panproto-parse 0.52.1

Tree-sitter full-AST parsers and emitters for panproto language protocols
Documentation
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
//! Parser registry mapping protocol names to full-AST parser implementations.

use std::path::Path;
use std::sync::Arc;

use panproto_schema::{AbstractSchema, DecoratedSchema, Schema};
use rustc_hash::FxHashMap;

use crate::error::ParseError;
use crate::layout_policy::LayoutPolicy;
use crate::theory_extract::ExtractedTheoryMeta;

/// A full-AST parser and emitter for a specific programming language.
///
/// Each implementation wraps a tree-sitter grammar and its auto-derived theory,
/// providing parse (source → Schema) and emit (Schema → source) operations.
pub trait AstParser: Send + Sync {
    /// The panproto protocol name (e.g. `"typescript"`, `"python"`).
    fn protocol_name(&self) -> &str;

    /// Parse source code into a full-AST [`Schema`].
    ///
    /// # Errors
    ///
    /// Returns [`ParseError`] if tree-sitter parsing fails or schema construction fails.
    fn parse(&self, source: &[u8], file_path: &str) -> Result<Schema, ParseError>;

    /// Emit a [`Schema`] back to source code bytes.
    ///
    /// The emitter walks the schema graph top-down, using formatting constraints
    /// (comment, indent, blank-lines-before) to reproduce the original formatting.
    ///
    /// # Errors
    ///
    /// Returns [`ParseError::EmitFailed`] if emission fails.
    fn emit(&self, schema: &Schema) -> Result<Vec<u8>, ParseError>;

    /// File extensions this parser handles (e.g. `["ts", "tsx"]`).
    fn supported_extensions(&self) -> &[&str];

    /// The auto-derived theory metadata for this language.
    fn theory_meta(&self) -> &ExtractedTheoryMeta;

    /// Render a by-construction [`Schema`] (one with no parse-recovered
    /// byte positions or interstitials) to source bytes.
    ///
    /// Unlike [`emit`](Self::emit), which reconstructs source from
    /// byte-position fragments stored on the schema during `parse`,
    /// `emit_pretty` walks tree-sitter `grammar.json` production rules
    /// to render schemas built from scratch via `SchemaBuilder`.
    ///
    /// # Errors
    ///
    /// Returns [`ParseError::EmitFailed`] when the language has no
    /// vendored `grammar.json`, when a vertex's kind is not a grammar
    /// rule, or when a required field has no corresponding schema edge.
    fn emit_pretty(&self, schema: &Schema) -> Result<Vec<u8>, ParseError> {
        self.emit_pretty_with_policy(schema, &crate::emit_pretty::FormatPolicy::default())
    }

    /// Render a by-construction [`Schema`] under a caller-supplied
    /// [`FormatPolicy`](crate::emit_pretty::FormatPolicy).
    ///
    /// The policy governs every configurable aspect of the rendered
    /// output: separator between glued tokens, newline byte sequence,
    /// indent width, line-break and indent-open/close token sets. The
    /// default policy (used by [`emit_pretty`](Self::emit_pretty))
    /// targets syntactic validity with ASCII conventions; callers
    /// supplying their own policy can pin idiomatic formatting.
    ///
    /// # Errors
    ///
    /// Returns [`ParseError::EmitFailed`] when the language has no
    /// vendored `grammar.json`, when a vertex's kind is not a grammar
    /// rule, or when a required field has no corresponding schema edge.
    fn emit_pretty_with_policy(
        &self,
        schema: &Schema,
        policy: &crate::emit_pretty::FormatPolicy,
    ) -> Result<Vec<u8>, ParseError> {
        let _ = (schema, policy);
        Err(ParseError::EmitFailed {
            protocol: self.protocol_name().to_owned(),
            reason: format!(
                "emit_pretty_with_policy not implemented for protocol '{}'",
                self.protocol_name()
            ),
        })
    }
}

/// Registry of all full-AST parsers, keyed by protocol name.
///
/// Provides language detection by file extension and dispatches parse/emit
/// operations to the appropriate language parser.
pub struct ParserRegistry {
    /// Parsers keyed by protocol name.
    ///
    /// Held by `Arc` (not `Box`) so the same handle can be shared with
    /// the layout-enrichment registry without re-wrapping at every
    /// lookup. Registration installs both: the parser into `parsers`
    /// and a thin adapter into the lens crate's enrichment registry.
    parsers: FxHashMap<String, Arc<dyn AstParser>>,
    /// Extension → protocol name mapping.
    extension_map: FxHashMap<String, String>,
}

impl ParserRegistry {
    /// Create a new registry populated with all enabled language parsers.
    ///
    /// With the `grammars` feature (default), this populates the registry from
    /// `panproto-grammars`, which provides up to 261 tree-sitter languages.
    /// Without the `grammars` feature, this returns an empty registry; call
    /// [`register`](Self::register) to add parsers manually using individual
    /// grammar crates.
    #[must_use]
    pub fn new() -> Self {
        let mut registry = Self {
            parsers: FxHashMap::default(),
            extension_map: FxHashMap::default(),
        };

        #[cfg(feature = "grammars")]
        for grammar in panproto_grammars::grammars() {
            let config = crate::languages::walker_configs::walker_config_for(grammar.name);
            match crate::languages::common::LanguageParser::from_language_with_grammar_json(
                grammar.name,
                grammar.extensions.to_vec(),
                grammar.language,
                grammar.node_types,
                grammar.tags_query,
                config,
                grammar.grammar_json,
            ) {
                Ok(p) => registry.register(Box::new(p)),
                Err(err) => {
                    let _ = err;
                    #[cfg(debug_assertions)]
                    eprintln!(
                        "warning: grammar '{}' theory extraction failed: {err}",
                        grammar.name
                    );
                }
            }
        }

        registry
    }

    /// Register a parser implementation.
    ///
    /// In addition to keying the parser by its protocol name, this
    /// installs a [`LayoutEnricher`](panproto_lens::enrichment_registry::LayoutEnricher)
    /// adapter into the global enrichment registry so that a
    /// `parse_emit_protolens(protocol, …)` instantiation finds a
    /// synthesis driver without any further wiring.
    pub fn register(&mut self, parser: Box<dyn AstParser>) {
        let name = parser.protocol_name().to_owned();
        for ext in parser.supported_extensions() {
            self.extension_map.insert((*ext).to_owned(), name.clone());
        }
        let arc: Arc<dyn AstParser> = Arc::from(parser);
        crate::decorate::register_layout_enricher(Arc::clone(&arc));
        self.parsers.insert(name, arc);
    }

    /// Register a tree-sitter language as a full-AST parser.
    ///
    /// Used by `panproto-grammars-*` companion crates that ship grammars
    /// outside the default `panproto-grammars` build. The byte-slice
    /// arguments must outlive this registry; the canonical pattern is
    /// for the companion to bake the data into `&'static` rodata at
    /// compile time and pass references that are valid for the process
    /// lifetime.
    ///
    /// `walker_config` is looked up by `name` from the bundled per-language
    /// configuration table. Languages without a tailored configuration
    /// fall back to the default walker config.
    ///
    /// # Errors
    ///
    /// Returns [`ParseError`] if theory extraction from `node_types_json`
    /// fails or if the tags query rejects compilation.
    pub fn register_external_grammar(
        &mut self,
        name: &'static str,
        extensions: Vec<&'static str>,
        language: tree_sitter::Language,
        node_types_json: &'static [u8],
        tags_query: Option<&'static str>,
        grammar_json: Option<&'static [u8]>,
    ) -> Result<(), crate::error::ParseError> {
        let config = crate::languages::walker_configs::walker_config_for(name);
        let parser = crate::languages::common::LanguageParser::from_language_with_grammar_json(
            name,
            extensions,
            language,
            node_types_json,
            tags_query,
            config,
            grammar_json,
        )?;
        self.register(Box::new(parser));
        Ok(())
    }

    /// Owned-data variant of [`register_external_grammar`](Self::register_external_grammar).
    ///
    /// Accepts `String` / `Vec<u8>` rather than `&'static` references. The
    /// caller is presumed not to have process-lifetime rodata available
    /// (typical dev-time use: bytes read from disk via the Python binding's
    /// override hook). To match the trait's `'static` lifetime requirement
    /// the inputs are leaked into the heap; the leak is one-time per
    /// override.
    ///
    /// This is the registration primitive for grammar-author workflows
    /// where a grammar's `parser.c` / `grammar.json` / `node-types.json`
    /// are evolving outside the panproto release cadence. Production
    /// builds should continue to use [`register_external_grammar`](Self::register_external_grammar) with
    /// `'static` data baked into the binary at compile time.
    ///
    /// # Errors
    ///
    /// Returns [`ParseError`] if theory extraction or tags-query
    /// compilation fails.
    pub fn register_external_grammar_owned(
        &mut self,
        name: String,
        extensions: Vec<String>,
        language: tree_sitter::Language,
        node_types_json: Vec<u8>,
        tags_query: Option<String>,
        grammar_json: Option<Vec<u8>>,
    ) -> Result<(), crate::error::ParseError> {
        let name_static: &'static str = Box::leak(name.into_boxed_str());
        let extensions_static: Vec<&'static str> = extensions
            .into_iter()
            .map(|s| Box::leak(s.into_boxed_str()) as &'static str)
            .collect();
        let node_types_static: &'static [u8] = Box::leak(node_types_json.into_boxed_slice());
        let tags_query_static: Option<&'static str> =
            tags_query.map(|s| Box::leak(s.into_boxed_str()) as &'static str);
        let grammar_json_static: Option<&'static [u8]> =
            grammar_json.map(|v| Box::leak(v.into_boxed_slice()) as &'static [u8]);

        self.register_external_grammar(
            name_static,
            extensions_static,
            language,
            node_types_static,
            tags_query_static,
            grammar_json_static,
        )
    }

    /// Remove a registration by protocol name.
    ///
    /// Drops the parser and any extension mappings that pointed at it.
    /// Returns `true` if a parser was removed, `false` if no such
    /// registration existed. Primarily intended for grammar-author
    /// workflows where a registered grammar is being replaced by a
    /// freshly-compiled version mid-process.
    pub fn unregister(&mut self, name: &str) -> bool {
        let removed = self.parsers.remove(name).is_some();
        if removed {
            self.extension_map.retain(|_, v| v != name);
        }
        removed
    }

    /// Override a registered grammar with new owned data.
    ///
    /// Equivalent to [`unregister`](Self::unregister) followed by
    /// [`register_external_grammar_owned`](Self::register_external_grammar_owned),
    /// and intended for the same grammar-author dev workflow. Any
    /// extension mappings previously bound to `name` are replaced by
    /// the new `extensions`.
    ///
    /// # Errors
    ///
    /// Returns [`ParseError`] if theory extraction or tags-query
    /// compilation fails on the new grammar; in that case the prior
    /// registration is already gone.
    pub fn override_grammar(
        &mut self,
        name: String,
        extensions: Vec<String>,
        language: tree_sitter::Language,
        node_types_json: Vec<u8>,
        tags_query: Option<String>,
        grammar_json: Option<Vec<u8>>,
    ) -> Result<(), crate::error::ParseError> {
        self.unregister(&name);
        self.register_external_grammar_owned(
            name,
            extensions,
            language,
            node_types_json,
            tags_query,
            grammar_json,
        )
    }

    /// Detect the language protocol for a file path by its extension.
    ///
    /// Returns `None` if the extension is not recognized (caller should
    /// fall back to the `raw_file` protocol).
    #[must_use]
    pub fn detect_language(&self, path: &Path) -> Option<&str> {
        path.extension()
            .and_then(|ext| ext.to_str())
            .and_then(|ext| self.extension_map.get(ext))
            .map(String::as_str)
    }

    /// Parse a file by detecting its language from the file path.
    ///
    /// # Errors
    ///
    /// Returns [`ParseError::UnknownLanguage`] if the file extension is not recognized.
    /// Returns other [`ParseError`] variants if parsing fails.
    pub fn parse_file(&self, path: &Path, content: &[u8]) -> Result<Schema, ParseError> {
        let protocol = self
            .detect_language(path)
            .ok_or_else(|| ParseError::UnknownLanguage {
                extension: path
                    .extension()
                    .and_then(|e| e.to_str())
                    .unwrap_or("")
                    .to_owned(),
            })?;

        self.parse_with_protocol(protocol, content, &path.display().to_string())
    }

    /// Parse source code with a specific protocol name.
    ///
    /// # Errors
    ///
    /// Returns [`ParseError::UnknownLanguage`] if the protocol is not registered.
    pub fn parse_with_protocol(
        &self,
        protocol: &str,
        content: &[u8],
        file_path: &str,
    ) -> Result<Schema, ParseError> {
        let parser = self
            .parsers
            .get(protocol)
            .ok_or_else(|| ParseError::UnknownLanguage {
                extension: protocol.to_owned(),
            })?;

        parser.parse(content, file_path)
    }

    /// Emit a schema back to source code bytes using the specified protocol.
    ///
    /// # Errors
    ///
    /// Returns [`ParseError::UnknownLanguage`] if the protocol is not registered.
    pub fn emit_with_protocol(
        &self,
        protocol: &str,
        schema: &Schema,
    ) -> Result<Vec<u8>, ParseError> {
        let parser = self
            .parsers
            .get(protocol)
            .ok_or_else(|| ParseError::UnknownLanguage {
                extension: protocol.to_owned(),
            })?;

        parser.emit(schema)
    }

    /// Render a by-construction schema using the named protocol.
    ///
    /// # Errors
    ///
    /// Returns [`ParseError::UnknownLanguage`] if the protocol is not
    /// registered, or [`ParseError::EmitFailed`] from the underlying
    /// parser's `emit_pretty`.
    pub fn emit_pretty_with_protocol(
        &self,
        protocol: &str,
        schema: &Schema,
    ) -> Result<Vec<u8>, ParseError> {
        let parser = self
            .parsers
            .get(protocol)
            .ok_or_else(|| ParseError::UnknownLanguage {
                extension: protocol.to_owned(),
            })?;

        parser.emit_pretty(schema)
    }

    /// Report the test-verification status of `emit_pretty` for a
    /// given protocol.
    ///
    /// The status is a programmatic check that downstream tooling
    /// (e.g. quivers, schema-migration pipelines) can use to refuse
    /// emit on protocols whose fixed-point law has never been
    /// exercised by panproto's test suite. The three tiers are:
    ///
    /// * [`EmitVerificationStatus::Verified`] — the protocol has an
    ///   explicit fixed-point or roundtrip test in panproto's suite.
    ///   `emit_pretty(parse(emit_pretty(s))) == emit_pretty(s)` is
    ///   known to hold on representative source.
    /// * [`EmitVerificationStatus::Generic`] — the protocol is
    ///   registered (a tree-sitter grammar is vendored) and the
    ///   generic dispatch path applies, but no per-language test
    ///   asserts emit correctness. Output is structurally derived
    ///   from `grammar.json` + the universal cassette layer and is
    ///   likely correct, but unverified.
    /// * [`EmitVerificationStatus::Unsupported`] — the protocol is
    ///   not registered, OR is registered but no `grammar.json` was
    ///   vendored at build time. `emit_pretty` will return
    ///   [`ParseError::EmitFailed`].
    #[must_use]
    pub fn emit_verification_status(&self, protocol: &str) -> EmitVerificationStatus {
        if !self.parsers.contains_key(protocol) {
            return EmitVerificationStatus::Unsupported;
        }
        if VERIFIED_EMIT_PROTOCOLS.binary_search(&protocol).is_ok() {
            EmitVerificationStatus::Verified
        } else {
            EmitVerificationStatus::Generic
        }
    }
}

/// Programmatic verification tier for [`ParserRegistry::emit_verification_status`].
#[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
pub enum EmitVerificationStatus {
    /// `emit_pretty` for this protocol has a test in panproto's suite
    /// asserting the fixed-point law on representative source.
    Verified,
    /// The protocol is registered and the generic dispatch path
    /// applies, but no per-language test asserts emit correctness.
    Generic,
    /// The protocol is not registered, or its grammar lacks the
    /// vendored `grammar.json` that `emit_pretty` requires.
    Unsupported,
}

/// Protocols whose `emit_pretty` is verified to a bar that justifies
/// downstream trust. A protocol qualifies on one of two bases:
///
/// 1. **Corpus-verified** — every entry in the grammar author's own
///    `test/corpus/` round-trips under the full oracle (byte fixed point +
///    vertex-kind multiset + edge-shape multiset), checked by the strict
///    `emit_corpus_audit` test. This is the strong bar: the corpus exercises
///    the whole grammar, not one hand-written sample.
/// 2. **Backend-verified** — a quivers transpile backend (`python`, `stan`,
///    `bugs`, `jags`, `julia`, `scheme`, `javascript`) covered by dedicated
///    emit regression tests for the construct surface quivers actually emits.
///    These are pinned by `emit_verification_status` tests as a downstream
///    contract; bringing each to full corpus-pass (basis 1) is tracked work.
///
/// A single hand-written round-trip sample is NOT sufficient: an earlier
/// expansion to 149 protocols on minimal samples was reverted after a corpus
/// audit showed most failed their own grammar's test corpus.
///
/// Names MUST be kept in sorted order so the binary-search lookup in
/// [`ParserRegistry::emit_verification_status`] works.
const VERIFIED_EMIT_PROTOCOLS: &[&str] = &[
    "abc",
    "actionscript",
    "ada",
    "agda",
    "al",
    "angular",
    "apex",
    "arduino",
    "asciidoc",
    "asm",
    "astro",
    "awk",
    "bash",
    "bass",
    "batch",
    "beancount",
    "bibtex",
    "bicep",
    "bitbake",
    "blade",
    "brightscript",
    "bsl",
    "bugs",
    "c",
    "caddy",
    "cairo",
    "capnp",
    "cedar",
    "cedarschema",
    "chatito",
    "chuck",
    "circom",
    "clarity",
    "clojure",
    "cmake",
    "cobol",
    "commonlisp",
    "cooklang",
    "corn",
    "cpon",
    "cpp",
    "crystal",
    "csharp",
    "csound",
    "css",
    "csv",
    "cuda",
    "cue",
    "cylc",
    "d",
    "dart",
    "desktop",
    "devicetree",
    "diff",
    "djot",
    "dockerfile",
    "dot",
    "doxygen",
    "dtd",
    "earthfile",
    "ebnf",
    "editorconfig",
    "eds",
    "eex",
    "elisp",
    "elixir",
    "elm",
    "elsa",
    "embedded_template",
    "enforce",
    "erlang",
    "facility",
    "faust",
    "fennel",
    "fidl",
    "firrtl",
    "fish",
    "foam",
    "forth",
    "fortran",
    "fsharp",
    "fsharp_signature",
    "func",
    "gdscript",
    "git_config",
    "git_rebase",
    "gitattributes",
    "gitcommit",
    "gitignore",
    "gleam",
    "glicol",
    "glsl",
    "gn",
    "go",
    "godot_resource",
    "gomod",
    "gosum",
    "graphql",
    "groovy",
    "gstlaunch",
    "hack",
    "hare",
    "haskell",
    "haxe",
    "hcl",
    "heex",
    "hlsl",
    "html",
    "http",
    "hurl",
    "hyprlang",
    "idris",
    "ini",
    "ispc",
    "jags",
    "janet",
    "java",
    "javascript",
    "jinja2",
    "jq",
    "jsdoc",
    "json",
    "jsonnet",
    "julia",
    "just",
    "kconfig",
    "kdl",
    "kotlin",
    "latex",
    "lean",
    "ledger",
    "lilypond",
    "linkerscript",
    "liquid",
    "llvm",
    "lua",
    "luadoc",
    "luap",
    "luau",
    "magik",
    "make",
    "markdown",
    "markdown_inline",
    "matlab",
    "mermaid",
    "meson",
    "mojo",
    "netlinx",
    "nginx",
    "nickel",
    "nim",
    "ninja",
    "nix",
    "norg",
    "nqc",
    "nushell",
    "objc",
    "ocaml",
    "ocaml_interface",
    "odin",
    "org",
    "pascal",
    "pem",
    "perl",
    "pgn",
    "php",
    "pkl",
    "po",
    "pony",
    "postscript",
    "powershell",
    "printf",
    "prisma",
    "prolog",
    "promql",
    "properties",
    "protobuf",
    "psv",
    "pug",
    "puppet",
    "purescript",
    "pymanifest",
    "python",
    "ql",
    "qml",
    "qmldir",
    "query",
    "qvr",
    "r",
    "racket",
    "re2c",
    "readline",
    "regex",
    "rego",
    "requirements",
    "rescript",
    "robot",
    "ron",
    "rst",
    "ruby",
    "rust",
    "scala",
    "scheme",
    "scss",
    "smali",
    "smithy",
    "solidity",
    "sparql",
    "sql",
    "squirrel",
    "ssh_config",
    "stan",
    "stanfunctions",
    "starlark",
    "strudel_mini",
    "supercollider",
    "svelte",
    "swift",
    "tablegen",
    "tcl",
    "teal",
    "templ",
    "terraform",
    "textproto",
    "thrift",
    "tidal_mini",
    "tlaplus",
    "tmux",
    "toml",
    "tsv",
    "tsx",
    "turtle",
    "twig",
    "typescript",
    "typst",
    "udev",
    "ungrammar",
    "uxntal",
    "v",
    "vb",
    "verilog",
    "vhdl",
    "vim",
    "vimdoc",
    "vue",
    "wast",
    "wat",
    "wgsl",
    "wit",
    "xcompose",
    "xml",
    "yaml",
    "yuck",
    "zig",
    "zsh",
];

impl ParserRegistry {
    /// Decorate an [`AbstractSchema`] with the layout enrichment
    /// fibre required by `emit_pretty_with_protocol` and friends.
    ///
    /// This is the put-direction of the parse / decorate / emit lens
    /// at `protocol`. The implementation routes through the same
    /// grammar walker as `emit_pretty` followed by `parse`, so the
    /// resulting [`DecoratedSchema`] carries a complete layout fibre
    /// recovered by the parse-side walker — `start-byte`, `end-byte`,
    /// every `interstitial-N`, `chose-alt-fingerprint`, and
    /// `chose-alt-child-kinds`.
    ///
    /// The section law holds up to kind- and edge-multiset
    /// equivalence: `forget_layout(decorate(a)) ≅ a` modulo vertex-id
    /// renaming. Grammars where parsing consolidates tokens that the
    /// emitter rendered as separate sequences (e.g. lilypond's `c'4`
    /// re-parses to a single note) do not preserve a one-to-one
    /// vertex correspondence, so the result's vertex IDs are always
    /// freshly minted by the parser.
    ///
    /// # Errors
    ///
    /// Returns [`ParseError::UnknownLanguage`] when `protocol` is not
    /// registered, [`ParseError::SchemaConstruction`] when the
    /// abstract schema was built for a different protocol than
    /// `protocol`, [`ParseError::EmitFailed`] when the grammar walker
    /// cannot render the abstract schema (missing `grammar.json`,
    /// vertex kind not a rule), or any other parser error if the
    /// re-parse step rejects the canonical bytes (a regression in the
    /// parse/emit pipeline, not a user bug).
    pub fn decorate(
        &self,
        protocol: &str,
        abstract_schema: &AbstractSchema,
        policy: &LayoutPolicy,
    ) -> Result<DecoratedSchema, ParseError> {
        let parser = self
            .parsers
            .get(protocol)
            .ok_or_else(|| ParseError::UnknownLanguage {
                extension: protocol.to_owned(),
            })?;
        // `decorate_with_parser` enforces the protocol-match invariant
        // between the parser and the abstract schema, so no extra guard
        // is needed here.
        crate::decorate::decorate_with_parser(parser.as_ref(), abstract_schema, policy)
    }

    /// Render an [`AbstractSchema`] to canonical source bytes under
    /// `policy`.
    ///
    /// Implementation note: this is exactly the first emit step of
    /// [`decorate`](Self::decorate) — `decorate` then re-parses to
    /// recover the layout fibre, but if all the caller wants is the
    /// bytes, the re-parse is wasted work. Going through
    /// `emit_pretty_with_policy` directly preserves every field of
    /// `policy` in the output (`separator`, `newline`, `indent_width`,
    /// `line_break_after`, `indent_open` / `indent_close`).
    ///
    /// # Errors
    ///
    /// See [`decorate`](Self::decorate).
    pub fn pretty_with_protocol(
        &self,
        protocol: &str,
        abstract_schema: &AbstractSchema,
        policy: &LayoutPolicy,
    ) -> Result<Vec<u8>, ParseError> {
        let parser = self
            .parsers
            .get(protocol)
            .ok_or_else(|| ParseError::UnknownLanguage {
                extension: protocol.to_owned(),
            })?;
        check_protocol_match(
            protocol,
            abstract_schema.as_schema(),
            "pretty_with_protocol",
        )?;
        parser.emit_pretty_with_policy(abstract_schema.as_schema(), policy)
    }

    /// Return the canonical [`Protolens`](panproto_lens::Protolens)
    /// describing the parse / decorate / emit relationship at
    /// `protocol`.
    ///
    /// The protolens encodes the schema-level structure of the
    /// relationship: source-side strips the layout enrichment fibre,
    /// target-side adds it via the registered
    /// [`LayoutEnricher`](panproto_lens::enrichment_registry::LayoutEnricher).
    /// It composes with the rest of the `panproto-lens` protolens
    /// algebra for chain-law reasoning. The operational entry points
    /// for running the relationship on real schemas are
    /// [`decorate`](Self::decorate),
    /// [`pretty_with_protocol`](Self::pretty_with_protocol), and
    /// [`emit_pretty_with_protocol`](Self::emit_pretty_with_protocol).
    ///
    /// # Errors
    ///
    /// Returns [`ParseError::UnknownLanguage`] when `protocol` is not
    /// registered.
    pub fn parse_emit_protolens(
        &self,
        protocol: &str,
        policy: &LayoutPolicy,
    ) -> Result<panproto_lens::Protolens, ParseError> {
        if !self.parsers.contains_key(protocol) {
            return Err(ParseError::UnknownLanguage {
                extension: protocol.to_owned(),
            });
        }
        Ok(crate::parse_emit_protolens::parse_emit_protolens(
            protocol, policy,
        ))
    }

    /// Get the theory metadata for a specific protocol.
    #[must_use]
    pub fn theory_meta(&self, protocol: &str) -> Option<&ExtractedTheoryMeta> {
        self.parsers.get(protocol).map(|p| p.theory_meta())
    }

    /// List all registered protocol names.
    pub fn protocol_names(&self) -> impl Iterator<Item = &str> {
        self.parsers.keys().map(String::as_str)
    }

    /// O(1) lookup: is a parser already registered for `protocol`?
    ///
    /// Useful for dedup at the registration boundary. The umbrella
    /// `panproto-grammars-all` companion pack overlaps with both the
    /// built-in core grammars and every per-group pack; callers can
    /// short-circuit before re-registering rather than scanning
    /// `protocol_names()` linearly.
    #[must_use]
    pub fn has_parser(&self, protocol: &str) -> bool {
        self.parsers.contains_key(protocol)
    }

    /// Get the number of registered parsers.
    #[must_use]
    pub fn len(&self) -> usize {
        self.parsers.len()
    }

    /// Check if the registry is empty.
    #[must_use]
    pub fn is_empty(&self) -> bool {
        self.parsers.is_empty()
    }
}

impl Default for ParserRegistry {
    fn default() -> Self {
        Self::new()
    }
}

/// Guard against running parser-tied operations on a schema built
/// for a different protocol. Catches the user-visible error of
/// passing (say) a JSON schema to a Python parser before the
/// underlying grammar walker would surface it as an opaque rule
/// mismatch.
fn check_protocol_match(
    expected: &str,
    schema: &Schema,
    operation: &'static str,
) -> Result<(), ParseError> {
    if schema.protocol == expected {
        Ok(())
    } else {
        Err(ParseError::SchemaConstruction {
            reason: format!(
                "{operation}: protocol mismatch — registry called with '{expected}' but \
                 schema carries protocol '{}'",
                schema.protocol,
            ),
        })
    }
}