not).
+
+There's so many problems with this idea, it's not even a little viable. But here
+goes:
+
+* Order of execution becomes super duper fuzzy. It would be really difficult to
+  think about how your program is actually going to work.
+
+* Having Statement sets just latch onto Contexts is super janky. They would get
+  registered I guess, and it would be pretty straightforward to differentiate
+  one Context from another, but what about conflicts? If two Statements want to
+  latch onto the same Context then what? If we wanted to keep the metaphor one
+  would just get randomly chosen over the other, but obviously that's insane.
+
+############
+
+I explained some of this to ibrahim already, but I might as well get it all
+down, cause I've expanded on it a bit since.
+
+Basically, ops (functions) are fucking everything up. The biggest reason for
+this is that they are really really hard to implement without a type annotation
+system. The previous big braindump is about that, but basically I can't figure
+out a way that feels clean and good enough to be called a "solution" to type
+inference. I really don't want to have to add type annotations just to support
+functions, at least not until I explore all of my options.
+
+The only other option I've come up with so far is the context thing. It's nice
+because it covers a lot of ground without adding a lot of complexity. Really the
+biggest problem with it is it doesn't allow for creating new things which look
+like operations. Instead, everything is done with the %do operator, which feels
+janky.
+
+One solution I just thought of is to get rid of the %do operator and simply make
+it so that a list of Statements can be used as the operator in another
+Statement. This would _probably_ allow for everything that I want to do. One
+outstanding problem I'm facing is figuring out if all Statements should take a
+Context or not.
+
+* If they did it would be a lot more explicit what's going on. There wouldn't be
+  an ethereal "this context" that would need to be managed and thought about. It
+  would also make things like using a set of Statements as an operator a lot
+  more straightforward, since without Contexts in the Statement it'll be weird
+  to "do" a set of Statements in another Context.
+
+* On the other hand, it's quite a bit more boilerplate. For the most part most
+  Statements are going to want to be run in "this" context. Also this wouldn't
+  really decrease the number of necessary macros, since one would still be
+  needed in order to retrieve the "root" Context.
+
+* One option would be for a Statement's Context to be optional. I don't really
+  like this option, it makes a very fundamental datatype (a Statement) a bit
+  fuzzier.
+
+* Another thing to think about is that I might just rethink how %bind works so
+  that it doesn't operate on an ethereal "this" Context. %ctxbind is one attempt
+  at this, but there's probably other ways.
+
+* One issue I just thought of with having a set of Statements be used as an
+  operator is that the argument to that Statement becomes.... weird. What even
+  is it? Something the set of Statements can access somehow? Then we still need
+  something like the %in operator.
+
+Let me backtrack a bit. What's the actual problem? The actual thing I'm
+struggling with is allowing for code re-use, specifically pure functions. I
+don't think there's any way anyone could argue that pure functions are not an
+effective building block in all of programming, so I think I can make that my
+statement of faith: pure functions are good and worthwhile, impure functions
+are.... fine.
+
+Implementing them, however, is quite difficult. Moreso than I thought it would
+be. The big inhibitor is the method by which I actually pass input data into the
+function's body. From an implementation standpoint it's difficult because I
+*need* to know how many bytes on the stack the arguments take up. From a syntax
+standpoint this is difficult without a type annotation system. And from a
+usability standpoint this is difficult because it's a task the programmer has to
+do which doesn't really have to do with the actual purpose or content of the
+function, it's just a book-keeping exercise.
+
+So the stack is what's screwing us over here. It's a nice idea, but ultimately
+makes what we're trying to do difficult. I'm not sure if there's ever going to
+be a method of implementing pure functions that doesn't involve argument/return
+value copying though, and therefore which doesn't involve knowing the byte size
+of your arguments ahead of time.
+
+It's probably not worth backtracking this much either. For starters, cpus are
+heavily optimized for stack based operations, and much of the way we currently
+think about programming is also based on the stack. It would take a lot of
+backtracking if we ever moved to something else, if there even is anything else
+worth moving to.
+
+If that's the case, how is the stack actually used then?
+
+* There's a stack pointer which points at an address on the stack, the stack
+  being a contiguous range of memory addresses. The place the stack points to is
+  the "top" of the stack, all higher addresses are considered unused (no matter
+  what's in them). All the values in the stack are available to the currently
+  executing code, it simply needs to know either their absolute address or their
+  relative position to the stack pointer.
+
+* When a function is "called" the arguments to it are copied onto the top of the
+  stack, the stack pointer is increased to reflect the new stack height, and the
+  function's body is jumped to. Inside the body the function need only pop
+  values off the stack as it expects them, as long as it was called properly it
+  doesn't matter how or when the function was called. Once it's done operating
+  the function ensures all the input values have been popped off the stack, and
+  subsequently pushes the return values onto the stack, and jumps back to the
+  caller (the return address was also stored on the stack).
+
+That's not quite right, but it's close enough for most cases. The more I'm
+reading about this the more I think it's not going to be worth it to backtrack
+passed the stack. There's a lot of compiler and machine specific crap that gets
+involved at that low of a level, and I don't think it's worth getting into it.
+LLVM did all of that for me, I should learn how to make use of that to make what
+I want happen.
+
+But what do I actually want? That's the hard part. I guess I've come full
+circle. I pretty much *need* to use llvm functions. But I can't do it without
+declaring the types ahead of time. Ugghh.
+
+################################
+
+So here's the current problem:
+
+I have the concept of a list of statements representing a code block. It's
+possible/probable that more than this will be needed to represent a code block,
+but we'll see.
+
+There's two different ways I think it's logical to use a block:
+
+* As a way of running statements within a new context which inherits all of its
+  bindings from the parent. This would be used for things like if statements and
+  loops, and behaves the way a code block behaves in most other languages.
+
+* To define a operator body. An operator's body is effectively the same as the
+  first use-case, except that it has input/output as well. An operator can be
+  bound to an identifier and used in any statement.
+
+So the hard part, really, is that second point. I have the first done already.
+The second one isn't too hard to "fake" using our current context system, but it
+can't be made to be used as an operator in a statement. Here's how to fake it
+though:
+
+* Define the list of statements
+* Make a new context
+* Bind the "input" bindings into the new context
+* Run %do with that new context and list of statements
+* Pull the "output" bindings out of that new context
+
+And that's it. It's a bit complicated but it ultimately works and effectively
+inlines a function call.
+
+It's important that this looks like a normal operator call though, because I
+believe in guy steele. Here's the current problems I'm having:
+
+* Defining the input/output values is the big one. In the inline method those
+  were defined implicitly based on what the statements actually use, and the
+  compiler would fail if any were missing or the wrong type. But here we ideally
+  want to define an actual llvm function and not inline everytime. So we need to
+  somehow "know" what the input/output is, and their types.
+
+    * The output value isn't actually *that* difficult. We just look at the
+      output type of the last statement in the list and use that.
+
+    * The input is where it gets tricky. One idea would be to use a statement
+      with no input as the first statement in the list, and that would define
+      the input type. The way macros work this could potentially "just work",
+      but it's tricky.
+
+    * It would also be kind of difficult to make work with operators that take
+      in multiple parameters too. For example, `bind A, 1` would be the normal
+      syntax for binding, but if we want to bind an input value it gets weirder.
+
+        * We could use a "future" kind of syntax, like `bind A, _` or something
+          like that, but that would requre a new expression type and also just
+          be kind of weird.
+
+        * We could have a single macro which always returns the input, like
+          `%in` or something. So the bind would become `bind A, %in` or
+          `bind (A, B), %in` if we ever get destructuring. This isn't a terrible
+          solution, though a bit unfortunate in that it could get confusing with
+          different operators all using the same input variable effectively. It
+          also might be a bit difficult to implement, since it kind of forces us
+          to only have a single argument to the LLVM function? Hard to say how
+          that would work. Possibly all llvm functions could be made to take in
+          a struct, but that would be ghetto af. Not doing a struct would take a
+          special interaction though.... It might not be possible to do this
+          without a struct =/
+
+* Somehow allowing to define the context which gets used on each call to the
+  operator, instead of always using a blank one, would be nice.
+
+    * The big part of this problem is actually the syntax for calling the
+      operator. It's pretty easy to have this handled within the operator by the
+      %thisctx macro. But we want the operator to be callable by the same syntax
+      as all other operator calls, and currently that doesn't have any way of
+      passing in a new context.
+
+    * Additionally, if we're implementing the operator as an LLVM function then
+      there's not really any way to pass in that context to it without making
+      those variables global or something, which is shitty.
+
+* So writing all this out it really feels like I'm dealing with two separate
+  types that just happen to look similar:
+
+    * Block: a list of statements which run with a variable context.
+
+    * Operator: a list of statements which run with a fixed (empty?) context,
+      and have input/output.
+
+* There's so very nearly a symmetry there. Things that are inconsistent:
+
+    * A block doesn't have input/output
+
+        * It sort of does, in the form of the context it's being run with and
+          %ctxget, but not an explicit input/output like the operator has.
+
+        * If this could be reconciled I think this whole shitshow could be made
+          to have some consistency.
+
+        * Using %in this pretty much "just works". But it's still weird. Really
+          we'd want to turn the block into a one-off operator everytime we use
+          it. This is possible.
+
+    * An operator's context must be empty
+
+        * It doesn't *have* to be, defining the ctx which goes with the operator
+          could be part of however an operator is created.
+
+* So after all of that, I think operators and blocks are kind of the same.
+
+    * They both use %in to take in input, and both output using the last statement
+      in their list of statements.
+
+    * They both have a context bound to them, operators are fixed but a block
+      changes.
+
+    * An operator is a block with a bound context.
+
+##############@@@@@@@@@#$%^&^%$#@#$%^&*
+
+* New problem: type inference. LLVM requires that a function's definition have
+  the type specified up-front. This kind of blows. Well actually, it blows a lot
+  more than kind of. There's two things that need to be infered from a List of
+  Statements then: the input type and the output type. There's two approaches
+  I've thought of in the current setup.
+
+    * There's two approaches to determining the type of an operator: analyze the
+      code as ginger expressions, or build the actual llvm structures and
+      analyze those.
+
+        * Looking at the ginger expressions is definitely somewhat fuzzy. We can
+          look at all the statements and sub-statements until we find an
+          instance of %in, then look at what that's in input into. But if it's
+          simply binding into an Identifier then we have to find the identifier.
+          If it's destructuring then that gets even *more* complicated.
+
+            * Destructuring is what really makes this approach difficult.
+              Presumably there's going to be a function that takes in an
+              Identifier (or %in I guess?) and a set of Statements and returns
+              the type for that Identifier. If we find that %in is destructured
+              into a tuple then we would run that function for each constituent
+              Identifier and put it all together. But then this inference
+              function is really coupled to %bind, which kind of blows. Also we
+              may one day want to support destructuring into non-tuples as well,
+              which would make this even harder.
+
+            * We could make it the job of the macro definition to know its input
+              and output types, as well as the types of any bindings it makes.
+              That places some burden on user macros in the future, but then
+              maybe it can be inferred for user macros? That's a lot of hope. It
+              would also mean the macro would need the full set of statements
+              that will ever run in the same Context as it, so it can determine
+              the types of any bindings it makes.
+
+       * The second method is to build the statements into LLVM structures and
+         then look at those structures. This has the benefit of being
+         non-ambiguous once we actually find the answer. LLVM is super strongly
+         typed, and re-iterates the types involved for every operation. So if
+         the llvm builder builds it then we need only look for the first usage
+         of every argument/return and we'll know the types involved.
+
+            * This requires us to use structs for tuples, and not actually use
+              multiple arguments. Otherwise it won't be possible to know the
+              difference between a 3 argument function and a 4 argument one
+              which doesn't use its 4th argument (which shouldn't really happen,
+              but could).
+
+            * The main hinderence is that the llvm builder is really not
+              designed for this sort of thing. We could conceivably create a
+              "dummy" function with bogus types and write the body, analyze the
+              body, erase the function, and start over with a non-dummy
+              function. But it's the "analyze the body" step that's difficult.
+              It's difficult to find the types of things without the llvm.Value
+              objects in hand, but since building is set up as a recursive
+              process that becomes non-trivial. This really feels like the way
+              to go though, I think it's actually doable.
+
+                * This could be something we tack onto llvmVal, and then make
+                  Build return extra data about what types the Statements it
+                  handled input and output.
+
+* For other setups that would enable this a bit better, the one that keeps
+  coming to mind is a more pipeline style system. Things like %bind would need
+  to be refactored from something that takes a Tuple to something that only
+  takes an Identifier and returns a macro which will bind to that Identifier.
+  This doesn't *really* solve the type problem I guess, since whatever is input
+  into the Identifier's bind doesn't necessarily have a type attached to it.
+  Sooo yeah nvm.

README.md

@@ -1,29 +1,118 @@
-# Ginger
+# Ginger - holy fuck again?
 
-A programming language utilizing a graph datastructure for syntax. Currently in
+## The final result. A language which can do X
-super-early-alpha-don't-actually-use-this-for-anything development.
 
-## Development
+- Support my OS
+    - Compile on many architectures
+    - Be low level and fast (effectively c-level)
+    - Be well defined, using a simple syntax
+    - Extensible based on which section of the OS I'm working on
+    - Good error messages
 
-Current efforts on ginger are focused on a golang-based virtual machine, which
+- Support other programmers and other programming areas
-will then be used to bootstrap the language.
+    - Effectively means able to be used in most purposes
+    - Able to be quickly learned
+    - Able to be shared
+        - Free
+        - New or improved components shared between computers/platforms/people
 
-If you are on a machine with nix installed, you can run:
+- Support itself
+    - Garner a team to work on the compiler
+        - Team must not require my help for day-to-day
+        - Team must stick to the manifesto, either through the design or through
+          trust
 
-```
+## The language: A manifesto, defines the concept of the language
-nix develop
-```
 
-from the repo root and you will be dropped into a shell with all dependencies
+- Quips
-(including the correct go version) in your PATH, ready to use.
+    - Easier is not better
 
-## Demo
+- Data as the language
+    - Differentiation between "syntax" and "language", parser vs compiler
+        - Syntax defines the form which is parsed
+            - The parser reads the syntax forms into data structures
+        - Language defines how the syntax is read into data structures and
+          "understood" (i.e. and what is done with those structures).
+            - A language maybe have multiple syntaxes, if they all parse into
+              the same underlying data structures they can be understood in the
+              same way.
+            - A compiler turns the parsed language into machine code. An
+              interpreter performs actions directly based off of the parsed
+              language.
 
-An example program which computes the Nth fibonacci number can be found at
+- Types, instances, and operations
-`examples/fib.gg`. You can try it out by doing:
+    - A language has a set of elemental types, and composite types
+        - "The type defines the [fundamental] operations that can be done on the
+          data, the meaning of the data, and the way values of that type can be
+          stored"
+        - Elemental types are all forms of numbers, since numbers are all a
+          computer really knows
+        - Composite types take two forms:
+            - Homogeneous: all composed values are the same type (arrays)
+            - Heterogeneous: all composed values are different
+                - If known size and known types per-index, tuples
+                - A 0-tuple is kind of special, and simply indicates absence of
+                  any value.
+        - A third type, Any, indicates that the type is unknown at compile-time.
+          Type information must be passed around with it at runtime.
+        - An operation has an input and output. It does some action on the input
+          to produce the output (presumably). An operation may be performed as
+          many times as needed, given any value of the input type. The types of
+          both the input and output are constant, and together they form the
+          operation's type.
+    - A value is an instance of a type, where the type is known at compile-time
+      (though the type may be Any). Multiple values may be instances of the same
+      type. E.g.: 1 and 2 are both instances of int
+        - A value is immutable
+        - TODO value is a weird word, since an instance of a type has both a
+          type and value. I need to think about this more. Instance might be a
+          better name
 
-```
+- Stack and scope
-go run ./cmd/eval/main.go "$(cat examples/fib.gg)" 5
+    - A function call operates within a scope. The scope had arguments passed
-```
+      into it.
+    - When a function calls another, that other's scope is said to be "inside"
+      the caller's scope.
+    - A pure function only operates on the arguments passed into it.
+    - A pointer allows for modification outside of the current scope, but only a
+      pointer into an outer scope. A function which does this is "impure"
 
-Where you can replace `5` with any number.
+- Built-in
+    - Elementals
+        - ints (n-bit)
+        - tuples
+        - stack arrays
+            - indexable
+            - head/tail
+            - reversible (?)
+            - appendable
+        - functions (?)
+        - pointers (?)
+        - Any (?)
+    - Elementals must be enough to define the type of a variable
+    - Ability to create and modify elmental types
+        - immutable, pure functions
+    - Other builtin functionality:
+        - Load/call linked libraries
+    - Comiletime macros
+        - Red/Blue
+
+- Questions
+    - Strings need to be defined in terms of the built-in types, which would be
+      an array of lists. But this means I'm married to that definition of a
+      string, it'd be difficult for anyone to define their own and have it
+      interop. Unless "int" was some kind of macro type that did some fancy
+      shit, but that's kind of gross.
+    - An idea of the "equality" of two variables being tied not just to their
+      value but to the context in which they were created. Would aid in things
+      like compiler tagging.
+    - There's a "requirement loop" of things which need figuring out:
+        - function structure
+        - types
+        - seq type
+        - stack/scope
+    - Most likely I'm going to need some kind of elemental type to indicate
+      something should happen at compile-time and not runtime, or the other way
+      around.
+
+## The roadmap: A plan of action for tackling the language

cmd/eval/main.go

@@ -1,37 +0,0 @@
-package main
-
-import (
-	"bytes"
-	"fmt"
-	"os"
-
-	"code.betamike.com/mediocregopher/ginger/gg"
-	"code.betamike.com/mediocregopher/ginger/vm"
-)
-
-func main() {
-
-	if len(os.Args) < 3 {
-		fmt.Printf(`Usage: %s <operation source> "in = <value>"\n`, os.Args[0])
-		return
-	}
-
-	opSrc := os.Args[1]
-	inSrc := os.Args[2]
-
-	inVal, err := gg.NewDecoder(bytes.NewBufferString(inSrc)).Next()
-	if err != nil {
-		panic(fmt.Sprintf("decoding input: %v", err))
-	}
-
-	res, err := vm.EvaluateSource(
-		bytes.NewBufferString(opSrc),
-		vm.Value{Value: inVal.Value},
-		vm.GlobalScope,
-	)
-	if err != nil {
-		panic(fmt.Sprintf("evaluating: %v", err))
-	}
-
-	fmt.Println(res)
-}

examples/examples_test.go

@@ -1,53 +0,0 @@
-package examples_test
-
-import (
-	"embed"
-	"fmt"
-	"testing"
-
-	"code.betamike.com/mediocregopher/ginger/gg"
-	"code.betamike.com/mediocregopher/ginger/vm"
-	"github.com/stretchr/testify/assert"
-)
-
-//go:embed *.gg
-var examplesFS embed.FS
-
-func TestAllExamples(t *testing.T) {
-
-	tests := []struct {
-		path string
-		in   vm.Value
-		exp  vm.Value
-	}{
-		{
-			path: "fib.gg",
-			in:   vm.Value{Value: gg.Number(5)},
-			exp:  vm.Value{Value: gg.Number(5)},
-		},
-		{
-			path: "fib.gg",
-			in:   vm.Value{Value: gg.Number(10)},
-			exp:  vm.Value{Value: gg.Number(55)},
-		},
-		{
-			path: "fib.gg",
-			in:   vm.Value{Value: gg.Number(69)},
-			exp:  vm.Value{Value: gg.Number(117669030460994)},
-		},
-	}
-
-	for _, test := range tests {
-		t.Run(fmt.Sprintf("%s_%v", test.path, test.in), func(t *testing.T) {
-			f, err := examplesFS.Open(test.path)
-			if err != nil {
-				t.Fatal(err)
-			}
-			defer f.Close()
-
-			got, err := vm.EvaluateSource(f, test.in, vm.GlobalScope)
-			assert.NoError(t, err)
-			assert.True(t, test.exp.Equal(got), "%v != %v", test.exp, got)
-		})
-	}
-}

examples/fib.gg

@@ -1,24 +0,0 @@
-* A function which accepts a number N and returns the Nth fibonacci number
-{
-    * We are passing a tuple of inputs into a graph here, such that the graph is
-    * evaluated as an anonymous function. That anonymous function uses !recur
-    * internally to compute the result.
-    !out = {
-
-        * A little helper function.
-        decr = { !out = !add < (!in, -1) };
-
-        * Deconstruct the input tuple into its individual elements, for clarity.
-        * There will be a more ergonomic way of doing this one day.
-        n = !tupEl < (!in, 0);
-        a = !tupEl < (!in, 1);
-        b = !tupEl < (!in, 2);
-
-        !out = !if < (
-            !isZero < n,
-            a,
-            !recur < ( decr<n, b, !add<(a,b) ),
-        );
-
-    } < (!in, 0, 1);
-}

expr/build.go

@@ -0,0 +1,219 @@
+package expr
+
+import (
+	"fmt"
+	"log"
+
+	"llvm.org/llvm/bindings/go/llvm"
+)
+
+func init() {
+	log.Printf("initializing llvm")
+	llvm.LinkInMCJIT()
+	llvm.InitializeNativeTarget()
+	llvm.InitializeNativeAsmPrinter()
+}
+
+type BuildCtx struct {
+	B llvm.Builder
+	M llvm.Module
+}
+
+func NewBuildCtx(moduleName string) BuildCtx {
+	return BuildCtx{
+		B: llvm.NewBuilder(),
+		M: llvm.NewModule(moduleName),
+	}
+}
+
+func (bctx BuildCtx) Build(ctx Ctx, stmts ...Statement) llvm.Value {
+	var lastVal llvm.Value
+	for _, stmt := range stmts {
+		if e := bctx.BuildStmt(ctx, stmt); e != nil {
+			if lv, ok := e.(llvmVal); ok {
+				lastVal = llvm.Value(lv)
+			} else {
+				log.Printf("BuildStmt returned non llvmVal from %v: %v (%T)", stmt, e, e)
+			}
+		}
+	}
+	if (lastVal == llvm.Value{}) {
+		lastVal = bctx.B.CreateRetVoid()
+	}
+	return lastVal
+}
+
+func (bctx BuildCtx) BuildStmt(ctx Ctx, s Statement) Expr {
+	log.Printf("building: %v", s)
+	switch o := s.Op.(type) {
+	case Macro:
+		return ctx.Macro(o)(bctx, ctx, s.Arg)
+	case Identifier:
+		s2 := s
+		s2.Op = ctx.Identifier(o).(llvmVal)
+		return bctx.BuildStmt(ctx, s2)
+	case Statement:
+		s2 := s
+		s2.Op = bctx.BuildStmt(ctx, o)
+		return bctx.BuildStmt(ctx, s2)
+	case llvmVal:
+		arg := bctx.buildExpr(ctx, s.Arg).(llvmVal)
+		out := bctx.B.CreateCall(llvm.Value(o), []llvm.Value{llvm.Value(arg)}, "")
+		return llvmVal(out)
+	default:
+		panic(fmt.Sprintf("non op type %v (%T)", s.Op, s.Op))
+	}
+}
+
+// may return nil if e is a Statement which has no return
+func (bctx BuildCtx) buildExpr(ctx Ctx, e Expr) Expr {
+	return bctx.buildExprTill(ctx, e, func(Expr) bool { return false })
+}
+
+// like buildExpr, but will stop short and stop recursing when the function
+// returns true
+func (bctx BuildCtx) buildExprTill(ctx Ctx, e Expr, fn func(e Expr) bool) Expr {
+	if fn(e) {
+		return e
+	}
+
+	switch ea := e.(type) {
+	case llvmVal:
+		return e
+	case Int:
+		return llvmVal(llvm.ConstInt(llvm.Int64Type(), uint64(ea), false))
+	case Identifier:
+		return ctx.Identifier(ea)
+	case Statement:
+		return bctx.BuildStmt(ctx, ea)
+	case Tuple:
+		// if the tuple is empty then it is a void
+		if len(ea) == 0 {
+			return llvmVal(llvm.Undef(llvm.VoidType()))
+		}
+
+		ea2 := make(Tuple, len(ea))
+		for i := range ea {
+			ea2[i] = bctx.buildExprTill(ctx, ea[i], fn)
+		}
+
+		// if the fields of the tuple are all llvmVal then we can make a proper
+		// struct
+		vals := make([]llvm.Value, len(ea2))
+		typs := make([]llvm.Type, len(ea2))
+		for i := range ea2 {
+			if v, ok := ea2[i].(llvmVal); ok {
+				val := llvm.Value(v)
+				vals[i] = val
+				typs[i] = val.Type()
+			} else {
+				return ea2
+			}
+		}
+
+		str := llvm.Undef(llvm.StructType(typs, false))
+		for i := range vals {
+			str = bctx.B.CreateInsertValue(str, vals[i], i, "")
+		}
+		return llvmVal(str)
+	case List:
+		ea2 := make(Tuple, len(ea))
+		for i := range ea {
+			ea2[i] = bctx.buildExprTill(ctx, ea[i], fn)
+		}
+		return ea2
+	case Ctx:
+		return ea
+	default:
+		panicf("%v (type %T) can't express a value", ea, ea)
+	}
+	panic("go is dumb")
+}
+
+func (bctx BuildCtx) buildVal(ctx Ctx, e Expr) llvm.Value {
+	return llvm.Value(bctx.buildExpr(ctx, e).(llvmVal))
+}
+
+// globalCtx describes what's available to *all* contexts, and is what all
+// contexts should have as the root parent in the tree.
+//
+// We define in this weird way cause NewCtx actually references globalCtx
+var globalCtx *Ctx
+var _ = func() bool {
+	globalCtx = &Ctx{
+		macros: map[Macro]MacroFn{
+			"add": func(bctx BuildCtx, ctx Ctx, e Expr) Expr {
+				tup := bctx.buildExpr(ctx, e).(llvmVal)
+				a := bctx.B.CreateExtractValue(llvm.Value(tup), 0, "")
+				b := bctx.B.CreateExtractValue(llvm.Value(tup), 1, "")
+				return llvmVal(bctx.B.CreateAdd(a, b, ""))
+			},
+
+			// TODO this chould be a user macro!!!! WUT this language is baller
+			"bind": func(bctx BuildCtx, ctx Ctx, e Expr) Expr {
+				tup := bctx.buildExprTill(ctx, e, isIdentifier).(Tuple)
+				id := bctx.buildExprTill(ctx, tup[0], isIdentifier).(Identifier)
+				val := bctx.buildExpr(ctx, tup[1])
+				ctx.Bind(id, val)
+				return NewTuple()
+			},
+
+			"ctxnew": func(bctx BuildCtx, ctx Ctx, e Expr) Expr {
+				return NewCtx()
+			},
+
+			"ctxthis": func(bctx BuildCtx, ctx Ctx, e Expr) Expr {
+				return ctx
+			},
+
+			"ctxbind": func(bctx BuildCtx, ctx Ctx, e Expr) Expr {
+				tup := bctx.buildExprTill(ctx, e, isIdentifier).(Tuple)
+				thisCtx := bctx.buildExpr(ctx, tup[0]).(Ctx)
+				id := bctx.buildExprTill(ctx, tup[1], isIdentifier).(Identifier)
+				thisCtx.Bind(id, bctx.buildExpr(ctx, tup[2]))
+				return NewTuple()
+			},
+
+			"ctxget": func(bctx BuildCtx, ctx Ctx, e Expr) Expr {
+				tup := bctx.buildExprTill(ctx, e, isIdentifier).(Tuple)
+				thisCtx := bctx.buildExpr(ctx, tup[0]).(Ctx)
+				id := bctx.buildExprTill(ctx, tup[1], isIdentifier).(Identifier)
+				return thisCtx.Identifier(id)
+			},
+
+			"do": func(bctx BuildCtx, ctx Ctx, e Expr) Expr {
+				tup := bctx.buildExprTill(ctx, e, isStmt).(Tuple)
+				thisCtx := tup[0].(Ctx)
+				for _, stmtE := range tup[1].(List) {
+					bctx.BuildStmt(thisCtx, stmtE.(Statement))
+				}
+				return NewTuple()
+			},
+
+			"op": func(bctx BuildCtx, ctx Ctx, e Expr) Expr {
+				l := bctx.buildExprTill(ctx, e, isList).(List)
+				stmts := make([]Statement, len(l))
+				for i := range l {
+					stmts[i] = l[i].(Statement)
+				}
+
+				// TODO obviously this needs to be fixed
+				fn := llvm.AddFunction(bctx.M, "", llvm.FunctionType(llvm.Int64Type(), []llvm.Type{llvm.Int64Type()}, false))
+				fnbl := llvm.AddBasicBlock(fn, "")
+
+				prevbl := bctx.B.GetInsertBlock()
+				bctx.B.SetInsertPoint(fnbl, fnbl.FirstInstruction())
+				out := bctx.Build(NewCtx(), stmts...)
+				bctx.B.CreateRet(out)
+				bctx.B.SetInsertPointAtEnd(prevbl)
+				return llvmVal(fn)
+			},
+
+			"in": func(bctx BuildCtx, ctx Ctx, e Expr) Expr {
+				fn := bctx.B.GetInsertBlock().Parent()
+				return llvmVal(fn.Param(0))
+			},
+		},
+	}
+	return false
+}()

expr/build_test.go

@@ -0,0 +1,99 @@
+package expr
+
+import (
+	"fmt"
+	. "testing"
+
+	"llvm.org/llvm/bindings/go/llvm"
+)
+
+func buildTest(t *T, expected int64, stmts ...Statement) {
+	fmt.Println("-----------------------------------------")
+	ctx := NewCtx()
+	bctx := NewBuildCtx("")
+
+	fn := llvm.AddFunction(bctx.M, "", llvm.FunctionType(llvm.Int64Type(), []llvm.Type{}, false))
+	fnbl := llvm.AddBasicBlock(fn, "")
+	bctx.B.SetInsertPoint(fnbl, fnbl.FirstInstruction())
+	out := bctx.Build(ctx, stmts...)
+	bctx.B.CreateRet(out)
+
+	fmt.Println("######## dumping IR")
+	bctx.M.Dump()
+	fmt.Println("######## done dumping IR")
+
+	if err := llvm.VerifyModule(bctx.M, llvm.ReturnStatusAction); err != nil {
+		t.Fatal(err)
+	}
+
+	eng, err := llvm.NewExecutionEngine(bctx.M)
+	if err != nil {
+		t.Fatal(err)
+	}
+
+	res := eng.RunFunction(fn, []llvm.GenericValue{}).Int(false)
+	if int64(res) != expected {
+		t.Errorf("expected:[%T]%v actual:[%T]%v", expected, expected, res, res)
+	}
+}
+
+func TestAdd(t *T) {
+	buildTest(t, 2,
+		NewStatement(Macro("add"), Int(1), Int(1)))
+	buildTest(t, 4,
+		NewStatement(Macro("add"), Int(1),
+			NewStatement(Macro("add"), Int(1), Int(2))))
+	buildTest(t, 6,
+		NewStatement(Macro("add"),
+			NewStatement(Macro("add"), Int(1), Int(2)),
+			NewStatement(Macro("add"), Int(1), Int(2))))
+}
+
+func TestBind(t *T) {
+	buildTest(t, 2,
+		NewStatement(Macro("bind"), Identifier("A"), Int(1)),
+		NewStatement(Macro("add"), Identifier("A"), Int(1)))
+	buildTest(t, 2,
+		NewStatement(Macro("bind"), Identifier("A"), Int(1)),
+		NewStatement(Macro("add"), Identifier("A"), Identifier("A")))
+	buildTest(t, 2,
+		NewStatement(Macro("bind"), Identifier("A"), NewTuple(Int(1), Int(1))),
+		NewStatement(Macro("add"), Identifier("A")))
+	buildTest(t, 3,
+		NewStatement(Macro("bind"), Identifier("A"), NewTuple(Int(1), Int(1))),
+		NewStatement(Macro("add"), Int(1),
+			NewStatement(Macro("add"), Identifier("A"))))
+	buildTest(t, 4,
+		NewStatement(Macro("bind"), Identifier("A"), NewTuple(Int(1), Int(1))),
+		NewStatement(Macro("add"),
+			NewStatement(Macro("add"), Identifier("A")),
+			NewStatement(Macro("add"), Identifier("A"))))
+}
+
+func TestOp(t *T) {
+	incr := NewStatement(Macro("op"),
+		NewList(
+			NewStatement(Macro("add"), Int(1), NewStatement(Macro("in"))),
+		),
+	)
+
+	// bound op
+	buildTest(t, 2,
+		NewStatement(Macro("bind"), Identifier("incr"), incr),
+		NewStatement(Identifier("incr"), Int(1)))
+
+	// double bound op
+	buildTest(t, 3,
+		NewStatement(Macro("bind"), Identifier("incr"), incr),
+		NewStatement(Identifier("incr"),
+			NewStatement(Identifier("incr"), Int(1))))
+
+	// anon op
+	buildTest(t, 2,
+		NewStatement(incr, Int(1)))
+
+	// double anon op
+	buildTest(t, 3,
+		NewStatement(incr,
+			NewStatement(incr, Int(1))))
+}

expr/ctx.go

@@ -0,0 +1,72 @@
+package expr
+
+// MacroFn is a compiler function which takes in an existing Expr and returns
+// the llvm Value for it
+type MacroFn func(BuildCtx, Ctx, Expr) Expr
+
+// Ctx contains all the Macros and Identifiers available. A Ctx also keeps a
+// reference to the global context, which has a number of macros available for
+// all contexts to use.
+type Ctx struct {
+	global *Ctx
+	macros map[Macro]MacroFn
+	idents map[Identifier]Expr
+}
+
+// NewCtx returns a blank context instance
+func NewCtx() Ctx {
+	return Ctx{
+		global: globalCtx,
+		macros: map[Macro]MacroFn{},
+		idents: map[Identifier]Expr{},
+	}
+}
+
+// Macro returns the MacroFn associated with the given identifier, or panics
+// if the macro isn't found
+func (c Ctx) Macro(m Macro) MacroFn {
+	if fn := c.macros[m]; fn != nil {
+		return fn
+	}
+	if fn := c.global.macros[m]; fn != nil {
+		return fn
+	}
+	panicf("macro %q not found in context", m)
+	return nil
+}
+
+// Identifier returns the llvm.Value for the Identifier, or panics
+func (c Ctx) Identifier(i Identifier) Expr {
+	if e := c.idents[i]; e != nil {
+		return e
+	}
+	// The global context doesn't have any identifiers, so don't bother checking
+	panicf("identifier %q not found", i)
+	panic("go is dumb")
+}
+
+// Copy returns a deep copy of the Ctx
+func (c Ctx) Copy() Ctx {
+	cc := Ctx{
+		global: c.global,
+		macros: make(map[Macro]MacroFn, len(c.macros)),
+		idents: make(map[Identifier]Expr, len(c.idents)),
+	}
+	for m, mfn := range c.macros {
+		cc.macros[m] = mfn
+	}
+	for i, e := range c.idents {
+		cc.idents[i] = e
+	}
+	return cc
+}
+
+// Bind returns a new Ctx which is a copy of this one, but with the given
+// Identifier bound to the given Expr. Will panic if the Identifier is already
+// bound
+func (c Ctx) Bind(i Identifier, e Expr) {
+	if _, ok := c.idents[i]; ok {
+		panicf("identifier %q is already bound", i)
+	}
+	c.idents[i] = e
+}

expr/expr.go

@@ -0,0 +1,210 @@
+package expr
+
+import (
+	"fmt"
+
+	"llvm.org/llvm/bindings/go/llvm"
+)
+
+// Expr represents the actual expression in question.
+type Expr interface{}
+
+// equaler is used to compare two expressions. The comparison should not take
+// into account Token values, only the actual value being represented
+type equaler interface {
+	equal(equaler) bool
+}
+
+// will panic if either Expr doesn't implement equaler
+func exprEqual(e1, e2 Expr) bool {
+	eq1, ok1 := e1.(equaler)
+	eq2, ok2 := e2.(equaler)
+	if !ok1 || !ok2 {
+		panic(fmt.Sprintf("can't compare %T and %T", e1, e2))
+	}
+	return eq1.equal(eq2)
+}
+
+////////////////////////////////////////////////////////////////////////////////
+
+// an Expr which simply wraps an existing llvm.Value
+type llvmVal llvm.Value
+
+/*
+func voidVal(lctx LLVMCtx) llvmVal {
+	return llvmVal{lctx.B.CreateRetVoid()}
+}
+*/
+
+////////////////////////////////////////////////////////////////////////////////
+
+/*
+// Void represents no data (size = 0)
+type Void struct{}
+
+func (v Void) equal(e equaler) bool {
+	_, ok := e.(Void)
+	return ok
+}
+*/
+
+////////////////////////////////////////////////////////////////////////////////
+/*
+// Bool represents a true or false value
+type Bool bool
+
+func (b Bool) equal(e equaler) bool {
+	bb, ok := e.(Bool)
+	if !ok {
+		return false
+	}
+	return bb == b
+}
+*/
+////////////////////////////////////////////////////////////////////////////////
+
+// Int represents an integer value
+type Int int64
+
+func (i Int) equal(e equaler) bool {
+	ii, ok := e.(Int)
+	return ok && ii == i
+}
+
+func (i Int) String() string {
+	return fmt.Sprintf("%d", i)
+}
+
+////////////////////////////////////////////////////////////////////////////////
+/*
+// String represents a string value
+type String string
+
+func (s String) equal(e equaler) bool {
+	ss, ok := e.(String)
+	if !ok {
+		return false
+	}
+	return ss == s
+}
+*/
+////////////////////////////////////////////////////////////////////////////////
+
+// Identifier represents a binding to some other value which has been given a
+// name
+type Identifier string
+
+func (id Identifier) equal(e equaler) bool {
+	idid, ok := e.(Identifier)
+	return ok && idid == id
+}
+
+func isIdentifier(e Expr) bool {
+	_, ok := e.(Identifier)
+	return ok
+}
+
+////////////////////////////////////////////////////////////////////////////////
+
+// Macro is an identifier for a macro which can be used to transform
+// expressions. The tokens for macros start with a '%', but the Macro identifier
+// itself has that stripped off
+type Macro string
+
+// String returns the Macro with a '%' prepended to it
+func (m Macro) String() string {
+	return "%" + string(m)
+}
+
+func (m Macro) equal(e equaler) bool {
+	mm, ok := e.(Macro)
+	return ok && m == mm
+}
+
+////////////////////////////////////////////////////////////////////////////////
+
+// Tuple represents a fixed set of expressions which are interacted with as if
+// they were a single value
+type Tuple []Expr
+
+// NewTuple returns a Tuple around the given list of Exprs
+func NewTuple(ee ...Expr) Tuple {
+	return Tuple(ee)
+}
+
+func (tup Tuple) String() string {
+	return "(" + exprsJoin(tup) + ")"
+}
+
+func (tup Tuple) equal(e equaler) bool {
+	tuptup, ok := e.(Tuple)
+	return ok && exprsEqual(tup, tuptup)
+}
+
+func isTuple(e Expr) bool {
+	_, ok := e.(Tuple)
+	return ok
+}
+
+////////////////////////////////////////////////////////////////////////////////
+
+// List represents an ordered set of Exprs, all of the same type. A List's size
+// does not affect its type signature, unlike a Tuple
+type List []Expr
+
+// NewList returns a List around the given list of Exprs
+func NewList(ee ...Expr) List {
+	return List(ee)
+}
+
+func (l List) String() string {
+	return "[" + exprsJoin(l) + "]"
+}
+
+func (l List) equal(e equaler) bool {
+	ll, ok := e.(List)
+	return ok && exprsEqual(l, ll)
+}
+
+func isList(e Expr) bool {
+	_, ok := e.(List)
+	return ok
+}
+
+////////////////////////////////////////////////////////////////////////////////
+
+// Statement represents an actual action which will be taken. The input value is
+// used as the input to the pipe, and the output of the pipe is the output of
+// the statement
+type Statement struct {
+	Op, Arg Expr
+}
+
+// NewStatement returns a Statement whose Op is the first Expr. If the given
+// list is empty Arg will be 0-tuple, if its length is one Arg will be that
+// single Expr, otherwise Arg will be a Tuple of the list
+func NewStatement(e Expr, ee ...Expr) Statement {
+	s := Statement{Op: e}
+	if len(ee) > 1 {
+		s.Arg = NewTuple(ee...)
+	} else if len(ee) == 1 {
+		s.Arg = ee[0]
+	} else if len(ee) == 0 {
+		s.Arg = NewTuple()
+	}
+	return s
+}
+
+func (s Statement) String() string {
+	return fmt.Sprintf("(%v %s)", s.Op, s.Arg)
+}
+
+func (s Statement) equal(e equaler) bool {
+	ss, ok := e.(Statement)
+	return ok && exprEqual(s.Op, ss.Op) && exprEqual(s.Arg, ss.Arg)
+}
+
+func isStmt(e Expr) bool {
+	_, ok := e.(Statement)
+	return ok
+}

expr/parse.go

@@ -0,0 +1,299 @@
+package expr
+
+//type exprErr struct {
+//	reason string
+//	err    error
+//	tok    lexer.Token
+//	tokCtx string // e.g. "block starting at" or "open paren at"
+//}
+//
+//func (e exprErr) Error() string {
+//	var msg string
+//	if e.err != nil {
+//		msg = e.err.Error()
+//	} else {
+//		msg = e.reason
+//	}
+//	if err := e.tok.Err(); err != nil {
+//		msg += " - token error: " + err.Error()
+//	} else if (e.tok != lexer.Token{}) {
+//		msg += " - "
+//		if e.tokCtx != "" {
+//			msg += e.tokCtx + ": "
+//		}
+//		msg = fmt.Sprintf("%s [line:%d col:%d]", msg, e.tok.Row, e.tok.Col)
+//	}
+//	return msg
+//}
+//
+//////////////////////////////////////////////////////////////////////////////////
+//
+//// toks[0] must be start
+//func sliceEnclosedToks(toks []lexer.Token, start, end lexer.Token) ([]lexer.Token, []lexer.Token, error) {
+//	c := 1
+//	ret := []lexer.Token{}
+//	first := toks[0]
+//	for i, tok := range toks[1:] {
+//		if tok.Err() != nil {
+//			return nil, nil, exprErr{
+//				reason: fmt.Sprintf("missing closing %v", end),
+//				tok:    tok,
+//			}
+//		}
+//
+//		if tok.Equal(start) {
+//			c++
+//		} else if tok.Equal(end) {
+//			c--
+//		}
+//		if c == 0 {
+//			return ret, toks[2+i:], nil
+//		}
+//		ret = append(ret, tok)
+//	}
+//
+//	return nil, nil, exprErr{
+//		reason: fmt.Sprintf("missing closing %v", end),
+//		tok:    first,
+//		tokCtx: "starting at",
+//	}
+//}
+//
+//// Parse reads in all expressions it can from the given io.Reader and returns
+//// them
+//func Parse(r io.Reader) ([]Expr, error) {
+//	toks := readAllToks(r)
+//	var ret []Expr
+//	var expr Expr
+//	var err error
+//	for len(toks) > 0 {
+//		if toks[0].TokenType == lexer.EOF {
+//			return ret, nil
+//		}
+//		expr, toks, err = parse(toks)
+//		if err != nil {
+//			return nil, err
+//		}
+//		ret = append(ret, expr)
+//	}
+//	return ret, nil
+//}
+//
+//// ParseAsBlock reads the given io.Reader as if it was implicitly surrounded by
+//// curly braces, making it into a Block. This means all expressions from the
+//// io.Reader *must* be statements. The returned Expr's Actual will always be a
+//// Block.
+//func ParseAsBlock(r io.Reader) (Expr, error) {
+//	return parseBlock(readAllToks(r))
+//}
+//
+//func readAllToks(r io.Reader) []lexer.Token {
+//	l := lexer.New(r)
+//	var toks []lexer.Token
+//	for l.HasNext() {
+//		toks = append(toks, l.Next())
+//	}
+//	return toks
+//}
+//
+//// For all parse methods it is assumed that toks is not empty
+//
+//var (
+//	openParen  = lexer.Token{TokenType: lexer.Wrapper, Val: "("}
+//	closeParen = lexer.Token{TokenType: lexer.Wrapper, Val: ")"}
+//	openCurly  = lexer.Token{TokenType: lexer.Wrapper, Val: "{"}
+//	closeCurly = lexer.Token{TokenType: lexer.Wrapper, Val: "}"}
+//	comma      = lexer.Token{TokenType: lexer.Punctuation, Val: ","}
+//	arrow      = lexer.Token{TokenType: lexer.Punctuation, Val: ">"}
+//)
+//
+//func parse(toks []lexer.Token) (Expr, []lexer.Token, error) {
+//	expr, toks, err := parseSingle(toks)
+//	if err != nil {
+//		return Expr{}, nil, err
+//	}
+//
+//	if len(toks) > 0 && toks[0].TokenType == lexer.Punctuation {
+//		return parseConnectingPunct(toks, expr)
+//	}
+//
+//	return expr, toks, nil
+//}
+//
+//func parseSingle(toks []lexer.Token) (Expr, []lexer.Token, error) {
+//	var expr Expr
+//	var err error
+//
+//	if toks[0].Err() != nil {
+//		return Expr{}, nil, exprErr{
+//			reason: "could not parse token",
+//			tok:    toks[0],
+//		}
+//	}
+//
+//	if toks[0].Equal(openParen) {
+//		starter := toks[0]
+//		var ptoks []lexer.Token
+//		ptoks, toks, err = sliceEnclosedToks(toks, openParen, closeParen)
+//		if err != nil {
+//			return Expr{}, nil, err
+//		}
+//
+//		if expr, ptoks, err = parse(ptoks); err != nil {
+//			return Expr{}, nil, err
+//		} else if len(ptoks) > 0 {
+//			return Expr{}, nil, exprErr{
+//				reason: "multiple expressions inside parenthesis",
+//				tok:    starter,
+//				tokCtx: "starting at",
+//			}
+//		}
+//		return expr, toks, nil
+//
+//	} else if toks[0].Equal(openCurly) {
+//		var btoks []lexer.Token
+//		btoks, toks, err = sliceEnclosedToks(toks, openCurly, closeCurly)
+//		if err != nil {
+//			return Expr{}, nil, err
+//		}
+//
+//		if expr, err = parseBlock(btoks); err != nil {
+//			return Expr{}, nil, err
+//		}
+//		return expr, toks, nil
+//	}
+//
+//	if expr, err = parseNonPunct(toks[0]); err != nil {
+//		return Expr{}, nil, err
+//	}
+//	return expr, toks[1:], nil
+//}
+//
+//func parseNonPunct(tok lexer.Token) (Expr, error) {
+//	if tok.TokenType == lexer.Identifier {
+//		return parseIdentifier(tok)
+//	} else if tok.TokenType == lexer.String {
+//		//return parseString(tok)
+//	}
+//
+//	return Expr{}, exprErr{
+//		reason: "unexpected non-punctuation token",
+//		tok:    tok,
+//	}
+//}
+//
+//func parseIdentifier(t lexer.Token) (Expr, error) {
+//	e := Expr{Token: t}
+//	if t.Val[0] == '-' || (t.Val[0] >= '0' && t.Val[0] <= '9') {
+//		n, err := strconv.ParseInt(t.Val, 10, 64)
+//		if err != nil {
+//			return Expr{}, exprErr{
+//				err: err,
+//				tok: t,
+//			}
+//		}
+//		e.Actual = Int(n)
+//
+//		/*
+//			} else if t.Val == "%true" {
+//				e.Actual = Bool(true)
+//
+//			} else if t.Val == "%false" {
+//				e.Actual = Bool(false)
+//		*/
+//
+//	} else if t.Val[0] == '%' {
+//		e.Actual = Macro(t.Val[1:])
+//
+//	} else {
+//		e.Actual = Identifier(t.Val)
+//	}
+//
+//	return e, nil
+//}
+//
+///*
+//func parseString(t lexer.Token) (Expr, error) {
+//	str, err := strconv.Unquote(t.Val)
+//	if err != nil {
+//		return Expr{}, exprErr{
+//			err: err,
+//			tok: t,
+//		}
+//	}
+//	return Expr{Token: t, Actual: String(str)}, nil
+//}
+//*/
+//
+//func parseConnectingPunct(toks []lexer.Token, root Expr) (Expr, []lexer.Token, error) {
+//	if toks[0].Equal(comma) {
+//		return parseTuple(toks, root)
+//
+//	} else if toks[0].Equal(arrow) {
+//		expr, toks, err := parse(toks[1:])
+//		if err != nil {
+//			return Expr{}, nil, err
+//		}
+//		return Expr{Token: root.Token, Actual: Statement{In: root, To: expr}}, toks, nil
+//	}
+//
+//	return root, toks, nil
+//}
+//
+//func parseTuple(toks []lexer.Token, root Expr) (Expr, []lexer.Token, error) {
+//	rootTup, ok := root.Actual.(Tuple)
+//	if !ok {
+//		rootTup = Tuple{root}
+//	}
+//
+//	// rootTup is modified throughout, be we need to make it into an Expr for
+//	// every return, which is annoying. so make a function to do it on the fly
+//	mkRoot := func() Expr {
+//		return Expr{Token: rootTup[0].Token, Actual: rootTup}
+//	}
+//
+//	if len(toks) < 2 {
+//		return mkRoot(), toks, nil
+//	} else if !toks[0].Equal(comma) {
+//		if toks[0].TokenType == lexer.Punctuation {
+//			return parseConnectingPunct(toks, mkRoot())
+//		}
+//		return mkRoot(), toks, nil
+//	}
+//
+//	var expr Expr
+//	var err error
+//	if expr, toks, err = parseSingle(toks[1:]); err != nil {
+//		return Expr{}, nil, err
+//	}
+//
+//	rootTup = append(rootTup, expr)
+//	return parseTuple(toks, mkRoot())
+//}
+//
+//// parseBlock assumes that the given token list is the entire block, already
+//// pulled from outer curly braces by sliceEnclosedToks, or determined to be the
+//// entire block in some other way.
+//func parseBlock(toks []lexer.Token) (Expr, error) {
+//	b := Block{}
+//	first := toks[0]
+//	var expr Expr
+//	var err error
+//	for {
+//		if len(toks) == 0 {
+//			return Expr{Token: first, Actual: b}, nil
+//		}
+//
+//		if expr, toks, err = parse(toks); err != nil {
+//			return Expr{}, err
+//		}
+//		if _, ok := expr.Actual.(Statement); !ok {
+//			return Expr{}, exprErr{
+//				reason: "blocks may only contain full statements",
+//				tok:    expr.Token,
+//				tokCtx: "non-statement here",
+//			}
+//		}
+//		b = append(b, expr)
+//	}
+//}

expr/parse_test.go

@@ -0,0 +1,149 @@
+package expr
+
+//import . "testing"
+
+//func TestSliceEnclosedToks(t *T) {
+//	doAssert := func(in, expOut, expRem []lexer.Token) {
+//		out, rem, err := sliceEnclosedToks(in, openParen, closeParen)
+//		require.Nil(t, err)
+//		assert.Equal(t, expOut, out)
+//		assert.Equal(t, expRem, rem)
+//	}
+//	foo := lexer.Token{TokenType: lexer.Identifier, Val: "foo"}
+//	bar := lexer.Token{TokenType: lexer.Identifier, Val: "bar"}
+//
+//	toks := []lexer.Token{openParen, closeParen}
+//	doAssert(toks, []lexer.Token{}, []lexer.Token{})
+//
+//	toks = []lexer.Token{openParen, foo, closeParen, bar}
+//	doAssert(toks, []lexer.Token{foo}, []lexer.Token{bar})
+//
+//	toks = []lexer.Token{openParen, foo, foo, closeParen, bar, bar}
+//	doAssert(toks, []lexer.Token{foo, foo}, []lexer.Token{bar, bar})
+//
+//	toks = []lexer.Token{openParen, foo, openParen, bar, closeParen, closeParen}
+//	doAssert(toks, []lexer.Token{foo, openParen, bar, closeParen}, []lexer.Token{})
+//
+//	toks = []lexer.Token{openParen, foo, openParen, bar, closeParen, bar, closeParen, foo}
+//	doAssert(toks, []lexer.Token{foo, openParen, bar, closeParen, bar}, []lexer.Token{foo})
+//}
+//
+//func assertParse(t *T, in []lexer.Token, expExpr Expr, expOut []lexer.Token) {
+//	expr, out, err := parse(in)
+//	require.Nil(t, err)
+//	assert.True(t, expExpr.equal(expr), "expr:%+v expExpr:%+v", expr, expExpr)
+//	assert.Equal(t, expOut, out, "out:%v expOut:%v", out, expOut)
+//}
+//
+//func TestParseSingle(t *T) {
+//	foo := lexer.Token{TokenType: lexer.Identifier, Val: "foo"}
+//	fooM := lexer.Token{TokenType: lexer.Identifier, Val: "%foo"}
+//	fooExpr := Expr{Actual: Identifier("foo")}
+//	fooMExpr := Expr{Actual: Macro("foo")}
+//
+//	toks := []lexer.Token{foo}
+//	assertParse(t, toks, fooExpr, []lexer.Token{})
+//
+//	toks = []lexer.Token{foo, foo}
+//	assertParse(t, toks, fooExpr, []lexer.Token{foo})
+//
+//	toks = []lexer.Token{openParen, foo, closeParen, foo}
+//	assertParse(t, toks, fooExpr, []lexer.Token{foo})
+//
+//	toks = []lexer.Token{openParen, openParen, foo, closeParen, closeParen, foo}
+//	assertParse(t, toks, fooExpr, []lexer.Token{foo})
+//
+//	toks = []lexer.Token{fooM, foo}
+//	assertParse(t, toks, fooMExpr, []lexer.Token{foo})
+//}
+//
+//func TestParseTuple(t *T) {
+//	tup := func(ee ...Expr) Expr {
+//		return Expr{Actual: Tuple(ee)}
+//	}
+//
+//	foo := lexer.Token{TokenType: lexer.Identifier, Val: "foo"}
+//	fooExpr := Expr{Actual: Identifier("foo")}
+//
+//	toks := []lexer.Token{foo, comma, foo}
+//	assertParse(t, toks, tup(fooExpr, fooExpr), []lexer.Token{})
+//
+//	toks = []lexer.Token{foo, comma, foo, foo}
+//	assertParse(t, toks, tup(fooExpr, fooExpr), []lexer.Token{foo})
+//
+//	toks = []lexer.Token{foo, comma, foo, comma, foo}
+//	assertParse(t, toks, tup(fooExpr, fooExpr, fooExpr), []lexer.Token{})
+//
+//	toks = []lexer.Token{foo, comma, foo, comma, foo, comma, foo}
+//	assertParse(t, toks, tup(fooExpr, fooExpr, fooExpr, fooExpr), []lexer.Token{})
+//
+//	toks = []lexer.Token{foo, comma, openParen, foo, comma, foo, closeParen, comma, foo}
+//	assertParse(t, toks, tup(fooExpr, tup(fooExpr, fooExpr), fooExpr), []lexer.Token{})
+//
+//	toks = []lexer.Token{foo, comma, openParen, foo, comma, foo, closeParen, comma, foo, foo}
+//	assertParse(t, toks, tup(fooExpr, tup(fooExpr, fooExpr), fooExpr), []lexer.Token{foo})
+//}
+//
+//func TestParseStatement(t *T) {
+//	stmt := func(in, to Expr) Expr {
+//		return Expr{Actual: Statement{In: in, To: to}}
+//	}
+//
+//	foo := lexer.Token{TokenType: lexer.Identifier, Val: "foo"}
+//	fooExpr := Expr{Actual: Identifier("foo")}
+//
+//	toks := []lexer.Token{foo, arrow, foo}
+//	assertParse(t, toks, stmt(fooExpr, fooExpr), []lexer.Token{})
+//
+//	toks = []lexer.Token{openParen, foo, arrow, foo, closeParen}
+//	assertParse(t, toks, stmt(fooExpr, fooExpr), []lexer.Token{})
+//
+//	toks = []lexer.Token{foo, arrow, openParen, foo, closeParen}
+//	assertParse(t, toks, stmt(fooExpr, fooExpr), []lexer.Token{})
+//
+//	toks = []lexer.Token{foo, arrow, foo}
+//	assertParse(t, toks, stmt(fooExpr, fooExpr), []lexer.Token{})
+//
+//	toks = []lexer.Token{foo, arrow, foo, foo}
+//	assertParse(t, toks, stmt(fooExpr, fooExpr), []lexer.Token{foo})
+//
+//	toks = []lexer.Token{foo, arrow, openParen, foo, closeParen, foo}
+//	assertParse(t, toks, stmt(fooExpr, fooExpr), []lexer.Token{foo})
+//
+//	toks = []lexer.Token{openParen, foo, closeParen, arrow, openParen, foo, closeParen, foo}
+//	assertParse(t, toks, stmt(fooExpr, fooExpr), []lexer.Token{foo})
+//
+//	fooTupExpr := Expr{Actual: Tuple{fooExpr, fooExpr}}
+//	toks = []lexer.Token{foo, arrow, openParen, foo, comma, foo, closeParen, foo}
+//	assertParse(t, toks, stmt(fooExpr, fooTupExpr), []lexer.Token{foo})
+//
+//	toks = []lexer.Token{foo, comma, foo, arrow, foo}
+//	assertParse(t, toks, stmt(fooTupExpr, fooExpr), []lexer.Token{})
+//
+//	toks = []lexer.Token{openParen, foo, comma, foo, closeParen, arrow, foo}
+//	assertParse(t, toks, stmt(fooTupExpr, fooExpr), []lexer.Token{})
+//}
+//
+//func TestParseBlock(t *T) {
+//	stmt := func(in, to Expr) Expr {
+//		return Expr{Actual: Statement{In: in, To: to}}
+//	}
+//	block := func(stmts ...Expr) Expr {
+//		return Expr{Actual: Block(stmts)}
+//	}
+//
+//	foo := lexer.Token{TokenType: lexer.Identifier, Val: "foo"}
+//	fooExpr := Expr{Actual: Identifier("foo")}
+//
+//	toks := []lexer.Token{openCurly, foo, arrow, foo, closeCurly}
+//	assertParse(t, toks, block(stmt(fooExpr, fooExpr)), []lexer.Token{})
+//
+//	toks = []lexer.Token{openCurly, foo, arrow, foo, closeCurly, foo}
+//	assertParse(t, toks, block(stmt(fooExpr, fooExpr)), []lexer.Token{foo})
+//
+//	toks = []lexer.Token{openCurly, foo, arrow, foo, openParen, foo, arrow, foo, closeParen, closeCurly, foo}
+//	assertParse(t, toks, block(stmt(fooExpr, fooExpr), stmt(fooExpr, fooExpr)), []lexer.Token{foo})
+//
+//	toks = []lexer.Token{openCurly, foo, arrow, foo, openParen, foo, arrow, foo, closeParen, closeCurly, foo}
+//	assertParse(t, toks, block(stmt(fooExpr, fooExpr), stmt(fooExpr, fooExpr)), []lexer.Token{foo})
+//}

expr/util.go

@@ -0,0 +1,40 @@
+package expr
+
+import (
+	"encoding/hex"
+	"fmt"
+	"math/rand"
+	"strings"
+)
+
+func randStr() string {
+	b := make([]byte, 16)
+	if _, err := rand.Read(b); err != nil {
+		panic(err)
+	}
+	return hex.EncodeToString(b)
+}
+
+func exprsJoin(ee []Expr) string {
+	strs := make([]string, len(ee))
+	for i := range ee {
+		strs[i] = fmt.Sprint(ee[i])
+	}
+	return strings.Join(strs, ", ")
+}
+
+func exprsEqual(ee1, ee2 []Expr) bool {
+	if len(ee1) != len(ee2) {
+		return false
+	}
+	for i := range ee1 {
+		if !exprEqual(ee1[i], ee2[i]) {
+			return false
+		}
+	}
+	return true
+}
+
+func panicf(msg string, args ...interface{}) {
+	panic(fmt.Sprintf(msg, args...))
+}

flake.lock

@@ -1,26 +0,0 @@
-{
-  "nodes": {
-    "nixpkgs": {
-      "locked": {
-        "lastModified": 1696983906,
-        "narHash": "sha256-L7GyeErguS7Pg4h8nK0wGlcUTbfUMDu+HMf1UcyP72k=",
-        "owner": "NixOS",
-        "repo": "nixpkgs",
-        "rev": "bd1cde45c77891214131cbbea5b1203e485a9d51",
-        "type": "github"
-      },
-      "original": {
-        "id": "nixpkgs",
-        "ref": "nixos-23.05",
-        "type": "indirect"
-      }
-    },
-    "root": {
-      "inputs": {
-        "nixpkgs": "nixpkgs"
-      }
-    }
-  },
-  "root": "root",
-  "version": 7
-}

flake.nix

@@ -1,44 +0,0 @@
-{
-  description = "gotc development environment";
-
-  # Nixpkgs / NixOS version to use.
-  inputs.nixpkgs.url = "nixpkgs/nixos-23.05";
-
-  outputs = { self, nixpkgs }:
-    let
-
-      # to work with older version of flakes
-      lastModifiedDate = self.lastModifiedDate or self.lastModified or "19700101";
-
-      # Generate a user-friendly version number.
-      version = builtins.substring 0 8 lastModifiedDate;
-
-      # System types to support.
-      supportedSystems = [ "x86_64-linux" "x86_64-darwin" "aarch64-linux" "aarch64-darwin" ];
-
-      # Helper function to generate an attrset '{ x86_64-linux = f "x86_64-linux"; ... }'.
-      forAllSystems = nixpkgs.lib.genAttrs supportedSystems;
-
-      # Nixpkgs instantiated for supported system types.
-      nixpkgsFor = forAllSystems (system: import nixpkgs {
-        inherit system;
-      });
-
-    in
-    {
-
-      # Add dependencies that are only needed for development
-      devShells = forAllSystems (system:
-        let
-          pkgs = nixpkgsFor.${system};
-        in {
-          default = pkgs.mkShell {
-            buildInputs = [
-              pkgs.go
-              pkgs.gotools
-              pkgs.golangci-lint
-            ];
-          };
-        });
-    };
-}

gg/decoder.go

@@ -1,297 +0,0 @@
-package gg
-
-import (
-	"fmt"
-	"io"
-	"strconv"
-	"unicode"
-
-	. "code.betamike.com/mediocregopher/ginger/gg/grammar"
-	"code.betamike.com/mediocregopher/ginger/graph"
-	"golang.org/x/exp/slices"
-)
-
-var (
-	notNewline = RuneFunc(
-		"not-newline", func(r rune) bool { return r != '\n' },
-	)
-
-	comment = Prefixed(
-		Prefixed(Rune('*'), ZeroOrMore(notNewline)), Rune('\n'),
-	)
-
-	whitespace = ZeroOrMore(FirstOf(
-		Discard(RuneFunc("whitespace", unicode.IsSpace)),
-		Discard(comment),
-	))
-)
-
-func trimmed[T any](sym Symbol[T]) Symbol[T] {
-	sym = PrefixDiscarded(whitespace, sym)
-	sym = Suffixed(sym, whitespace)
-	return sym
-}
-
-func trimmedRune(r rune) Symbol[Located[rune]] {
-	return trimmed(Rune(r))
-}
-
-var (
-	digit = RuneFunc(
-		"digit", func(r rune) bool { return '0' <= r && r <= '9' },
-	)
-
-	positiveNumber = StringFromRunes(OneOrMore(digit))
-
-	negativeNumber = Reduction(
-		Rune('-'),
-		positiveNumber,
-		func(neg Located[rune], posNum Located[string]) Located[string] {
-			return Locate(neg.Location, string(neg.Value)+posNum.Value)
-		},
-	)
-
-	number = Named(
-		"number",
-		Mapping(
-			FirstOf(negativeNumber, positiveNumber),
-			func(str Located[string]) Located[Value] {
-				i, err := strconv.ParseInt(str.Value, 10, 64)
-				if err != nil {
-					panic(fmt.Errorf("parsing %q as int: %w", str, err))
-				}
-
-				return Locate(str.Location, Number(i))
-			},
-		),
-	)
-)
-
-var (
-	nameHead = FirstOf(
-		RuneFunc("letter", unicode.IsLetter),
-		RuneFunc("mark", unicode.IsMark),
-		Rune('!'),
-	)
-
-	nameTail = ZeroOrMore(FirstOf(nameHead, digit))
-
-	name = Named(
-		"name",
-		Reduction(
-			nameHead,
-			nameTail,
-			func(head Located[rune], tail []Located[rune]) Located[Value] {
-				name := make([]rune, 0, len(tail)+1)
-				name = append(name, head.Value)
-				for _, r := range tail {
-					name = append(name, r.Value)
-				}
-				return Locate(head.Location, Name(string(name)))
-			},
-		),
-	)
-)
-
-func openEdgeIntoValue(val Value, oe *OpenEdge) *OpenEdge {
-	switch {
-	case oe == nil:
-		return graph.ValueOut(None, val)
-	case !oe.EdgeValue().Valid:
-		return oe.WithEdgeValue(Some(val))
-	default:
-		return graph.TupleOut(Some(val), oe)
-	}
-}
-
-var graphSym, value = func() (
-	Symbol[Located[Value]], Symbol[Located[Value]],
-) {
-
-	type tupleState struct {
-		ins []*OpenEdge
-		oe  *OpenEdge
-	}
-
-	type graphState struct {
-		g  *Graph
-		oe *OpenEdge
-	}
-
-	var (
-		tupleEnd = Mapping(
-			trimmedRune(')'),
-			func(Located[rune]) tupleState {
-				// if ')', then map that to an empty state. This acts as a
-				// sentinel value to indicate "end of tuple".
-				return tupleState{}
-			},
-		)
-
-		graphEnd = Mapping(
-			trimmedRune('}'),
-			func(Located[rune]) graphState {
-				// if '}', then map that to an empty state. This acts as a
-				// sentinel value to indicate "end of graph".
-				return graphState{}
-			},
-		)
-	)
-
-	var (
-		// pre-define these, and then fill in the pointers after, in order to
-		// deal with recursive dependencies between them.
-		value = new(SymbolPtr[Located[Value]])
-
-		tuple                  = new(SymbolPtr[*OpenEdge])
-		tupleTail              = new(SymbolPtr[tupleState])
-		tupleOpenEdge          = new(SymbolPtr[tupleState])
-		tupleOpenEdgeTail      = new(SymbolPtr[tupleState])
-		tupleOpenEdgeValueTail = new(SymbolPtr[tupleState])
-
-		graphSym               = new(SymbolPtr[Located[Value]])
-		graphTail              = new(SymbolPtr[graphState])
-		graphOpenEdge          = new(SymbolPtr[graphState])
-		graphOpenEdgeTail      = new(SymbolPtr[graphState])
-		graphOpenEdgeValueTail = new(SymbolPtr[graphState])
-	)
-
-	tuple.Symbol = Named(
-		"tuple",
-		Reduction[Located[rune], tupleState, *OpenEdge](
-			trimmedRune('('),
-			tupleTail,
-			func(_ Located[rune], ts tupleState) *OpenEdge {
-				slices.Reverse(ts.ins)
-				return graph.TupleOut(None, ts.ins...)
-			},
-		),
-	)
-
-	tupleTail.Symbol = FirstOf(
-		tupleEnd,
-		Mapping[tupleState, tupleState](
-			tupleOpenEdge,
-			func(ts tupleState) tupleState {
-				ts.ins = append(ts.ins, ts.oe)
-				ts.oe = nil
-				return ts
-			},
-		),
-	)
-
-	tupleOpenEdge.Symbol = FirstOf(
-		Reduction[Located[Value], tupleState, tupleState](
-			value,
-			tupleOpenEdgeValueTail,
-			func(val Located[Value], ts tupleState) tupleState {
-				ts.oe = openEdgeIntoValue(val.Value, ts.oe)
-				return ts
-			},
-		),
-		Reduction[*OpenEdge, tupleState, tupleState](
-			tuple,
-			tupleOpenEdgeTail,
-			func(oe *OpenEdge, ts tupleState) tupleState {
-				ts.oe = oe
-				return ts
-			},
-		),
-	)
-
-	tupleOpenEdgeTail.Symbol = FirstOf(
-		tupleEnd,
-		Prefixed[Located[rune], tupleState](trimmedRune(','), tupleTail),
-	)
-
-	tupleOpenEdgeValueTail.Symbol = FirstOf[tupleState](
-		tupleOpenEdgeTail,
-		Prefixed[Located[rune], tupleState](trimmedRune('<'), tupleOpenEdge),
-	)
-
-	graphSym.Symbol = Named(
-		"graph",
-		Reduction[Located[rune], graphState, Located[Value]](
-			trimmedRune('{'),
-			graphTail,
-			func(r Located[rune], gs graphState) Located[Value] {
-				if gs.g == nil {
-					gs.g = new(Graph)
-				}
-
-				return Locate(r.Location, Value{Graph: gs.g})
-			},
-		),
-	)
-
-	graphTail.Symbol = FirstOf(
-		graphEnd,
-		Reduction(
-			name,
-			Prefixed[Located[rune], graphState](
-				trimmedRune('='), graphOpenEdge,
-			),
-			func(name Located[Value], gs graphState) graphState {
-				if gs.g == nil {
-					gs.g = new(Graph)
-				}
-
-				gs.g = gs.g.AddValueIn(name.Value, gs.oe)
-				gs.oe = nil
-				return gs
-			},
-		),
-	)
-
-	graphOpenEdge.Symbol = FirstOf(
-		Reduction[Located[Value], graphState, graphState](
-			value,
-			graphOpenEdgeValueTail,
-			func(val Located[Value], gs graphState) graphState {
-				gs.oe = openEdgeIntoValue(val.Value, gs.oe)
-				return gs
-			},
-		),
-		Reduction[*OpenEdge, graphState, graphState](
-			tuple,
-			graphOpenEdgeTail,
-			func(oe *OpenEdge, gs graphState) graphState {
-				gs.oe = oe
-				return gs
-			},
-		),
-	)
-
-	graphOpenEdgeTail.Symbol = FirstOf(
-		graphEnd,
-		Prefixed[Located[rune], graphState](trimmedRune(';'), graphTail),
-	)
-
-	graphOpenEdgeValueTail.Symbol = FirstOf[graphState](
-		graphOpenEdgeTail,
-		Prefixed[Located[rune], graphState](trimmedRune('<'), graphOpenEdge),
-	)
-
-	value.Symbol = trimmed(FirstOf[Located[Value]](name, number, graphSym))
-
-	return graphSym, value
-}()
-
-// Decoder reads Values off of an io.Reader, or return io.EOF.
-type Decoder interface {
-	Next() (Located[Value], error)
-}
-
-type decoder struct {
-	r Reader
-}
-
-// NewDecoder returns a Decoder which reads off the given io.Reader. The
-// io.Reader should not be read from after this call.
-func NewDecoder(r io.Reader) Decoder {
-	return &decoder{r: NewReader(r)}
-}
-
-func (d *decoder) Next() (Located[Value], error) {
-	return value.Decode(d.r)
-}

gg/decoder_test.go

@@ -1,365 +0,0 @@
-package gg
-
-import (
-	"bytes"
-	"strconv"
-	"testing"
-
-	. "code.betamike.com/mediocregopher/ginger/gg/grammar"
-	"code.betamike.com/mediocregopher/ginger/graph"
-	"github.com/stretchr/testify/assert"
-)
-
-func TestDecoder(t *testing.T) {
-	type test struct {
-		in     string
-		exp    Located[Value]
-		expErr string
-	}
-
-	runTests := func(
-		t *testing.T, name string, sym Symbol[Located[Value]], tests []test,
-	) {
-		t.Run(name, func(t *testing.T) {
-			for i, test := range tests {
-				t.Run(strconv.Itoa(i), func(t *testing.T) {
-					r := NewReader(bytes.NewBufferString(test.in))
-					got, err := sym.Decode(r)
-					if test.expErr != "" {
-						assert.Error(t, err)
-						assert.Equal(t, test.expErr, err.Error())
-					} else if assert.NoError(t, err) {
-						assert.True(t,
-							test.exp.Value.Equal(got.Value),
-							"\nexp:%v\ngot:%v", test.exp, got,
-						)
-						assert.Equal(t, test.exp.Location, got.Location)
-					}
-				})
-			}
-		})
-	}
-
-	expNum := func(row, col int, n int64) Located[Value] {
-		return Locate(Location{Row: row, Col: col}, Number(n))
-	}
-
-	runTests(t, "number", number, []test{
-		{in: `0`, exp: expNum(1, 1, 0)},
-		{in: `100`, exp: expNum(1, 1, 100)},
-		{in: `-100`, exp: expNum(1, 1, -100)},
-		{in: `0foo`, exp: expNum(1, 1, 0)},
-		{in: `100foo`, exp: expNum(1, 1, 100)},
-	})
-
-	expName := func(row, col int, name string) Located[Value] {
-		return Locate(Location{Row: row, Col: col}, Name(name))
-	}
-
-	expGraph := func(row, col int, g *Graph) Located[Value] {
-		return Locate(Location{Row: row, Col: col}, Value{Graph: g})
-	}
-
-	runTests(t, "name", name, []test{
-		{in: `a`, exp: expName(1, 1, "a")},
-		{in: `ab`, exp: expName(1, 1, "ab")},
-		{in: `ab2c`, exp: expName(1, 1, "ab2c")},
-		{in: `ab2c,`, exp: expName(1, 1, "ab2c")},
-		{in: `!ab2c,`, exp: expName(1, 1, "!ab2c")},
-	})
-
-	runTests(t, "graph", graphSym, []test{
-		{in: `{}`, exp: expGraph(1, 1, new(Graph))},
-		{in: `{`, expErr: `1:2: expected '}' or name`},
-		{in: `{a}`, expErr: `1:3: expected '='`},
-		{in: `{a=}`, expErr: `1:4: expected name or number or graph or tuple`},
-		{
-			in: `{foo=a}`,
-			exp: expGraph(
-				1, 1, new(Graph).
-					AddValueIn(Name("foo"), graph.ValueOut(None, Name("a"))),
-			),
-		},
-		{
-			in: `{ foo = a }`,
-			exp: expGraph(
-				1, 1, new(Graph).
-					AddValueIn(Name("foo"), graph.ValueOut(None, Name("a"))),
-			),
-		},
-		{in: `{1=a}`, expErr: `1:2: expected '}' or name`},
-		{in: `{foo=a ,}`, expErr: `1:8: expected '}' or ';' or '<'`},
-		{in: `{foo=a`, expErr: `1:7: expected '}' or ';' or '<'`},
-		{
-			in: `{foo=a<b}`,
-			exp: expGraph(
-				1, 1, new(Graph).
-					AddValueIn(
-						Name("foo"),
-						graph.ValueOut(Some(Name("a")), Name("b")),
-					),
-			),
-		},
-		{
-			in: `{foo=a< b <c}`,
-			exp: expGraph(
-				1, 1, new(Graph).
-					AddValueIn(
-						Name("foo"),
-						graph.TupleOut(
-							Some(Name("a")),
-							graph.ValueOut(
-								Some(Name("b")),
-								Name("c"),
-							),
-						),
-					),
-			),
-		},
-		{
-			in: `{foo =a<b<c<1 }`,
-			exp: expGraph(
-				1, 1, new(Graph).
-					AddValueIn(
-						Name("foo"),
-						graph.TupleOut(
-							Some(Name("a")),
-							graph.TupleOut(
-								Some(Name("b")),
-								graph.ValueOut(
-									Some(Name("c")),
-									Number(1),
-								),
-							),
-						),
-					),
-			),
-		},
-		{
-			in: `{foo=a<b ; }`,
-			exp: expGraph(
-				1, 1, new(Graph).
-					AddValueIn(
-						Name("foo"),
-						graph.ValueOut(
-							Some(Name("a")),
-							Name("b"),
-						),
-					),
-			),
-		},
-		{
-			in: `{foo=a<b;bar=c}`,
-			exp: expGraph(
-				1, 1, new(Graph).
-					AddValueIn(
-						Name("foo"),
-						graph.ValueOut(
-							Some(Name("a")),
-							Name("b"),
-						),
-					).
-					AddValueIn(
-						Name("bar"),
-						graph.ValueOut(None, Name("c")),
-					),
-			),
-		},
-		{
-			in: `{foo= a<{ baz=1 } ; bar=c}`,
-			exp: expGraph(
-				1, 1, new(Graph).
-					AddValueIn(
-						Name("foo"),
-						graph.ValueOut(
-							Some(Name("a")),
-							Value{Graph: new(Graph).AddValueIn(
-								Name("baz"),
-								graph.ValueOut(None, Number(1)),
-							)},
-						),
-					).
-					AddValueIn(
-						Name("bar"),
-						graph.ValueOut(None, Name("c")),
-					),
-			),
-		},
-		{
-			in: `{foo= {baz=1} <a; bar=c}`,
-			exp: expGraph(
-				1, 1, new(Graph).
-					AddValueIn(
-						Name("foo"),
-						graph.ValueOut(
-							Some(Value{Graph: new(Graph).AddValueIn(
-								Name("baz"),
-								graph.ValueOut(None, Number(1)),
-							)}),
-							Name("a"),
-						),
-					).
-					AddValueIn(
-						Name("bar"),
-						graph.ValueOut(None, Name("c")),
-					),
-			),
-		},
-	})
-
-	runTests(t, "tuple", graphSym, []test{
-		{
-			in: `{foo=(a)}`,
-			exp: expGraph(
-				1, 1, new(Graph).
-					AddValueIn(
-						Name("foo"),
-						graph.ValueOut(None, Name("a")),
-					),
-			),
-		},
-		{
-			in: `{foo=(a<b)}`,
-			exp: expGraph(
-				1, 1, new(Graph).
-					AddValueIn(
-						Name("foo"),
-						graph.ValueOut(
-							Some(Name("a")),
-							Name("b"),
-						),
-					),
-			),
-		},
-		{
-			in: `{foo=a<(b)}`,
-			exp: expGraph(
-				1, 1, new(Graph).
-					AddValueIn(
-						Name("foo"),
-						graph.ValueOut(
-							Some(Name("a")),
-							Name("b"),
-						),
-					),
-			),
-		},
-		{
-			in: `{foo=a<(b,c)}`,
-			exp: expGraph(
-				1, 1, new(Graph).
-					AddValueIn(
-						Name("foo"),
-						graph.TupleOut(
-							Some(Name("a")),
-							graph.ValueOut(None, Name("b")),
-							graph.ValueOut(None, Name("c")),
-						),
-					),
-			),
-		},
-		{
-			in: `{foo=a<(b<c)}`,
-			exp: expGraph(
-				1, 1, new(Graph).
-					AddValueIn(
-						Name("foo"),
-						graph.TupleOut(
-							Some(Name("a")),
-							graph.TupleOut(
-								Some(Name("b")),
-								graph.ValueOut(None, Name("c")),
-							),
-						),
-					),
-			),
-		},
-		{
-			in: `{foo=a<(b<(c))}`,
-			exp: expGraph(
-				1, 1, new(Graph).
-					AddValueIn(
-						Name("foo"),
-						graph.TupleOut(
-							Some(Name("a")),
-							graph.TupleOut(
-								Some(Name("b")),
-								graph.ValueOut(None, Name("c")),
-							),
-						),
-					),
-			),
-		},
-		{
-			in: `{foo=a<(b<(c,d<1))}`,
-			exp: expGraph(
-				1, 1, new(Graph).
-					AddValueIn(
-						Name("foo"),
-						graph.TupleOut(
-							Some(Name("a")),
-							graph.TupleOut(
-								Some(Name("b")),
-								graph.ValueOut(None, Name("c")),
-								graph.ValueOut(
-									Some(Name("d")),
-									Number(1),
-								),
-							),
-						),
-					),
-			),
-		},
-		{
-			in: `{foo=a<(b<( ( (c) ) ))}`,
-			exp: expGraph(
-				1, 1, new(Graph).
-					AddValueIn(
-						Name("foo"),
-						graph.TupleOut(
-							Some(Name("a")),
-							graph.TupleOut(
-								Some(Name("b")),
-								graph.ValueOut(None, Name("c")),
-							),
-						),
-					),
-			),
-		},
-	})
-
-	runTests(t, "comment", graphSym, []test{
-		{
-			in:  "*\n{}",
-			exp: expGraph(2, 1, new(Graph)),
-		},
-		{
-			in:  "* ignore me!\n{}",
-			exp: expGraph(2, 1, new(Graph)),
-		},
-		{
-			in:  "{* ignore me!\n}",
-			exp: expGraph(1, 1, new(Graph)),
-		},
-		{
-			in: "{foo* ignore me!\n  =  a}",
-			exp: expGraph(
-				1, 1, new(Graph).
-					AddValueIn(
-						Name("foo"),
-						graph.ValueOut(None, Name("a")),
-					),
-			),
-		},
-		{
-			in: "{foo  =  a* ignore me!\n}",
-			exp: expGraph(
-				1, 1, new(Graph).
-					AddValueIn(
-						Name("foo"),
-						graph.ValueOut(None, Name("a")),
-					),
-			),
-		},
-	})
-}

gg/gg.bnf

@@ -1,25 +0,0 @@
-<digit>           ::= "0" | "1" | "2" | "3" | "4" | "5" | "6" | "7" | "8" | "9"
-<positive-number> ::= <digit>+
-<negative-number> ::= "-" <positive-number>
-<number>          ::= <negative-number> | <positive-number>
-
-<name-head> ::= <letter> | <mark> | "!"
-<name-tail> ::= <name-head> | <digit>
-<name> ::= <name-head> <name-tail>*
-
-<tuple>                      ::= "(" <tuple-tail>
-<tuple-tail>                 ::= ")" | <tuple-open-edge>
-<tuple-open-edge>            ::= <value> <tuple-open-edge-value-tail>
-                               | <tuple> <tuple-open-edge-tail>
-<tuple-open-edge-tail>       ::= ")" | "," <tuple-tail>
-<tuple-open-edge-value-tail> ::= <tuple-open-edge-tail> | "<" <tuple-open-edge>
-
-<graph>                      ::= "{" <graph-tail>
-<graph-tail>                 ::= "}" | <name> "=" <graph-open-edge>
-<graph-open-edge>            ::= <value> <graph-open-edge-value-tail>
-                               | <tuple> <graph-open-edge-tail>
-<graph-open-edge-tail>       ::= "}" | ";" <graph-tail>
-<graph-open-edge-value-tail> ::= <graph-open-edge-tail> | "<" <graph-open-edge>
-
-<value> ::= <name> | <number> | <graph>
-<gg>    ::= <eof> | <value> <gg>

gg/gg.go

@@ -1,114 +1,554 @@
-// Package gg implements graph serialization to/from the gg text format.
+// Package gg implements ginger graph creation, traversal, and (de)serialization
 package gg
 
 import (
+	"crypto/rand"
+	"encoding/hex"
 	"fmt"
-
+	"strings"
-	"code.betamike.com/mediocregopher/ginger/graph"
 )
 
-// Type aliases for convenience
+// Value wraps a go value in a way such that it will be uniquely identified
-type (
+// within any Graph and between Graphs. Use NewValue to create a Value instance.
-	Graph    = graph.Graph[OptionalValue, Value]
+// You can create an instance manually as long as ID is globally unique.
-	OpenEdge = graph.OpenEdge[OptionalValue, Value]
-)
-
-// Value represents a value which can be serialized by the gg text format.
 type Value struct {
-	// Only one of these fields may be set
+	ID string
-	Name   *string
+	V  interface{}
-	Number *int64
-	Graph  *Graph
 }
 
-// Name returns a name Value.
+// NewValue returns a Value instance wrapping any go value. The Value returned
-func Name(name string) Value {
+// will be independent of the passed in go value. So if the same go value is
-	return Value{Name: &name}
+// passed in twice then the two returned Value instances will be treated as
+// being different values by Graph.
+func NewValue(V interface{}) Value {
+	b := make([]byte, 8)
+	if _, err := rand.Read(b); err != nil {
+		panic(err)
+	}
+	return Value{
+		ID: hex.EncodeToString(b),
+		V:  V,
+	}
 }
 
-// Number returns a number Value.
+// VertexType enumerates the different possible vertex types
-func Number(n int64) Value {
+type VertexType string
-	return Value{Number: &n}
+
+const (
+	// ValueVertex is a Vertex which contains exactly one value and has at least
+	// one edge (either input or output)
+	ValueVertex VertexType = "value"
+
+	// JunctionVertex is a Vertex which contains two or more in edges and
+	// exactly one out edge
+	JunctionVertex VertexType = "junction"
+)
+
+// Edge is a uni-directional connection between two vertices with an attribute
+// value
+type Edge struct {
+	From  *Vertex
+	Value Value
+	To    *Vertex
 }
 
-// Equal returns true if the passed in Value is equivalent, ignoring the
+// Vertex is a vertex in a Graph. No fields should be modified directly, only
-// LexerToken on either Value.
+// through method calls
+type Vertex struct {
+	ID string
+	VertexType
+	Value   Value // Value is valid if-and-only-if VertexType is ValueVertex
+	In, Out []Edge
+}
+
+////////////////////////////////////////////////////////////////////////////////
+
+// OpenEdge is an un-realized Edge which can't be used for anything except
+// constructing graphs. It has no meaning on its own.
+type OpenEdge struct {
+	// fromV will be the source vertex as-if the vertex (and any sub-vertices of
+	// it) doesn't already exist in the graph. If it or it's sub-vertices does
+	// already that will need to be taken into account when persisting into the
+	// graph
+	fromV vertex
+	val   Value
+}
+
+func (oe OpenEdge) id() string {
+	return fmt.Sprintf("(%s,%s)", oe.fromV.id, oe.val.ID)
+}
+
+// vertex is a representation of a vertex in the graph. Each Graph contains a
+// set of all the Value vertex instances it knows about. Each of these contains
+// all the input OpenEdges which are known for it. So you can think of these
+// "top-level" Value vertex instances as root nodes in a tree, and each OpenEdge
+// as a branch.
 //
-// Will panic if the passed in v2 is not a Value from this package.
+// If a OpenEdge contains a fromV which is a Value that vertex won't have its in
-func (v Value) Equal(v2g graph.Value) bool {
+// slice populated no matter what. If fromV is a Junction it will be populated,
+// with any sub-Value's not being populated and so-on recursively
+//
+// When a view is constructed in makeView these Value instances are deduplicated
+// and the top-level one's in value is used to properly connect it.
+type vertex struct {
+	id string
+	VertexType
+	val Value
+	in  []OpenEdge
+}
 
-	v2 := v2g.(Value)
+func (v vertex) cp() vertex {
+	cp := v
+	cp.in = make([]OpenEdge, len(v.in))
+	copy(cp.in, v.in)
+	return cp
+}
 
-	switch {
+func (v vertex) hasOpenEdge(oe OpenEdge) bool {
+	oeID := oe.id()
+	for _, in := range v.in {
+		if in.id() == oeID {
+			return true
+		}
+	}
+	return false
+}
 
-	case v.Name != nil && v2.Name != nil && *v.Name == *v2.Name:
+func (v vertex) cpAndDelOpenEdge(oe OpenEdge) (vertex, bool) {
-		return true
+	oeID := oe.id()
+	for i, in := range v.in {
+		if in.id() == oeID {
+			v = v.cp()
+			v.in = append(v.in[:i], v.in[i+1:]...)
+			return v, true
+		}
+	}
+	return v, false
+}
 
-	case v.Number != nil && v2.Number != nil && *v.Number == *v2.Number:
+// Graph is a wrapper around a set of connected Vertices
-		return true
+type Graph struct {
+	vM map[string]vertex // only contains value vertices
 
-	case v.Graph != nil && v2.Graph != nil && v.Graph.Equal(v2.Graph):
+	// generated by makeView on-demand
-		return true
+	byVal map[string]*Vertex
+	all   map[string]*Vertex
+}
 
+// Null is the root empty graph, and is the base off which all graphs are built
+var Null = &Graph{
+	vM:    map[string]vertex{},
+	byVal: map[string]*Vertex{},
+	all:   map[string]*Vertex{},
+}
+
+// this does _not_ copy the view, as it's assumed the only reason to copy a
+// graph is to modify it anyway
+func (g *Graph) cp() *Graph {
+	cp := &Graph{
+		vM: make(map[string]vertex, len(g.vM)),
+	}
+	for vID, v := range g.vM {
+		cp.vM[vID] = v
+	}
+	return cp
+}
+
+////////////////////////////////////////////////////////////////////////////////
+// Graph creation
+
+func mkVertex(typ VertexType, val Value, ins ...OpenEdge) vertex {
+	v := vertex{VertexType: typ, in: ins}
+	switch typ {
+	case ValueVertex:
+		v.id = val.ID
+		v.val = val
+	case JunctionVertex:
+		inIDs := make([]string, len(ins))
+		for i := range ins {
+			inIDs[i] = ins[i].id()
+		}
+		v.id = "[" + strings.Join(inIDs, ",") + "]"
 	default:
+		panic(fmt.Sprintf("unknown vertex type %q", typ))
+	}
+	return v
+}
+
+// ValueOut creates a OpenEdge which, when used to construct a Graph, represents
+// an edge (with edgeVal attached to it) coming from the ValueVertex containing
+// val.
+//
+// When constructing Graphs, Value vertices are de-duplicated on their Value. So
+// multiple ValueOut OpenEdges constructed with the same val will be leaving the
+// same Vertex instance in the constructed Graph.
+func ValueOut(val, edgeVal Value) OpenEdge {
+	return OpenEdge{fromV: mkVertex(ValueVertex, val), val: edgeVal}
+}
+
+// JunctionOut creates a OpenEdge which, when used to construct a Graph,
+// represents an edge (with edgeVal attached to it) coming from the
+// JunctionVertex comprised of the given ordered-set of input edges.
+//
+// When constructing Graphs Junction vertices are de-duplicated on their input
+// edges. So multiple Junction OpenEdges constructed with the same set of input
+// edges will be leaving the same Junction instance in the constructed Graph.
+func JunctionOut(ins []OpenEdge, edgeVal Value) OpenEdge {
+	return OpenEdge{
+		fromV: mkVertex(JunctionVertex, Value{}, ins...),
+		val:   edgeVal,
+	}
+}
+
+// AddValueIn takes a OpenEdge and connects it to the Value Vertex containing
+// val, returning the new Graph which reflects that connection. Any Vertices
+// referenced within toe OpenEdge which do not yet exist in the Graph will also
+// be created in this step.
+func (g *Graph) AddValueIn(oe OpenEdge, val Value) *Graph {
+	to := mkVertex(ValueVertex, val)
+	toID := to.id
+
+	// if to is already in the graph, pull it out, as it might have existing in
+	// edges we want to keep
+	if exTo, ok := g.vM[toID]; ok {
+		to = exTo
+	}
+
+	// if the incoming edge already exists in to then there's nothing to do
+	if to.hasOpenEdge(oe) {
+		return g
+	}
+
+	to = to.cp()
+	to.in = append(to.in, oe)
+	g = g.cp()
+
+	// starting with to (which we always overwrite) go through vM and
+	// recursively add in any vertices which aren't already there
+	var persist func(vertex)
+	persist = func(v vertex) {
+		if v.VertexType == ValueVertex {
+			vID := v.id
+			if _, ok := g.vM[vID]; !ok {
+				g.vM[vID] = v
+			}
+		} else {
+			for _, e := range v.in {
+				persist(e.fromV)
+			}
+		}
+	}
+	delete(g.vM, toID)
+	persist(to)
+	for _, e := range to.in {
+		persist(e.fromV)
+	}
+
+	return g
+}
+
+// DelValueIn takes a OpenEdge and disconnects it from the Value Vertex
+// containing val, returning the new Graph which reflects the disconnection. If
+// the Value Vertex doesn't exist within the graph, or it doesn't have the given
+// OpenEdge, no changes are made. Any vertices referenced by toe OpenEdge for
+// which that edge is their only outgoing edge will be removed from the Graph.
+func (g *Graph) DelValueIn(oe OpenEdge, val Value) *Graph {
+	to := mkVertex(ValueVertex, val)
+	toID := to.id
+
+	// pull to out of the graph. if it's not there then bail
+	var ok bool
+	if to, ok = g.vM[toID]; !ok {
+		return g
+	}
+
+	// get new copy of to without the half-edge, or return if the half-edge
+	// wasn't even in to
+	to, ok = to.cpAndDelOpenEdge(oe)
+	if !ok {
+		return g
+	}
+	g = g.cp()
+	g.vM[toID] = to
+
+	// connectedTo returns whether the vertex has any connections with the
+	// vertex of the given id, descending recursively
+	var connectedTo func(string, vertex) bool
+	connectedTo = func(vID string, curr vertex) bool {
+		for _, in := range curr.in {
+			if in.fromV.VertexType == ValueVertex && in.fromV.id == vID {
+				return true
+			} else if in.fromV.VertexType == JunctionVertex && connectedTo(vID, in.fromV) {
+				return true
+			}
+		}
 		return false
 	}
-}
 
-func (v Value) String() string {
+	// isOrphaned returns whether the given vertex has any connections to other
-
+	// nodes in the graph
-	switch {
+	isOrphaned := func(v vertex) bool {
-
+		vID := v.id
-	case v.Name != nil:
+		if v, ok := g.vM[vID]; ok && len(v.in) > 0 {
-		return *v.Name
+			return false
-
+		}
-	case v.Number != nil:
+		for vID2, v2 := range g.vM {
-		return fmt.Sprint(*v.Number)
+			if vID2 == vID {
-
+				continue
-	case v.Graph != nil:
+			} else if connectedTo(vID, v2) {
-		return v.Graph.String()
+				return false
-
+			}
-	default:
+		}
-		panic("no fields set on Value")
-	}
-}
-
-// OptionalValue is a Value which may be unset. This is used for edge values,
-// since edges might not have a value.
-type OptionalValue struct {
-	Value
-	Valid bool
-}
-
-// None is the zero OptionalValue (hello rustaceans).
-var None OptionalValue
-
-// Some wraps a Value to be an OptionalValue.
-func Some(v Value) OptionalValue {
-	return OptionalValue{Valid: true, Value: v}
-}
-
-func (v OptionalValue) String() string {
-	if !v.Valid {
-		return "<none>"
-	}
-	return v.Value.String()
-}
-
-func (v OptionalValue) Equal(v2g graph.Value) bool {
-	var v2 OptionalValue
-
-	if v2Val, ok := v2g.(Value); ok {
-		v2 = Some(v2Val)
-	} else {
-		v2 = v2g.(OptionalValue)
-	}
-
-	if v.Valid != v2.Valid {
-		return false
-	} else if !v.Valid {
 		return true
 	}
 
-	return v.Value.Equal(v2.Value)
+	// if to is orphaned get rid of it
+	if isOrphaned(to) {
+		delete(g.vM, toID)
+	}
+
+	// rmOrphaned descends down the given OpenEdge and removes any Value
+	// Vertices referenced in it which are now orphaned
+	var rmOrphaned func(OpenEdge)
+	rmOrphaned = func(oe OpenEdge) {
+		if oe.fromV.VertexType == ValueVertex && isOrphaned(oe.fromV) {
+			delete(g.vM, oe.fromV.id)
+		} else if oe.fromV.VertexType == JunctionVertex {
+			for _, juncOe := range oe.fromV.in {
+				rmOrphaned(juncOe)
+			}
+		}
+	}
+	rmOrphaned(oe)
+
+	return g
+}
+
+// Union takes in another Graph and returns a new one which is the union of the
+// two. Value vertices which are shared between the two will be merged so that
+// the new vertex has the input edges of both.
+//
+// TODO it bothers me that the opposite of Disjoin is Union and not "Join"
+func (g *Graph) Union(g2 *Graph) *Graph {
+	g = g.cp()
+	for vID, v2 := range g2.vM {
+		v, ok := g.vM[vID]
+		if !ok {
+			v = v2
+		} else {
+			for _, v2e := range v2.in {
+				if !v.hasOpenEdge(v2e) {
+					v.in = append(v.in, v2e)
+				}
+			}
+		}
+		g.vM[vID] = v
+	}
+	return g
+}
+
+// Disjoin splits the Graph into as many independently connected Graphs as it
+// contains. Each Graph returned will have vertices connected only within itself
+// and not across to the other Graphs, and the Union of all returned Graphs will
+// be the original again.
+//
+// The order of the Graphs returned is not deterministic.
+//
+// Null.Disjoin() returns empty slice.
+func (g *Graph) Disjoin() []*Graph {
+	m := map[string]*Graph{}    // maps each id to the Graph it belongs to
+	mG := map[*Graph]struct{}{} // tracks unique Graphs created
+
+	var connectedTo func(vertex) *Graph
+	connectedTo = func(v vertex) *Graph {
+		if v.VertexType == ValueVertex {
+			if g := m[v.id]; g != nil {
+				return g
+			}
+		}
+		for _, oe := range v.in {
+			if g := connectedTo(oe.fromV); g != nil {
+				return g
+			}
+		}
+		return nil
+	}
+
+	// used upon finding out that previously-thought-to-be disconnected vertices
+	// aren't. Merges the two graphs they're connected into together into one
+	// and updates all state internal to this function accordingly.
+	rejoin := func(gDst, gSrc *Graph) {
+		for id, v := range gSrc.vM {
+			gDst.vM[id] = v
+			m[id] = gDst
+		}
+		delete(mG, gSrc)
+	}
+
+	var connectTo func(vertex, *Graph)
+	connectTo = func(v vertex, g *Graph) {
+		if v.VertexType == ValueVertex {
+			if g2, ok := m[v.id]; ok && g != g2 {
+				rejoin(g, g2)
+			}
+			m[v.id] = g
+		}
+		for _, oe := range v.in {
+			connectTo(oe.fromV, g)
+		}
+	}
+
+	for id, v := range g.vM {
+		gV := connectedTo(v)
+
+		// if gV is nil it means this vertex is part of a new Graph which
+		// nothing else has been connected to yet.
+		if gV == nil {
+			gV = Null.cp()
+			mG[gV] = struct{}{}
+		}
+		gV.vM[id] = v
+
+		// do this no matter what, because we want to descend in to the in edges
+		// and mark all of those as being part of this graph too
+		connectTo(v, gV)
+	}
+
+	gg := make([]*Graph, 0, len(mG))
+	for g := range mG {
+		gg = append(gg, g)
+	}
+	return gg
+}
+
+////////////////////////////////////////////////////////////////////////////////
+// Graph traversal
+
+func (g *Graph) makeView() {
+	if g.byVal != nil {
+		return
+	}
+
+	g.byVal = make(map[string]*Vertex, len(g.vM))
+	g.all = map[string]*Vertex{}
+
+	var getV func(vertex, bool) *Vertex
+	getV = func(v vertex, top bool) *Vertex {
+		V, ok := g.all[v.id]
+		if !ok {
+			V = &Vertex{ID: v.id, VertexType: v.VertexType, Value: v.val}
+			g.all[v.id] = V
+		}
+
+		// we can be sure all Value vertices will be called with top==true at
+		// some point, so we only need to descend into the input edges if:
+		// * top is true
+		// * this is a junction's first time being gotten
+		if !top && (ok || v.VertexType != JunctionVertex) {
+			return V
+		}
+
+		V.In = make([]Edge, 0, len(v.in))
+		for i := range v.in {
+			fromV := getV(v.in[i].fromV, false)
+			e := Edge{From: fromV, Value: v.in[i].val, To: V}
+			fromV.Out = append(fromV.Out, e)
+			V.In = append(V.In, e)
+		}
+
+		if v.VertexType == ValueVertex {
+			g.byVal[v.val.ID] = V
+		}
+
+		return V
+	}
+
+	for _, v := range g.vM {
+		getV(v, true)
+	}
+}
+
+// ValueVertex returns the Value Vertex for the given value. If the Graph
+// doesn't contain a vertex for the value then nil is returned
+func (g *Graph) ValueVertex(val Value) *Vertex {
+	g.makeView()
+	return g.byVal[val.ID]
+}
+
+// ValueVertices returns all Value Vertices in the Graph
+func (g *Graph) ValueVertices() []*Vertex {
+	g.makeView()
+	vv := make([]*Vertex, 0, len(g.byVal))
+	for _, v := range g.byVal {
+		vv = append(vv, v)
+	}
+	return vv
+}
+
+// Equal returns whether or not the two Graphs are equivalent in value
+func Equal(g1, g2 *Graph) bool {
+	if len(g1.vM) != len(g2.vM) {
+		return false
+	}
+	for v1ID, v1 := range g1.vM {
+		v2, ok := g2.vM[v1ID]
+		if !ok {
+			return false
+		}
+
+		// since the vertices are values we must make sure their input sets are
+		// the same (which is tricky since they're unordered, unlike a
+		// junction's)
+		if len(v1.in) != len(v2.in) {
+			return false
+		}
+		for _, in := range v1.in {
+			if !v2.hasOpenEdge(in) {
+				return false
+			}
+		}
+	}
+	return true
+}
+
+// TODO Walk, but by edge
+// TODO Walk, but without end. AKA FSM
+
+// Iter will iterate through the Graph's vertices, calling the callback on every
+// Vertex in the Graph once. The vertex order used is non-deterministic.  If the
+// callback returns false the iteration is stopped.
+func (g *Graph) Iter(callback func(*Vertex) bool) {
+	g.makeView()
+	if len(g.byVal) == 0 {
+		return
+	}
+
+	seen := make(map[*Vertex]bool, len(g.byVal))
+	var innerWalk func(*Vertex) bool
+	innerWalk = func(v *Vertex) bool {
+		if seen[v] {
+			return true
+		} else if !callback(v) {
+			return false
+		}
+		seen[v] = true
+		for _, e := range v.In {
+			if !innerWalk(e.From) {
+				return false
+			}
+		}
+		return true
+	}
+
+	for _, v := range g.byVal {
+		if !innerWalk(v) {
+			return
+		}
+	}
+}
+
+// ByID returns all vertices indexed by their ID field
+func (g *Graph) ByID() map[string]*Vertex {
+	g.makeView()
+	return g.all
 }

gg/gg_test.go

@@ -0,0 +1,665 @@
+package gg
+
+import (
+	"fmt"
+	"sort"
+	"strings"
+	. "testing"
+
+	"github.com/stretchr/testify/assert"
+)
+
+func edge(val Value, from *Vertex) Edge {
+	return Edge{Value: val, From: from}
+}
+
+func value(val Value, in ...Edge) *Vertex {
+	return &Vertex{
+		VertexType: ValueVertex,
+		Value:      val,
+		In:         in,
+	}
+}
+
+func junction(val Value, in ...Edge) Edge {
+	return Edge{
+		From: &Vertex{
+			VertexType: JunctionVertex,
+			In:         in,
+		},
+		Value: val,
+	}
+}
+
+func assertVertexEqual(t *T, exp, got *Vertex, msgAndArgs ...interface{}) bool {
+	var assertInner func(*Vertex, *Vertex, map[*Vertex]bool) bool
+	assertInner = func(exp, got *Vertex, m map[*Vertex]bool) bool {
+		// if got is already in m then we've already looked at it
+		if m[got] {
+			return true
+		}
+		m[got] = true
+
+		assert.Equal(t, exp.VertexType, got.VertexType, msgAndArgs...)
+		assert.Equal(t, exp.Value, got.Value, msgAndArgs...)
+		if !assert.Len(t, got.In, len(exp.In), msgAndArgs...) {
+			return false
+		}
+		for i := range exp.In {
+			assertInner(exp.In[i].From, got.In[i].From, m)
+			assert.Equal(t, exp.In[i].Value, got.In[i].Value, msgAndArgs...)
+			assert.Equal(t, got, got.In[i].To)
+			assert.Contains(t, got.In[i].From.Out, got.In[i])
+		}
+		return true
+
+	}
+	return assertInner(exp, got, map[*Vertex]bool{})
+}
+
+func assertIter(t *T, expVals, expJuncs int, g *Graph, msgAndArgs ...interface{}) {
+	seen := map[*Vertex]bool{}
+	var gotVals, gotJuncs int
+	g.Iter(func(v *Vertex) bool {
+		assert.NotContains(t, seen, v, msgAndArgs...)
+		seen[v] = true
+		if v.VertexType == ValueVertex {
+			gotVals++
+		} else {
+			gotJuncs++
+		}
+		return true
+	})
+	assert.Equal(t, expVals, gotVals, msgAndArgs...)
+	assert.Equal(t, expJuncs, gotJuncs, msgAndArgs...)
+}
+
+type graphTest struct {
+	name              string
+	out               func() *Graph
+	exp               []*Vertex
+	numVals, numJuncs int
+}
+
+func mkTest(name string, out func() *Graph, numVals, numJuncs int, exp ...*Vertex) graphTest {
+	return graphTest{
+		name:    name,
+		out:     out,
+		exp:     exp,
+		numVals: numVals, numJuncs: numJuncs,
+	}
+}
+
+func TestGraph(t *T) {
+	var (
+		v0  = NewValue("v0")
+		v1  = NewValue("v1")
+		v2  = NewValue("v2")
+		v3  = NewValue("v3")
+		e0  = NewValue("e0")
+		e00 = NewValue("e00")
+		e01 = NewValue("e01")
+		e1  = NewValue("e1")
+		e10 = NewValue("e10")
+		e11 = NewValue("e11")
+		e2  = NewValue("e2")
+		e20 = NewValue("e20")
+		e21 = NewValue("e21")
+		ej0 = NewValue("ej0")
+		ej1 = NewValue("ej1")
+		ej2 = NewValue("ej2")
+	)
+	tests := []graphTest{
+		mkTest(
+			"values-basic",
+			func() *Graph {
+				return Null.AddValueIn(ValueOut(v0, e0), v1)
+			},
+			2, 0,
+			value(v0),
+			value(v1, edge(e0, value(v0))),
+		),
+
+		mkTest(
+			"values-2edges",
+			func() *Graph {
+				g0 := Null.AddValueIn(ValueOut(v0, e0), v2)
+				return g0.AddValueIn(ValueOut(v1, e1), v2)
+			},
+			3, 0,
+			value(v0),
+			value(v1),
+			value(v2,
+				edge(e0, value(v0)),
+				edge(e1, value(v1)),
+			),
+		),
+
+		mkTest(
+			"values-separate",
+			func() *Graph {
+				g0 := Null.AddValueIn(ValueOut(v0, e0), v1)
+				return g0.AddValueIn(ValueOut(v2, e2), v3)
+			},
+			4, 0,
+			value(v0),
+			value(v1, edge(e0, value(v0))),
+			value(v2),
+			value(v3, edge(e2, value(v2))),
+		),
+
+		mkTest(
+			"values-circular",
+			func() *Graph {
+				return Null.AddValueIn(ValueOut(v0, e0), v0)
+			},
+			1, 0,
+			value(v0, edge(e0, value(v0))),
+		),
+
+		mkTest(
+			"values-circular2",
+			func() *Graph {
+				g0 := Null.AddValueIn(ValueOut(v0, e0), v1)
+				return g0.AddValueIn(ValueOut(v1, e1), v0)
+			},
+			2, 0,
+			value(v0, edge(e1, value(v1, edge(e0, value(v0))))),
+			value(v1, edge(e0, value(v0, edge(e1, value(v1))))),
+		),
+
+		mkTest(
+			"values-circular3",
+			func() *Graph {
+				g0 := Null.AddValueIn(ValueOut(v0, e0), v1)
+				g1 := g0.AddValueIn(ValueOut(v1, e1), v2)
+				return g1.AddValueIn(ValueOut(v2, e2), v1)
+			},
+			3, 0,
+			value(v0),
+			value(v1,
+				edge(e0, value(v0)),
+				edge(e2, value(v2, edge(e1, value(v1)))),
+			),
+			value(v2, edge(e1, value(v1,
+				edge(e0, value(v0)),
+				edge(e2, value(v2)),
+			))),
+		),
+
+		mkTest(
+			"junction-basic",
+			func() *Graph {
+				e0 := ValueOut(v0, e0)
+				e1 := ValueOut(v1, e1)
+				ej0 := JunctionOut([]OpenEdge{e0, e1}, ej0)
+				return Null.AddValueIn(ej0, v2)
+			},
+			3, 1,
+			value(v0), value(v1),
+			value(v2, junction(ej0,
+				edge(e0, value(v0)),
+				edge(e1, value(v1)),
+			)),
+		),
+
+		mkTest(
+			"junction-basic2",
+			func() *Graph {
+				e00 := ValueOut(v0, e00)
+				e10 := ValueOut(v1, e10)
+				ej0 := JunctionOut([]OpenEdge{e00, e10}, ej0)
+				e01 := ValueOut(v0, e01)
+				e11 := ValueOut(v1, e11)
+				ej1 := JunctionOut([]OpenEdge{e01, e11}, ej1)
+				ej2 := JunctionOut([]OpenEdge{ej0, ej1}, ej2)
+				return Null.AddValueIn(ej2, v2)
+			},
+			3, 3,
+			value(v0), value(v1),
+			value(v2, junction(ej2,
+				junction(ej0,
+					edge(e00, value(v0)),
+					edge(e10, value(v1)),
+				),
+				junction(ej1,
+					edge(e01, value(v0)),
+					edge(e11, value(v1)),
+				),
+			)),
+		),
+
+		mkTest(
+			"junction-circular",
+			func() *Graph {
+				e0 := ValueOut(v0, e0)
+				e1 := ValueOut(v1, e1)
+				ej0 := JunctionOut([]OpenEdge{e0, e1}, ej0)
+				g0 := Null.AddValueIn(ej0, v2)
+				e20 := ValueOut(v2, e20)
+				g1 := g0.AddValueIn(e20, v0)
+				e21 := ValueOut(v2, e21)
+				return g1.AddValueIn(e21, v1)
+			},
+			3, 1,
+			value(v0, edge(e20, value(v2, junction(ej0,
+				edge(e0, value(v0)),
+				edge(e1, value(v1, edge(e21, value(v2)))),
+			)))),
+			value(v1, edge(e21, value(v2, junction(ej0,
+				edge(e0, value(v0, edge(e20, value(v2)))),
+				edge(e1, value(v1)),
+			)))),
+			value(v2, junction(ej0,
+				edge(e0, value(v0, edge(e20, value(v2)))),
+				edge(e1, value(v1, edge(e21, value(v2)))),
+			)),
+		),
+	}
+
+	for i := range tests {
+		t.Logf("test[%d]:%q", i, tests[i].name)
+		out := tests[i].out()
+		for j, exp := range tests[i].exp {
+			msgAndArgs := []interface{}{
+				"tests[%d].name:%q exp[%d].val:%q",
+				i, tests[i].name, j, exp.Value.V.(string),
+			}
+			v := out.ValueVertex(exp.Value)
+			if !assert.NotNil(t, v, msgAndArgs...) {
+				continue
+			}
+			assertVertexEqual(t, exp, v, msgAndArgs...)
+		}
+
+		msgAndArgs := []interface{}{
+			"tests[%d].name:%q",
+			i, tests[i].name,
+		}
+
+		// sanity check that graphs are equal to themselves
+		assert.True(t, Equal(out, out), msgAndArgs...)
+
+		// test the Iter method in here too
+		assertIter(t, tests[i].numVals, tests[i].numJuncs, out, msgAndArgs...)
+	}
+}
+
+func TestGraphImmutability(t *T) {
+	v0 := NewValue("v0")
+	v1 := NewValue("v1")
+	e0 := NewValue("e0")
+	oe0 := ValueOut(v0, e0)
+	g0 := Null.AddValueIn(oe0, v1)
+	assert.Nil(t, Null.ValueVertex(v0))
+	assert.Nil(t, Null.ValueVertex(v1))
+	assert.NotNil(t, g0.ValueVertex(v0))
+	assert.NotNil(t, g0.ValueVertex(v1))
+
+	// half-edges should be re-usable
+	v2 := NewValue("v2")
+	v3a, v3b := NewValue("v3a"), NewValue("v3b")
+	e1 := NewValue("e1")
+	oe1 := ValueOut(v2, e1)
+	g1a := g0.AddValueIn(oe1, v3a)
+	g1b := g0.AddValueIn(oe1, v3b)
+	assertVertexEqual(t, value(v3a, edge(e1, value(v2))), g1a.ValueVertex(v3a))
+	assert.Nil(t, g1a.ValueVertex(v3b))
+	assertVertexEqual(t, value(v3b, edge(e1, value(v2))), g1b.ValueVertex(v3b))
+	assert.Nil(t, g1b.ValueVertex(v3a))
+
+	// ... even re-usable twice in succession
+	v3 := NewValue("v3")
+	v4 := NewValue("v4")
+	g2 := g0.AddValueIn(oe1, v3).AddValueIn(oe1, v4)
+	assert.Nil(t, g2.ValueVertex(v3b))
+	assert.Nil(t, g2.ValueVertex(v3a))
+	assertVertexEqual(t, value(v3, edge(e1, value(v2))), g2.ValueVertex(v3))
+	assertVertexEqual(t, value(v4, edge(e1, value(v2))), g2.ValueVertex(v4))
+}
+
+func TestGraphDelValueIn(t *T) {
+	v0 := NewValue("v0")
+	v1 := NewValue("v1")
+	e0 := NewValue("e0")
+	{ // removing from null
+		g := Null.DelValueIn(ValueOut(v0, e0), v1)
+		assert.True(t, Equal(Null, g))
+	}
+
+	e1 := NewValue("e1")
+	{ // removing edge from vertex which doesn't have that edge
+		g0 := Null.AddValueIn(ValueOut(v0, e0), v1)
+		g1 := g0.DelValueIn(ValueOut(v0, e1), v1)
+		assert.True(t, Equal(g0, g1))
+	}
+
+	{ // removing only edge
+		oe := ValueOut(v0, e0)
+		g0 := Null.AddValueIn(oe, v1)
+		g1 := g0.DelValueIn(oe, v1)
+		assert.True(t, Equal(Null, g1))
+	}
+
+	ej0 := NewValue("ej0")
+	v2 := NewValue("v2")
+	{ // removing only edge (junction)
+		oe := JunctionOut([]OpenEdge{
+			ValueOut(v0, e0),
+			ValueOut(v1, e1),
+		}, ej0)
+		g0 := Null.AddValueIn(oe, v2)
+		g1 := g0.DelValueIn(oe, v2)
+		assert.True(t, Equal(Null, g1))
+	}
+
+	{ // removing one of two edges
+		oe := ValueOut(v1, e0)
+		g0 := Null.AddValueIn(ValueOut(v0, e0), v2)
+		g1 := g0.AddValueIn(oe, v2)
+		g2 := g1.DelValueIn(oe, v2)
+		assert.True(t, Equal(g0, g2))
+		assert.NotNil(t, g2.ValueVertex(v0))
+		assert.Nil(t, g2.ValueVertex(v1))
+		assert.NotNil(t, g2.ValueVertex(v2))
+	}
+
+	e2 := NewValue("e2")
+	eja, ejb := NewValue("eja"), NewValue("ejb")
+	v3 := NewValue("v3")
+	{ // removing one of two edges (junction)
+		e0 := ValueOut(v0, e0)
+		e1 := ValueOut(v1, e1)
+		e2 := ValueOut(v2, e2)
+		oeA := JunctionOut([]OpenEdge{e0, e1}, eja)
+		oeB := JunctionOut([]OpenEdge{e1, e2}, ejb)
+		g0a := Null.AddValueIn(oeA, v3)
+		g0b := Null.AddValueIn(oeB, v3)
+		g1 := g0a.Union(g0b).DelValueIn(oeA, v3)
+		assert.True(t, Equal(g1, g0b))
+		assert.Nil(t, g1.ValueVertex(v0))
+		assert.NotNil(t, g1.ValueVertex(v1))
+		assert.NotNil(t, g1.ValueVertex(v2))
+		assert.NotNil(t, g1.ValueVertex(v3))
+	}
+
+	{ // removing one of two edges in circular graph
+		e0 := ValueOut(v0, e0)
+		e1 := ValueOut(v1, e1)
+		g0 := Null.AddValueIn(e0, v1).AddValueIn(e1, v0)
+		g1 := g0.DelValueIn(e0, v1)
+		assert.True(t, Equal(Null.AddValueIn(e1, v0), g1))
+		assert.NotNil(t, g1.ValueVertex(v0))
+		assert.NotNil(t, g1.ValueVertex(v1))
+	}
+
+	ej := NewValue("ej")
+	{ // removing to's only edge, sub-nodes have edge to each other
+		oej := JunctionOut([]OpenEdge{
+			ValueOut(v0, ej0),
+			ValueOut(v1, ej0),
+		}, ej)
+		g0 := Null.AddValueIn(oej, v2)
+		e0 := ValueOut(v0, e0)
+		g1 := g0.AddValueIn(e0, v1)
+		g2 := g1.DelValueIn(oej, v2)
+		assert.True(t, Equal(Null.AddValueIn(e0, v1), g2))
+		assert.NotNil(t, g2.ValueVertex(v0))
+		assert.NotNil(t, g2.ValueVertex(v1))
+		assert.Nil(t, g2.ValueVertex(v2))
+	}
+}
+
+// deterministically hashes a Graph
+func graphStr(g *Graph) string {
+	var vStr func(vertex) string
+	var oeStr func(OpenEdge) string
+	vStr = func(v vertex) string {
+		if v.VertexType == ValueVertex {
+			return fmt.Sprintf("v:%q\n", v.val.V.(string))
+		}
+		s := fmt.Sprintf("j:%d\n", len(v.in))
+		ssOE := make([]string, len(v.in))
+		for i := range v.in {
+			ssOE[i] = oeStr(v.in[i])
+		}
+		sort.Strings(ssOE)
+		return s + strings.Join(ssOE, "")
+	}
+	oeStr = func(oe OpenEdge) string {
+		s := fmt.Sprintf("oe:%q\n", oe.val.V.(string))
+		return s + vStr(oe.fromV)
+	}
+	sVV := make([]string, 0, len(g.vM))
+	for _, v := range g.vM {
+		sVV = append(sVV, vStr(v))
+	}
+	sort.Strings(sVV)
+	return strings.Join(sVV, "")
+}
+
+func assertEqualSets(t *T, exp, got []*Graph) bool {
+	if !assert.Equal(t, len(exp), len(got)) {
+		return false
+	}
+
+	m := map[*Graph]string{}
+	for _, g := range exp {
+		m[g] = graphStr(g)
+	}
+	for _, g := range got {
+		m[g] = graphStr(g)
+	}
+
+	sort.Slice(exp, func(i, j int) bool {
+		return m[exp[i]] < m[exp[j]]
+	})
+	sort.Slice(got, func(i, j int) bool {
+		return m[got[i]] < m[got[j]]
+	})
+
+	b := true
+	for i := range exp {
+		b = b || assert.True(t, Equal(exp[i], got[i]), "i:%d exp:%q got:%q", i, m[exp[i]], m[got[i]])
+	}
+	return b
+}
+
+func TestGraphUnion(t *T) {
+	assertUnion := func(g1, g2 *Graph) *Graph {
+		ga := g1.Union(g2)
+		gb := g2.Union(g1)
+		assert.True(t, Equal(ga, gb))
+		return ga
+	}
+
+	assertDisjoin := func(g *Graph, exp ...*Graph) {
+		ggDisj := g.Disjoin()
+		assertEqualSets(t, exp, ggDisj)
+	}
+
+	v0 := NewValue("v0")
+	v1 := NewValue("v1")
+	e0 := NewValue("e0")
+	{ // Union with Null
+		assert.True(t, Equal(Null, Null.Union(Null)))
+
+		g := Null.AddValueIn(ValueOut(v0, e0), v1)
+		assert.True(t, Equal(g, assertUnion(g, Null)))
+
+		assertDisjoin(g, g)
+	}
+
+	v2 := NewValue("v2")
+	v3 := NewValue("v3")
+	e1 := NewValue("e1")
+	{ // Two disparate graphs union'd
+		g0 := Null.AddValueIn(ValueOut(v0, e0), v1)
+		g1 := Null.AddValueIn(ValueOut(v2, e1), v3)
+		g := assertUnion(g0, g1)
+		assertVertexEqual(t, value(v0), g.ValueVertex(v0))
+		assertVertexEqual(t, value(v1, edge(e0, value(v0))), g.ValueVertex(v1))
+		assertVertexEqual(t, value(v2), g.ValueVertex(v2))
+		assertVertexEqual(t, value(v3, edge(e1, value(v2))), g.ValueVertex(v3))
+
+		assertDisjoin(g, g0, g1)
+	}
+
+	va0, vb0 := NewValue("va0"), NewValue("vb0")
+	va1, vb1 := NewValue("va1"), NewValue("vb1")
+	va2, vb2 := NewValue("va2"), NewValue("vb2")
+	ea0, eb0 := NewValue("ea0"), NewValue("eb0")
+	ea1, eb1 := NewValue("ea1"), NewValue("eb1")
+	eaj, ebj := NewValue("eaj"), NewValue("ebj")
+	{ // Two disparate graphs with junctions
+		ga := Null.AddValueIn(JunctionOut([]OpenEdge{
+			ValueOut(va0, ea0),
+			ValueOut(va1, ea1),
+		}, eaj), va2)
+		gb := Null.AddValueIn(JunctionOut([]OpenEdge{
+			ValueOut(vb0, eb0),
+			ValueOut(vb1, eb1),
+		}, ebj), vb2)
+		g := assertUnion(ga, gb)
+		assertVertexEqual(t, value(va0), g.ValueVertex(va0))
+		assertVertexEqual(t, value(va1), g.ValueVertex(va1))
+		assertVertexEqual(t,
+			value(va2, junction(eaj,
+				edge(ea0, value(va0)),
+				edge(ea1, value(va1)))),
+			g.ValueVertex(va2),
+		)
+		assertVertexEqual(t, value(vb0), g.ValueVertex(vb0))
+		assertVertexEqual(t, value(vb1), g.ValueVertex(vb1))
+		assertVertexEqual(t,
+			value(vb2, junction(ebj,
+				edge(eb0, value(vb0)),
+				edge(eb1, value(vb1)))),
+			g.ValueVertex(vb2),
+		)
+
+		assertDisjoin(g, ga, gb)
+	}
+
+	{ // Two partially overlapping graphs
+		g0 := Null.AddValueIn(ValueOut(v0, e0), v2)
+		g1 := Null.AddValueIn(ValueOut(v1, e1), v2)
+		g := assertUnion(g0, g1)
+		assertVertexEqual(t, value(v0), g.ValueVertex(v0))
+		assertVertexEqual(t, value(v1), g.ValueVertex(v1))
+		assertVertexEqual(t,
+			value(v2,
+				edge(e0, value(v0)),
+				edge(e1, value(v1)),
+			),
+			g.ValueVertex(v2),
+		)
+
+		assertDisjoin(g, g)
+	}
+
+	ej0 := NewValue("ej0")
+	ej1 := NewValue("ej1")
+	{ // two partially overlapping graphs with junctions
+		g0 := Null.AddValueIn(JunctionOut([]OpenEdge{
+			ValueOut(v0, e0),
+			ValueOut(v1, e1),
+		}, ej0), v2)
+		g1 := Null.AddValueIn(JunctionOut([]OpenEdge{
+			ValueOut(v0, e0),
+			ValueOut(v1, e1),
+		}, ej1), v2)
+		g := assertUnion(g0, g1)
+		assertVertexEqual(t, value(v0), g.ValueVertex(v0))
+		assertVertexEqual(t, value(v1), g.ValueVertex(v1))
+		assertVertexEqual(t,
+			value(v2,
+				junction(ej0, edge(e0, value(v0)), edge(e1, value(v1))),
+				junction(ej1, edge(e0, value(v0)), edge(e1, value(v1))),
+			),
+			g.ValueVertex(v2),
+		)
+
+		assertDisjoin(g, g)
+	}
+
+	{ // Two equal graphs
+		g0 := Null.AddValueIn(ValueOut(v0, e0), v1)
+		g := assertUnion(g0, g0)
+		assertVertexEqual(t, value(v0), g.ValueVertex(v0))
+		assertVertexEqual(t,
+			value(v1, edge(e0, value(v0))),
+			g.ValueVertex(v1),
+		)
+	}
+
+	{ // Two equal graphs with junctions
+		g0 := Null.AddValueIn(JunctionOut([]OpenEdge{
+			ValueOut(v0, e0),
+			ValueOut(v1, e1),
+		}, ej0), v2)
+		g := assertUnion(g0, g0)
+		assertVertexEqual(t, value(v0), g.ValueVertex(v0))
+		assertVertexEqual(t, value(v1), g.ValueVertex(v1))
+		assertVertexEqual(t,
+			value(v2,
+				junction(ej0, edge(e0, value(v0)), edge(e1, value(v1))),
+			),
+			g.ValueVertex(v2),
+		)
+	}
+}
+
+func TestGraphEqual(t *T) {
+	assertEqual := func(g1, g2 *Graph) {
+		assert.True(t, Equal(g1, g2))
+		assert.True(t, Equal(g2, g1))
+	}
+
+	assertNotEqual := func(g1, g2 *Graph) {
+		assert.False(t, Equal(g1, g2))
+		assert.False(t, Equal(g2, g1))
+	}
+
+	assertEqual(Null, Null) // duh
+
+	v0 := NewValue("v0")
+	v1 := NewValue("v1")
+	v2 := NewValue("v2")
+	e0 := NewValue("e0")
+	e1 := NewValue("e1")
+	e1a, e1b := NewValue("e1a"), NewValue("e1b")
+	{
+		// graph is equal to itself, not to null
+		e0 := ValueOut(v0, e0)
+		g0 := Null.AddValueIn(e0, v1)
+		assertNotEqual(g0, Null)
+		assertEqual(g0, g0)
+
+		// adding the an existing edge again shouldn't do anything
+		assertEqual(g0, g0.AddValueIn(e0, v1))
+
+		// g1a and g1b have the same vertices, but the edges are different
+		g1a := g0.AddValueIn(ValueOut(v0, e1a), v2)
+		g1b := g0.AddValueIn(ValueOut(v0, e1b), v2)
+		assertNotEqual(g1a, g1b)
+	}
+
+	{ // equal construction should yield equality, even if out of order
+		ga := Null.AddValueIn(ValueOut(v0, e0), v1)
+		ga = ga.AddValueIn(ValueOut(v1, e1), v2)
+		gb := Null.AddValueIn(ValueOut(v1, e1), v2)
+		gb = gb.AddValueIn(ValueOut(v0, e0), v1)
+		assertEqual(ga, gb)
+	}
+
+	ej := NewValue("ej")
+	{ // junction basic test
+		e0 := ValueOut(v0, e0)
+		e1 := ValueOut(v1, e1)
+		ga := Null.AddValueIn(JunctionOut([]OpenEdge{e0, e1}, ej), v2)
+		gb := Null.AddValueIn(JunctionOut([]OpenEdge{e1, e0}, ej), v2)
+		assertEqual(ga, ga)
+		assertNotEqual(ga, gb)
+	}
+}

gg/grammar/example_test.go

@@ -1,154 +0,0 @@
-package grammar_test
-
-import (
-	"bytes"
-	"fmt"
-	"strconv"
-	"strings"
-
-	"code.betamike.com/mediocregopher/ginger/gg/grammar"
-	"golang.org/x/exp/slices"
-)
-
-/*
-This example demonstrates how to describe the following EBNF using the grammar
-package:
-
-```
-<digit>           ::= "0" | "1" | "2" | "3" | "4" | "5" | "6" | "7" | "8" | "9"
-<positive-number> ::= <digit>+
-<negative-number> ::= "-" <positive-number>
-<number>          ::= <negative-number> | <positive-number>
-
-<list-el-tail> ::= <element> <list-tail>
-<list-tail>    ::= ")" | "," <list-el-tail>
-<list-head>    ::= ")" | <list-el-tail>
-<list>         ::= "(" <list-head>
-
-<element> ::= <number> | <list>
-```
-*/
-
-// Element represents an element of a list, which can be either a number or a
-// sub-list.
-type Element struct {
-	Number int64
-	List   []Element
-}
-
-func (e Element) String() string {
-	if e.List != nil {
-		listElStrs := make([]string, len(e.List))
-		for i := range e.List {
-			listElStrs[i] = e.List[i].String()
-		}
-		return fmt.Sprintf("(%s)", strings.Join(listElStrs, ","))
-	}
-
-	return fmt.Sprint(e.Number)
-}
-
-var (
-	digit = grammar.RuneFunc(
-		"digit", func(r rune) bool { return '0' <= r && r <= '9' },
-	)
-
-	positiveNumber = grammar.StringFromRunes(grammar.OneOrMore(digit))
-
-	negativeNumber = grammar.Reduction(
-		grammar.Rune('-'),
-		positiveNumber,
-		func(
-			neg grammar.Located[rune], posNum grammar.Located[string],
-		) grammar.Located[string] {
-			return grammar.Locate(neg.Location, string(neg.Value)+posNum.Value)
-		},
-	)
-
-	number = grammar.Named(
-		"number",
-		grammar.Mapping(
-			grammar.FirstOf(negativeNumber, positiveNumber),
-			func(str grammar.Located[string]) Element {
-				i, err := strconv.ParseInt(str.Value, 10, 64)
-				if err != nil {
-					panic(fmt.Errorf("parsing %q as int: %w", str, err))
-				}
-				return Element{Number: i}
-			},
-		),
-	)
-
-	// Because the list/element definitions are recursive it requires using
-	// SymbolPtrs, which is easier to do via a global initialization function
-	// like this.
-	list = func() grammar.Symbol[Element] {
-
-		type listState []Element
-
-		var (
-			listTail = new(grammar.SymbolPtr[listState])
-			list     = new(grammar.SymbolPtr[Element])
-			element  = new(grammar.SymbolPtr[Element])
-
-			// Right parenthesis indicates the end of a list, at which point we
-			// can initialize the state which gets returned down the stack.
-			listTerm = grammar.Mapping(
-				grammar.Rune(')'),
-				func(grammar.Located[rune]) listState { return listState{} },
-			)
-
-			listElTail = grammar.Reduction[Element, listState, listState](
-				element,
-				listTail,
-				func(el Element, ls listState) listState {
-					ls = append(ls, el)
-					return ls
-				},
-			)
-
-			listHead = grammar.FirstOf(listTerm, listElTail)
-		)
-
-		listTail.Symbol = grammar.FirstOf(
-			listTerm,
-			grammar.Prefixed(grammar.Rune(','), listElTail),
-		)
-
-		list.Symbol = grammar.Named(
-			"list",
-			grammar.Reduction[grammar.Located[rune], listState, Element](
-				grammar.Rune('('),
-				listHead,
-				func(_ grammar.Located[rune], ls listState) Element {
-					slices.Reverse(ls)
-					return Element{List: []Element(ls)}
-				},
-			),
-		)
-
-		element.Symbol = grammar.FirstOf[Element](number, list)
-
-		return list
-	}()
-)
-
-func Example() {
-	r := grammar.NewReader(bytes.NewBufferString(
-		`()` + `(1,(2,-3),4)` + `(ERROR`,
-	))
-
-	l1, err := list.Decode(r)
-	fmt.Println(l1, err)
-
-	l2, err := list.Decode(r)
-	fmt.Println(l2, err)
-
-	_, err = list.Decode(r)
-	fmt.Println(err)
-
-	// Output:
-	// () <nil>
-	// (1,(2,-3),4) <nil>
-	// 1:16: expected ')' or number or list
-}

gg/grammar/grammar.go

@@ -1,27 +0,0 @@
-// Package grammar is used for parsing a stream of runes according to a set of
-// grammatical rules. This package only supports context-free grammars.
-package grammar
-
-import "fmt"
-
-// Stringer is a convenience tool for working with fmt.Stringer. Exactly one of
-// the fields must be set, and will be used to implement the fmt.Stringer
-// interface.
-type Stringer struct {
-	I fmt.Stringer
-	F func() string
-	S string
-}
-
-func (s Stringer) String() string {
-	switch {
-	case s.I != nil:
-		return s.I.String()
-	case s.F != nil:
-		return s.F()
-	case s.S != "":
-		return s.S
-	default:
-		panic("no fields set on Stringer")
-	}
-}

gg/grammar/location.go

@@ -1,32 +0,0 @@
-package grammar
-
-import "fmt"
-
-// Location indicates a position in a stream of runes identified by column
-// within newline-separated rows.
-type Location struct {
-	Row, Col int
-}
-
-func (l Location) errf(str string, args ...any) LocatedError {
-	return LocatedError{l, fmt.Errorf(str, args...)}
-}
-
-// Located wraps a value so that it has a Location attached to it.
-type Located[T any] struct {
-	Location
-	Value T
-}
-
-// Locate returns a Located instance combining the given values.
-func Locate[T any](l Location, v T) Located[T] {
-	return Located[T]{l, v}
-}
-
-// LocatedError is an error related to a specific point within a stream of
-// runes.
-type LocatedError Located[error]
-
-func (e LocatedError) Error() string {
-	return fmt.Sprintf("%d:%d: %v", e.Row, e.Col, e.Value)
-}

gg/grammar/reader.go

@@ -1,74 +0,0 @@
-package grammar
-
-import (
-	"bufio"
-	"io"
-)
-
-// Reader is used for reading Runes from a stream.
-type Reader interface {
-
-	// ReadRune reads the next Rune off the stream, or returns io.EOF.
-	ReadRune() (Located[rune], error)
-
-	// UnreadRune can be used to place a Rune onto an internal buffer, such that
-	// the Rune will be the next to be read using ReadRune. If called multiple
-	// times then ReadRune will produce the given Runes in LIFO order.
-	UnreadRune(Located[rune])
-
-	// NextLocation returns the Location of the next Rune which will be returned
-	// with ReadRune.
-	NextLocation() Location
-}
-
-type reader struct {
-	br        *bufio.Reader
-	brNextLoc Location
-
-	unread []Located[rune]
-}
-
-// NewReader wraps the io.Reader as a Reader. The given Reader should not be
-// read from after this call.
-func NewReader(r io.Reader) Reader {
-	return &reader{
-		br:        bufio.NewReader(r),
-		brNextLoc: Location{Row: 1, Col: 1},
-	}
-}
-
-func (rr *reader) ReadRune() (Located[rune], error) {
-	if len(rr.unread) > 0 {
-		r := rr.unread[len(rr.unread)-1]
-		rr.unread = rr.unread[:len(rr.unread)-1]
-		return r, nil
-	}
-
-	loc := rr.brNextLoc
-
-	r, _, err := rr.br.ReadRune()
-	if err != nil {
-		return Located[rune]{}, err
-	}
-
-	if r == '\n' {
-		rr.brNextLoc.Row++
-		rr.brNextLoc.Col = 1
-	} else {
-		rr.brNextLoc.Col++
-	}
-
-	return Located[rune]{loc, r}, nil
-}
-
-func (rr *reader) UnreadRune(r Located[rune]) {
-	rr.unread = append(rr.unread, r)
-}
-
-func (rr *reader) NextLocation() Location {
-	if len(rr.unread) > 0 {
-		return rr.unread[len(rr.unread)-1].Location
-	}
-
-	return rr.brNextLoc
-}

gg/grammar/symbol.go

@@ -1,306 +0,0 @@
-package grammar
-
-import (
-	"errors"
-	"fmt"
-	"io"
-	"strings"
-)
-
-// ErrNoMatch is used by Symbol's Decode method, see that method's docs for more
-// details.
-var ErrNoMatch = errors.New("no match")
-
-// Symbol represents a symbol in the grammar. A Symbol is expected to be
-// stateless, and is usually constructed from other Symbols using functions in
-// this package.
-type Symbol[T any] interface {
-	fmt.Stringer // Used when generating errors related to this Symbol, e.g. "number"
-
-	// Decode reads and parses a value represented by this Symbol off the
-	// Reader.
-	//
-	// This may return ErrNoMatch to indicate that the upcoming data on the
-	// Reader is rejected by this Symbol. In this case the Symbol should leave
-	// the Reader in the same state it was passed.
-	Decode(Reader) (T, error)
-}
-
-type symbol[T any] struct {
-	fmt.Stringer
-	decodeFn func(Reader) (T, error)
-}
-
-func (s *symbol[T]) Decode(r Reader) (T, error) { return s.decodeFn(r) }
-
-// SymbolPtr wraps a Symbol in a such a way as to make lazily initializing a
-// Symbol variable possible. This allows for recursion amongst different
-// Symbols.
-//
-// Example:
-//
-//	a := new(SymbolPtr)
-//	b := new(SymbolPtr)
-//	a.Symbol = FirstOf(Rune('a'), b)
-//	b.Symbol = FirstOf(Rune('b'), a)
-type SymbolPtr[T any] struct {
-	Symbol[T]
-}
-
-func named[T any](stringer fmt.Stringer, sym Symbol[T]) Symbol[T] {
-	return &symbol[T]{
-		stringer,
-		sym.Decode,
-	}
-}
-
-// Named wraps the given Symbol such that its String method returns the given
-// name.
-func Named[T any](name string, sym Symbol[T]) Symbol[T] {
-	return named(Stringer{S: name}, sym)
-}
-
-// RuneFunc matches and produces any rune for which the given function returns
-// true.
-func RuneFunc(name string, fn func(rune) bool) Symbol[Located[rune]] {
-	return &symbol[Located[rune]]{
-		Stringer{S: name},
-		func(rr Reader) (Located[rune], error) {
-			var zero Located[rune]
-
-			r, err := rr.ReadRune()
-			if errors.Is(err, io.EOF) {
-				return zero, ErrNoMatch
-			} else if err != nil {
-				return zero, err
-			}
-
-			if !fn(r.Value) {
-				rr.UnreadRune(r)
-				return zero, ErrNoMatch
-			}
-
-			return r, nil
-		},
-	}
-}
-
-// Rune matches and produces the given rune.
-func Rune(r rune) Symbol[Located[rune]] {
-	return RuneFunc(
-		fmt.Sprintf("'%c'", r),
-		func(r2 rune) bool { return r == r2 },
-	)
-}
-
-// StringFromRunes produces a string from the slice of runes produced by the
-// given Symbol. The slice must not be empty. StringFromRunes does not match if
-// the given Symbol does not match.
-func StringFromRunes(sym Symbol[[]Located[rune]]) Symbol[Located[string]] {
-	return Mapping(sym, func(runes []Located[rune]) Located[string] {
-		if len(runes) == 0 {
-			panic("StringFromRunes used on empty set of runes")
-		}
-
-		str := make([]rune, len(runes))
-		for i := range runes {
-			str[i] = runes[i].Value
-		}
-		return Located[string]{runes[0].Location, string(str)}
-	})
-}
-
-// Mapping produces a value of type Tb by decoding a value from the given
-// Symbol and passing it through the given mapping function. If the given Symbol
-// doesn't match then neither does Map.
-func Mapping[Ta, Tb any](
-	sym Symbol[Ta], fn func(Ta) Tb,
-) Symbol[Tb] {
-	return &symbol[Tb]{
-		sym,
-		func(rr Reader) (Tb, error) {
-			var zero Tb
-			va, err := sym.Decode(rr)
-			if err != nil {
-				return zero, err
-			}
-			return fn(va), nil
-		},
-	}
-}
-
-// OneOrMore will produce as many of the given Symbol's value as can be found
-// sequentially, up until a non-matching value is encountered. If no matches are
-// found then OneOrMore does not match.
-func OneOrMore[T any](sym Symbol[T]) Symbol[[]T] {
-	return &symbol[[]T]{
-		Stringer{F: func() string {
-			return fmt.Sprintf("one or more %v", sym)
-		}},
-		func(rr Reader) ([]T, error) {
-			var vv []T
-			for {
-				v, err := sym.Decode(rr)
-				if errors.Is(err, ErrNoMatch) {
-					break
-				} else if err != nil {
-					return nil, err
-				}
-
-				vv = append(vv, v)
-			}
-
-			if len(vv) == 0 {
-				return nil, ErrNoMatch
-			}
-
-			return vv, nil
-		},
-	}
-}
-
-// ZeroOrMore will produce as many of the given Symbol's value as can be found
-// sequentially, up until a non-matching value is encountered. If no matches are
-// found then an empty slice is produced.
-func ZeroOrMore[T any](sym Symbol[T]) Symbol[[]T] {
-	return &symbol[[]T]{
-		Stringer{F: func() string {
-			return fmt.Sprintf("zero or more %v", sym)
-		}},
-		func(rr Reader) ([]T, error) {
-			var vv []T
-			for {
-				v, err := sym.Decode(rr)
-				if errors.Is(err, ErrNoMatch) {
-					break
-				} else if err != nil {
-					return nil, err
-				}
-
-				vv = append(vv, v)
-			}
-
-			return vv, nil
-		},
-	}
-}
-
-func firstOf[T any](stringer fmt.Stringer, syms ...Symbol[T]) Symbol[T] {
-	return &symbol[T]{
-		stringer,
-		func(rr Reader) (T, error) {
-			var zero T
-			for _, sym := range syms {
-				v, err := sym.Decode(rr)
-				if errors.Is(err, ErrNoMatch) {
-					continue
-				} else if err != nil {
-					return zero, err
-				}
-
-				return v, nil
-			}
-
-			return zero, ErrNoMatch
-		},
-	}
-}
-
-// FirstOf matches and produces the value for the first Symbol in the list which
-// matches. FirstOf does not match if none of the given Symbols match.
-func FirstOf[T any](syms ...Symbol[T]) Symbol[T] {
-	return firstOf(
-		Stringer{F: func() string {
-			descrs := make([]string, len(syms))
-			for i := range syms {
-				descrs[i] = syms[i].String()
-			}
-			return strings.Join(descrs, " or ")
-		}},
-		syms...,
-	)
-}
-
-// Reduction produces a value of type Tc by first reading a value from symA,
-// then symB, and then running those through the given function.
-//
-// If symA does not match then Reduction does not match. If symA matches but
-// symB does not then also match then Reduction produces a LocatedError.
-func Reduction[Ta, Tb, Tc any](
-	symA Symbol[Ta],
-	symB Symbol[Tb],
-	fn func(Ta, Tb) Tc,
-) Symbol[Tc] {
-	return &symbol[Tc]{
-		symA,
-		func(rr Reader) (Tc, error) {
-			var zero Tc
-
-			va, err := symA.Decode(rr)
-			if err != nil {
-				return zero, err
-			}
-
-			vb, err := symB.Decode(rr)
-			if errors.Is(err, ErrNoMatch) {
-				return zero, rr.NextLocation().errf("expected %v", symB)
-			} else if err != nil {
-				return zero, err
-			}
-
-			return fn(va, vb), nil
-
-		},
-	}
-}
-
-// Prefixed matches on prefixSym, discards its value, then produces the value
-// produced by sym.
-//
-// If prefixSym does not match then Prefixed does not match. If prefixSym
-// matches but sym does not also match then Prefixed produces a LocatedError.
-func Prefixed[Ta, Tb any](prefixSym Symbol[Ta], sym Symbol[Tb]) Symbol[Tb] {
-	return named(prefixSym, Reduction(prefixSym, sym, func(_ Ta, b Tb) Tb {
-		return b
-	}))
-}
-
-// PrefixDiscarded is similar to Prefixed, except that if sym does not match
-// then PrefixDiscarded does not match, whereas Prefixed produces a LocatedError
-// in that case.
-//
-// NOTE PrefixDiscarded does not fully honor the contract of Symbol. If
-// prefixSym matches, but sym does not, then only sym will restore Reader to its
-// prior state; prefixSym cannot return whatever data it read back onto the
-// Reader. Therefore ErrNoMatch can be returned without Reader being fully back
-// in its original state. In practice this isn't a big deal, given the common
-// use-cases of PrefixDiscarded, but it may prove tricky.
-func PrefixDiscarded[Ta, Tb any](prefixSym Symbol[Ta], sym Symbol[Tb]) Symbol[Tb] {
-	return &symbol[Tb]{
-		sym,
-		func(rr Reader) (Tb, error) {
-			var zero Tb
-			if _, err := prefixSym.Decode(rr); err != nil {
-				return zero, err
-			}
-			return sym.Decode(rr)
-		},
-	}
-}
-
-// Suffixed matchs on sym and then suffixSym, returning the value produced by
-// sym and discarding the one produced by suffixSym.
-//
-// If sym does not match then Suffixed does not match. If sym matches but
-// suffixSym does not also match then Suffixed produces a LocatedError.
-func Suffixed[Ta, Tb any](sym Symbol[Ta], suffixSym Symbol[Tb]) Symbol[Ta] {
-	return named(sym, Reduction(sym, suffixSym, func(a Ta, _ Tb) Ta {
-		return a
-	}))
-}
-
-// Discard matches if the given Symbol does, but discards the value it produces,
-// producing an empty value instead.
-func Discard[T any](sym Symbol[T]) Symbol[struct{}] {
-	return Mapping(sym, func(T) struct{} { return struct{}{} })
-}

gg/json.go

@@ -0,0 +1,186 @@
+package gg
+
+import (
+	"encoding/json"
+	"fmt"
+)
+
+type openEdgeJSON struct {
+	From    vertexJSON `json:"from"`
+	ValueID string     `json:"valueID"`
+}
+
+type vertexJSON struct {
+	Type    VertexType     `json:"type"`
+	ValueID string         `json:"valueID,omitempty"`
+	In      []openEdgeJSON `json:"in"`
+}
+
+type graphJSON struct {
+	Values        map[string]json.RawMessage `json:"values"`
+	ValueVertices []vertexJSON               `json:"valueVertices"`
+}
+
+// MarshalJSON implements the json.Marshaler interface for a Graph. All Values
+// in the Graph will have json.Marshal called on them as-is in order to marshal
+// them.
+func (g *Graph) MarshalJSON() ([]byte, error) {
+	gJ := graphJSON{
+		Values:        map[string]json.RawMessage{},
+		ValueVertices: make([]vertexJSON, 0, len(g.vM)),
+	}
+
+	withVal := func(val Value) (string, error) {
+		if _, ok := gJ.Values[val.ID]; !ok {
+			valJ, err := json.Marshal(val.V)
+			if err != nil {
+				return "", err
+			}
+			gJ.Values[val.ID] = json.RawMessage(valJ)
+		}
+		return val.ID, nil
+	}
+
+	// two locally defined, mutually recursive functions. This kind of thing
+	// could probably be abstracted out, I feel like it happens frequently with
+	// graph code.
+	var mkIns func([]OpenEdge) ([]openEdgeJSON, error)
+	var mkVert func(vertex) (vertexJSON, error)
+
+	mkIns = func(in []OpenEdge) ([]openEdgeJSON, error) {
+		inJ := make([]openEdgeJSON, len(in))
+		for i := range in {
+			valID, err := withVal(in[i].val)
+			if err != nil {
+				return nil, err
+			}
+			vJ, err := mkVert(in[i].fromV)
+			if err != nil {
+				return nil, err
+			}
+			inJ[i] = openEdgeJSON{From: vJ, ValueID: valID}
+		}
+		return inJ, nil
+	}
+
+	mkVert = func(v vertex) (vertexJSON, error) {
+		ins, err := mkIns(v.in)
+		if err != nil {
+			return vertexJSON{}, err
+		}
+		vJ := vertexJSON{
+			Type: v.VertexType,
+			In:   ins,
+		}
+		if v.VertexType == ValueVertex {
+			valID, err := withVal(v.val)
+			if err != nil {
+				return vJ, err
+			}
+			vJ.ValueID = valID
+		}
+		return vJ, nil
+	}
+
+	for _, v := range g.vM {
+		vJ, err := mkVert(v)
+		if err != nil {
+			return nil, err
+		}
+		gJ.ValueVertices = append(gJ.ValueVertices, vJ)
+	}
+
+	return json.Marshal(gJ)
+}
+
+type jsonUnmarshaler struct {
+	g  *Graph
+	fn func(json.RawMessage) (interface{}, error)
+}
+
+// JSONUnmarshaler returns a json.Unmarshaler instance which, when used, will
+// unmarshal a json string into the Graph instance being called on here.
+//
+// The passed in function is used to unmarshal Values (used in both ValueVertex
+// vertices and edges) from json strings into go values. The returned inteface{}
+// should have already had the unmarshal from the given json string performed on
+// it.
+//
+// The json.Unmarshaler returned can be used many times, but will reset the
+// Graph completely before each use.
+func (g *Graph) JSONUnmarshaler(fn func(json.RawMessage) (interface{}, error)) json.Unmarshaler {
+	return jsonUnmarshaler{g: g, fn: fn}
+}
+
+func (jm jsonUnmarshaler) UnmarshalJSON(b []byte) error {
+	*(jm.g) = Graph{}
+	jm.g.vM = map[string]vertex{}
+
+	var gJ graphJSON
+	if err := json.Unmarshal(b, &gJ); err != nil {
+		return err
+	}
+
+	vals := map[string]Value{}
+	getVal := func(valID string) (Value, error) {
+		if val, ok := vals[valID]; ok {
+			return val, nil
+		}
+
+		j, ok := gJ.Values[valID]
+		if !ok {
+			return Value{}, fmt.Errorf("unmarshaling malformed graph, value with ID %q not defined", valID)
+		}
+
+		V, err := jm.fn(j)
+		if err != nil {
+			return Value{}, err
+		}
+
+		val := Value{ID: valID, V: V}
+		vals[valID] = val
+		return val, nil
+	}
+
+	var mkIns func([]openEdgeJSON) ([]OpenEdge, error)
+	var mkVert func(vertexJSON) (vertex, error)
+
+	mkIns = func(inJ []openEdgeJSON) ([]OpenEdge, error) {
+		in := make([]OpenEdge, len(inJ))
+		for i := range inJ {
+			val, err := getVal(inJ[i].ValueID)
+			if err != nil {
+				return nil, err
+			}
+			v, err := mkVert(inJ[i].From)
+			if err != nil {
+				return nil, err
+			}
+			in[i] = OpenEdge{fromV: v, val: val}
+		}
+		return in, nil
+	}
+
+	mkVert = func(vJ vertexJSON) (vertex, error) {
+		ins, err := mkIns(vJ.In)
+		if err != nil {
+			return vertex{}, err
+		}
+		var val Value
+		if vJ.Type == ValueVertex {
+			if val, err = getVal(vJ.ValueID); err != nil {
+				return vertex{}, err
+			}
+		}
+		return mkVertex(vJ.Type, val, ins...), nil
+	}
+
+	for _, v := range gJ.ValueVertices {
+		v, err := mkVert(v)
+		if err != nil {
+			return err
+		}
+		jm.g.vM[v.id] = v
+	}
+	return nil
+}

gim/NOTES

@@ -0,0 +1,111 @@
+Notes from reading https://www.graphviz.org/Documentation/TSE93.pdf, which
+describes an algorithm for drawing an acyclic graph in basically the way which I
+want.
+
+This document assumes the primary flow of drawing is downward, and secondary is
+right.
+
+For all of this it might be easier to not even consider edge values yet, as
+those could be done by converting them into vertices themselves after the
+cyclic-edge-reversal and then converting them back later.
+
+Drawing the graph is a four step process:
+
+1) Rank nodes in the Y axis
+    - Graph must be acyclic.
+        - This can be accomplished by strategically reversing edges which cause
+          a cycle, and then reversing them back as a post-processing step.
+        - Edges can be found by:
+            - walking out from a particular node depth-first from some arbitrary
+              node.
+            - As you do so you assign a rank based on depth to each node you
+              encounter.
+            - If any edge is destined for a node which has already been seen you
+              look at the ranks of the source and destination, and if the source
+              is _greater_ than the destination you reverse the edge's
+              direction.
+        - I think that algorithm only works if there's a source/sink? might have
+          to be modified, or the walk must traverse both to & from.
+    - Assign all edges a weight, default 1, but possibly externally assigned to
+      be greater.
+    - Take a "feasible" minimum spanning tree (MST) of the graph
+        - Feasibility is defined as each edge being "tight", meaning, once you
+          rank each node by their distance from the root and define the length
+          of an edge as the difference of rank of its head and tail, that each
+          tree edge will have a length of 1.
+    - Perform the following on the MST:
+        - For each edge of the graph assign the cut value
+            - If you were to remove any edge of an MST it would create two
+              separate MSTs. The side the edge was pointing from is the tail,
+              the side it was pointing to is the head.
+            - Looking at edges _in the original graph_, sum the weights of all
+              edges directed from the tail to the head (including the one
+              removed) and subtract from that the sum of the weights of the
+              edges directed from the head to the tail. This is the cut value.
+            - "...note that the cut values can be computed using information
+              local to an edge if the search is ordered from the leaves of the
+              feasible tree inward. It is trivial to compute the cut value of a
+              tree edge with one of its endpoints a leaf in the tree, since
+              either the head or the tail component consists of a single node.
+              Now, assuming the cut values are known for all the edges incident
+              on a given node except one, the cut value of the remaining edge is
+              the sum of the known cut values plus a term dependent only on the
+              edges incident to the given node."
+        - Take an edge with a negative cut value and remove it. Find the graph
+          edge between the remaining head and tail MSTs with the smallest
+          "slack" (distance in rank between its ends) and add that edge to the
+          MST to make it connected again.
+        - Repeat until there are no negative cut values.
+        - Apparently searching "cyclically" through the negative edges, rather
+          than iterating from the start each time, is worthwhile.
+    - Normalize the MST by assigning the root node the rank of 0 (and so on), if
+      it changed.
+    - All edges in the MST are of length 1, and the rest can be inferred from
+      that.
+    - To reduce crowding, nodes with equal in/out edge weights and which could
+      be placed on multiple rankings are moved to the ranking with the fewest
+      nodes.
+
+2) Order nodes in the X axis to reduce edge crossings
+    - Add ephemeral vertices along edges with lengths greater than 1, so all
+      "spaces" are filled.
+    - If any vertices have edges to vertices on their same rank, those are
+      ordered so that all these "flag edges" are pointed in the same direction
+      across that rank, and the ordering of those particular vertices is always
+      kept.
+    - Iterate over the graph some fixed number of times (the paper recommends
+      24)
+        - possibly with some heuristic which looks at percentage improvement
+          each time to determine if it's worth the effort.
+        - on one iteration move "down" the graph, on the next move "up", etc...
+          shaker style
+        - On each iteration:
+            - For each vertex look at the median position of all of the vertices
+              it has edges to in the previous rank
+            - If the number of previous vertices is even do this complicated
+              thing (P is the set of positions previous):
+              ```
+              if |P| = 2 then
+                return (P[0] + P[1])/2;
+              else
+                left = P[m-1] - P[0];
+                right = P[|P| -1] - P[m];
+                return (P[m-1]*right + P[m]*left)/(left+right);
+              endif
+              ```
+            - Sort the vertices by their median position
+                - vertices with no previous vertices remain fixed
+            - Then, for each vertex in the rank attempt to transpose it with its
+              neighbor and see if that reduces the number of edge crossings
+              between the rank and its previous.
+            - If equality is found during these two steps (same median, or same
+              number of crossings) the vertices in question should be flipped.
+
+3) Compute node coordinates
+    - Determining the Y coordinates is considered trivial: find the maxHeight of
+      each rank, and ensure they are separated by that much plus whatever the
+      separation value is.
+    - For the X coordinates: do some insane shit involving the network simplex
+      again.
+
+4) Determine edge splines

gim/geo/geo.go

@@ -0,0 +1,139 @@
+// Package geo implements basic geometric concepts used by gim
+package geo
+
+import "math"
+
+// XY describes a 2-dimensional position or vector. The origin of the
+// 2-dimensional space is a 0,0, with the x-axis going to the left and the
+// y-axis going down.
+type XY [2]int
+
+// Zero is the zero point, or a zero vector, depending on what you're doing
+var Zero = XY{0, 0}
+
+// Unit vectors
+var (
+	Up    = XY{0, -1}
+	Down  = XY{0, 1}
+	Left  = XY{-1, 0}
+	Right = XY{1, 0}
+)
+
+// Units is the set of unit vectors
+var Units = []XY{
+	Up,
+	Down,
+	Left,
+	Right,
+}
+
+func (xy XY) toF64() [2]float64 {
+	return [2]float64{
+		float64(xy[0]),
+		float64(xy[1]),
+	}
+}
+
+func abs(i int) int {
+	if i < 0 {
+		return i * -1
+	}
+	return i
+}
+
+// Abs returns the XY with all fields made positive, if they weren't already
+func (xy XY) Abs() XY {
+	return XY{abs(xy[0]), abs(xy[1])}
+}
+
+// Unit returns the XY with each field divided by its absolute value (i.e.
+// scaled down to 1 or -1). Fields which are 0 are left alone
+func (xy XY) Unit() XY {
+	for i := range xy {
+		if xy[i] > 0 {
+			xy[i] = 1
+		} else if xy[i] < 0 {
+			xy[i] = -1
+		}
+	}
+	return xy
+}
+
+// Len returns the length (aka magnitude) of the XY as a vector.
+func (xy XY) Len() int {
+	if xy[0] == 0 {
+		return abs(xy[1])
+	} else if xy[1] == 0 {
+		return abs(xy[0])
+	}
+
+	xyf := xy.toF64()
+	lf := math.Sqrt((xyf[0] * xyf[0]) + (xyf[1] * xyf[1]))
+	return Rounder.Round(lf)
+}
+
+// Add returns the result of adding the two XYs' fields individually
+func (xy XY) Add(xy2 XY) XY {
+	xy[0] += xy2[0]
+	xy[1] += xy2[1]
+	return xy
+}
+
+// Mul returns the result of multiplying the two XYs' fields individually
+func (xy XY) Mul(xy2 XY) XY {
+	xy[0] *= xy2[0]
+	xy[1] *= xy2[1]
+	return xy
+}
+
+// Div returns the results of dividing the two XYs' field individually.
+func (xy XY) Div(xy2 XY) XY {
+	xyf, xy2f := xy.toF64(), xy2.toF64()
+	return XY{
+		Rounder.Round(xyf[0] / xy2f[0]),
+		Rounder.Round(xyf[1] / xy2f[1]),
+	}
+}
+
+// Scale returns the result of multiplying both of the XY's fields by the scalar
+func (xy XY) Scale(scalar int) XY {
+	return xy.Mul(XY{scalar, scalar})
+}
+
+// Inv inverses the XY, a shortcut for xy.Scale(-1)
+func (xy XY) Inv() XY {
+	return xy.Scale(-1)
+}
+
+// Sub subtracts xy2 from xy and returns the result. A shortcut for
+// xy.Add(xy2.Inv())
+func (xy XY) Sub(xy2 XY) XY {
+	return xy.Add(xy2.Inv())
+}
+
+// Midpoint returns the midpoint between the two XYs.
+func (xy XY) Midpoint(xy2 XY) XY {
+	return xy.Add(xy2.Sub(xy).Div(XY{2, 2}))
+}
+
+// Min returns an XY whose fields are the minimum values of the two XYs'
+// fields compared individually
+func (xy XY) Min(xy2 XY) XY {
+	for i := range xy {
+		if xy2[i] < xy[i] {
+			xy[i] = xy2[i]
+		}
+	}
+	return xy
+}
+
+// Max returns an XY whose fields are the Maximum values of the two XYs'
+// fields compared individually
+func (xy XY) Max(xy2 XY) XY {
+	for i := range xy {
+		if xy2[i] > xy[i] {
+			xy[i] = xy2[i]
+		}
+	}
+	return xy
+}

gim/geo/rect.go

@@ -0,0 +1,127 @@
+package geo
+
+import (
+	"fmt"
+)
+
+// Rect describes a rectangle based on the position of its top-left corner and
+// size
+type Rect struct {
+	TopLeft XY
+	Size    XY
+}
+
+// Edge describes a straight edge starting at its first XY and ending at its
+// second
+type Edge [2]XY
+
+// EdgeCoord returns the coordinate of the edge indicated by the given direction
+// (Up, Down, Left, or Right). The coordinate will be for the axis applicable to
+// the direction, so for Left/Right it will be the x coordinate and for Up/Down
+// the y.
+func (r Rect) EdgeCoord(dir XY) int {
+	switch dir {
+	case Up:
+		return r.TopLeft[1]
+	case Down:
+		return r.TopLeft[1] + r.Size[1] - 1
+	case Left:
+		return r.TopLeft[0]
+	case Right:
+		return r.TopLeft[0] + r.Size[0] - 1
+	default:
+		panic(fmt.Sprintf("unsupported direction: %#v", dir))
+	}
+}
+
+// Corner returns the position of the corner identified by the given directions
+// (Left/Right, Up/Down)
+func (r Rect) Corner(xDir, yDir XY) XY {
+	switch {
+	case r.Size[0] == 0 || r.Size[1] == 0:
+		panic(fmt.Sprintf("rectangle with non-multidimensional size has no corners: %v", r.Size))
+	case xDir == Left && yDir == Up:
+		return r.TopLeft
+	case xDir == Right && yDir == Up:
+		return r.TopLeft.Add(r.Size.Mul(Right)).Add(XY{-1, 0})
+	case xDir == Left && yDir == Down:
+		return r.TopLeft.Add(r.Size.Mul(Down)).Add(XY{0, -1})
+	case xDir == Right && yDir == Down:
+		return r.TopLeft.Add(r.Size).Add(XY{-1, -1})
+	default:
+		panic(fmt.Sprintf("unsupported Corner args: %v, %v", xDir, yDir))
+	}
+}
+
+// Edge returns an Edge instance for the edge of the Rect indicated by the given
+// direction (Up, Down, Left, or Right). secDir indicates the direction the
+// returned Edge should be pointing (i.e. the order of its XY's) and must be
+// perpendicular to dir
+func (r Rect) Edge(dir, secDir XY) Edge {
+	var e Edge
+	switch dir {
+	case Up:
+		e[0], e[1] = r.Corner(Left, Up), r.Corner(Right, Up)
+	case Down:
+		e[0], e[1] = r.Corner(Left, Down), r.Corner(Right, Down)
+	case Left:
+		e[0], e[1] = r.Corner(Left, Up), r.Corner(Left, Down)
+	case Right:
+		e[0], e[1] = r.Corner(Right, Up), r.Corner(Right, Down)
+	default:
+		panic(fmt.Sprintf("unsupported direction: %#v", dir))
+	}
+
+	switch secDir {
+	case Left, Up:
+		e[0], e[1] = e[1], e[0]
+	default:
+		// do nothing
+	}
+	return e
+}
+
+// Midpoint returns the point which is the midpoint of the Edge
+func (e Edge) Midpoint() XY {
+	return e[0].Midpoint(e[1])
+}
+
+func (r Rect) halfSize() XY {
+	return r.Size.Div(XY{2, 2})
+}
+
+// Center returns the centerpoint of the rectangle.
+func (r Rect) Center() XY {
+	return r.TopLeft.Add(r.halfSize())
+}
+
+// Translate returns an instance of Rect which is the same as this one but
+// translated by the given amount.
+func (r Rect) Translate(by XY) Rect {
+	r.TopLeft = r.TopLeft.Add(by)
+	return r
+}
+
+// Centered returns an instance of Rect which is this one but translated to be
+// centered on the given point.
+func (r Rect) Centered(on XY) Rect {
+	r.TopLeft = on.Sub(r.halfSize())
+	return r
+}
+
+// Union returns the smallest Rect which encompasses the given Rect and the one
+// being called upon.
+func (r Rect) Union(r2 Rect) Rect {
+	if r.Size == Zero {
+		return r2
+	} else if r2.Size == Zero {
+		return r
+	}
+
+	tl := r.TopLeft.Min(r2.TopLeft)
+	br := r.Corner(Right, Down).Max(r2.Corner(Right, Down))
+	return Rect{
+		TopLeft: tl,
+		Size:    br.Sub(tl).Add(XY{1, 1}),
+	}
+}

gim/geo/rect_test.go

@@ -0,0 +1,113 @@
+package geo
+
+import (
+	. "testing"
+
+	"github.com/stretchr/testify/assert"
+)
+
+func TestRect(t *T) {
+	r := Rect{
+		TopLeft: XY{1, 2},
+		Size:    XY{2, 2},
+	}
+
+	assert.Equal(t, 2, r.EdgeCoord(Up))
+	assert.Equal(t, 3, r.EdgeCoord(Down))
+	assert.Equal(t, 1, r.EdgeCoord(Left))
+	assert.Equal(t, 2, r.EdgeCoord(Right))
+
+	lu := XY{1, 2}
+	ld := XY{1, 3}
+	ru := XY{2, 2}
+	rd := XY{2, 3}
+
+	assert.Equal(t, lu, r.Corner(Left, Up))
+	assert.Equal(t, ld, r.Corner(Left, Down))
+	assert.Equal(t, ru, r.Corner(Right, Up))
+	assert.Equal(t, rd, r.Corner(Right, Down))
+
+	assert.Equal(t, Edge{lu, ld}, r.Edge(Left, Down))
+	assert.Equal(t, Edge{ru, rd}, r.Edge(Right, Down))
+	assert.Equal(t, Edge{lu, ru}, r.Edge(Up, Right))
+	assert.Equal(t, Edge{ld, rd}, r.Edge(Down, Right))
+	assert.Equal(t, Edge{ld, lu}, r.Edge(Left, Up))
+	assert.Equal(t, Edge{rd, ru}, r.Edge(Right, Up))
+	assert.Equal(t, Edge{ru, lu}, r.Edge(Up, Left))
+	assert.Equal(t, Edge{rd, ld}, r.Edge(Down, Left))
+}
+
+func TestRectCenter(t *T) {
+	assertCentered := func(exp, given Rect, center XY) {
+		got := given.Centered(center)
+		assert.Equal(t, exp, got)
+		assert.Equal(t, center, got.Center())
+	}
+
+	{
+		r := Rect{
+			Size: XY{4, 4},
+		}
+		assert.Equal(t, XY{2, 2}, r.Center())
+		assertCentered(
+			Rect{TopLeft: XY{1, 1}, Size: XY{4, 4}},
+			r, XY{3, 3},
+		)
+	}
+
+	{
+		r := Rect{
+			Size: XY{5, 5},
+		}
+		assert.Equal(t, XY{3, 3}, r.Center())
+		assertCentered(
+			Rect{TopLeft: XY{0, 0}, Size: XY{5, 5}},
+			r, XY{3, 3},
+		)
+	}
+}
+
+func TestRectUnion(t *T) {
+	assertUnion := func(exp, r1, r2 Rect) {
+		assert.Equal(t, exp, r1.Union(r2))
+		assert.Equal(t, exp, r2.Union(r1))
+	}
+
+	{ // Zero
+		r := Rect{TopLeft: XY{1, 1}, Size: XY{2, 2}}
+		assertUnion(r, r, Rect{})
+	}
+
+	{ // Equal
+		r := Rect{Size: XY{2, 2}}
+		assertUnion(r, r, r)
+	}
+
+	{ // Overlapping corner
+		r1 := Rect{TopLeft: XY{0, 0}, Size: XY{2, 2}}
+		r2 := Rect{TopLeft: XY{1, 1}, Size: XY{2, 2}}
+		ex := Rect{TopLeft: XY{0, 0}, Size: XY{3, 3}}
+		assertUnion(ex, r1, r2)
+	}
+
+	{ // 2 overlapping corners
+		r1 := Rect{TopLeft: XY{0, 0}, Size: XY{4, 4}}
+		r2 := Rect{TopLeft: XY{1, 1}, Size: XY{4, 2}}
+		ex := Rect{TopLeft: XY{0, 0}, Size: XY{5, 4}}
+		assertUnion(ex, r1, r2)
+	}
+
+	{ // Shared edge
+		r1 := Rect{TopLeft: XY{0, 0}, Size: XY{2, 1}}
+		r2 := Rect{TopLeft: XY{1, 0}, Size: XY{1, 2}}
+		ex := Rect{TopLeft: XY{0, 0}, Size: XY{2, 2}}
+		assertUnion(ex, r1, r2)
+	}
+
+	{ // Adjacent edge
+		r1 := Rect{TopLeft: XY{0, 0}, Size: XY{2, 2}}
+		r2 := Rect{TopLeft: XY{2, 0}, Size: XY{2, 2}}
+		ex := Rect{TopLeft: XY{0, 0}, Size: XY{4, 2}}
+		assertUnion(ex, r1, r2)
+	}
+}

gim/geo/round.go

@@ -0,0 +1,33 @@
+package geo
+
+import (
+	"math"
+)
+
+// RounderFunc is a function which converts a floating point number into an
+// integer.
+type RounderFunc func(float64) int64
+
+// Round is helper for calling the RounderFunc and converting the result to an
+// int.
+func (rf RounderFunc) Round(f float64) int {
+	return int(rf(f))
+}
+
+// A few RounderFuncs which can be used. Set the Rounder global variable to pick
+// one.
+var (
+	Floor RounderFunc = func(f float64) int64 { return int64(math.Floor(f)) }
+	Ceil  RounderFunc = func(f float64) int64 { return int64(math.Ceil(f)) }
+	Round RounderFunc = func(f float64) int64 {
+		if f < 0 {
+			f = math.Ceil(f - 0.5)
+		}
+		f = math.Floor(f + 0.5)
+		return int64(f)
+	}
+)
+
+// Rounder is the RounderFunc which will be used by all functions and methods in
+// this package when needed.
+var Rounder = Ceil

gim/main.go

@@ -0,0 +1,91 @@
+package main
+
+import (
+	"math/rand"
+	"time"
+
+	"github.com/mediocregopher/ginger/gg"
+	"github.com/mediocregopher/ginger/gim/geo"
+	"github.com/mediocregopher/ginger/gim/terminal"
+	"github.com/mediocregopher/ginger/gim/view"
+)
+
+// TODO be able to draw circular graphs
+// TODO audit all steps, make sure everything is deterministic
+// TODO self-edges
+
+//const (
+//	framerate   = 10
+//	frameperiod = time.Second / time.Duration(framerate)
+//)
+
+//func debugf(str string, args ...interface{}) {
+//	if !strings.HasSuffix(str, "\n") {
+//		str += "\n"
+//	}
+//	fmt.Fprintf(os.Stderr, str, args...)
+//}
+
+func mkGraph() (*gg.Graph, gg.Value) {
+	a := gg.NewValue("a")
+	aE0 := gg.NewValue("aE0")
+	aE1 := gg.NewValue("aE1")
+	aE2 := gg.NewValue("aE2")
+	aE3 := gg.NewValue("aE3")
+	b0 := gg.NewValue("b0")
+	b1 := gg.NewValue("b1")
+	b2 := gg.NewValue("b2")
+	b3 := gg.NewValue("b3")
+	oaE0 := gg.ValueOut(a, aE0)
+	oaE1 := gg.ValueOut(a, aE1)
+	oaE2 := gg.ValueOut(a, aE2)
+	oaE3 := gg.ValueOut(a, aE3)
+	g := gg.Null
+	g = g.AddValueIn(oaE0, b0)
+	g = g.AddValueIn(oaE1, b1)
+	g = g.AddValueIn(oaE2, b2)
+	g = g.AddValueIn(oaE3, b3)
+
+	c := gg.NewValue("c")
+	empty := gg.NewValue("")
+	jE := gg.JunctionOut([]gg.OpenEdge{
+		gg.ValueOut(b0, empty),
+		gg.ValueOut(b1, empty),
+		gg.ValueOut(b2, empty),
+		gg.ValueOut(b3, empty),
+	}, gg.NewValue("jE"))
+	g = g.AddValueIn(jE, c)
+
+	// TODO this really fucks it up
+	//d := gg.NewValue("d")
+	//deE := gg.ValueOut(d, gg.NewValue("deE"))
+	//g = g.AddValueIn(deE, gg.NewValue("e"))
+
+	return g, c
+}
+
+//func mkGraph() *gg.Graph {
+//	g := gg.Null
+//	g = g.AddValueIn(gg.ValueOut(str("a"), str("e")), str("b"))
+//	return g
+//}
+
+func main() {
+	rand.Seed(time.Now().UnixNano())
+	term := terminal.New()
+	wSize := term.WindowSize()
+	center := geo.Zero.Midpoint(wSize)
+
+	g, start := mkGraph()
+	view := view.New(g, start, geo.Right, geo.Down)
+	viewBuf := terminal.NewBuffer()
+	view.Draw(viewBuf)
+
+	buf := terminal.NewBuffer()
+	buf.DrawBufferCentered(center, viewBuf)
+
+	term.Clear()
+	term.WriteBuffer(geo.Zero, buf)
+	term.SetPos(wSize.Add(geo.XY{0, -1}))
+	term.Draw()
+}

gim/terminal/buffer.go

@@ -0,0 +1,217 @@
+package terminal
+
+import (
+	"fmt"
+	"strconv"
+	"unicode"
+
+	"github.com/mediocregopher/ginger/gim/geo"
+)
+
+// Reset all custom styles
+const ansiReset = "\033[0m"
+
+// Color describes the foreground or background color of text
+type Color int
+
+// Available Color values
+const (
+	// whatever the terminal's default color scheme is
+	Default = iota
+
+	Black
+	Red
+	Green
+	Yellow
+	Blue
+	Magenta
+	Cyan
+	White
+)
+
+type bufStyle struct {
+	fgColor Color
+	bgColor Color
+}
+
+// returns foreground and background ansi codes
+func (bf bufStyle) ansi() (string, string) {
+	var fg, bg string
+	if bf.fgColor != Default {
+		fg = "\033[0;3" + strconv.Itoa(int(bf.fgColor)-1) + "m"
+	}
+	if bf.bgColor != Default {
+		bg = "\033[0;4" + strconv.Itoa(int(bf.bgColor)-1) + "m"
+	}
+	return fg, bg
+}
+
+// returns the ansi sequence which would modify the style to the given one
+func (bf bufStyle) diffTo(bf2 bufStyle) string {
+	// this implementation is naive, but whatever
+	if bf == bf2 {
+		return ""
+	}
+
+	fg, bg := bf2.ansi()
+	if (bf == bufStyle{}) {
+		return fg + bg
+	}
+	return ansiReset + fg + bg
+}
+
+type bufPoint struct {
+	r rune
+	bufStyle
+}
+
+// Buffer describes an infinitely sized terminal buffer to which anything may be
+// drawn, and which will efficiently generate strings representing the drawn
+// text.
+type Buffer struct {
+	currStyle bufStyle
+	currPos   geo.XY
+	m         *mat
+	max       geo.XY
+}
+
+// NewBuffer initializes and returns a new empty buffer. The proper way to clear
+// a buffer is to toss the old one and generate a new one.
+func NewBuffer() *Buffer {
+	return &Buffer{
+		m:   newMat(),
+		max: geo.XY{-1, -1},
+	}
+}
+
+// Copy creates a new identical instance of this Buffer and returns it.
+func (b *Buffer) Copy() *Buffer {
+	b2 := NewBuffer()
+	b.m.iter(func(x, y int, v interface{}) bool {
+		b2.setRune(geo.XY{x, y}, v.(bufPoint))
+		return true
+	})
+	b2.currStyle = b.currStyle
+	b2.currPos = b.currPos
+	return b2
+}
+
+func (b *Buffer) setRune(at geo.XY, p bufPoint) {
+	b.m.set(at[0], at[1], p)
+	b.max = b.max.Max(at)
+}
+
+// WriteRune writes the given rune to the Buffer at whatever the current
+// position is, with whatever the current styling is.
+func (b *Buffer) WriteRune(r rune) {
+	if r == '\n' {
+		b.currPos[0], b.currPos[1] = 0, b.currPos[1]+1
+		return
+	} else if r == '\r' {
+		b.currPos[0] = 0
+	} else if !unicode.IsPrint(r) {
+		panic(fmt.Sprintf("character %q is not supported by terminal.Buffer", r))
+	}
+
+	b.setRune(b.currPos, bufPoint{
+		r:        r,
+		bufStyle: b.currStyle,
+	})
+	b.currPos[0]++
+}
+
+// WriteString writes the given string to the Buffer at whatever the current
+// position is, with whatever the current styling is.
+func (b *Buffer) WriteString(s string) {
+	for _, r := range s {
+		b.WriteRune(r)
+	}
+}
+
+// SetPos sets the cursor position in the Buffer, so Print operations will begin
+// at that point. Remember that the origin is at point (0, 0).
+func (b *Buffer) SetPos(xy geo.XY) {
+	b.currPos = xy
+}
+
+// SetFGColor sets subsequent text's foreground color.
+func (b *Buffer) SetFGColor(c Color) {
+	b.currStyle.fgColor = c
+}
+
+// SetBGColor sets subsequent text's background color.
+func (b *Buffer) SetBGColor(c Color) {
+	b.currStyle.bgColor = c
+}
+
+// ResetStyle unsets all text styling options which have been set.
+func (b *Buffer) ResetStyle() {
+	b.currStyle = bufStyle{}
+}
+
+// String renders and returns a string which, when printed to a terminal, will
+// print the Buffer's contents at the terminal's current cursor position.
+func (b *Buffer) String() string {
+	s := ansiReset // always start with a reset
+	var style bufStyle
+	var pos geo.XY
+	move := func(to geo.XY) {
+		diff := to.Sub(pos)
+		if diff[0] > 0 {
+			s += "\033[" + strconv.Itoa(diff[0]) + "C"
+		} else if diff[0] < 0 {
+			s += "\033[" + strconv.Itoa(-diff[0]) + "D"
+		}
+		if diff[1] > 0 {
+			s += "\033[" + strconv.Itoa(diff[1]) + "B"
+		} else if diff[1] < 0 {
+			s += "\033[" + strconv.Itoa(-diff[1]) + "A"
+		}
+		pos = to
+	}
+
+	b.m.iter(func(x, y int, v interface{}) bool {
+		p := v.(bufPoint)
+		move(geo.XY{x, y})
+		s += style.diffTo(p.bufStyle)
+		style = p.bufStyle
+		s += string(p.r)
+		pos[0]++
+		return true
+	})
+	return s
+}
+
+// DrawBuffer copies the given Buffer onto this one, with the given's top-left
+// corner being at the given position. The given buffer may be the same as this
+// one.
+//
+// Calling this method does not affect this Buffer's current cursor position or
+// style.
+func (b *Buffer) DrawBuffer(at geo.XY, b2 *Buffer) {
+	if b == b2 {
+		b2 = b2.Copy()
+	}
+	b2.m.iter(func(x, y int, v interface{}) bool {
+		x += at[0]
+		y += at[1]
+		if x < 0 || y < 0 {
+			return true
+		}
+		b.setRune(geo.XY{x, y}, v.(bufPoint))
+		return true
+	})
+}
+
+// DrawBufferCentered is like DrawBuffer, but centered around the given point
+// instead of translated by it.
+func (b *Buffer) DrawBufferCentered(around geo.XY, b2 *Buffer) {
+	b2rect := geo.Rect{Size: b2.Size()}
+	b.DrawBuffer(b2rect.Centered(around).TopLeft, b2)
+}
+
+// Size returns the dimensions of the Buffer's current area which has been
+// written to.
+func (b *Buffer) Size() geo.XY {
+	return b.max.Add(geo.XY{1, 1})
+}

gim/terminal/example/main.go

@@ -0,0 +1,59 @@
+package main
+
+import (
+	"log"
+	"time"
+
+	"github.com/mediocregopher/ginger/gim/geo"
+	"github.com/mediocregopher/ginger/gim/terminal"
+)
+
+func main() {
+	b := terminal.NewBuffer()
+	b.WriteString("this is fun")
+
+	b.SetFGColor(terminal.Blue)
+	b.SetBGColor(terminal.Green)
+	b.SetPos(geo.XY{18, 0})
+	b.WriteString("blue and green")
+
+	b.ResetStyle()
+	b.SetFGColor(terminal.Red)
+	b.SetPos(geo.XY{3, 3})
+	b.WriteString("red!!!")
+
+	b.ResetStyle()
+	b.SetFGColor(terminal.Blue)
+	b.SetPos(geo.XY{20, 0})
+	b.WriteString("boo")
+
+	bcp := b.Copy()
+	b.DrawBuffer(geo.XY{2, 2}, bcp)
+	b.DrawBuffer(geo.XY{-1, 1}, bcp)
+
+	brect := terminal.NewBuffer()
+	brect.DrawRect(geo.Rect{Size: b.Size().Add(geo.XY{2, 2})}, terminal.SingleLine)
+	log.Printf("b.Size:%v", b.Size())
+	brect.DrawBuffer(geo.XY{1, 1}, b)
+
+	t := terminal.New()
+	p := geo.XY{0, 0}
+	dirH, dirV := geo.Right, geo.Down
+	wsize := t.WindowSize()
+	for range time.Tick(time.Second / 15) {
+		t.Clear()
+		t.WriteBuffer(p, brect)
+		t.Draw()
+
+		brectSize := brect.Size()
+		p = p.Add(dirH).Add(dirV)
+		if p[0] < 0 || p[0]+brectSize[0] > wsize[0] {
+			dirH = dirH.Scale(-1)
+			p = p.Add(dirH.Scale(2))
+		}
+		if p[1] < 0 || p[1]+brectSize[1] > wsize[1] {
+			dirV = dirV.Scale(-1)
+			p = p.Add(dirV.Scale(2))
+		}
+	}
+}

gim/terminal/mat.go

@@ -0,0 +1,117 @@
+package terminal
+
+import (
+	"container/list"
+)
+
+type matEl struct {
+	x int
+	v interface{}
+}
+
+type matRow struct {
+	y int
+	l *list.List
+}
+
+// a 2-d sparse matrix
+type mat struct {
+	rows *list.List
+
+	currY     int
+	currRowEl *list.Element
+	currEl    *list.Element
+}
+
+func newMat() *mat {
+	return &mat{
+		rows: list.New(),
+	}
+}
+
+func (m *mat) getRow(y int) *list.List {
+	m.currY = y             // this will end up being true no matter what
+	if m.currRowEl == nil { // first call
+		l := list.New()
+		m.currRowEl = m.rows.PushFront(matRow{y: y, l: l})
+		return l
+
+	} else if m.currRowEl.Value.(matRow).y > y {
+		m.currRowEl = m.rows.Front()
+	}
+
+	for {
+		currRow := m.currRowEl.Value.(matRow)
+		switch {
+		case currRow.y == y:
+			return currRow.l
+		case currRow.y < y:
+			if m.currRowEl = m.currRowEl.Next(); m.currRowEl == nil {
+				l := list.New()
+				m.currRowEl = m.rows.PushBack(matRow{y: y, l: l})
+				return l
+			}
+		default: // currRow.y > y
+			l := list.New()
+			m.currRowEl = m.rows.InsertBefore(matRow{y: y, l: l}, m.currRowEl)
+			return l
+		}
+	}
+}
+
+func (m *mat) getEl(x, y int) *matEl {
+	var rowL *list.List
+	if m.currRowEl == nil || m.currY != y {
+		rowL = m.getRow(y)
+		m.currEl = rowL.Front()
+	} else {
+		rowL = m.currRowEl.Value.(matRow).l
+	}
+
+	if m.currEl == nil || m.currEl.Value.(*matEl).x > x {
+		if m.currEl = rowL.Front(); m.currEl == nil {
+			// row is empty
+			mel := &matEl{x: x}
+			m.currEl = rowL.PushFront(mel)
+			return mel
+		}
+	}
+
+	for {
+		currEl := m.currEl.Value.(*matEl)
+		switch {
+		case currEl.x == x:
+			return currEl
+		case currEl.x < x:
+			if m.currEl = m.currEl.Next(); m.currEl == nil {
+				mel := &matEl{x: x}
+				m.currEl = rowL.PushBack(mel)
+				return mel
+			}
+		default: // currEl.x > x
+			mel := &matEl{x: x}
+			m.currEl = rowL.InsertBefore(mel, m.currEl)
+			return mel
+		}
+	}
+}
+
+func (m *mat) get(x, y int) interface{} {
+	return m.getEl(x, y).v
+}
+
+func (m *mat) set(x, y int, v interface{}) {
+	m.getEl(x, y).v = v
+}
+
+func (m *mat) iter(f func(x, y int, v interface{}) bool) {
+	for rowEl := m.rows.Front(); rowEl != nil; rowEl = rowEl.Next() {
+		row := rowEl.Value.(matRow)
+		for el := row.l.Front(); el != nil; el = el.Next() {
+			mel := el.Value.(*matEl)
+			if !f(mel.x, row.y, mel.v) {
+				return
+			}
+		}
+	}
+}

gim/terminal/mat_test.go

@@ -0,0 +1,59 @@
+package terminal
+
+import (
+	"fmt"
+	"math/rand"
+	"strings"
+	. "testing"
+	"time"
+)
+
+func TestMat(t *T) {
+	r := rand.New(rand.NewSource(time.Now().UnixNano()))
+
+	type xy struct {
+		x, y int
+	}
+
+	type action struct {
+		xy
+		set int
+	}
+
+	run := func(aa []action) {
+		aaStr := func(i int) string {
+			s := fmt.Sprintf("%#v", aa[:i+1])
+			return strings.Replace(s, "terminal.", "", -1)
+		}
+
+		m := newMat()
+		mm := map[xy]int{}
+		for i, a := range aa {
+			if a.set > 0 {
+				mm[a.xy] = a.set
+				m.set(a.xy.x, a.xy.y, a.set)
+				continue
+			}
+
+			expI, expOk := mm[a.xy]
+			gotI, gotOk := m.get(a.xy.x, a.xy.y).(int)
+			if expOk != gotOk {
+				t.Fatalf("get failed: expOk:%v gotOk:%v actions:%#v", expOk, gotOk, aaStr(i))
+			} else if expI != gotI {
+				t.Fatalf("get failed: expI:%v gotI:%v actions:%#v", expI, gotI, aaStr(i))
+			}
+		}
+	}
+
+	for i := 0; i < 10000; i++ {
+		var actions []action
+		for j := r.Intn(1000); j > 0; j-- {
+			a := action{xy: xy{x: r.Intn(5), y: r.Intn(5)}}
+			if r.Intn(3) == 0 {
+				a.set = r.Intn(10000) + 1
+			}
+			actions = append(actions, a)
+		}
+		run(actions)
+	}
+}

gim/terminal/shape.go

@@ -0,0 +1,189 @@
+package terminal
+
+import (
+	"fmt"
+	"strings"
+
+	"github.com/mediocregopher/ginger/gim/geo"
+)
+
+// SingleLine is a set of single-pixel-width lines.
+var SingleLine = LineStyle{
+	Horiz:       '─',
+	Vert:        '│',
+	TopLeft:     '┌',
+	TopRight:    '┐',
+	BottomLeft:  '└',
+	BottomRight: '┘',
+	PerpUp:      '┴',
+	PerpDown:    '┬',
+	PerpLeft:    '┤',
+	PerpRight:   '├',
+	ArrowUp:     '^',
+	ArrowDown:   'v',
+	ArrowLeft:   '<',
+	ArrowRight:  '>',
+}
+
+// LineStyle defines a set of characters to use together when drawing lines and
+// corners.
+type LineStyle struct {
+	Horiz, Vert rune
+
+	// Corner characters, identified as corners of a rectangle
+	TopLeft, TopRight, BottomLeft, BottomRight rune
+
+	// Characters for a straight segment a perpendicular attached
+	PerpUp, PerpDown, PerpLeft, PerpRight rune
+
+	// Characters for pointing arrows
+	ArrowUp, ArrowDown, ArrowLeft, ArrowRight rune
+}
+
+// Segment takes two different directions (i.e. geo.Up/Down/Left/Right) and
+// returns the line character which points in both of those directions.
+//
+// For example, SingleLine.Segment(geo.Up, geo.Left) returns '┘'.
+func (ls LineStyle) Segment(a, b geo.XY) rune {
+	inner := func(a, b geo.XY) rune {
+		type c struct{ a, b geo.XY }
+		switch (c{a, b}) {
+		case c{geo.Up, geo.Down}:
+			return ls.Vert
+		case c{geo.Left, geo.Right}:
+			return ls.Horiz
+		case c{geo.Down, geo.Right}:
+			return ls.TopLeft
+		case c{geo.Down, geo.Left}:
+			return ls.TopRight
+		case c{geo.Up, geo.Right}:
+			return ls.BottomLeft
+		case c{geo.Up, geo.Left}:
+			return ls.BottomRight
+		default:
+			return 0
+		}
+	}
+	if r := inner(a, b); r != 0 {
+		return r
+	} else if r = inner(b, a); r != 0 {
+		return r
+	}
+	panic(fmt.Sprintf("invalid LineStyle.Segment directions: %v, %v", a, b))
+}
+
+// Perpendicular returns the line character for a perpendicular segment
+// traveling in the given direction.
+func (ls LineStyle) Perpendicular(dir geo.XY) rune {
+	switch dir {
+	case geo.Up:
+		return ls.PerpUp
+	case geo.Down:
+		return ls.PerpDown
+	case geo.Left:
+		return ls.PerpLeft
+	case geo.Right:
+		return ls.PerpRight
+	default:
+		panic(fmt.Sprintf("invalid LineStyle.Perpendicular direction: %v", dir))
+	}
+}
+
+// Arrow returns the arrow character for an arrow pointing in the given
+// direction.
+func (ls LineStyle) Arrow(dir geo.XY) rune {
+	switch dir {
+	case geo.Up:
+		return ls.ArrowUp
+	case geo.Down:
+		return ls.ArrowDown
+	case geo.Left:
+		return ls.ArrowLeft
+	case geo.Right:
+		return ls.ArrowRight
+	default:
+		panic(fmt.Sprintf("invalid LineStyle.Arrow direction: %v", dir))
+	}
+}
+
+// DrawRect draws the given Rect to the Buffer with the given LineStyle. The
+// Rect's TopLeft field is used for its position.
+//
+// If Rect's Size is not at least 2x2 this does nothing.
+func (b *Buffer) DrawRect(r geo.Rect, ls LineStyle) {
+	if r.Size[0] < 2 || r.Size[1] < 2 {
+		return
+	}
+	horiz := strings.Repeat(string(ls.Horiz), r.Size[0]-2)
+
+	b.SetPos(r.TopLeft)
+	b.WriteRune(ls.TopLeft)
+	b.WriteString(horiz)
+	b.WriteRune(ls.TopRight)
+
+	for i := 0; i < r.Size[1]-2; i++ {
+		b.SetPos(r.TopLeft.Add(geo.XY{0, i + 1}))
+		b.WriteRune(ls.Vert)
+		b.SetPos(r.TopLeft.Add(geo.XY{r.Size[0] - 1, i + 1}))
+		b.WriteRune(ls.Vert)
+	}
+
+	b.SetPos(r.TopLeft.Add(geo.XY{0, r.Size[1] - 1}))
+	b.WriteRune(ls.BottomLeft)
+	b.WriteString(horiz)
+	b.WriteRune(ls.BottomRight)
+}
+
+// DrawLine draws a line from the start point to the ending one, primarily
+// moving in the given direction, using the given LineStyle to do so.
+func (b *Buffer) DrawLine(start, end, dir geo.XY, ls LineStyle) {
+	// given the "primary" direction the line should be headed, pick a possible
+	// secondary one which may be used to detour along the path in order to
+	// reach the destination (in the case that the two boxes are diagonal from
+	// each other)
+	var perpDir geo.XY
+	perpDir[0], perpDir[1] = dir[1], dir[0]
+	dirSec := end.Sub(start).Mul(perpDir.Abs()).Unit()
+	mid := start.Midpoint(end)
+
+	along := func(xy, dir geo.XY) int {
+		if dir[0] != 0 {
+			return xy[0]
+		}
+		return xy[1]
+	}
+
+	// collect the points along the line into an array
+	var pts []geo.XY
+	var curr geo.XY
+	midPrim := along(mid, dir)
+	endSec := along(end, dirSec)
+	for curr = start; curr != end; {
+		pts = append(pts, curr)
+		if prim := along(curr, dir); prim == midPrim {
+			if sec := along(curr, dirSec); sec != endSec {
+				curr = curr.Add(dirSec)
+				continue
+			}
+		}
+		curr = curr.Add(dir)
+	}
+	pts = append(pts, curr) // appending end
+
+	// draw each point
+	for i, pt := range pts {
+		var prev, next geo.XY
+		switch {
+		case i == 0:
+			prev = pt.Add(dir.Inv())
+			next = pts[i+1]
+		case i == len(pts)-1:
+			prev = pts[i-1]
+			next = pt.Add(dir)
+		default:
+			prev, next = pts[i-1], pts[i+1]
+		}
+		b.SetPos(pt)
+		b.WriteRune(ls.Segment(prev.Sub(pt), next.Sub(pt)))
+	}
+}

gim/terminal/terminal.go

@@ -0,0 +1,108 @@
+// Package terminal implements functionality related to interacting with a
+// terminal. Using this package takes the place of using stdout directly
+package terminal
+
+import (
+	"bytes"
+	"fmt"
+	"io"
+	"os"
+	"syscall"
+	"unsafe"
+
+	"github.com/mediocregopher/ginger/gim/geo"
+)
+
+// Terminal provides an interface to a terminal which allows for "drawing"
+// rather than just writing. Note that all operations on a Terminal aren't
+// actually drawn to the screen until Flush is called.
+//
+// The coordinate system described by Terminal looks like this:
+//
+// 0,0 ------------------> x
+//  |
+//  |
+//  |
+//  |
+//  |
+//  |
+//  |
+//  |
+//  v
+//  y
+//
+type Terminal struct {
+	buf *bytes.Buffer
+
+	// When initialized this will be set to os.Stdout, but can be set to
+	// anything
+	Out io.Writer
+}
+
+// New initializes and returns a usable Terminal
+func New() *Terminal {
+	return &Terminal{
+		buf: new(bytes.Buffer),
+		Out: os.Stdout,
+	}
+}
+
+// WindowSize returns the size of the terminal window (width/height)
+// TODO this doesn't support winblows
+func (t *Terminal) WindowSize() geo.XY {
+	var sz struct {
+		rows    uint16
+		cols    uint16
+		xpixels uint16
+		ypixels uint16
+	}
+	_, _, err := syscall.Syscall(
+		syscall.SYS_IOCTL,
+		uintptr(syscall.Stdin),
+		uintptr(syscall.TIOCGWINSZ),
+		uintptr(unsafe.Pointer(&sz)),
+	)
+	if err != 0 {
+		panic(err.Error())
+	}
+	return geo.XY{int(sz.cols), int(sz.rows)}
+}
+
+// SetPos sets the terminal's actual cursor position to the given coordinates.
+func (t *Terminal) SetPos(to geo.XY) {
+	// actual terminal uses 1,1 as top-left, because 1-indexing is a great idea
+	fmt.Fprintf(t.buf, "\033[%d;%dH", to[1]+1, to[0]+1)
+}
+
+// HideCursor causes the cursor to not actually be shown
+func (t *Terminal) HideCursor() {
+	fmt.Fprintf(t.buf, "\033[?25l")
+}
+
+// ShowCursor causes the cursor to be shown, if it was previously hidden
+func (t *Terminal) ShowCursor() {
+	fmt.Fprintf(t.buf, "\033[?25h")
+}
+
+// Clear completely clears all drawn characters on the screen and returns the
+// cursor to the origin. This implicitly calls Draw.
+func (t *Terminal) Clear() {
+	t.buf.Reset()
+	fmt.Fprintf(t.buf, "\033[2J")
+	t.Draw()
+}
+
+// WriteBuffer writes the contents to the Buffer to the Terminal's buffer,
+// starting at the given coordinate.
+func (t *Terminal) WriteBuffer(at geo.XY, b *Buffer) {
+	t.SetPos(at)
+	t.buf.WriteString(b.String())
+}
+
+// Draw writes all buffered changes to the screen
+func (t *Terminal) Draw() {
+	if _, err := io.Copy(t.Out, t.buf); err != nil {
+		panic(err)
+	}
+	t.buf.Reset()
+}

gim/view/box.go

@@ -0,0 +1,66 @@
+package view
+
+import (
+	"fmt"
+
+	"github.com/mediocregopher/ginger/gg"
+	"github.com/mediocregopher/ginger/gim/geo"
+	"github.com/mediocregopher/ginger/gim/terminal"
+)
+
+type box struct {
+	topLeft       geo.XY
+	flowDir       geo.XY
+	numIn, numOut int
+	buf           *terminal.Buffer
+	bodyBuf       *terminal.Buffer
+}
+
+func boxFromVertex(v *gg.Vertex, flowDir geo.XY) box {
+	b := box{
+		flowDir: flowDir,
+		numIn:   len(v.In),
+		numOut:  len(v.Out),
+	}
+	if v.VertexType == gg.ValueVertex {
+		b.bodyBuf = terminal.NewBuffer()
+		b.bodyBuf.WriteString(v.Value.V.(string))
+	}
+	return b
+}
+
+func (b box) rect() geo.Rect {
+	var bodyRect geo.Rect
+	if b.bodyBuf != nil {
+		bodyRect.Size = b.bodyBuf.Size().Add(geo.XY{2, 2})
+	}
+
+	var edgesRect geo.Rect
+	{
+		var neededByEdges int
+		if b.numIn > b.numOut {
+			neededByEdges = b.numIn*2 + 1
+		} else {
+			neededByEdges = b.numOut*2 + 1
+		}
+
+		switch b.flowDir {
+		case geo.Left, geo.Right:
+			edgesRect.Size = geo.XY{2, neededByEdges}
+		case geo.Up, geo.Down:
+			edgesRect.Size = geo.XY{neededByEdges, 2}
+		default:
+			panic(fmt.Sprintf("unknown flowDir: %#v", b.flowDir))
+		}
+	}
+
+	return bodyRect.Union(edgesRect).Translate(b.topLeft)
+}
+
+func (b box) draw(buf *terminal.Buffer) {
+	rect := b.rect()
+	buf.DrawRect(rect, terminal.SingleLine)
+	if b.bodyBuf != nil {
+		buf.DrawBufferCentered(rect.Center(), b.bodyBuf)
+	}
+}

gim/view/constraint/constraint.go

@@ -0,0 +1,119 @@
+// Package constraint implements an extremely simple constraint engine.
+// Elements, and constraints on those elements, are given to the engine, which
+// uses those constraints to generate an output. Elements are defined as a
+// string
+package constraint
+
+import (
+	"github.com/mediocregopher/ginger/gg"
+)
+
+// Constraint describes a constraint on an element. The Elem field must be
+// filled in, as well as exactly one other field
+type Constraint struct {
+	Elem string
+
+	// LT says that Elem is less than this element
+	LT string
+}
+
+var ltEdge = gg.NewValue("lt")
+
+// Engine processes sets of constraints to generate an output
+type Engine struct {
+	g    *gg.Graph
+	vals map[string]gg.Value
+}
+
+// NewEngine initializes and returns an empty Engine
+func NewEngine() *Engine {
+	return &Engine{g: gg.Null, vals: map[string]gg.Value{}}
+}
+
+func (e *Engine) getVal(elem string) gg.Value {
+	if val, ok := e.vals[elem]; ok {
+		return val
+	}
+	val := gg.NewValue(elem)
+	e.vals[elem] = val
+	return val
+}
+
+// AddConstraint adds the given constraint to the engine's set, returns false if
+// the constraint couldn't be added due to a conflict with a previous constraint
+func (e *Engine) AddConstraint(c Constraint) bool {
+	elem := e.getVal(c.Elem)
+	g := e.g.AddValueIn(gg.ValueOut(elem, ltEdge), e.getVal(c.LT))
+
+	// Check for loops in g starting at c.Elem, bail if there are any
+	{
+		seen := map[*gg.Vertex]bool{}
+		start := g.ValueVertex(elem)
+		var hasLoop func(v *gg.Vertex) bool
+		hasLoop = func(v *gg.Vertex) bool {
+			if seen[v] {
+				return v == start
+			}
+			seen[v] = true
+			for _, out := range v.Out {
+				if hasLoop(out.To) {
+					return true
+				}
+			}
+			return false
+		}
+		if hasLoop(start) {
+			return false
+		}
+	}
+
+	e.g = g
+	return true
+}
+
+// Solve uses the constraints which have been added to the engine to give a
+// possible solution. The given element is one which has been added to the
+// engine and whose value is known to be zero.
+func (e *Engine) Solve() map[string]int {
+	m := map[string]int{}
+	if len(e.g.ValueVertices()) == 0 {
+		return m
+	}
+
+	vElem := func(v *gg.Vertex) string {
+		return v.Value.V.(string)
+	}
+
+	// first the roots are determined to be the elements with no In edges, which
+	// _must_ exist since the graph presumably has no loops
+	var roots []*gg.Vertex
+	e.g.Iter(func(v *gg.Vertex) bool {
+		if len(v.In) == 0 {
+			roots = append(roots, v)
+			m[vElem(v)] = 0
+		}
+		return true
+	})
+
+	// sanity check
+	if len(roots) == 0 {
+		panic("no roots found in graph somehow")
+	}
+
+	// a vertex's value is then the length of the longest path from it to one of
+	// the roots
+	var walk func(*gg.Vertex, int)
+	walk = func(v *gg.Vertex, val int) {
+		if elem := vElem(v); val > m[elem] {
+			m[elem] = val
+		}
+		for _, out := range v.Out {
+			walk(out.To, val+1)
+		}
+	}
+	for _, root := range roots {
+		walk(root, 0)
+	}
+
+	return m
+}

gim/view/constraint/constraint_test.go

@@ -0,0 +1,94 @@
+package constraint
+
+import (
+	. "testing"
+
+	"github.com/stretchr/testify/assert"
+)
+
+func TestEngineAddConstraint(t *T) {
+	{
+		e := NewEngine()
+		assert.True(t, e.AddConstraint(Constraint{Elem: "0", LT: "1"}))
+		assert.True(t, e.AddConstraint(Constraint{Elem: "1", LT: "2"}))
+		assert.True(t, e.AddConstraint(Constraint{Elem: "-1", LT: "0"}))
+		assert.False(t, e.AddConstraint(Constraint{Elem: "1", LT: "0"}))
+		assert.False(t, e.AddConstraint(Constraint{Elem: "2", LT: "0"}))
+		assert.False(t, e.AddConstraint(Constraint{Elem: "2", LT: "-1"}))
+	}
+
+	{
+		e := NewEngine()
+		assert.True(t, e.AddConstraint(Constraint{Elem: "0", LT: "1"}))
+		assert.True(t, e.AddConstraint(Constraint{Elem: "0", LT: "2"}))
+		assert.True(t, e.AddConstraint(Constraint{Elem: "1", LT: "2"}))
+		assert.True(t, e.AddConstraint(Constraint{Elem: "2", LT: "3"}))
+	}
+}
+
+func TestEngineSolve(t *T) {
+	assertSolve := func(exp map[string]int, cc ...Constraint) {
+		e := NewEngine()
+		for _, c := range cc {
+			assert.True(t, e.AddConstraint(c), "c:%#v", c)
+		}
+		assert.Equal(t, exp, e.Solve())
+	}
+
+	// basic
+	assertSolve(
+		map[string]int{"a": 0, "b": 1, "c": 2},
+		Constraint{Elem: "a", LT: "b"},
+		Constraint{Elem: "b", LT: "c"},
+	)
+
+	// "triangle" graph
+	assertSolve(
+		map[string]int{"a": 0, "b": 1, "c": 2},
+		Constraint{Elem: "a", LT: "b"},
+		Constraint{Elem: "a", LT: "c"},
+		Constraint{Elem: "b", LT: "c"},
+	)
+
+	// "hexagon" graph
+	assertSolve(
+		map[string]int{"a": 0, "b": 1, "c": 1, "d": 2, "e": 2, "f": 3},
+		Constraint{Elem: "a", LT: "b"},
+		Constraint{Elem: "a", LT: "c"},
+		Constraint{Elem: "b", LT: "d"},
+		Constraint{Elem: "c", LT: "e"},
+		Constraint{Elem: "d", LT: "f"},
+		Constraint{Elem: "e", LT: "f"},
+	)
+
+	// "hexagon" with centerpoint graph
+	assertSolve(
+		map[string]int{"a": 0, "b": 1, "c": 1, "center": 2, "d": 3, "e": 3, "f": 4},
+		Constraint{Elem: "a", LT: "b"},
+		Constraint{Elem: "a", LT: "c"},
+		Constraint{Elem: "b", LT: "d"},
+		Constraint{Elem: "c", LT: "e"},
+		Constraint{Elem: "d", LT: "f"},
+		Constraint{Elem: "e", LT: "f"},
+
+		Constraint{Elem: "c", LT: "center"},
+		Constraint{Elem: "b", LT: "center"},
+		Constraint{Elem: "center", LT: "e"},
+		Constraint{Elem: "center", LT: "d"},
+	)
+
+	// multi-root, using two triangles which end up connecting
+	assertSolve(
+		map[string]int{"a": 0, "b": 1, "c": 2, "d": 0, "e": 1, "f": 2, "g": 3},
+		Constraint{Elem: "a", LT: "b"},
+		Constraint{Elem: "a", LT: "c"},
+		Constraint{Elem: "b", LT: "c"},
+
+		Constraint{Elem: "d", LT: "e"},
+		Constraint{Elem: "d", LT: "f"},
+		Constraint{Elem: "e", LT: "f"},
+
+		Constraint{Elem: "f", LT: "g"},
+	)
+
+}

gim/view/line.go

@@ -0,0 +1,31 @@
+package view
+
+import (
+	"github.com/mediocregopher/ginger/gim/geo"
+	"github.com/mediocregopher/ginger/gim/terminal"
+)
+
+type line struct {
+	from, to   *box
+	fromI, toI int
+	bodyBuf    *terminal.Buffer
+}
+
+func (l line) draw(buf *terminal.Buffer, flowDir, secFlowDir geo.XY) {
+	from, to := *(l.from), *(l.to)
+	start := from.rect().Edge(flowDir, secFlowDir)[0].Add(secFlowDir.Scale(l.fromI*2 + 1))
+	end := to.rect().Edge(flowDir.Inv(), secFlowDir)[0]
+	end = end.Add(flowDir.Inv())
+	end = end.Add(secFlowDir.Scale(l.toI*2 + 1))
+
+	buf.SetPos(start)
+	buf.WriteRune(terminal.SingleLine.Perpendicular(flowDir))
+	buf.DrawLine(start.Add(flowDir), end.Add(flowDir.Inv()), flowDir, terminal.SingleLine)
+	buf.SetPos(end)
+	buf.WriteRune(terminal.SingleLine.Arrow(flowDir))
+
+	// draw the body
+	if l.bodyBuf != nil {
+		buf.DrawBufferCentered(start.Midpoint(end), l.bodyBuf)
+	}
+}

gim/view/vertex.go

@@ -0,0 +1,27 @@
+package view
+
+import (
+	"github.com/mediocregopher/ginger/gim/geo"
+	"github.com/mediocregopher/ginger/gim/terminal"
+)
+
+type edge struct {
+	from, to   *vertex
+	tail, head rune // if empty do directional segment char
+	body       string
+	switchback bool
+
+	lineStyle terminal.LineStyle
+}
+
+type vertex struct {
+	coord, pos geo.XY
+	in, out    [][]*edge // top level is port index
+	body       string
+
+	// means it won't be drawn, and will be removed and have its in/out edges
+	// spliced together into a single edge.
+	ephemeral bool
+
+	lineStyle terminal.LineStyle // if zero value don't draw border
+}

gim/view/view.go

@@ -0,0 +1,274 @@
+// Package view implements rendering a graph to a terminal.
+//
+// Steps for rendering
+//
+//	- Preprocessing: Disjoin Graph into multiple Graphs, and decide how to
+//	  arrange them (maybe sort by number of vertices or number of edges (or the
+//	  sum of both) or something).
+//
+//	- Convert Graph into internal representation.
+//		- Still uses gg.Graph, but vertices and edge values are wrapped in types
+//		  internal to this package, and on which further mapping will be done.
+//		- Positions unknown at this point.
+//		- Junctions are converted to value vertices with set edge order.
+//		- Edges contain both their body and their tail/head rune.
+//
+//	- Find eligible "root" vertex, probably by one which has the fewest input
+//	  edges.
+//
+//	- Find cycles and reverse edges as needed.
+//		- The to/from vertices are reversed, as are the head/tail runes, so the
+//		  direction will appear consistent with the original graph
+//		- TODO this might not be necessary? Or at least may need to be modified.
+//		  In the paper this is done, but that algorithm allows for edges upward
+//		  from their tail, whereas this one doesn't. It might only be necessary
+//		  for the MST stuff, in which case this might only need to take place
+//		  within Positioning-Part1.
+//
+//	- Replace edge bodies with a vertex with a single input/output edge.
+//
+//	- Position all vertices
+//		- `coord` field on vertices used as row/column coordinates.
+//		- Positioning will be done with down being the primary direction and
+//		  right being the secondary direction.
+//		- Part 1) find vertical positions for all vertices (aka assign rows)
+//			- This step uses some fancy MST stuff as outlined by (TODO refer to
+//			  paper here).
+//		- Part 2) find horizontal positions within rows (aka assign columns)
+//			- Part of this will include creating ephemeral vertices where an
+//			  edge spans a row without having a vertex on it. These will be
+//			  removed as the final part of this step.
+//			- The jist of this step is to find vertex ordering which reduces
+//			  number of edge crossings between adjacent rows.
+//			- Some extra care is taken for cases where an edge's from vertex is
+//			  not a lower row than its to vertex.
+//				- This is an unavoidable case, as at the least a vertex may
+//				  connect to itself.
+//				- These edges will have their `switchback` field set to true.
+//				- For the purposes of calculating edge crossings these edges
+//				  should be ignored. During the absolute positioning and drawing
+//				  steps they will be accounted for and dealt with.
+//		- Part 3) row/column positions into terminal positions, which are
+//		  stored on the vertices in the `pos` field. Primary/secondary
+//		  direction are taken into account here.
+//
+//	- Post-processing: any additional absolute positioning and other formatting
+//	  given by the user for the Graph should be done here
+//
+//	- Draw vertices and their edges to buffer
+//		- At this point drawing vertices is easy. Edges is more complicated but
+//		  the start/end positions of each edge should already be known, so while
+//		  drawing may be complex it's not difficult.
+//
+package view
+
+import (
+	"sort"
+
+	"github.com/mediocregopher/ginger/gg"
+	"github.com/mediocregopher/ginger/gim/geo"
+	"github.com/mediocregopher/ginger/gim/terminal"
+	"github.com/mediocregopher/ginger/gim/view/constraint"
+)
+
+// View wraps a single Graph instance and a set of display options for it, and
+// generates renderable terminal output for it.
+type View struct {
+	g     *gg.Graph
+	start gg.Value // TODO shouldn't need this
+
+	primFlowDir, secFlowDir geo.XY
+}
+
+// New instantiates and returns a view around the given Graph instance, with
+// start indicating the value vertex to consider the "root" of the graph.
+//
+// Drawing is done by aligning the vertices into rows and columns in such a way
+// as to reduce edge crossings. primaryDir indicates the direction edges will
+// primarily be pointed in. For example, if it is geo.Down then adjacent
+// vertices will be arranged into columns.
+//
+// secondaryDir indicates the direction vertices should be arranged when they
+// end up in the same "rank" (e.g. when primaryDir is geo.Down, all vertices on
+// the same row will be the same "rank").
+//
+// A primaryDir/secondaryDir of either geo.Down/geo.Right or geo.Right/geo.Down
+// are recommended, but any combination of perpendicular directions is allowed.
+func New(g *gg.Graph, start gg.Value, primaryDir, secondaryDir geo.XY) *View {
+	return &View{
+		g:           g,
+		start:       start,
+		primFlowDir: primaryDir,
+		secFlowDir:  secondaryDir,
+	}
+}
+
+// Draw renders and draws the View's Graph to the Buffer.
+func (view *View) Draw(buf *terminal.Buffer) {
+	relPos, _, secSol := posSolve(view.g)
+
+	// create boxes
+	var boxes []*box
+	boxesM := map[*box]*gg.Vertex{}
+	boxesMr := map[*gg.Vertex]*box{}
+	const (
+		primPadding = 5
+		secPadding  = 1
+	)
+	var primPos int
+	for _, vv := range relPos {
+		var primBoxes []*box // boxes on just this level
+		var maxPrim int
+		var secPos int
+		for _, v := range vv {
+			primVec := view.primFlowDir.Scale(primPos)
+			secVec := view.secFlowDir.Scale(secPos)
+
+			b := boxFromVertex(v, view.primFlowDir)
+			b.topLeft = primVec.Add(secVec)
+			boxes = append(boxes, &b)
+			primBoxes = append(primBoxes, &b)
+			boxesM[&b] = v
+			boxesMr[v] = &b
+
+			bSize := b.rect().Size
+			primBoxLen := bSize.Mul(view.primFlowDir).Len()
+			secBoxLen := bSize.Mul(view.secFlowDir).Len()
+			if primBoxLen > maxPrim {
+				maxPrim = primBoxLen
+			}
+			secPos += secBoxLen + secPadding
+		}
+		for _, b := range primBoxes {
+			b.topLeft = b.topLeft.Add(view.primFlowDir.Scale(primPos))
+		}
+		primPos += maxPrim + primPadding
+	}
+
+	// maps a vertex to all of its to edges, sorted by secSol
+	findFromIM := map[*gg.Vertex][]gg.Edge{}
+	// returns the index of this edge in from's Out
+	findFromI := func(from *gg.Vertex, e gg.Edge) int {
+		edges, ok := findFromIM[from]
+		if !ok {
+			edges = make([]gg.Edge, len(from.Out))
+			copy(edges, from.Out)
+			sort.Slice(edges, func(i, j int) bool {
+				// TODO if two edges go to the same vertex, how are they sorted?
+				return secSol[edges[i].To.ID] < secSol[edges[j].To.ID]
+			})
+			findFromIM[from] = edges
+		}
+
+		for i, fe := range edges {
+			if fe == e {
+				return i
+			}
+		}
+		panic("edge not found in from.Out")
+	}
+
+	// create lines
+	var lines []line
+	for _, b := range boxes {
+		v := boxesM[b]
+		for i, e := range v.In {
+			bFrom := boxesMr[e.From]
+			fromI := findFromI(e.From, e)
+			buf := terminal.NewBuffer()
+			buf.WriteString(e.Value.V.(string))
+			lines = append(lines, line{
+				from:    bFrom,
+				fromI:   fromI,
+				to:      b,
+				toI:     i,
+				bodyBuf: buf,
+			})
+		}
+	}
+
+	// actually draw the boxes and lines
+	for _, b := range boxes {
+		b.draw(buf)
+	}
+	for _, line := range lines {
+		line.draw(buf, view.primFlowDir, view.secFlowDir)
+	}
+}
+
+// "Solves" vertex position by detemining relative positions of vertices in
+// primary and secondary directions (independently), with relative positions
+// being described by "levels", where multiple vertices can occupy one level.
+//
+// Primary determines relative position in the primary direction by trying
+// to place vertices before their outs and after their ins.
+//
+// Secondary determines relative position in the secondary direction by
+// trying to place vertices relative to vertices they share an edge with in
+// the order that the edges appear on the shared node.
+func posSolve(g *gg.Graph) ([][]*gg.Vertex, map[string]int, map[string]int) {
+	primEng := constraint.NewEngine()
+	secEng := constraint.NewEngine()
+
+	strM := g.ByID()
+	for _, v := range strM {
+		var prevIn *gg.Vertex
+		for _, e := range v.In {
+			primEng.AddConstraint(constraint.Constraint{
+				Elem: e.From.ID,
+				LT:   v.ID,
+			})
+			if prevIn != nil {
+				secEng.AddConstraint(constraint.Constraint{
+					Elem: prevIn.ID,
+					LT:   e.From.ID,
+				})
+			}
+			prevIn = e.From
+		}
+
+		var prevOut *gg.Vertex
+		for _, e := range v.Out {
+			if prevOut == nil {
+				continue
+			}
+			secEng.AddConstraint(constraint.Constraint{
+				Elem: prevOut.ID,
+				LT:   e.To.ID,
+			})
+			prevOut = e.To
+		}
+	}
+	prim := primEng.Solve()
+	sec := secEng.Solve()
+
+	// determine maximum primary level
+	var maxPrim int
+	for _, lvl := range prim {
+		if lvl > maxPrim {
+			maxPrim = lvl
+		}
+	}
+
+	outStr := make([][]string, maxPrim+1)
+	for v, lvl := range prim {
+		outStr[lvl] = append(outStr[lvl], v)
+	}
+
+	// sort each primary level
+	for _, vv := range outStr {
+		sort.Slice(vv, func(i, j int) bool {
+			return sec[vv[i]] < sec[vv[j]]
+		})
+	}
+
+	// convert to vertices
+	out := make([][]*gg.Vertex, len(outStr))
+	for i, vv := range outStr {
+		out[i] = make([]*gg.Vertex, len(outStr[i]))
+		for j, v := range vv {
+			out[i][j] = strM[v]
+		}
+	}
+	return out, prim, sec
+}

go.mod

@@ -1,14 +0,0 @@
-module code.betamike.com/mediocregopher/ginger
-
-go 1.18
-
-require (
-	github.com/stretchr/testify v1.7.0
-	golang.org/x/exp v0.0.0-20231006140011-7918f672742d
-)
-
-require (
-	github.com/davecgh/go-spew v1.1.0 // indirect
-	github.com/pmezard/go-difflib v1.0.0 // indirect
-	gopkg.in/yaml.v3 v3.0.0-20200313102051-9f266ea9e77c // indirect
-)

go.sum

@@ -1,13 +0,0 @@
-github.com/davecgh/go-spew v1.1.0 h1:ZDRjVQ15GmhC3fiQ8ni8+OwkZQO4DARzQgrnXU1Liz8=
-github.com/davecgh/go-spew v1.1.0/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
-github.com/pmezard/go-difflib v1.0.0 h1:4DBwDE0NGyQoBHbLQYPwSUPoCMWR5BEzIk/f1lZbAQM=
-github.com/pmezard/go-difflib v1.0.0/go.mod h1:iKH77koFhYxTK1pcRnkKkqfTogsbg7gZNVY4sRDYZ/4=
-github.com/stretchr/objx v0.1.0/go.mod h1:HFkY916IF+rwdDfMAkV7OtwuqBVzrE8GR6GFx+wExME=
-github.com/stretchr/testify v1.7.0 h1:nwc3DEeHmmLAfoZucVR881uASk0Mfjw8xYJ99tb5CcY=
-github.com/stretchr/testify v1.7.0/go.mod h1:6Fq8oRcR53rry900zMqJjRRixrwX3KX962/h/Wwjteg=
-golang.org/x/exp v0.0.0-20231006140011-7918f672742d h1:jtJma62tbqLibJ5sFQz8bKtEM8rJBtfilJ2qTU199MI=
-golang.org/x/exp v0.0.0-20231006140011-7918f672742d/go.mod h1:ldy0pHrwJyGW56pPQzzkH36rKxoZW1tw7ZJpeKx+hdo=
-gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405 h1:yhCVgyC4o1eVCa2tZl7eS0r+SDo693bJlVdllGtEeKM=
-gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0=
-gopkg.in/yaml.v3 v3.0.0-20200313102051-9f266ea9e77c h1:dUUwHk2QECo/6vqA44rthZ8ie2QXMNeKRTHCNY2nXvo=
-gopkg.in/yaml.v3 v3.0.0-20200313102051-9f266ea9e77c/go.mod h1:K4uyk7z7BCEPqu6E+C64Yfv1cQ7kz7rIZviUmN+EgEM=

graph/graph.go

@@ -1,432 +0,0 @@
-// Package graph implements a generic directed graph type, with support for
-// tuple vertices in addition to traditional "value" vertices.
-package graph
-
-import (
-	"fmt"
-	"strings"
-)
-
-// Value is any value which can be stored within a Graph. Values should be
-// considered immutable, ie once used with the graph package their internal
-// value does not change.
-type Value interface {
-	Equal(Value) bool
-	String() string
-}
-
-// OpenEdge consists of the edge value (E) and source vertex value (V) of an
-// edge in a Graph. When passed into the AddValueIn method a full edge is
-// created. An OpenEdge can also be sourced from a tuple vertex, whose value is
-// an ordered set of OpenEdges of this same type.
-type OpenEdge[E, V Value] struct {
-	val *V
-	tup []*OpenEdge[E, V]
-
-	edgeVal E
-}
-
-func (oe *OpenEdge[E, V]) equal(oe2 *OpenEdge[E, V]) bool {
-	if !oe.edgeVal.Equal(oe2.edgeVal) {
-		return false
-	}
-
-	if oe.val != nil {
-		return oe2.val != nil && (*oe.val).Equal(*oe2.val)
-	}
-
-	if len(oe.tup) != len(oe2.tup) {
-		return false
-	}
-
-	for i := range oe.tup {
-		if !oe.tup[i].equal(oe2.tup[i]) {
-			return false
-		}
-	}
-
-	return true
-}
-
-func (oe *OpenEdge[E, V]) String() string {
-
-	vertexType := "tup"
-
-	var fromStr string
-
-	if oe.val != nil {
-
-		vertexType = "val"
-		fromStr = (*oe.val).String()
-
-	} else {
-
-		strs := make([]string, len(oe.tup))
-
-		for i := range oe.tup {
-			strs[i] = oe.tup[i].String()
-		}
-
-		fromStr = fmt.Sprintf("[%s]", strings.Join(strs, ", "))
-	}
-
-	return fmt.Sprintf("%s(%s, %s)", vertexType, fromStr, oe.edgeVal.String())
-}
-
-// WithEdgeValue returns a copy of the OpenEdge with the given Value replacing
-// the previous edge value.
-//
-// NOTE I _think_ this can be factored out once Graph is genericized.
-func (oe *OpenEdge[E, V]) WithEdgeValue(val E) *OpenEdge[E, V] {
-	oeCp := *oe
-	oeCp.edgeVal = val
-	return &oeCp
-}
-
-// EdgeValue returns the Value which lies on the edge itself.
-func (oe OpenEdge[E, V]) EdgeValue() E {
-	return oe.edgeVal
-}
-
-// FromValue returns the Value from which the OpenEdge was created via ValueOut,
-// or false if it wasn't created via ValueOut.
-func (oe OpenEdge[E, V]) FromValue() (V, bool) {
-	if oe.val == nil {
-		var zero V
-		return zero, false
-	}
-
-	return *oe.val, true
-}
-
-// FromTuple returns the tuple of OpenEdges from which the OpenEdge was created
-// via TupleOut, or false if it wasn't created via TupleOut.
-func (oe OpenEdge[E, V]) FromTuple() ([]*OpenEdge[E, V], bool) {
-	if oe.val != nil {
-		return nil, false
-	}
-
-	return oe.tup, true
-}
-
-// ValueOut creates a OpenEdge which, when used to construct a Graph, represents
-// an edge (with edgeVal attached to it) coming from the vertex containing val.
-func ValueOut[E, V Value](edgeVal E, val V) *OpenEdge[E, V] {
-	return &OpenEdge[E, V]{
-		val:     &val,
-		edgeVal: edgeVal,
-	}
-}
-
-// TupleOut creates an OpenEdge which, when used to construct a Graph,
-// represents an edge (with edgeVal attached to it) coming from the vertex
-// comprised of the given ordered-set of input edges.
-func TupleOut[E, V Value](edgeVal E, ins ...*OpenEdge[E, V]) *OpenEdge[E, V] {
-
-	if len(ins) == 1 {
-
-		var (
-			zero E
-			in   = ins[0]
-		)
-
-		if edgeVal.Equal(zero) {
-			return in
-		}
-
-		if in.edgeVal.Equal(zero) {
-			return in.WithEdgeValue(edgeVal)
-		}
-
-	}
-
-	return &OpenEdge[E, V]{
-		tup:     ins,
-		edgeVal: edgeVal,
-	}
-}
-
-type graphValueIn[E, V Value] struct {
-	val  V
-	edge *OpenEdge[E, V]
-}
-
-func (valIn graphValueIn[E, V]) equal(valIn2 graphValueIn[E, V]) bool {
-	return valIn.val.Equal(valIn2.val) && valIn.edge.equal(valIn2.edge)
-}
-
-// Graph is an immutable container of a set of vertices. The Graph keeps track
-// of all Values which terminate an OpenEdge. E indicates the type of edge
-// values, while V indicates the type of vertex values.
-//
-// NOTE The current implementation of Graph is incredibly inefficient, there's
-// lots of O(N) operations, unnecessary copying on changes, and duplicate data
-// in memory.
-type Graph[E, V Value] struct {
-	edges  []*OpenEdge[E, V]
-	valIns []graphValueIn[E, V]
-}
-
-func (g *Graph[E, V]) cp() *Graph[E, V] {
-	cp := &Graph[E, V]{
-		edges:  make([]*OpenEdge[E, V], len(g.edges)),
-		valIns: make([]graphValueIn[E, V], len(g.valIns)),
-	}
-	copy(cp.edges, g.edges)
-	copy(cp.valIns, g.valIns)
-	return cp
-}
-
-func (g *Graph[E, V]) String() string {
-
-	var strs []string
-
-	for _, valIn := range g.valIns {
-		strs = append(
-			strs,
-			fmt.Sprintf("valIn(%s, %s)", valIn.edge.String(), valIn.val.String()),
-		)
-	}
-
-	return fmt.Sprintf("graph(%s)", strings.Join(strs, ", "))
-}
-
-// NOTE this method is used more for its functionality than for any performance
-// reasons... it's incredibly inefficient in how it deduplicates edges, but by
-// doing the deduplication we enable the graph map operation to work correctly.
-func (g *Graph[E, V]) dedupeEdge(edge *OpenEdge[E, V]) *OpenEdge[E, V] {
-
-	// check if there's an existing edge which is fully equivalent in the graph
-	// already, and if so return that.
-	for i := range g.edges {
-		if g.edges[i].equal(edge) {
-			return g.edges[i]
-		}
-	}
-
-	// if this edge is a value edge then there's nothing else to do, return it.
-	if _, ok := edge.FromValue(); ok {
-		return edge
-	}
-
-	// this edge is a tuple edge, it's possible that one of its sub-edges is
-	// already in the graph. dedupe each sub-edge individually.
-
-	tupEdges := make([]*OpenEdge[E, V], len(edge.tup))
-
-	for i := range edge.tup {
-		tupEdges[i] = g.dedupeEdge(edge.tup[i])
-	}
-
-	return TupleOut(edge.EdgeValue(), tupEdges...)
-}
-
-// ValueIns returns, if any, all OpenEdges which lead to the given Value in the
-// Graph (ie, all those added via AddValueIn).
-//
-// The returned slice should not be modified.
-//
-// TODO better name might be OpenEdgesInto.
-func (g *Graph[E, V]) ValueIns(val Value) []*OpenEdge[E, V] {
-
-	var edges []*OpenEdge[E, V]
-
-	for _, valIn := range g.valIns {
-		if valIn.val.Equal(val) {
-			edges = append(edges, valIn.edge)
-		}
-	}
-
-	return edges
-}
-
-// AddValueIn takes a OpenEdge and connects it to the Value vertex containing
-// val, returning the new Graph which reflects that connection.
-func (g *Graph[E, V]) AddValueIn(val V, oe *OpenEdge[E, V]) *Graph[E, V] {
-
-	valIn := graphValueIn[E, V]{
-		val:  val,
-		edge: oe,
-	}
-
-	for i := range g.valIns {
-		if g.valIns[i].equal(valIn) {
-			return g
-		}
-	}
-
-	valIn.edge = g.dedupeEdge(valIn.edge)
-
-	g = g.cp()
-	g.valIns = append(g.valIns, valIn)
-
-	return g
-}
-
-// AllValueIns returns all values which have had incoming edges added to them
-// using AddValueIn.
-func (g *Graph[E, V]) AllValueIns() []V {
-	vals := make([]V, len(g.valIns))
-	for i := range g.valIns {
-		vals[i] = g.valIns[i].val
-	}
-	return vals
-}
-
-// Equal returns whether or not the two Graphs are equivalent in value.
-func (g *Graph[E, V]) Equal(g2 *Graph[E, V]) bool {
-
-	if len(g.valIns) != len(g2.valIns) {
-		return false
-	}
-
-outer:
-	for _, valIn := range g.valIns {
-
-		for _, valIn2 := range g2.valIns {
-
-			if valIn.equal(valIn2) {
-				continue outer
-			}
-		}
-
-		return false
-	}
-
-	return true
-}
-
-func mapReduce[Ea, Va Value, Vb any](
-	root *OpenEdge[Ea, Va],
-	mapVal func(Va) (Vb, error),
-	reduceEdge func(*OpenEdge[Ea, Va], []Vb) (Vb, error),
-) (
-	Vb, error,
-) {
-
-	if valA, ok := root.FromValue(); ok {
-
-		valB, err := mapVal(valA)
-
-		if err != nil {
-			var zero Vb
-			return zero, err
-		}
-
-		return reduceEdge(root, []Vb{valB})
-	}
-
-	tupA, _ := root.FromTuple()
-
-	valsB := make([]Vb, len(tupA))
-
-	for i := range tupA {
-
-		valB, err := mapReduce[Ea, Va, Vb](
-			tupA[i], mapVal, reduceEdge,
-		)
-
-		if err != nil {
-			var zero Vb
-			return zero, err
-		}
-
-		valsB[i] = valB
-	}
-
-	return reduceEdge(root, valsB)
-}
-
-type mappedVal[Va Value, Vb any] struct {
-	valA Va
-	valB Vb // result
-}
-
-type reducedEdge[Ea, Va Value, Vb any] struct {
-	edgeA *OpenEdge[Ea, Va]
-	valB  Vb // result
-}
-
-// MapReduce recursively computes a resultant Value of type Vb from an
-// OpenEdge[Ea, Va].
-//
-// Tuple edges which are encountered will have Reduce called on each OpenEdge
-// branch of the tuple, to obtain a Vb for each branch. The edge value of the
-// tuple edge (Ea) and the just obtained Vbs are then passed to reduceEdge to
-// obtain a Vb for that edge.
-//
-// The values of value edges (Va) which are encountered are mapped to Vb using
-// the mapVal function. The edge value of those value edges (Ea) and the just
-// obtained Vb value are then passed to reduceEdge to obtain a Vb for that edge.
-//
-// If either the map or reduce function returns an error then processing is
-// immediately cancelled and that error is returned directly.
-//
-// If a value or edge is connected to multiple times within the root OpenEdge it
-// will only be mapped/reduced a single time, and the result of that single
-// map/reduction will be passed to each dependant operation.
-func MapReduce[Ea, Va Value, Vb any](
-	root *OpenEdge[Ea, Va],
-	mapVal func(Va) (Vb, error),
-	reduceEdge func(Ea, []Vb) (Vb, error),
-) (
-	Vb, error,
-) {
-
-	var (
-		zeroB Vb
-
-		// we use these to memoize reductions on values and edges, so a
-		// reduction is only performed a single time for each value/edge.
-		//
-		// NOTE this is not implemented very efficiently.
-		mappedVals   []mappedVal[Va, Vb]
-		reducedEdges []reducedEdge[Ea, Va, Vb]
-	)
-
-	return mapReduce[Ea, Va, Vb](
-		root,
-		func(valA Va) (Vb, error) {
-
-			for _, mappedVal := range mappedVals {
-				if mappedVal.valA.Equal(valA) {
-					return mappedVal.valB, nil
-				}
-			}
-
-			valB, err := mapVal(valA)
-
-			if err != nil {
-				return zeroB, err
-			}
-
-			mappedVals = append(mappedVals, mappedVal[Va, Vb]{
-				valA: valA,
-				valB: valB,
-			})
-
-			return valB, nil
-		},
-		func(edgeA *OpenEdge[Ea, Va], valBs []Vb) (Vb, error) {
-
-			for _, reducedEdge := range reducedEdges {
-				if reducedEdge.edgeA.equal(edgeA) {
-					return reducedEdge.valB, nil
-				}
-			}
-
-			valB, err := reduceEdge(edgeA.EdgeValue(), valBs)
-
-			if err != nil {
-				return zeroB, err
-			}
-
-			reducedEdges = append(reducedEdges, reducedEdge[Ea, Va, Vb]{
-				edgeA: edgeA,
-				valB:  valB,
-			})
-
-			return valB, nil
-		},
-	)
-}

graph/graph_test.go

@@ -1,268 +0,0 @@
-package graph
-
-import (
-	"errors"
-	"fmt"
-	"strconv"
-	"testing"
-
-	"github.com/stretchr/testify/assert"
-)
-
-type S string
-
-func (s S) Equal(s2 Value) bool { return s == s2.(S) }
-
-func (s S) String() string { return string(s) }
-
-type I int
-
-func (i I) Equal(i2 Value) bool { return i == i2.(I) }
-
-func (i I) String() string { return strconv.Itoa(int(i)) }
-
-func TestEqual(t *testing.T) {
-
-	var (
-		zeroValue S
-		zeroGraph = new(Graph[S, S])
-	)
-
-	tests := []struct {
-		a, b *Graph[S, S]
-		exp  bool
-	}{
-		{
-			a:   zeroGraph,
-			b:   zeroGraph,
-			exp: true,
-		},
-		{
-			a:   zeroGraph,
-			b:   zeroGraph.AddValueIn("out", ValueOut[S, S]("incr", "in")),
-			exp: false,
-		},
-		{
-			a:   zeroGraph.AddValueIn("out", ValueOut[S, S]("incr", "in")),
-			b:   zeroGraph.AddValueIn("out", ValueOut[S, S]("incr", "in")),
-			exp: true,
-		},
-		{
-			a: zeroGraph.AddValueIn("out", ValueOut[S, S]("incr", "in")),
-			b: zeroGraph.AddValueIn("out", TupleOut[S, S](
-				"add",
-				ValueOut[S, S]("ident", "in"),
-				ValueOut[S, S]("ident", "1"),
-			)),
-			exp: false,
-		},
-		{
-			// tuples are different order
-			a: zeroGraph.AddValueIn("out", TupleOut[S, S](
-				"add",
-				ValueOut[S, S]("ident", "1"),
-				ValueOut[S, S]("ident", "in"),
-			)),
-			b: zeroGraph.AddValueIn("out", TupleOut[S, S](
-				"add",
-				ValueOut[S, S]("ident", "in"),
-				ValueOut[S, S]("ident", "1"),
-			)),
-			exp: false,
-		},
-		{
-			// tuple with no edge value and just a single input edge should be
-			// equivalent to just that edge.
-			a: zeroGraph.AddValueIn("out", TupleOut[S, S](
-				zeroValue,
-				ValueOut[S, S]("ident", "1"),
-			)),
-			b:   zeroGraph.AddValueIn("out", ValueOut[S, S]("ident", "1")),
-			exp: true,
-		},
-		{
-			// tuple with an edge value and just a single input edge that has no
-			// edgeVal should be equivalent to just that edge with the tuple's
-			// edge value.
-			a: zeroGraph.AddValueIn("out", TupleOut[S, S](
-				"ident",
-				ValueOut[S, S](zeroValue, "1"),
-			)),
-			b:   zeroGraph.AddValueIn("out", ValueOut[S, S]("ident", "1")),
-			exp: true,
-		},
-		{
-			a: zeroGraph.
-				AddValueIn("out", ValueOut[S, S]("incr", "in")).
-				AddValueIn("out2", ValueOut[S, S]("incr2", "in2")),
-			b: zeroGraph.
-				AddValueIn("out", ValueOut[S, S]("incr", "in")),
-			exp: false,
-		},
-		{
-			a: zeroGraph.
-				AddValueIn("out", ValueOut[S, S]("incr", "in")).
-				AddValueIn("out2", ValueOut[S, S]("incr2", "in2")),
-			b: zeroGraph.
-				AddValueIn("out", ValueOut[S, S]("incr", "in")).
-				AddValueIn("out2", ValueOut[S, S]("incr2", "in2")),
-			exp: true,
-		},
-		{
-			// order of value ins shouldn't matter
-			a: zeroGraph.
-				AddValueIn("out", ValueOut[S, S]("incr", "in")).
-				AddValueIn("out2", ValueOut[S, S]("incr2", "in2")),
-			b: zeroGraph.
-				AddValueIn("out2", ValueOut[S, S]("incr2", "in2")).
-				AddValueIn("out", ValueOut[S, S]("incr", "in")),
-			exp: true,
-		},
-	}
-
-	for i, test := range tests {
-		t.Run(strconv.Itoa(i), func(t *testing.T) {
-			assert.Equal(t, test.exp, test.a.Equal(test.b))
-		})
-	}
-}
-
-type mapReduceTestEdge struct {
-	name string
-	fn   func([]int) int
-	done bool
-}
-
-func (e *mapReduceTestEdge) Equal(e2i Value) bool {
-
-	e2, _ := e2i.(*mapReduceTestEdge)
-
-	if e == nil || e2 == nil {
-		return e == e2
-	}
-
-	return e.name == e2.name
-}
-
-func (e *mapReduceTestEdge) String() string {
-	return e.name
-}
-
-func (e *mapReduceTestEdge) do(ii []int) int {
-
-	if e.done {
-		panic(fmt.Sprintf("%q already done", e.name))
-	}
-
-	e.done = true
-
-	return e.fn(ii)
-}
-
-func TestMapReduce(t *testing.T) {
-
-	type (
-		Va   = I
-		Vb   = int
-		Ea   = *mapReduceTestEdge
-		edge = OpenEdge[Ea, Va]
-	)
-
-	var (
-		zeroVb Vb
-	)
-
-	vOut := func(edge Ea, val Va) *edge {
-		return ValueOut[Ea, Va](edge, val)
-	}
-
-	tOut := func(edge Ea, ins ...*edge) *edge {
-		return TupleOut[Ea, Va](edge, ins...)
-	}
-
-	add := func() *mapReduceTestEdge {
-		return &mapReduceTestEdge{
-			name: "add",
-			fn: func(ii []int) int {
-				var n int
-				for _, i := range ii {
-					n += i
-				}
-				return n
-			},
-		}
-	}
-
-	mul := func() *mapReduceTestEdge {
-		return &mapReduceTestEdge{
-			name: "mul",
-			fn: func(ii []int) int {
-				n := 1
-				for _, i := range ii {
-					n *= i
-				}
-				return n
-			},
-		}
-	}
-
-	mapVal := func(valA Va) (Vb, error) {
-		return Vb(valA * 10), nil
-	}
-
-	reduceEdge := func(edgeA Ea, valBs []Vb) (Vb, error) {
-
-		if edgeA == nil {
-
-			if len(valBs) == 1 {
-				return valBs[0], nil
-			}
-
-			return zeroVb, errors.New("tuple edge must have edge value")
-		}
-
-		return edgeA.do(valBs), nil
-	}
-
-	tests := []struct {
-		in  *edge
-		exp int
-	}{
-		{
-			in:  vOut(nil, 1),
-			exp: 10,
-		},
-		{
-			in:  vOut(add(), 1),
-			exp: 10,
-		},
-		{
-			in: tOut(
-				add(),
-				vOut(nil, 1),
-				vOut(add(), 2),
-				vOut(mul(), 3),
-			),
-			exp: 60,
-		},
-		{
-			// duplicate edges and values getting used twice, each should only
-			// get eval'd once
-			in: tOut(
-				add(),
-				tOut(add(), vOut(nil, 1), vOut(nil, 2)),
-				tOut(add(), vOut(nil, 1), vOut(nil, 2)),
-				tOut(add(), vOut(nil, 3), vOut(nil, 3)),
-			),
-			exp: 120,
-		},
-	}
-
-	for i, test := range tests {
-		t.Run(strconv.Itoa(i), func(t *testing.T) {
-			got, err := MapReduce(test.in, mapVal, reduceEdge)
-			assert.NoError(t, err)
-			assert.Equal(t, test.exp, got)
-		})
-	}
-}

lang/lang.go

@@ -0,0 +1,118 @@
+package lang
+
+import (
+	"fmt"
+	"reflect"
+	"strings"
+)
+
+// Commonly used Terms
+var (
+	// Language structure types
+	AAtom  = Atom("atom")
+	AConst = Atom("const")
+	ATuple = Atom("tup")
+	AList  = Atom("list")
+
+	// Match shortcuts
+	AUnder    = Atom("_")
+	TDblUnder = Tuple{AUnder, AUnder}
+)
+
+// Term is a unit of language which carries some meaning. Some Terms are
+// actually comprised of multiple sub-Terms.
+type Term interface {
+	fmt.Stringer // for debugging
+
+	// Type returns a Term which describes the type of this Term, i.e.  the
+	// components this Term is comprised of.
+	Type() Term
+}
+
+// Equal returns whether or not two Terms are of equal value
+func Equal(t1, t2 Term) bool {
+	return reflect.DeepEqual(t1, t2)
+}
+
+////////////////////////////////////////////////////////////////////////////////
+
+// Atom is a constant with no other meaning than that it can be equal or not
+// equal to another Atom.
+type Atom string
+
+func (a Atom) String() string {
+	return string(a)
+}
+
+// Type implements the method for Term
+func (a Atom) Type() Term {
+	return AAtom
+}
+
+////////////////////////////////////////////////////////////////////////////////
+
+// Const is a constant whose meaning depends on the context in which it is used
+type Const string
+
+func (a Const) String() string {
+	return string(a)
+}
+
+// Type implements the method for Term
+func (a Const) Type() Term {
+	return AConst
+}
+
+////////////////////////////////////////////////////////////////////////////////
+
+// Tuple is a compound Term of zero or more sub-Terms, each of which may have a
+// different Type. Both the length of the Tuple and the Type of each of it's
+// sub-Terms are components in the Tuple's Type.
+type Tuple []Term
+
+func (t Tuple) String() string {
+	ss := make([]string, len(t))
+	for i := range t {
+		ss[i] = t[i].String()
+	}
+	return "(" + strings.Join(ss, " ") + ")"
+}
+
+// Type implements the method for Term
+func (t Tuple) Type() Term {
+	tt := make(Tuple, len(t))
+	for i := range t {
+		tt[i] = t[i].Type()
+	}
+	return Tuple{ATuple, tt}
+}
+
+////////////////////////////////////////////////////////////////////////////////
+
+type list struct {
+	typ Term
+	ll  []Term
+}
+
+// List is a compound Term of zero or more sub-Terms, each of which must have
+// the same Type (the one given as the first argument to this function). Only
+// the Type of the sub-Terms is a component in the List's Type.
+func List(typ Term, elems ...Term) Term {
+	return list{
+		typ: typ,
+		ll:  elems,
+	}
+}
+
+func (l list) String() string {
+	ss := make([]string, len(l.ll))
+	for i := range l.ll {
+		ss[i] = l.ll[i].String()
+	}
+	return "[" + strings.Join(ss, " ") + "]"
+}
+
+// Type implements the method for Term
+func (l list) Type() Term {
+	return Tuple{AList, l.typ}
+}

lang/match.go

@@ -0,0 +1,54 @@
+package lang
+
+import "fmt"
+
+// Match is used to pattern match an arbitrary Term against a pattern. A pattern
+// is a 2-tuple of the type (as an atom, e.g. AAtom, AConst) and a matching
+// value.
+//
+// If the value is AUnder the pattern will match all Terms of the type,
+// regardless of their value. If the pattern's type and value are both AUnder
+// the pattern will match all Terms.
+//
+// If the pattern's value is a Tuple or a List, each of its values will be used
+// as a sub-pattern to match against the corresponding value in the value.
+func Match(pat Tuple, t Term) bool {
+	if len(pat) != 2 {
+		return false
+	}
+	pt, pv := pat[0], pat[1]
+
+	switch pt {
+	case AAtom:
+		a, ok := t.(Atom)
+		return ok && (Equal(pv, AUnder) || Equal(pv, a))
+	case AConst:
+		c, ok := t.(Const)
+		return ok && (Equal(pv, AUnder) || Equal(pv, c))
+	case ATuple:
+		tt, ok := t.(Tuple)
+		if !ok {
+			return false
+		} else if Equal(pv, AUnder) {
+			return true
+		}
+
+		pvt := pv.(Tuple)
+		if len(tt) != len(pvt) {
+			return false
+		}
+		for i := range tt {
+			pvti, ok := pvt[i].(Tuple)
+			if !ok || !Match(pvti, tt[i]) {
+				return false
+			}
+		}
+		return true
+	case AList:
+		panic("TODO")
+	case AUnder:
+		return true
+	default:
+		panic(fmt.Sprintf("unknown type %T", pt))
+	}
+}

lang/match_test.go

@@ -0,0 +1,66 @@
+package lang
+
+import (
+	. "testing"
+
+	"github.com/stretchr/testify/assert"
+)
+
+func TestMatch(t *T) {
+	pat := func(typ, val Term) Tuple {
+		return Tuple{typ, val}
+	}
+
+	tests := []struct {
+		pattern Tuple
+		t       Term
+		exp     bool
+	}{
+		{pat(AAtom, Atom("foo")), Atom("foo"), true},
+		{pat(AAtom, Atom("foo")), Atom("bar"), false},
+		{pat(AAtom, Atom("foo")), Const("foo"), false},
+		{pat(AAtom, Atom("foo")), Tuple{Atom("a"), Atom("b")}, false},
+		{pat(AAtom, Atom("_")), Atom("bar"), true},
+		{pat(AAtom, Atom("_")), Const("bar"), false},
+
+		{pat(AConst, Const("foo")), Const("foo"), true},
+		{pat(AConst, Const("foo")), Atom("foo"), false},
+		{pat(AConst, Const("foo")), Const("bar"), false},
+		{pat(AConst, Atom("_")), Const("bar"), true},
+		{pat(AConst, Atom("_")), Atom("foo"), false},
+
+		{
+			pat(ATuple, Tuple{
+				pat(AAtom, Atom("foo")),
+				pat(AAtom, Atom("bar")),
+			}),
+			Tuple{Atom("foo"), Atom("bar")},
+			true,
+		},
+		{
+			pat(ATuple, Tuple{
+				pat(AAtom, Atom("_")),
+				pat(AAtom, Atom("bar")),
+			}),
+			Tuple{Atom("foo"), Atom("bar")},
+			true,
+		},
+		{
+			pat(ATuple, Tuple{
+				pat(AAtom, Atom("_")),
+				pat(AAtom, Atom("_")),
+				pat(AAtom, Atom("_")),
+			}),
+			Tuple{Atom("foo"), Atom("bar")},
+			false,
+		},
+
+		{pat(AUnder, AUnder), Atom("foo"), true},
+		{pat(AUnder, AUnder), Const("foo"), true},
+		{pat(AUnder, AUnder), Tuple{Atom("a"), Atom("b")}, true},
+	}
+
+	for _, testCase := range tests {
+		assert.Equal(t, testCase.exp, Match(testCase.pattern, testCase.t), "%#v", testCase)
+	}
+}

lexer/lexer.go

@@ -0,0 +1,349 @@
+package lexer
+
+import (
+	"bufio"
+	"bytes"
+	"errors"
+	"fmt"
+	"io"
+	"strings"
+)
+
+// TokenType indicates the type of a token
+type TokenType string
+
+// Different token types
+const (
+	Identifier TokenType = "identifier"
+
+	// Punctuation are tokens which connect two other tokens
+	Punctuation TokenType = "punctuation"
+
+	// Wrapper wraps one or more tokens
+	Wrapper TokenType = "wrapper"
+	String  TokenType = "string"
+	Err     TokenType = "err"
+	EOF     TokenType = "eof"
+)
+
+// Token is a single token which has been read in. All Tokens have a non-empty
+// Val
+type Token struct {
+	TokenType
+	Val      string
+	Row, Col int
+}
+
+// Equal returns whether two tokens are of equal type and value
+func (tok Token) Equal(tok2 Token) bool {
+	return tok.TokenType == tok2.TokenType && tok.Val == tok2.Val
+}
+
+// Err returns the error contained by the token, if any. Only returns non-nil if
+// TokenType is Err or EOF
+func (tok Token) Err() error {
+	if tok.TokenType == Err || tok.TokenType == EOF {
+		return fmt.Errorf("[line:%d col:%d] %s", tok.Row, tok.Col, tok.Val)
+	}
+	return nil
+}
+
+func (tok Token) String() string {
+	var typ string
+	switch tok.TokenType {
+	case Identifier:
+		typ = "ident"
+	case Punctuation:
+		typ = "punct"
+	case String:
+		typ = "str"
+	case Err, EOF:
+		typ = "err"
+	}
+	return fmt.Sprintf("%s(%q)", typ, tok.Val)
+}
+
+type lexerFn func(*Lexer) lexerFn
+
+// Lexer is used to read in ginger tokens from a source. HasNext() must be
+// called before every call to Next()
+type Lexer struct {
+	in  *bufio.Reader
+	out *bytes.Buffer
+	cur lexerFn
+
+	next []Token
+
+	row, col       int
+	absRow, absCol int
+}
+
+// New returns a Lexer which will read tokens from the given source.
+func New(r io.Reader) *Lexer {
+	return &Lexer{
+		in:  bufio.NewReader(r),
+		out: new(bytes.Buffer),
+		cur: lex,
+
+		row: -1,
+		col: -1,
+	}
+}
+
+func (l *Lexer) emit(t TokenType) {
+	str := l.out.String()
+	if str == "" {
+		panic("cannot emit empty token")
+	}
+	l.out.Reset()
+
+	l.emitTok(Token{
+		TokenType: t,
+		Val:       str,
+		Row:       l.row,
+		Col:       l.col,
+	})
+}
+
+func (l *Lexer) emitErr(err error) {
+	tok := Token{
+		TokenType: Err,
+		Val:       err.Error(),
+		Row:       l.absRow,
+		Col:       l.absCol,
+	}
+	if err == io.EOF {
+		tok.TokenType = EOF
+	}
+	l.emitTok(tok)
+}
+
+func (l *Lexer) emitTok(tok Token) {
+	l.next = append(l.next, tok)
+	l.row = -1
+	l.col = -1
+}
+
+func (l *Lexer) readRune() (rune, error) {
+	r, _, err := l.in.ReadRune()
+	if err != nil {
+		return r, err
+	}
+
+	if r == '\n' {
+		l.absRow++
+		l.absCol = 0
+	} else {
+		l.absCol++
+	}
+
+	return r, err
+}
+
+func (l *Lexer) peekRune() (rune, error) {
+	r, _, err := l.in.ReadRune()
+	if err != nil {
+		return r, err
+	}
+
+	if err := l.in.UnreadRune(); err != nil {
+		return r, err
+	}
+	return r, nil
+}
+
+func (l *Lexer) readAndPeek() (rune, rune, error) {
+	r, err := l.readRune()
+	if err != nil {
+		return r, 0, err
+	}
+
+	n, err := l.peekRune()
+	return r, n, err
+}
+
+func (l *Lexer) bufferRune(r rune) {
+	l.out.WriteRune(r)
+	if l.row < 0 && l.col < 0 {
+		l.row, l.col = l.absRow, l.absCol
+	}
+}
+
+// HasNext returns true if Next should be called, and false if it should not be
+// called and Err should be called instead. When HasNext returns false the Lexer
+// is considered to be done
+func (l *Lexer) HasNext() bool {
+	for {
+		if len(l.next) > 0 {
+			return true
+		} else if l.cur == nil {
+			return false
+		}
+		l.cur = l.cur(l)
+	}
+}
+
+// Next returns the next available token. HasNext must be called before every
+// call to Next
+func (l *Lexer) Next() Token {
+	t := l.next[0]
+	l.next = l.next[1:]
+	if len(l.next) == 0 {
+		l.next = nil
+	}
+	return t
+}
+
+////////////////////////////////////////////////////////////////////////////////
+// the actual fsm
+
+var whitespaceSet = " \n\r\t\v\f"
+var punctuationSet = ",>"
+var wrapperSet = "{}()"
+var identifierSepSet = whitespaceSet + punctuationSet + wrapperSet
+
+func lex(l *Lexer) lexerFn {
+	r, err := l.readRune()
+	if err != nil {
+		l.emitErr(err)
+		return nil
+	}
+
+	// handle comments first, cause we have to peek for those. We ignore errors,
+	// and assume that any error that would happen here will happen again the
+	// next read
+	if n, _ := l.peekRune(); r == '/' && n == '/' {
+		return lexLineComment
+	} else if r == '/' && n == '*' {
+		return lexBlockComment
+	}
+
+	return lexSingleRune(l, r)
+}
+
+func lexSingleRune(l *Lexer, r rune) lexerFn {
+	switch {
+	case strings.ContainsRune(whitespaceSet, r):
+		return lex
+	case strings.ContainsRune(punctuationSet, r):
+		l.bufferRune(r)
+		l.emit(Punctuation)
+		return lex
+	case strings.ContainsRune(wrapperSet, r):
+		l.bufferRune(r)
+		l.emit(Wrapper)
+		return lex
+	case r == '"' || r == '\'' || r == '`':
+		canEscape := r != '`'
+		return lexStrStart(l, r, makeLexStr(r, canEscape))
+	default:
+		l.bufferRune(r)
+		return lexIdentifier
+	}
+}
+
+func lexIdentifier(l *Lexer) lexerFn {
+	r, err := l.readRune()
+	if err != nil {
+		l.emit(Identifier)
+		l.emitErr(err)
+		return nil
+	}
+
+	if strings.ContainsRune(identifierSepSet, r) {
+		l.emit(Identifier)
+		return lexSingleRune(l, r)
+	}
+
+	l.bufferRune(r)
+
+	return lexIdentifier
+}
+
+func lexLineComment(l *Lexer) lexerFn {
+	r, err := l.readRune()
+	if err != nil {
+		l.emitErr(err)
+		return nil
+	}
+	if r == '\n' {
+		return lex
+	}
+	return lexLineComment
+}
+
+// assumes the starting / has been read already
+func lexBlockComment(l *Lexer) lexerFn {
+	depth := 1
+
+	var recurse lexerFn
+	recurse = func(l *Lexer) lexerFn {
+		r, err := l.readRune()
+		if err != nil {
+			l.emitErr(err)
+			return nil
+		}
+		n, _ := l.peekRune()
+
+		if r == '/' && n == '*' {
+			depth++
+		} else if r == '*' && n == '/' {
+			depth--
+		}
+
+		if depth == 0 {
+			return lexSkipThen(lex)
+		}
+		return recurse
+	}
+	return recurse
+}
+
+func lexStrStart(lexer *Lexer, r rune, then lexerFn) lexerFn {
+	lexer.bufferRune(r)
+	return then
+}
+
+func makeLexStr(quoteC rune, canEscape bool) lexerFn {
+	var fn lexerFn
+	fn = func(l *Lexer) lexerFn {
+		r, n, err := l.readAndPeek()
+		if err != nil {
+			if err == io.EOF {
+				if r == quoteC {
+					l.bufferRune(r)
+					l.emit(String)
+					l.emitErr(err)
+					return nil
+				}
+				l.emitErr(errors.New("expected end of string, got end of file"))
+				return nil
+			}
+		}
+
+		if canEscape && r == '\\' && n == quoteC {
+			l.bufferRune(r)
+			l.bufferRune(n)
+			return lexSkipThen(fn)
+		}
+
+		l.bufferRune(r)
+		if r == quoteC {
+			l.emit(String)
+			return lex
+		}
+
+		return fn
+	}
+	return fn
+}
+
+func lexSkipThen(then lexerFn) lexerFn {
+	return func(l *Lexer) lexerFn {
+		if _, err := l.readRune(); err != nil {
+			l.emitErr(err)
+			return nil
+		}
+		return then
+	}
+}

lexer/lexer_test.go

@@ -0,0 +1,82 @@
+package lexer
+
+import (
+	"bytes"
+	. "testing"
+
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+)
+
+var lexTestSrc = `
+	// this is a comment
+	// // this is also a comment
+	a
+	anIdentifier
+	1
+	100
+	1.5
+	1.5e9
+
+	/*
+		some stuff
+	*/
+
+	/* this should actually work */
+	/*/
+
+	/*
+		nested!
+		/*
+			wtf this is crazy
+		*/
+	*/
+
+	(punctuation,is{cool}> )
+	-tab
+
+	"this is a string", "and so is this one"
+	"\"foo"
+	"bar\"baz\""
+	"buz\0"
+`
+
+func TestLex(t *T) {
+	l := New(bytes.NewBufferString(lexTestSrc))
+
+	assertNext := func(typ TokenType, val string, row, col int) {
+		t.Logf("asserting %s %q [row:%d col:%d]", typ, val, row, col)
+		require.True(t, l.HasNext())
+		tok := l.Next()
+		assert.Equal(t, typ, tok.TokenType)
+		assert.Equal(t, val, tok.Val)
+		assert.Equal(t, row, tok.Row)
+		assert.Equal(t, col, tok.Col)
+	}
+
+	assertNext(Identifier, "a", 3, 2)
+	assertNext(Identifier, "anIdentifier", 4, 2)
+	assertNext(Identifier, "1", 5, 2)
+	assertNext(Identifier, "100", 6, 2)
+	assertNext(Identifier, "1.5", 7, 2)
+	assertNext(Identifier, "1.5e9", 8, 2)
+	assertNext(Wrapper, "(", 24, 2)
+	assertNext(Identifier, "punctuation", 24, 3)
+	assertNext(Punctuation, ",", 24, 14)
+	assertNext(Identifier, "is", 24, 15)
+	assertNext(Wrapper, "{", 24, 17)
+	assertNext(Identifier, "cool", 24, 18)
+	assertNext(Wrapper, "}", 24, 22)
+	assertNext(Punctuation, ">", 24, 23)
+	assertNext(Wrapper, ")", 24, 25)
+	assertNext(Identifier, "-tab", 25, 2)
+	assertNext(String, `"this is a string"`, 27, 2)
+	assertNext(Punctuation, ",", 27, 20)
+	assertNext(String, `"and so is this one"`, 27, 22)
+	assertNext(String, `"\"foo"`, 28, 2)
+	assertNext(String, `"bar\"baz\""`, 29, 2)
+	assertNext(String, `"buz\0"`, 30, 2)
+	assertNext(EOF, "EOF", 31, 0)
+
+	assert.False(t, l.HasNext())
+}

main.go

@@ -0,0 +1,47 @@
+package main
+
+import (
+	"fmt"
+
+	"github.com/mediocregopher/ginger/lang"
+	"github.com/mediocregopher/ginger/vm"
+)
+
+func main() {
+	mkcmd := func(a lang.Atom, args ...lang.Term) lang.Tuple {
+		if len(args) == 1 {
+			return lang.Tuple{a, args[0]}
+		}
+		return lang.Tuple{a, lang.Tuple(args)}
+	}
+	mkint := func(i string) lang.Tuple {
+		return lang.Tuple{vm.Int, lang.Const(i)}
+	}
+
+	//foo := lang.Atom("foo")
+	//tt := []lang.Term{
+	//	mkcmd(vm.Assign, foo, mkint("1")),
+	//	mkcmd(vm.Add, mkcmd(vm.Tuple, mkcmd(vm.Var, foo), mkint("2"))),
+	//}
+
+	foo := lang.Atom("foo")
+	bar := lang.Atom("bar")
+	baz := lang.Atom("baz")
+	tt := []lang.Term{
+		mkcmd(vm.Assign, foo, mkcmd(vm.Tuple, mkint("1"), mkint("2"))),
+		mkcmd(vm.Assign, bar, mkcmd(vm.Add, mkcmd(vm.Var, foo))),
+		mkcmd(vm.Assign, baz, mkcmd(vm.Add, mkcmd(vm.Var, foo))),
+		mkcmd(vm.Add, mkcmd(vm.Tuple, mkcmd(vm.Var, bar), mkcmd(vm.Var, baz))),
+	}
+
+	mod, err := vm.Build(tt...)
+	if err != nil {
+		panic(err)
+	}
+	defer mod.Dispose()
+
+	mod.Dump()
+
+	out, err := mod.Run()
+	fmt.Printf("\n\n########\nout: %v %v\n", out, err)
+}

sandbox/list/list.go

@@ -0,0 +1,39 @@
+package list
+
+import "fmt"
+
+/*
+	+ size isn't really _necessary_ unless O(1) Len is wanted
+	+ append doesn't work well on stack
+*/
+
+type List struct {
+	// in practice this would be a constant size, with the compiler knowing the
+	// size
+	underlying []int
+	head, size int
+}
+
+func New(ii ...int) List {
+	l := List{
+		underlying: make([]int, ii),
+		size:       len(ii),
+	}
+	copy(l.underlying, ii)
+	return l
+}
+
+func (l List) Len() int {
+	return l.size
+}
+
+func (l List) HeadTail() (int, List) {
+	if l.size == 0 {
+		panic(fmt.Sprintf("can't take HeadTail of empty list"))
+	}
+	return l.underlying[l.head], List{
+		underlying: l.underlying,
+		head:       l.head + 1,
+		size:       l.size - 1,
+	}
+}

sandbox/syntax.txt

@@ -1,108 +0,0 @@
-# 2021/08/26
-#
-# output of godoc on gg is this:
-
-var ZeroGraph = &Graph{ ... }
-func Equal(g1, g2 *Graph) bool
-type Edge struct{ ... }
-type Graph struct{ ... }
-type OpenEdge struct{ ... }
-    func TupleOut(ins []OpenEdge, edgeVal Value) OpenEdge
-    func ValueOut(val, edgeVal Value) OpenEdge
-type Value struct{ ... }
-    func NewValue(V interface{}) Value
-type Vertex struct{ ... }
-type VertexType string
-    const ValueVertex VertexType = "value" ...
-
-We just need to formulate a syntax which describes these operations and
-entities.
-
-Based on an old note I found in this file it seems like it reads
-better to actually order everything "backwards" in the syntax, so I'm going to
-go with that. I left the note at the bottom, commented out
-
--(<openEdge>,...)           TupleOut                 (an openEdge)
--<edgeVal>-(<openEdge>,...) TupleOut with an edgeVal (an openEdge)
--<val>                      ValueOut                 (an openEdge)
--<edgeVal>-<val>            ValueOut with an edgeVal (an openEdge)
-{<val> <openEdge>, ...}     ValueIn                  (a graph)
-
-values can only be alphanumeric, or graphs.
-TODO what to do about negative numbers? -1 is ambiguous
-
-This means the below fibonnaci can be done using:
-
-{
-    decr -{ out -sub-(-in, -1) }
-
-    out -{
-        n -0-in,
-        a -1-in,
-        b -2-in,
-
-        out -if-(
-            -zero?-n,
-            -a,
-            -recur-(
-                -decr-n,
-                -b,
-                -add-(-a,-b)
-            ),
-        )
-    }-(-in, -0, -1)
-}
-
-###
-
-Let's try to get rid of all the ugly prefix dashes (and maybe solve the -1
-question). We ditch the dashes all-together; TupleOut with an edgeVal can be
-done by just joining the two, and ValueOut with an edgeVal we can just make look
-like a TupleOut with a single openEdge (which... it kind of is anyway).
-
-(<openEdge>,...)          TupleOut                 (an openEdge)
-<edgeVal>(<openEdge>,...) TupleOut with an edgeVal (an openEdge)
-<val>                     ValueOut                 (an openEdge)
-<edgeVal>(<val>)          ValueOut with an edgeVal (an openEdge)
-{<val> <openEdge>[, ...]} ValueIn                  (a graph)
-
-values can only be alphanumeric, or graphs.
-
-```
-{
-    decr { out add(in, -1) }
-
-    out {
-        n 0(in),
-        a 1(in),
-        b 2(in),
-
-        out if(
-            zero?(n),
-            a,
-            recur(decr(n), b, add(a,b))
-        )
-
-    }(in, 0, 1)
-}
-```
-
-################
-# The Old Note #
-################
-#
-#decr- add- |- in
-#           |- (-1)
-#
-#fib- (
-#    fibInner- (
-#        {n, a, b}- in
-#        out- if- |- zero?- n
-#                 |- a
-#                 |- fibInner- |- decr- n
-#                              |- b
-#                              |- add- {a,b}
-#    )
-#)
-#
-#out- fib- atoi- first- in

vm/cmds.go

@@ -0,0 +1,280 @@
+package vm
+
+import (
+	"errors"
+	"fmt"
+	"strconv"
+
+	"github.com/mediocregopher/ginger/lang"
+	"llvm.org/llvm/bindings/go/llvm"
+)
+
+type op interface {
+	inType() valType
+	outType() valType
+	build(*Module) (llvm.Value, error)
+}
+
+type valType struct {
+	term lang.Term
+	llvm llvm.Type
+}
+
+func (vt valType) isInt() bool {
+	return lang.Equal(Int, vt.term)
+}
+
+func (vt valType) eq(vt2 valType) bool {
+	return lang.Equal(vt.term, vt2.term) && vt.llvm == vt2.llvm
+}
+
+// primitive valTypes
+var (
+	valTypeVoid = valType{term: lang.Tuple{}, llvm: llvm.VoidType()}
+	valTypeInt  = valType{term: Int, llvm: llvm.Int64Type()}
+)
+
+////////////////////////////////////////////////////////////////////////////////
+
+// most types don't have an input, so we use this as a shortcut
+type voidIn struct{}
+
+func (voidIn) inType() valType {
+	return valTypeVoid
+}
+
+////////////////////////////////////////////////////////////////////////////////
+
+type intOp struct {
+	voidIn
+	c lang.Const
+}
+
+func (io intOp) outType() valType {
+	return valTypeInt
+}
+
+func (io intOp) build(mod *Module) (llvm.Value, error) {
+	ci, err := strconv.ParseInt(string(io.c), 10, 64)
+	if err != nil {
+		return llvm.Value{}, err
+	}
+	return llvm.ConstInt(llvm.Int64Type(), uint64(ci), false), nil
+}
+
+////////////////////////////////////////////////////////////////////////////////
+
+type tupOp struct {
+	voidIn
+	els []op
+}
+
+func (to tupOp) outType() valType {
+	termTypes := make(lang.Tuple, len(to.els))
+	llvmTypes := make([]llvm.Type, len(to.els))
+	for i := range to.els {
+		elValType := to.els[i].outType()
+		termTypes[i] = elValType.term
+		llvmTypes[i] = elValType.llvm
+	}
+	vt := valType{term: lang.Tuple{Tuple, termTypes}}
+	if len(llvmTypes) == 0 {
+		vt.llvm = llvm.VoidType()
+	} else {
+		vt.llvm = llvm.StructType(llvmTypes, false)
+	}
+	return vt
+}
+
+func (to tupOp) build(mod *Module) (llvm.Value, error) {
+	str := llvm.Undef(to.outType().llvm)
+	var val llvm.Value
+	var err error
+	for i := range to.els {
+		if val, err = to.els[i].build(mod); err != nil {
+			return llvm.Value{}, err
+		}
+		str = mod.b.CreateInsertValue(str, val, i, "")
+	}
+	return str, err
+}
+
+////////////////////////////////////////////////////////////////////////////////
+
+type tupElOp struct {
+	voidIn
+	tup op
+	i   int
+}
+
+func (teo tupElOp) outType() valType {
+	tupType := teo.tup.outType()
+	return valType{
+		llvm: tupType.llvm.StructElementTypes()[teo.i],
+		term: tupType.term.(lang.Tuple)[1].(lang.Tuple)[1],
+	}
+}
+
+func (teo tupElOp) build(mod *Module) (llvm.Value, error) {
+	if to, ok := teo.tup.(tupOp); ok {
+		return to.els[teo.i].build(mod)
+	}
+
+	tv, err := teo.tup.build(mod)
+	if err != nil {
+		return llvm.Value{}, err
+	}
+	return mod.b.CreateExtractValue(tv, teo.i, ""), nil
+}
+
+////////////////////////////////////////////////////////////////////////////////
+
+type varOp struct {
+	op
+	v     llvm.Value
+	built bool
+}
+
+func (vo *varOp) build(mod *Module) (llvm.Value, error) {
+	if !vo.built {
+		var err error
+		if vo.v, err = vo.op.build(mod); err != nil {
+			return llvm.Value{}, err
+		}
+		vo.built = true
+	}
+	return vo.v, nil
+}
+
+type varCtx map[string]*varOp
+
+func (c varCtx) assign(name string, vo *varOp) error {
+	if _, ok := c[name]; ok {
+		return fmt.Errorf("var %q already assigned", name)
+	}
+	c[name] = vo
+	return nil
+}
+
+func (c varCtx) get(name string) (*varOp, error) {
+	if o, ok := c[name]; ok {
+		return o, nil
+	}
+	return nil, fmt.Errorf("var %q referenced before assignment", name)
+}
+
+////////////////////////////////////////////////////////////////////////////////
+
+type addOp struct {
+	voidIn
+	a, b op
+}
+
+func (ao addOp) outType() valType {
+	return ao.a.outType()
+}
+
+func (ao addOp) build(mod *Module) (llvm.Value, error) {
+	av, err := ao.a.build(mod)
+	if err != nil {
+		return llvm.Value{}, err
+	}
+	bv, err := ao.b.build(mod)
+	if err != nil {
+		return llvm.Value{}, err
+	}
+	return mod.b.CreateAdd(av, bv, ""), nil
+}
+
+////////////////////////////////////////////////////////////////////////////////
+
+func termToOp(ctx varCtx, t lang.Term) (op, error) {
+	aPat := func(a lang.Atom) lang.Tuple {
+		return lang.Tuple{lang.AAtom, a}
+	}
+	cPat := func(t lang.Term) lang.Tuple {
+		return lang.Tuple{lang.AConst, t}
+	}
+	tPat := func(el ...lang.Term) lang.Tuple {
+		return lang.Tuple{Tuple, lang.Tuple(el)}
+	}
+
+	if !lang.Match(tPat(aPat(lang.AUnder), lang.TDblUnder), t) {
+		return nil, fmt.Errorf("term %v does not look like a vm command", t)
+	}
+	k := t.(lang.Tuple)[0].(lang.Atom)
+	v := t.(lang.Tuple)[1]
+
+	// for when v is a Tuple argument, convenience function for casting
+	vAsTup := func(n int) ([]op, error) {
+		vop, err := termToOp(ctx, v)
+		if err != nil {
+			return nil, err
+		}
+		ops := make([]op, n)
+		for i := range ops {
+			ops[i] = tupElOp{tup: vop, i: i}
+		}
+
+		return ops, nil
+	}
+
+	switch k {
+	case Int:
+		if !lang.Match(cPat(lang.AUnder), v) {
+			return nil, errors.New("int requires constant arg")
+		}
+		return intOp{c: v.(lang.Const)}, nil
+	case Tuple:
+		if !lang.Match(lang.Tuple{Tuple, lang.AUnder}, v) {
+			return nil, errors.New("tup requires tuple arg")
+		}
+		tup := v.(lang.Tuple)
+		tc := tupOp{els: make([]op, len(tup))}
+		var err error
+		for i := range tup {
+			if tc.els[i], err = termToOp(ctx, tup[i]); err != nil {
+				return nil, err
+			}
+		}
+		return tc, nil
+	case Var:
+		if !lang.Match(aPat(lang.AUnder), v) {
+			return nil, errors.New("var requires atom arg")
+		}
+		name := v.(lang.Atom).String()
+		return ctx.get(name)
+
+	case Assign:
+		if !lang.Match(tPat(tPat(aPat(Var), aPat(lang.AUnder)), lang.TDblUnder), v) {
+			return nil, errors.New("assign requires 2-tuple arg, the first being a var")
+		}
+		tup := v.(lang.Tuple)
+		name := tup[0].(lang.Tuple)[1].String()
+		o, err := termToOp(ctx, tup[1])
+		if err != nil {
+			return nil, err
+		}
+
+		vo := &varOp{op: o}
+		if err := ctx.assign(name, vo); err != nil {
+			return nil, err
+		}
+		return vo, nil
+
+	// Add is special in some way, I think it's a function not a compiler op,
+	// not sure yet though
+	case Add:
+		els, err := vAsTup(2)
+		if err != nil {
+			return nil, err
+		} else if !els[0].outType().eq(valTypeInt) {
+			return nil, errors.New("add args must be numbers of the same type")
+		} else if !els[1].outType().eq(valTypeInt) {
+			return nil, errors.New("add args must be numbers of the same type")
+		}
+		return addOp{a: els[0], b: els[1]}, nil
+	default:
+		return nil, fmt.Errorf("op %v unknown, or its args are malformed", t)
+	}
+}

vm/function.go

@@ -1,246 +0,0 @@
-package vm
-
-import (
-	"errors"
-	"fmt"
-
-	"code.betamike.com/mediocregopher/ginger/gg"
-	"code.betamike.com/mediocregopher/ginger/graph"
-)
-
-// Function is an entity which accepts an argument Value and performs some
-// internal processing on that argument to return a resultant Value.
-type Function interface {
-	Perform(Value) Value
-}
-
-// FunctionFunc is a function which implements the Function interface.
-type FunctionFunc func(Value) Value
-
-// Perform calls the underlying FunctionFunc directly.
-func (f FunctionFunc) Perform(arg Value) Value {
-	return f(arg)
-}
-
-// Identity returns an Function which always returns the given Value,
-// regardless of the input argument.
-//
-// TODO this might not be the right name
-func Identity(val Value) Function {
-	return FunctionFunc(func(Value) Value {
-		return val
-	})
-}
-
-var (
-	valNameIn     = Value{Value: gg.Name("!in")}
-	valNameOut    = Value{Value: gg.Name("!out")}
-	valNameIf     = Value{Value: gg.Name("!if")}
-	valNameRecur  = Value{Value: gg.Name("!recur")}
-	valNumberZero = Value{Value: gg.Number(0)}
-)
-
-func checkGraphForFunction(g *gg.Graph) error {
-	for _, val := range g.AllValueIns() {
-		if val.Name == nil {
-			return fmt.Errorf("non-name %v cannot have incoming edges", val)
-		}
-
-		if !(Value{Value: val}).Equal(valNameOut) && (*val.Name)[0] == '!' {
-			return fmt.Errorf("name %v cannot start with a '!'", val)
-		}
-	}
-
-	// TODO check for acyclic-ness
-
-	return nil
-}
-
-// FunctionFromGraph wraps the given Graph such that it can be used as an
-// Function. The given Scope determines what values outside of the Graph are
-// available for use within the Function.
-func FunctionFromGraph(g *gg.Graph, scope Scope) (Function, error) {
-
-	if err := checkGraphForFunction(g); err != nil {
-		return nil, err
-	}
-
-	// edgeFn is distinct from a generic Function in that the Value passed into
-	// Perform will _always_ be the value of "in" for the overall Function.
-	//
-	// edgeFns will wrap each other, passing "in" downwards to the leaf edgeFns.
-	type edgeFn Function
-
-	var compileEdge func(*gg.OpenEdge) (edgeFn, error)
-
-	// TODO memoize?
-	valToEdgeFn := func(val Value) (edgeFn, error) {
-
-		if val.Name == nil {
-			return edgeFn(Identity(val)), nil
-		}
-
-		if val.Equal(valNameIn) {
-			return edgeFn(FunctionFunc(func(inArg Value) Value {
-				return inArg
-			})), nil
-		}
-
-		edgesIn := g.ValueIns(val.Value)
-
-		name := *val.Name
-
-		if l := len(edgesIn); l == 0 {
-			resolvedVal, err := scope.Resolve(name)
-			if errors.Is(err, ErrNameNotDefined) {
-				return edgeFn(Identity(val)), nil
-			} else if err != nil {
-				return nil, fmt.Errorf("resolving name %q from the outer scope: %w", name, err)
-			}
-
-			return edgeFn(Identity(resolvedVal)), nil
-
-		} else if l != 1 {
-			return nil, fmt.Errorf("resolved name %q to %d input edges, rather than one", name, l)
-		}
-
-		edge := edgesIn[0]
-
-		return compileEdge(edge)
-	}
-
-	// "out" resolves to more than a static value, treat the graph as a full
-	// operation.
-
-	// thisFn is used to support recur. It will get filled in with the Function
-	// which is returned by this function, once that Function is created.
-	thisFn := new(Function)
-
-	compileEdge = func(edge *gg.OpenEdge) (edgeFn, error) {
-
-		return graph.MapReduce[gg.OptionalValue, gg.Value, edgeFn](
-			edge,
-			func(ggVal gg.Value) (edgeFn, error) {
-				return valToEdgeFn(Value{Value: ggVal})
-			},
-			func(ggEdgeVal gg.OptionalValue, inEdgeFns []edgeFn) (edgeFn, error) {
-				if ggEdgeVal.Equal(valNameIf.Value) {
-
-					if len(inEdgeFns) != 3 {
-						return nil, fmt.Errorf("'!if' requires a 3-tuple argument")
-					}
-
-					return edgeFn(FunctionFunc(func(inArg Value) Value {
-
-						if pred := inEdgeFns[0].Perform(inArg); pred.Equal(valNumberZero) {
-							return inEdgeFns[2].Perform(inArg)
-						}
-
-						return inEdgeFns[1].Perform(inArg)
-
-					})), nil
-				}
-
-				// "!if" statements (above) are the only case where we want the
-				// input edges to remain separated, otherwise they should always
-				// be combined into a single edge whose value is a tuple. Do
-				// that here.
-
-				inEdgeFn := inEdgeFns[0]
-
-				if len(inEdgeFns) > 1 {
-					inEdgeFn = edgeFn(FunctionFunc(func(inArg Value) Value {
-						tupVals := make([]Value, len(inEdgeFns))
-
-						for i := range inEdgeFns {
-							tupVals[i] = inEdgeFns[i].Perform(inArg)
-						}
-
-						return Tuple(tupVals...)
-					}))
-				}
-
-				var edgeVal Value
-				if ggEdgeVal.Valid {
-					edgeVal.Value = ggEdgeVal.Value
-				}
-
-				if edgeVal.IsZero() {
-					return inEdgeFn, nil
-				}
-
-				if edgeVal.Equal(valNameRecur) {
-					return edgeFn(FunctionFunc(func(inArg Value) Value {
-						return (*thisFn).Perform(inEdgeFn.Perform(inArg))
-					})), nil
-				}
-
-				if edgeVal.Graph != nil {
-
-					opFromGraph, err := FunctionFromGraph(edgeVal.Graph, scope)
-					if err != nil {
-						return nil, fmt.Errorf("compiling graph to operation: %w", err)
-					}
-
-					edgeVal = Value{Function: opFromGraph}
-				}
-
-				// The Function is known at compile-time, so we can wrap it
-				// directly into an edgeVal using the existing inEdgeFn as the
-				// input.
-				if edgeVal.Function != nil {
-					return edgeFn(FunctionFunc(func(inArg Value) Value {
-						return edgeVal.Function.Perform(inEdgeFn.Perform(inArg))
-					})), nil
-				}
-
-				// the edgeVal is not a Function at compile time, and so
-				// it must become one at runtime. We must resolve edgeVal to an
-				// edgeFn as well (edgeEdgeFn), and then at runtime that is
-				// given the inArg and (hopefully) the resultant Function is
-				// called.
-
-				edgeEdgeFn, err := valToEdgeFn(edgeVal)
-
-				if err != nil {
-					return nil, err
-				}
-
-				return edgeFn(FunctionFunc(func(inArg Value) Value {
-
-					runtimeEdgeVal := edgeEdgeFn.Perform(inArg)
-
-					if runtimeEdgeVal.Graph != nil {
-
-						runtimeFn, err := FunctionFromGraph(
-							runtimeEdgeVal.Graph, scope,
-						)
-
-						if err != nil {
-							panic(fmt.Sprintf("compiling graph to operation: %v", err))
-						}
-
-						runtimeEdgeVal = Value{Function: runtimeFn}
-					}
-
-					if runtimeEdgeVal.Function == nil {
-						panic("edge value must be an operation")
-					}
-
-					return runtimeEdgeVal.Function.Perform(inEdgeFn.Perform(inArg))
-
-				})), nil
-			},
-		)
-	}
-
-	graphFn, err := valToEdgeFn(valNameOut)
-
-	if err != nil {
-		return nil, err
-	}
-
-	*thisFn = Function(graphFn)
-
-	return Function(graphFn), nil
-}

vm/scope.go

@@ -1,58 +0,0 @@
-package vm
-
-import (
-	"errors"
-)
-
-// ErrNameNotDefined is returned from Scope.Resolve when a name could not be
-// resolved within a Scope.
-var ErrNameNotDefined = errors.New("not defined")
-
-// Scope encapsulates a set of name->Value mappings.
-type Scope interface {
-
-	// Resolve accepts a name and returns an Value, or returns
-	// ErrNameNotDefined.
-	Resolve(string) (Value, error)
-}
-
-// ScopeMap implements the Scope interface.
-type ScopeMap map[string]Value
-
-var _ Scope = ScopeMap{}
-
-// Resolve uses the given name as a key into the ScopeMap map, and
-// returns the Operation held there for the key, if any.
-func (m ScopeMap) Resolve(name string) (Value, error) {
-
-	v, ok := m[name]
-
-	if !ok {
-		return Value{}, ErrNameNotDefined
-	}
-
-	return v, nil
-}
-
-type scopeWith struct {
-	Scope // parent
-	name  string
-	val   Value
-}
-
-// ScopeWith returns a copy of the given Scope, except that evaluating the given
-// name will always return the given Value.
-func ScopeWith(scope Scope, name string, val Value) Scope {
-	return &scopeWith{
-		Scope: scope,
-		name:  name,
-		val:   val,
-	}
-}
-
-func (s *scopeWith) Resolve(name string) (Value, error) {
-	if name == s.name {
-		return s.val, nil
-	}
-	return s.Scope.Resolve(name)
-}

vm/scope_global.go

@@ -1,61 +0,0 @@
-package vm
-
-import (
-	"fmt"
-
-	"code.betamike.com/mediocregopher/ginger/gg"
-)
-
-func globalFn(fn func(Value) (Value, error)) Value {
-	return Value{
-		Function: FunctionFunc(func(in Value) Value {
-			res, err := fn(in)
-			if err != nil {
-				panic(err)
-			}
-			return res
-		}),
-	}
-}
-
-// GlobalScope contains operations and values which are available from within
-// any operation in a ginger program.
-var GlobalScope = ScopeMap{
-
-	"!add": globalFn(func(val Value) (Value, error) {
-
-		var sum int64
-
-		for _, tupVal := range val.Tuple {
-
-			if tupVal.Number == nil {
-				return Value{}, fmt.Errorf("add requires a non-zero tuple of numbers as an argument")
-			}
-
-			sum += *tupVal.Number
-		}
-
-		return Value{Value: gg.Value{Number: &sum}}, nil
-
-	}),
-
-	"!tupEl": globalFn(func(val Value) (Value, error) {
-
-		tup, i := val.Tuple[0], val.Tuple[1]
-
-		return tup.Tuple[int(*i.Number)], nil
-
-	}),
-
-	"!isZero": globalFn(func(val Value) (Value, error) {
-
-		if *val.Number == 0 {
-			one := int64(1)
-			return Value{Value: gg.Value{Number: &one}}, nil
-		}
-
-		zero := int64(0)
-		return Value{Value: gg.Value{Number: &zero}}, nil
-
-	}),
-}

vm/vm.go

@@ -1,123 +1,129 @@
-// Package vm implements the execution of gg.Graphs as programs.
 package vm
 
 import (
 	"errors"
 	"fmt"
-	"io"
+	"sync"
-	"strings"
 
-	"code.betamike.com/mediocregopher/ginger/gg"
+	"github.com/mediocregopher/ginger/lang"
-	"code.betamike.com/mediocregopher/ginger/graph"
+
+	"llvm.org/llvm/bindings/go/llvm"
 )
 
-// ZeroValue is a Value with no fields set. It is equivalent to the 0-tuple.
+// Types supported by the vm in addition to those which are part of lang
-var ZeroValue Value
+var (
+	Atom  = lang.AAtom
+	Tuple = lang.ATuple
+	Int   = lang.Atom("int")
+)
 
-// Value extends a gg.Value to include Functions and Tuples as a possible
+// Ops supported by the vm
-// types.
+var (
-type Value struct {
+	Add    = lang.Atom("add")
-	gg.Value
+	Assign = lang.Atom("assign")
+	Var    = lang.Atom("var")
+)
 
-	Function
+////////////////////////////////////////////////////////////////////////////////
-	Tuple []Value
+
+// Module contains a compiled set of code which can be run, dumped in IR form,
+// or compiled. A Module should be Dispose()'d of once it's no longer being
+// used.
+type Module struct {
+	b      llvm.Builder
+	m      llvm.Module
+	ctx    varCtx
+	mainFn llvm.Value
 }
 
-// Tuple returns a tuple Value comprising the given Values. Calling Tuple with
+var initOnce sync.Once
-// no arguments returns ZeroValue.
+
-func Tuple(vals ...Value) Value {
+// Build creates a new Module by compiling the given Terms as code
-	return Value{Tuple: vals}
+// TODO only take in a single Term, implement List and use that with a do op
+func Build(tt ...lang.Term) (*Module, error) {
+	initOnce.Do(func() {
+		llvm.LinkInMCJIT()
+		llvm.InitializeNativeTarget()
+		llvm.InitializeNativeAsmPrinter()
+	})
+	mod := &Module{
+		b:   llvm.NewBuilder(),
+		m:   llvm.NewModule(""),
+		ctx: varCtx{},
+	}
+
+	var err error
+	if mod.mainFn, err = mod.buildFn(tt...); err != nil {
+		mod.Dispose()
+		return nil, err
+	}
+
+	if err := llvm.VerifyModule(mod.m, llvm.ReturnStatusAction); err != nil {
+		mod.Dispose()
+		return nil, fmt.Errorf("could not verify module: %s", err)
+	}
+
+	return mod, nil
 }
 
-// IsZero returns true if the Value is the zero value (aka the 0-tuple).
+// Dispose cleans up all resources held by the Module
-// LexerToken (within the gg.Value) is ignored for this check.
+func (mod *Module) Dispose() {
-func (v Value) IsZero() bool {
+	// TODO this panics for some reason...
-	return v.Equal(ZeroValue)
+	//mod.m.Dispose()
+	//mod.b.Dispose()
 }
 
-// Equal returns true if the passed in Value is equivalent, ignoring the
+// TODO make this return a val once we get function types
-// LexerToken on either Value.
+func (mod *Module) buildFn(tt ...lang.Term) (llvm.Value, error) {
-//
+	if len(tt) == 0 {
-// Will panic if the passed in v2 is not a Value from this package.
+		return llvm.Value{}, errors.New("function cannot be empty")
-func (v Value) Equal(v2g graph.Value) bool {
+	}
 
-	v2 := v2g.(Value)
+	ops := make([]op, len(tt))
-
+	var err error
-	switch {
+	for i := range tt {
-
+		if ops[i], err = termToOp(mod.ctx, tt[i]); err != nil {
-	case (v.Value != (gg.Value{}) || v2.Value != (gg.Value{})):
+			return llvm.Value{}, err
-		return v.Value.Equal(v2.Value)
-
-	case v.Function != nil || v2.Function != nil:
-		// for now we say that Functions can't be compared. This will probably
-		// get revisted later.
-		return false
-
-	case len(v.Tuple) == len(v2.Tuple):
-
-		for i := range v.Tuple {
-			if !v.Tuple[i].Equal(v2.Tuple[i]) {
-				return false
-			}
 		}
-
-		return true
-
-	default:
-
-		// if both were the zero value then both tuples would have the same
-		// length (0), which is covered by the previous check. So anything left
-		// over must be tuples with differing lengths.
-		return false
 	}
 
-}
+	var llvmIns []llvm.Type
+	if in := ops[0].inType(); in.llvm.TypeKind() == llvm.VoidTypeKind {
+		llvmIns = []llvm.Type{}
+	} else {
+		llvmIns = []llvm.Type{in.llvm}
+	}
+	llvmOut := ops[len(ops)-1].outType().llvm
 
-func (v Value) String() string {
+	fn := llvm.AddFunction(mod.m, "", llvm.FunctionType(llvmOut, llvmIns, false))
+	block := llvm.AddBasicBlock(fn, "")
+	mod.b.SetInsertPoint(block, block.FirstInstruction())
 
-	switch {
+	var out llvm.Value
-
+	for i := range ops {
-	case v.Function != nil:
+		if out, err = ops[i].build(mod); err != nil {
-
+			return llvm.Value{}, err
-		// We can try to get better strings for ops later
-		return "<fn>"
-
-	case v.Value != (gg.Value{}):
-		return v.Value.String()
-
-	default:
-
-		// we consider zero value to be the 0-tuple
-
-		strs := make([]string, len(v.Tuple))
-
-		for i := range v.Tuple {
-			strs[i] = v.Tuple[i].String()
 		}
-
-		return fmt.Sprintf("(%s)", strings.Join(strs, ", "))
-
 	}
-
+	mod.b.CreateRet(out)
+	return fn, nil
 }
 
-// EvaluateSource reads and parses the io.Reader as an operation, input is used
+// Dump dumps the Module's IR to stdout
-// as the argument to the operation, and the resultant value is returned.
+func (mod *Module) Dump() {
-//
+	mod.m.Dump()
-// scope contains pre-defined operations and values which are available during
+}
-// the evaluation.
+
-func EvaluateSource(opSrc io.Reader, input Value, scope Scope) (Value, error) {
+// Run executes the Module
-	v, err := gg.NewDecoder(opSrc).Next()
+// TODO input and output?
-	if err != nil {
+func (mod *Module) Run() (interface{}, error) {
-		return Value{}, err
+	engine, err := llvm.NewExecutionEngine(mod.m)
-	} else if v.Value.Graph == nil {
+	if err != nil {
-		return Value{}, errors.New("value must be a graph")
+		return nil, err
-	}
+	}
-
+	defer engine.Dispose()
-	fn, err := FunctionFromGraph(v.Value.Graph, scope)
+
-	if err != nil {
+	funcResult := engine.RunFunction(mod.mainFn, []llvm.GenericValue{})
-		return Value{}, err
+	defer funcResult.Dispose()
-	}
+	return funcResult.Int(false), nil
-
-	return fn.Perform(input), nil
 }

vm/vm_test.go

@@ -1,58 +1,84 @@
 package vm
 
 import (
-	"bytes"
+	. "testing"
-	"strconv"
-	"testing"
 
-	"code.betamike.com/mediocregopher/ginger/gg"
+	"github.com/mediocregopher/ginger/lang"
-	"github.com/stretchr/testify/assert"
 )
 
-func TestVM(t *testing.T) {
+func TestCompiler(t *T) {
-	tests := []struct {
+	mkcmd := func(a lang.Atom, args ...lang.Term) lang.Tuple {
-		src    string
+		// TODO a 1-tuple should be the same as its element?
-		in     Value
+		if len(args) == 1 {
-		exp    Value
+			return lang.Tuple{a, args[0]}
-		expErr string
+		}
-	}{
+		return lang.Tuple{a, lang.Tuple(args)}
+	}
+	mkint := func(i string) lang.Tuple {
+		return lang.Tuple{Int, lang.Const(i)}
+	}
+
+	type test struct {
+		in  []lang.Term
+		exp uint64
+	}
+
+	one := mkint("1")
+	two := mkint("2")
+	foo := mkcmd(Var, lang.Atom("foo"))
+	bar := mkcmd(Var, lang.Atom("bar"))
+	baz := mkcmd(Var, lang.Atom("baz"))
+
+	tests := []test{
 		{
-			src: `{
+			in:  []lang.Term{one},
-				incr = { !out = !add < (1, !in); };
+			exp: 1,
-				!out = incr < incr < !in;
-			}`,
-			in:  Value{Value: gg.Number(5)},
-			exp: Value{Value: gg.Number(7)},
 		},
 		{
-			src: `{
+			in: []lang.Term{
-				!foo = in;
+				mkcmd(Add, mkcmd(Tuple, one, two)),
-				!out = !foo;
+			},
-			}`,
+			exp: 3,
-			in:     Value{Value: gg.Number(1)},
-			expErr: "name !foo cannot start with a '!'",
 		},
 		{
-			src: `{foo = bar; !out = foo;}`,
+			in: []lang.Term{
-			in:  Value{},
+				mkcmd(Assign, foo, one),
-			exp: Value{Value: gg.Name("bar")},
+				mkcmd(Add, mkcmd(Tuple, foo, two)),
+			},
+			exp: 3,
+		},
+		{
+			in: []lang.Term{
+				mkcmd(Assign, foo, mkcmd(Tuple, one, two)),
+				mkcmd(Add, foo),
+			},
+			exp: 3,
+		},
+		{
+			in: []lang.Term{
+				mkcmd(Assign, foo, mkcmd(Tuple, one, two)),
+				mkcmd(Assign, bar, mkcmd(Add, foo)),
+				mkcmd(Assign, baz, mkcmd(Add, foo)),
+				mkcmd(Add, mkcmd(Tuple, bar, baz)),
+			},
+			exp: 6,
 		},
 	}
 
-	for i, test := range tests {
+	for _, test := range tests {
-		t.Run(strconv.Itoa(i), func(t *testing.T) {
+		t.Logf("testing program: %v", test.in)
-			t.Log(test.src)
+		mod, err := Build(test.in...)
-			val, err := EvaluateSource(
+		if err != nil {
-				bytes.NewBufferString(test.src), test.in, GlobalScope,
+			t.Fatalf("building failed: %s", err)
-			)
+		}
 
-			if test.expErr != "" {
+		out, err := mod.Run()
-				assert.Error(t, err)
+		if err != nil {
-				assert.Equal(t, test.expErr, err.Error())
+			mod.Dump()
-			} else {
+			t.Fatalf("running failed: %s", err)
-				assert.NoError(t, err)
+		} else if out != test.exp {
-				assert.True(t, val.Equal(test.exp), "%v != %v", test.exp, val)
+			mod.Dump()
-			}
+			t.Fatalf("expected result %T:%v, got %T:%v", test.exp, test.exp, out, out)
-		})
+		}
 	}
 }