Polymorphic defs are now actually emitted: each unique instantiation
gets its own specialised LLVM fn, mangled @ail_<m>_<def>__<descriptor>.
The typechecker's Forall instantiations from Iter 12a now have a real
backend.
Strategy:
- Pass 1 of lower_workspace splits fn-typed defs into mono and poly:
module_user_fns keeps LLVM-typed FnSig only for monomorphic fns;
module_polymorphic_fns holds the full FnDef for poly defs;
module_def_ail_types is a unified AILang-type lookup for both.
- Direct calls to poly defs (current-module or qualified cross-module)
flow through lower_polymorphic_call: it derives the type
substitution from the actual arg AILang types, mangles the symbol,
and queues (def, subst) for specialisation.
- emit_module drains the queue after the regular defs. Each entry
produces a specialised FnDef with rigid vars substituted in both
the type and the body (incl. Lam param/ret annotations), then
emits via the existing emit_fn pipeline. The synthetic name embeds
the descriptor; standard mangling produces the right symbol.
Codegen-side type tracking:
- locals tuple grew from (name, ssa, llvm_type) to 4-tuple with
ail_type. Updated all 6 push sites: fn entry, Let, match-ctor field,
match-open-arm, lambda capture (cap_meta also extended), lambda
param.
- CtorRef gained ail_fields parallel to fields, used by match-arm
bindings to inherit AILang types.
- New synth_arg_type / synth_with_extras: a small recursive walker
that derives the AILang type of an expression in the current scope.
Used at poly call sites for arg-type inference. The `extras`
parameter shadows locals during Let recursion without &mut self.
Helpers added:
- derive_substitution(vars, params, arg_tys): unifies the param
shapes against the actual args, binding rigid vars; reports
unbound vars as an internal error.
- apply_subst_to_type / apply_subst_to_term: substitute rigid vars
throughout a Type or Term (the Term variant only matters for Lam,
the only Term arm carrying types).
- descriptor_for_subst / type_descriptor: stable identifier-safe
name suffix. `Int → I`, `Bool → B`, `Unit → U`, `Str → S`, ADT
`Foo → FFoo`, `Fn → Fn_<...>__r_<ret>`. So `id__I` for id at Int,
`apply__I_I` for apply at (Int, Int).
- builtin_ail_type / builtin_effect_op_ret: AILang types for the
builtin operators and effect ops, used by the codegen-side type
tracker.
Examples + tests:
- examples/poly_id.ail.json: id used at Int (42) and Bool (true).
Output: 42 / true. Two specialised fns + adapters + static
closures get emitted (visible in the IR).
- examples/poly_apply.ail.json: apply(succ, 41) == 42. Exercises
the harder case where one of the polymorphic params is itself
a function value — the closure-pair ABI survives substitution.
- crates/ail/tests/e2e.rs: polymorphic_id_at_int_and_bool and
polymorphic_apply_with_fn_param.
Tests: 58/58 (was 56/56). Hash invariant holds: sum.ail.json keeps
db33f57cb329935e / d9a916a0ed10a3d3.
Limitations (known, deferred):
- Polymorphic fns can only be DIRECTLY called. Passing a poly fn as
a value (let f = id in f(42)) fails — resolve_top_level_fn looks
in module_user_fns which doesn't include poly defs. Adding this
needs per-instantiation closure-pair globals.
- No higher-rank polymorphism: a polymorphic arg passed to another
polymorphic call (apply(id, 42)) trips the simple unify_for_subst
which doesn't recurse into Forall. Acceptable for the MVP.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
examples/sort.ail.json: insertion sort over IntList. Defines `insert
:: Int -> IntList -> IntList` and `sort :: IntList -> IntList`
recursively, plus print_list using Iter 10's seq. Sorts an 11-element
input and prints `1 1 2 3 3 4 5 5 5 6 9` one-per-line.
Validates that the language handles deeper ADT recursion + branching
(if + <=) + ctor construction + IO sequencing without surprises.
Wrote, typechecked, ran first try.
The Iter 11 plan called for polymorphism, but on reflection the right
move was one more validation cycle before disturbing the pipeline.
Polymorphism is now queued explicitly as Iter 12 (12a typechecker
substitution, 12b codegen monomorphisation, 12c docs + generic
example).
Tests: 52 green (was 51). New e2e `insertion_sort_orders_list`.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds `Term::Seq { lhs, rhs }` (serde tag "seq") as a first-class
AST node for sequencing effectful expressions. Equivalent in
behaviour to `let _ = lhs in rhs`, but the dedicated node gives the
pretty-printer and diagnostics a cleaner shape and surfaces the
intent ("run for effect, then yield rhs") to future tooling.
Typecheck: lhs must be Unit; rhs's type is the result; effects
accumulate.
Codegen: lower lhs (drop SSA), lower rhs (return).
Capture / deps walkers: recurse into both sides.
Refactored examples/list_map.ail.json's print_list to use seq
instead of `let _ = ...`. Output unchanged (2/4/6).
Hash stability: existing examples without Term::Seq serialise
bit-identical; only list_map.ail.json's hashes changed (deliberate
refactor).
Tests: 51 green (was 50). New unit test
`ailang_check::tests::seq_lhs_must_be_unit` covers the type-error
path; existing list_map e2e covers the happy path.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Iter 9a — dogfood. examples/list_map.ail.json exercises everything
Iter 1-8 shipped in one program: ADTs (IntList = Nil | Cons Int IntList),
recursive fns over the ADT (map_int, print_list), pattern matching with
nested Var fields, a closure (`\\x. x * 2`, no captures), fn-typed
parameters, IO effects propagating through recursion, and `let`-
sequencing of an effectful sub-expression inside a match arm.
Result: nothing broke. The full pipeline (typecheck → IR emit →
clang → run) produces "2\\n4\\n6\\n". The language is now sufficient
for "small but real" programs.
Iter 9b — `ail run`. Convenience subcommand that builds into a
tempdir and executes, propagating the binary's exit code. Equivalent
to `ail build && ./bin` in one step. Build logic factored out of
`Cmd::Build` into a shared `build_to` helper used by both Build and
Run.
Tests: 50 green (was 49). New e2e `list_map_doubles_then_prints`.
No test for `ail run` itself — the existing build_and_run helper in
e2e.rs already exercises the same build+exec sequence path.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds anonymous functions to AILang. A lambda value is constructed by
malloc'ing an env struct that holds its captured locals, plus a
closure pair `{ thunk_ptr, env_ptr }` (Iter 8a's ABI). The lambda's
body lifts to a top-level thunk `@ail_<m>_<def>_lam<id>(ptr %env,
params...)` that unpacks captures back into named locals before
running.
AST: new Term::Lam { params, paramTypes, retType, effects, body }.
Serde tag is "lam"; existing modules (no Lam) serialize identically,
so all current hashes stay stable.
Pretty-printer: renders `(\\ (x: T ...) -> R . body)`.
Typecheck (ailang-check): a lambda's type is the declared Type::Fn.
Body's effect set must be a subset of declared effects (no row
polymorphism). Constructing a lambda is pure — only calling it picks
up the declared effects, via the existing App branch.
Codegen (ailang-codegen):
- collect_captures: free-var walk; excludes builtins, current-module
top-level fns, and qualified names.
- lower_lambda: per-fn lam counter, save/restore emitter state, emit
thunk into a deferred queue (flushed after the parent fn), pack env
+ closure pair on heap, register the resulting closure-pair SSA in
the sidetable so subsequent indirect calls find it.
- Capture sigs propagate: a fn-typed capture keeps its FnSig in the
thunk's sidetable so `f(args)` inside the body still indirect-calls.
- Env layout: 8 bytes per capture (typed load/store reads only the
needed bytes; padding wasted but uniform).
Tests: 49 green (was 48). New examples/closure.ail.json + e2e
`closure_captures_let_n` exercises `let n = 3 in apply(\\x. x + n, 39)`
end-to-end and asserts the binary prints 42.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Top-level fn names are now usable as values, fn-typed parameters can
be called as `f(args)`. KISS slice of "closures + HOFs + TIR": no
TIR, no heap, no ABI shift — that bundle stays in Iter 8 where
closure-with-capture and lambdas land together.
Codegen:
- Type::Fn lowers to LLVM `ptr`; sig travels via a per-fn-body
`ssa_fn_sigs` sidetable on `Emitter`.
- Term::Var falls through to `resolve_top_level_fn` when the name is
not a local; emits `@ail_<m>_<def>` and registers the sig.
- Term::App splits: static callee (builtin / current-module / qualified)
keeps the existing direct path; otherwise lower the callee, look up
the sig, emit indirect `call <ret> (<param-tys>) %fn(args...)`.
- emit_fn registers fn-typed params in the sidetable on entry.
- If branches propagate a fn-pointer sig to their phi when both arms
match.
Typechecker: unchanged. It already accepted `synth(callee)` against
Type::Fn; the only blocker was codegen rejecting non-Var callees.
Tests: 48 green (was 47). New `examples/hof.ail.json` and e2e
`higher_order_apply_inc` exercise `apply(inc, 41) == 42` end-to-end.
DESIGN.md: "What is not (yet) supported" rewritten — closures-with-
capture and lambdas remain pending, first-class fn-refs added as
positive bullet. Smoke-test list extended with `hof.ail.json`.
JOURNAL.md: Iter 7 entry with rationale, scope choice, architecture
self-check, Iter 8 plan.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Finishes the project-wide English convention:
- CLAUDE.md (the user opted in to translate it too).
- examples/hello.ail.json: "Hallo, AILang." -> "Hello, AILang.".
- E2E test assertion and IR snapshot for hello regenerated to match.
- DESIGN.md "Project language: English" section: the previous
CLAUDE.md exception is dropped.
After this commit, no German remains in any committed file
(grep -P '[äöüÄÖÜß]' is empty across .rs / .md / .json / .ll).
Verified: cargo test --workspace passes (44/44).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Neue API check_workspace(&Workspace) -> Vec<Diagnostic>; check_module
hebt das Modul intern in einen Trivial-Workspace. Term::Var mit genau
einem Punkt im Namen ist ein qualifizierter Verweis <prefix>.<def>,
aufgelöst über Import-Map (alias-oder-modulname → echter Modulname).
Drei neue Diagnostic-Codes: unknown-module, unknown-import,
invalid-def-name. ail check lädt jetzt immer via load_workspace;
Loader-Fehler werden im JSON-Modus zu strukturierten Diagnostics
(module-not-found, module-cycle, …). DESIGN.md und JOURNAL.md
dokumentieren die Konvention. Hash-Stabilität: alle ir_snapshot_*-
Tests bitidentisch grün, kein neuer AST-Knoten.
Neue Top-Level-API check_module(&Module) -> Vec<Diagnostic> in
ailang-check, plus stabile Codes (unbound-var, type-mismatch,
arity-mismatch, non-exhaustive-match, unknown-ctor-in-pattern,
duplicate-def, …). CLI bekommt --json-Flag für maschinenlesbares
Output, Exit 1 bei Errors. Text-Modus unverändert. E2E-Test
check_json_unbound_var sichert das Format ab.
- AST: Def::Type mit Ctors; Term::Ctor (Konstruktion) und Term::Match
mit Arm/Pattern. Patterns: Wild, Var, Lit, Ctor { ctor, fields } —
Sub-Patterns im MVP auf Var/Wild beschränkt.
- Typchecker: Type-Registry, ctor_index für O(1)-Resolution, Pattern-
Bindings, Exhaustiveness-Check gegen volle Konstruktormenge plus
Negativ-Tests.
- Codegen: Boxed-Heap-Layout via malloc; Tag in Offset 0, Felder ab
Offset 8 in 8-Byte-Slots. Match lowert zu load tag + switch + Phi
am Join. Default-Block ist unreachable, wenn vom Typchecker geprüft.
- examples/list.ail.json: rekursive Int-Liste mit sum_list via match.
E2E-Test + Exhaustiveness-Tests. 19/19 Tests grün.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- Codegen: current_block-Tracking ersetzt die Heuristik im phi-Lowering;
verschachtelte if-Ausdrücke produzieren jetzt korrekte LLVM IR.
examples/max3.ail.json + Test schützt gegen Regression.
- Strings: Lit::Str / Type Str / io/print_str Effekt-Op; Strings sind im
MVP immutable Konstanten. examples/hello.ail.json als zweiter E2E-Test.
- CLI: --json für manifest und builtins; neuer deps-Subcommand listet
statische Symbol-Referenzen pro Definition. Effekt-Ops mit Prefix
effect: markiert.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Erste lauffähige Iteration. examples/sum.ail.json wird zu nativem Binary
kompiliert und druckt 55 (Summe 1..10) als End-to-End-Test.
Architektur:
- ailang-core: hashbares JSON-AST + canonical-form + pretty-printer
- ailang-check: monomorpher HM-Subset + Effekt-Set-Tracking
- ailang-codegen: LLVM-IR-Text-Emitter (kein libllvm-link)
- ail: CLI mit check/manifest/render/describe/emit-ir/build/builtins
Designentscheidungen sind in docs/DESIGN.md dokumentiert; der Verlauf
in docs/JOURNAL.md.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Brummel