4bd9125b86
Fix delay of arguments to be exactly the same as codegen tests
2024-06-25 18:50:00 -04:00
f695276bf7
Fix codegen tree traversal to be updated for the otherwise field and future proof the traversal function
2024-06-25 18:50:00 -04:00
cc9df04093
Fix missing delay in list_access_to_uplc. Also fix one of the unit tests.
2024-06-25 18:50:00 -04:00
41b941e0e3
Fix castfromData in record access cases
2024-06-25 18:50:00 -04:00
e09f6bbc87
delay otherwise branch to prevent premature errors
2024-06-25 18:50:00 -04:00
df939e20ce
missed a Air Op Code and updated how we pass in otherwise for assignment
2024-06-25 18:50:00 -04:00
5024bd3f9c
feat: code gen support for if/is
...
Co-authored-by: Kasey White <kwhitemsg@gmail.com>
2024-06-25 18:50:00 -04:00
c48f15a957
revert #903 'feat: Emit keyword'
...
While we agree on the idea of having some ways of emitting events, the
design hasn't been completely fleshed out and it is unclear whether
events should have a well-defined format independent of the framework
/ compiler and what this format should be.
So we need more time discussing and agreeing about what use case we
are actually trying to solve with that.
Irrespective of that, some cleanup was also needed on the UPLC side
anyway since the PR introduced a lot of needless duplications.
2024-05-23 17:22:12 +02:00
fc0e88018e
Chore:
...
Refactor get_uplc_type to account for constr types that don't exactly resolve to a uplc type
Check arg_stack in uplc generator has only 1 argument at the end of the generation
warning fixes
2024-05-04 14:04:12 -04:00
ebe415cfc9
fix errors
2024-05-04 14:04:12 -04:00
91a7e77ab4
Add 'Pair' pattern and rework internals to use it.
...
Currently, pattern-matching on 'Pair' is handled by treating Pair as a
record, which comes as slightly odd given that it isn't actually a
record and isn't user-defined. Thus now, every use of a record must
distinguish between Pairs and other kind of records -- which screams
for another variant constructor instead.
We cannot use `Tuple` either for this, because then we have no ways to
tell 2-tuples apart from pairs, which is the whole point here. So the
most sensical thing to do is to define a new pattern `Pair` which is
akin to tuples, but simpler since we know the number of elements and
it's always 2.
2024-05-04 14:04:12 -04:00
75b076552c
feat: Do a major overhaul on how we check types to allow for match patterns instead of if statements
...
Also fix one more test
2024-05-04 14:04:12 -04:00
3c332ca42a
Few more places in codegen where we need to be able to deal with Pair records
2024-05-04 14:04:12 -04:00
26f68c2fb4
fix: found various unify and type issues while running tests
2024-05-04 14:04:12 -04:00
21b60896f0
remove wild card match from tree functions
...
Start working on supporting Pair clauses
2024-05-04 14:04:12 -04:00
f950ae7d3d
WIP: add new opcodes to Air and AirTree and update parts of codegen to handle the new pair type
2024-05-04 14:04:12 -04:00
d05d8e7de6
Start working on separating pairs from 2 tuples in Aiken
...
co-authored-by: KtorZ <matthias.benkort@gmail.com>
2024-05-04 14:04:12 -04:00
ff4ddfbe1b
Simplifying PR per reviewers request
2024-04-12 21:40:27 -04:00
d25b8f91c7
feat: Emit keyword
2024-04-12 21:40:27 -04:00
a5a0734629
fix: casting a field type to Data with expect and traces on was assuming the raw Data was of type constr
2024-03-27 15:52:23 -04:00
61936cb91e
fix(codegen): Add tracing when checking for a constr vs another primitive
2024-03-17 16:25:17 -04:00
435dd0d213
Refactor AssignmentKind to allow backpassing on both let and expect.
...
The 3rd kind of assignment kind (Bind) is gone and now reflected through a boolean parameter. Note that this parameter is completely erased by the type-checker so that the rest of the pipeline (i.e. code-generation) doesn't have to make any assumption. They simply can't see a backpassing let or expect.
2024-03-11 00:16:23 +01:00
ed9f5c6ef7
Preserve TypeAlias in types for better context/feedback.
2024-03-08 15:59:33 +01:00
bf429fbdbf
remove unused import
2024-03-08 00:12:44 -05:00
dcec8ecfe6
remove unused special function
2024-03-08 00:12:44 -05:00
23a22a65cb
Handle (recursive) generic types during reification.
...
Also moved a bunch of functions from code-gen back into _tipo_, as
they're better suited and generic enough to be reused elsewhere.
2024-03-07 18:07:52 +01:00
a3fbe6c155
fix tests
2024-03-06 23:27:10 -05:00
bdd84dc952
fixing the tests lead to me create a new function for converting from data
2024-03-06 23:27:10 -05:00
892da06e14
add more runtime checking for a few of the data to primitive conversions
2024-03-06 23:27:10 -05:00
c9dd281b45
disable assert that was blocking issue #844
2024-03-04 11:48:22 -05:00
c2dc47fa0b
Refactor creation of CodeGenerator and management of known data_types and functions.
...
This was a mess to say to the least. The mess started when we wanted
to make all definitions in codegen use immutable maps of references --
which was and still is a good idea. Yet, the population of the data
types and functions definitions was done somehow in a separate step,
in a rather ad-hoc manner.
This commit changes that to ensure the project's data_types and
functions are populated while type checking the AST such that we need
not to redo it after.
The code for registering the data type definitions and function
definitions was also duplicated in at least 3 places. It is now a
method of the TypedModule.
Note: this change isn't only just cosmetic, it's also necessary for
the commit that follows which aims at adding tests to the set of
available function definitions, thus allowing to make property tests
callable.
2024-03-03 19:33:26 +01:00
26e563a9be
Hardened property-based testing framework. More tests, less bugs.
...
Those end-to-end tests are useful. Both for controlling the behavior of the shrinker, but also to double check the reification of Plutus Data back into untyped expressions.
I had to work-around a few things to get opaque type and private types play nice. Also found a weird bug due to how we apply parameters after unique debruijn indexes have been also applied. A work-around is to re-intern the program.
2024-03-03 19:33:26 +01:00
70ea3c9598
Write boilerplate code for being able to easily test properties.
...
Loads of plumbing, but we now have at least some nice ways to test property execution and shrinking.
2024-03-03 19:33:26 +01:00
14f1025f0b
Display counterexamples as Aiken values instead of raw UPLC.
2024-03-03 19:33:24 +01:00
c766f44601
Allow Fuzzer with type parameter
...
Also fix shrinker first reduction, as well as passing of List/Tuples to fuzzer.
2024-03-03 19:33:24 +01:00
d698f76e3c
fix(codegen): builtin calls for g1 and g2 where flipped
...
closes #840
2024-02-29 12:13:51 -05:00
6c6be3f53d
got past the errors and warnings
2024-02-07 12:37:37 -05:00
51f1da2505
Removed AirStatements and AirExpressions
...
Still a WIP
2024-02-07 12:37:37 -05:00
8702c736d0
fix warning
2024-02-07 12:36:35 -05:00
806a74c192
fix all current errors
2024-02-07 12:36:35 -05:00
a12c374258
Start turning AirTree statements into expressions (let)
...
Still many places to fix, WIP.
2024-02-07 12:36:35 -05:00
3b4c6cb2aa
fix: missing message error term in one place for type is void
2024-02-07 11:48:52 -05:00
551941392e
fix: assert had one minor edge case due to final clauses with lists
2024-01-31 00:05:09 -05:00
444bccf19c
fix: change list_access_to_uplc to properly handle list discards
2024-01-30 23:53:33 -05:00
78d2049d7b
fix: Using the wrong match string for discards in FieldsExpose
...
Also need to return a lambda wrapped term from list_access_to_uplc under all conditions
2024-01-25 14:18:36 -05:00
36a0b317ad
fix condition to account for tail presence
2024-01-24 16:29:40 -05:00
3a44c45b48
fix: one builtin error wasn't be caught with a messaged exception
2024-01-24 16:29:40 -05:00
eda4e259d6
minor fix and some refactoring on the if statements
2024-01-24 16:29:40 -05:00
956c3d6cf0
feat: refactor code gen to avoid builtin errors when tracing is turned on
2024-01-24 16:29:40 -05:00
6fa272bd34
Remove compiler-generated helper 'global' traces in compact mode.
...
Since there's no line number to show here, we don't have much choice. And the alternative of showing shorter traces as code is just ugly as hell.
2024-01-19 14:30:44 +01:00
59c784778e
Convert span's start to line number + col
...
This requires to make line numbers a first-class citizen in the module
hierarchy but it is fortunately not _too involved_.
2024-01-19 14:30:15 +01:00
e67d5863a1
Introduce 'compact' trace level verbosity
...
For now, it only shows the span start. We'll change that in the next commit to show a line number and a column.
2024-01-19 14:30:15 +01:00
f79b37d551
Replace 'bool' with 'TraceLevel' in codegen
...
Co-authored-by: KtorZ <matthias.benkort@gmail.com>
2024-01-19 14:30:15 +01:00
d26524048e
fix: headlist builtin on assoc lists
...
implement chooseunit for 0 args
2024-01-13 19:29:34 -05:00
c7af27a6ba
fix: generic edge case with tuples that allowed 2 tuples and 3 tuples to use the same monomorphized function.
...
Also massively reduced the space taken up by generics in scripts when using generics with list and tuples
2024-01-13 17:46:32 -05:00
4a8fecb70a
fix: satisfy clippy's demands
2024-01-11 14:53:02 -05:00
ff462fa8ea
don't need clippy macro anymore :)
2024-01-08 18:08:50 -05:00
f722af1149
fix: accidentally put quotes around tail_name
2024-01-08 18:08:50 -05:00
2216f387c3
refactor: change codegen uplc to have more type safety
...
Also refactor list_access_to_uplc
2024-01-08 18:08:50 -05:00
30a6b77116
Get rid of 'VoidMsg' in favor of an 'Option'.
2024-01-04 16:03:51 -05:00
c50a9cb5bd
refactor: convert msgs to use AirMsg type instead of AirTree
2024-01-04 16:03:51 -05:00
394cac86b8
feat: expect on a type now can take in a msg when in trace mode
2024-01-04 16:03:51 -05:00
c7a1ff0959
refactor how tracing is popped off to be in one location in uplc_gen
2024-01-04 16:03:51 -05:00
355e38d6e2
feat: expects now print the line of code that failed
2024-01-04 16:03:51 -05:00
aa51ce3e3e
feat: add code messages when using expects on constrs
2024-01-04 16:03:51 -05:00
71cfb6f6af
feat: Add specific messages for using expect with booleans
...
TODO: fill out the rest of the expects with messages
2024-01-04 16:03:51 -05:00
45177cd08b
fix: add missing type checks for the new bls primitives
2023-11-23 13:00:24 -05:00
8b89ba3b93
feat: implement bls primitives in code gen
2023-11-15 15:55:56 -05:00
d51374aac1
feat: add conversion to data and from data for new primitive types
2023-11-15 15:55:56 -05:00
0d2ac952d0
feat: implemented the Bls381-12 builtins and types
...
also implemented Keccak256 and Blake2b_224
TODO: cost model, flat serialization, pretty
Co-authored-by: rvcas <x@rvcas.dev>
2023-11-15 15:55:56 -05:00
4dd17dacf3
chore: rename uplc builder builtins
2023-10-22 17:08:48 -04:00
4649a5a9d2
fix: don't reset the used special functions
2023-10-07 19:42:24 -04:00
8964675670
feat: Add tracing for common multivalidator issues
2023-10-07 19:42:24 -04:00
7a6ddc45a0
feat: Rework codegen traces to prevent repeated messages from taking up uplc script space
2023-10-07 19:42:24 -04:00
8a1c824dd7
chore: add specific error for when the validator returns false
2023-10-04 11:08:22 -04:00
47596f0324
feat: Remove tuple_index and record_access in favor of faster more direct functions for
...
accessing an item in a tuple or a field in a record
2023-09-28 01:05:05 -04:00
1ca3499128
chore: rename type
2023-09-25 21:16:19 -04:00
2f80d07132
fix: minor stuff including fixing the var name used in modify_cyclic_calls
...
and carefully controling the functions we add to sorted dependencies
2023-09-25 21:16:19 -04:00
f4310bcf33
feat: finished up mutual recursion
...
Now we "handle" vars that call the cyclic function.
That includes vars in the cyclic function as well as in other functions
"handle" meaning we modify the var to be a call that takes in more arguments.
2023-09-25 21:16:19 -04:00
ae3053522e
feat: Update cyclic functions to be aware of being in a cycle.
...
Finish the creation of cyclic functions
The last part is to update vars that call into a function in the cycle
2023-09-25 21:16:19 -04:00
794fc93084
remove unused structs
2023-09-25 21:16:19 -04:00
0b38855ce4
add new enum for hoistablefunctions
2023-09-25 21:16:19 -04:00
ced818c455
checkpoint commit
2023-09-25 21:16:19 -04:00
0fb9837ddf
chore: change UserFunction to HoistableFunction to prepare for mututal recursion
2023-09-25 21:16:19 -04:00
a4aa51ed2d
WIP: first part of mutual recursion is done.
...
This involves creating the function definition and detecting cycles.
The remaining part is to "fix" the call sites
of the mutually recursive functions
2023-09-25 21:16:19 -04:00
d808197507
chore: clippy fix
2023-09-13 18:17:59 -04:00
a45001376d
fix: is_record was used incorrectly in code gen,
...
the real solution was to look up the datatype and check constructors length
2023-09-13 00:33:02 -04:00
51c44c6a30
fix: add an assert for better error messages when doing empty types
2023-08-30 13:50:37 -04:00
084b900b2a
change: traverse_with_tree now has a boolean to determine when with is called
...
fix: Opaque types are now properly handled in code gen (i.e. code gen functions, in datums/redeemers, in from data casts)
chore: add specific nested opaque type tests to code gen
2023-08-19 20:07:37 -04:00
2456801b17
fix list clauses with guards and add more tests
2023-08-16 21:59:25 -04:00
eda388fb29
test(aiken-lang): add a new test for list edge case in when clause patterns
2023-08-08 20:47:35 -04:00
252f68de17
fix clippy
2023-08-07 19:08:18 -04:00
0d99afe5e2
Cargo fmt
2023-08-07 19:00:39 -04:00
65984ed15b
fix: move where we call the with in traverse_tree_with
2023-08-07 19:00:39 -04:00
dba0e11ba7
Add other shadowing cases
2023-08-07 19:00:39 -04:00
fc948f0029
Add the same optimization to dependent functions
...
I originally didn't add this because I thought this was mutually
recursive functions, which I couldn't picture how that would work;
I refactored all this logic into modify_self_calls, which maybe needs a
better name now.
Perf gain on some stdlib tests (line concat tests) is 93%!!
2023-08-07 19:00:39 -04:00
c45caaefc8
Rudimentary implementation
...
Adds an identify_recursive_static_params; doesn't handle all shadowing cases yet
2023-08-07 19:00:39 -04:00
09f889b121
Add codegen for recursive statics case
...
We also flip the recursive_statics fields to recursive_nonstatics; This makes the codegen a little easier. It also has a hacky way to hard code in some recursive statics for testing
2023-08-07 19:00:39 -04:00
586a2d7972
Add recursive_static_params to AIR
...
Any methods to a recursive function that are unchanged and forwarded
don't need to be applied each time we recurse; instead, you can
define a containing lambda, reducing the number of applications
dramatically when recursing
2023-08-07 19:00:39 -04:00
bfa4cc2efc
fix: some function dependency tree path was not being updated in order
...
fix: revert to old implicit way of casting to data for now
2023-08-07 12:02:44 -04:00
microproofs