symforge 8.9.7

SymForge — in-memory code intelligence MCP server for Claude Code, Codex, and Gemini CLI
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
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
use std::collections::HashMap;
use std::fmt;

#[derive(
    Clone, Debug, PartialEq, Eq, Hash, PartialOrd, Ord, serde::Serialize, serde::Deserialize,
)]
pub enum LanguageId {
    Rust,
    Python,
    JavaScript,
    TypeScript,
    Go,
    Java,
    C,
    Cpp,
    CSharp,
    Ruby,
    Php,
    Swift,
    Kotlin,
    Dart,
    Perl,
    Elixir,
    Json,
    Toml,
    Yaml,
    Markdown,
    Env,
    Html,
    Css,
    Scss,
}

impl LanguageId {
    pub fn from_extension(ext: &str) -> Option<Self> {
        match ext {
            "rs" => Some(Self::Rust),
            "py" => Some(Self::Python),
            "js" | "jsx" => Some(Self::JavaScript),
            "ts" | "tsx" => Some(Self::TypeScript),
            "go" => Some(Self::Go),
            "java" => Some(Self::Java),
            "c" | "h" => Some(Self::C),
            "cpp" | "cxx" | "cc" | "hpp" | "hxx" | "hh" => Some(Self::Cpp),
            "cs" => Some(Self::CSharp),
            "rb" => Some(Self::Ruby),
            "php" => Some(Self::Php),
            "swift" => Some(Self::Swift),
            "dart" => Some(Self::Dart),
            "kt" | "kts" => Some(Self::Kotlin),
            "pl" | "pm" => Some(Self::Perl),
            "ex" | "exs" => Some(Self::Elixir),
            "json" => Some(Self::Json),
            "toml" => Some(Self::Toml),
            "yaml" | "yml" => Some(Self::Yaml),
            "md" => Some(Self::Markdown),
            "env" => Some(Self::Env),
            "html" => Some(Self::Html),
            "css" => Some(Self::Css),
            "scss" => Some(Self::Scss),
            _ => None,
        }
    }

    /// Returns `true` when `relative_path` is a TSX source file (`.tsx`).
    ///
    /// `.tsx` and `.ts` both map to [`LanguageId::TypeScript`], but they require
    /// different tree-sitter grammars: `.tsx` needs the TSX grammar (JSX-aware,
    /// rejects legacy `<T>expr` casts) while `.ts` needs the plain TypeScript
    /// grammar (no JSX, accepts angle-bracket casts). The grammar is therefore
    /// selected from the file extension, not the [`LanguageId`], so this helper
    /// threads the TSX flavor to the parse sites.
    pub fn is_tsx_path(relative_path: &str) -> bool {
        relative_path
            .rsplit(['/', '\\'])
            .next()
            .and_then(|name| name.rsplit_once('.'))
            .is_some_and(|(_, ext)| ext.eq_ignore_ascii_case("tsx"))
    }

    /// Returns `true` when `relative_path` is a `.h` header file.
    ///
    /// `.h` maps to [`LanguageId::C`] by extension, but in cross-platform repos
    /// it routinely holds C++ (`class`/`namespace`/`::`) or Objective-C
    /// (`@interface`/`@property`) — none of which the C grammar can parse,
    /// causing total symbol loss (SF-STRESS-005). The grammar for a `.h` file is
    /// therefore disambiguated from CONTENT at parse time (see
    /// [`crate::parsing`]), not from the extension alone; this helper flags the
    /// `.h` extension so the parse site knows to apply that disambiguation.
    ///
    /// Unambiguous C++ header extensions (`.hpp`/`.hxx`/`.hh`) already map to
    /// [`LanguageId::Cpp`] and are NOT `.h`, so they are not flagged here.
    pub fn is_c_header_path(relative_path: &str) -> bool {
        relative_path
            .rsplit(['/', '\\'])
            .next()
            .and_then(|name| name.rsplit_once('.'))
            .is_some_and(|(_, ext)| ext.eq_ignore_ascii_case("h"))
    }

    pub fn extensions(&self) -> &[&str] {
        match self {
            Self::Rust => &["rs"],
            Self::Python => &["py"],
            Self::JavaScript => &["js", "jsx"],
            Self::TypeScript => &["ts", "tsx"],
            Self::Go => &["go"],
            Self::Java => &["java"],
            Self::C => &["c", "h"],
            Self::Cpp => &["cpp", "cxx", "cc", "hpp", "hxx", "hh"],
            Self::CSharp => &["cs"],
            Self::Ruby => &["rb"],
            Self::Php => &["php"],
            Self::Swift => &["swift"],
            Self::Kotlin => &["kt", "kts"],
            Self::Dart => &["dart"],
            Self::Perl => &["pl", "pm"],
            Self::Elixir => &["ex", "exs"],
            Self::Json => &["json"],
            Self::Toml => &["toml"],
            Self::Yaml => &["yaml", "yml"],
            Self::Markdown => &["md"],
            Self::Env => &["env"],
            Self::Html => &["html"],
            Self::Css => &["css"],
            Self::Scss => &["scss"],
        }
    }

    pub fn support_tier(&self) -> SupportTier {
        match self {
            Self::Rust | Self::Python | Self::JavaScript | Self::TypeScript | Self::Go => {
                SupportTier::QualityFocus
            }
            Self::Java
            | Self::C
            | Self::Cpp
            | Self::CSharp
            | Self::Ruby
            | Self::Php
            | Self::Swift
            | Self::Kotlin
            | Self::Dart
            | Self::Perl
            | Self::Elixir
            | Self::Json
            | Self::Toml
            | Self::Yaml
            | Self::Markdown
            | Self::Env
            | Self::Html
            | Self::Css
            | Self::Scss => SupportTier::Broader,
        }
    }
}

impl fmt::Display for LanguageId {
    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
        let name = match self {
            Self::Rust => "Rust",
            Self::Python => "Python",
            Self::JavaScript => "JavaScript",
            Self::TypeScript => "TypeScript",
            Self::Go => "Go",
            Self::Java => "Java",
            Self::C => "C",
            Self::Cpp => "C++",
            Self::CSharp => "C#",
            Self::Ruby => "Ruby",
            Self::Php => "PHP",
            Self::Swift => "Swift",
            Self::Kotlin => "Kotlin",
            Self::Dart => "Dart",
            Self::Perl => "Perl",
            Self::Elixir => "Elixir",
            Self::Json => "JSON",
            Self::Toml => "TOML",
            Self::Yaml => "YAML",
            Self::Markdown => "Markdown",
            Self::Env => "Env",
            Self::Html => "HTML",
            Self::Css => "CSS",
            Self::Scss => "SCSS",
        };
        write!(f, "{name}")
    }
}

#[derive(Clone, Debug, PartialEq, Eq)]
pub enum SupportTier {
    QualityFocus,
    Broader,
    Unsupported,
}

#[derive(Clone, Copy, Debug, PartialEq, Eq, Hash, serde::Serialize, serde::Deserialize)]
pub enum FileClass {
    Code,
    Text,
    Binary,
}

/// Canonical path-segment source of truth for noise classification, shared by
/// [`FileClassification::for_code_path`] (gates search_symbols/search_text) and
/// `NoiseClass::classify_path` (gates explore/repo_map/search_files) so the two
/// classifiers cannot diverge (SF-STRESS-011). A segment is a single
/// `/`-delimited path component, already lowercased.
///
/// Vendored / third-party dependency roots. `deps` and `extern`/`externals`
/// added per SF-STRESS-011 corpus evidence (redis `deps/`, node `deps/`).
pub const VENDOR_PATH_SEGMENTS: &[&str] = &[
    "vendor",
    "third_party",
    "third-party",
    "node_modules",
    ".venv",
    "venv",
    "site-packages",
    "pods",
    "bower_components",
    "deps",
    "extern",
    "externals",
];

/// Build-output / machine-generated directory roots. `dist` is shared here so
/// `FileClassification` agrees with `NoiseClass` (previously `dist` lived only
/// in `NoiseClass`, causing the laravel `dist/` divergence in SF-STRESS-011).
pub const GENERATED_PATH_SEGMENTS: &[&str] =
    &["dist", "generated", "__generated__", "generated-sources"];

/// Test-directory roots. `test_data`/`testdata`/`fixtures`/`__fixtures__`/
/// `__snapshots__` added per SF-STRESS-011 (rust-analyzer `test_data` fixtures,
/// snapshot trees) and SF-STRESS-009 (test-fixture bucket).
pub const TEST_PATH_SEGMENTS: &[&str] = &[
    "tests",
    "test",
    "__tests__",
    "spec",
    "test_data",
    "testdata",
    "fixtures",
    "__fixtures__",
    "__snapshots__",
];

#[derive(Clone, Copy, Debug, PartialEq, Eq, Hash, serde::Serialize, serde::Deserialize)]
pub struct FileClassification {
    pub class: FileClass,
    pub is_generated: bool,
    pub is_test: bool,
    pub is_vendor: bool,
    #[serde(default)]
    pub is_config: bool,
}

impl FileClassification {
    pub fn for_code_path(relative_path: &str) -> Self {
        let lower = relative_path.replace('\\', "/").to_ascii_lowercase();
        let segments: Vec<&str> = lower
            .split('/')
            .filter(|segment| !segment.is_empty())
            .collect();
        let basename = segments.last().copied().unwrap_or("");
        let stem = basename
            .rsplit_once('.')
            .map(|(name, _)| name)
            .unwrap_or(basename);

        let is_test = segments
            .iter()
            .any(|segment| TEST_PATH_SEGMENTS.contains(segment))
            || stem.starts_with("test_")
            || stem.ends_with("_test")
            || stem.ends_with(".test")
            || stem.ends_with("_spec")
            || stem.ends_with(".spec");

        let is_vendor = segments
            .iter()
            .any(|segment| VENDOR_PATH_SEGMENTS.contains(segment));

        let is_generated = segments
            .iter()
            .any(|segment| GENERATED_PATH_SEGMENTS.contains(segment))
            || basename.contains(".generated.")
            || basename.contains(".gen.")
            || basename.ends_with(".g.dart")
            || basename.ends_with(".pb.go")
            || basename.ends_with(".designer.cs")
            || basename.ends_with(".min.js")
            || basename.ends_with(".min.css")
            || basename.ends_with(".map");

        let ext = basename.rsplit_once('.').map(|(_, e)| e).unwrap_or("");
        let is_config = matches!(ext, "json" | "toml" | "yaml" | "yml" | "md" | "env");

        Self {
            class: FileClass::Code,
            is_generated,
            is_test,
            is_vendor,
            is_config,
        }
    }

    pub const fn is_code(&self) -> bool {
        matches!(self.class, FileClass::Code)
    }

    pub const fn is_text(&self) -> bool {
        matches!(self.class, FileClass::Text)
    }

    pub const fn is_binary(&self) -> bool {
        matches!(self.class, FileClass::Binary)
    }
}

#[derive(Clone, Debug, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
pub struct ParseDiagnostic {
    pub parser: String,
    pub message: String,
    pub line: Option<u32>,
    pub column: Option<u32>,
    pub byte_span: Option<(u32, u32)>,
    pub fallback_used: bool,
}

impl ParseDiagnostic {
    pub fn location_display(&self) -> Option<String> {
        match (self.line, self.column) {
            (Some(line), Some(column)) => Some(format!("line {line}, column {column}")),
            (Some(line), None) => Some(format!("line {line}")),
            (None, Some(column)) => Some(format!("column {column}")),
            (None, None) => None,
        }
    }

    pub fn summary(&self) -> String {
        let mut summary = format!("{}: {}", self.parser, self.message);
        if let Some(location) = self.location_display() {
            summary.push_str(&format!(" ({location})"));
        }
        if self.fallback_used {
            summary.push_str(" [fallback symbol extraction used]");
        }
        summary
    }
}

#[derive(Clone, Debug, PartialEq, Eq)]
pub struct FileProcessingResult {
    pub relative_path: String,
    pub language: LanguageId,
    pub classification: FileClassification,
    pub outcome: FileOutcome,
    pub parse_diagnostic: Option<ParseDiagnostic>,
    pub symbols: Vec<SymbolRecord>,
    pub byte_len: u64,
    pub content_hash: String,
    /// Cross-references extracted by `parsing::xref::extract_references`.
    /// Empty until Task 2 wires xref extraction into the parse pipeline.
    pub references: Vec<ReferenceRecord>,
    /// Import alias map for this file: alias -> original name (e.g. "Map" -> "HashMap").
    pub alias_map: HashMap<String, String>,
}

#[derive(Clone, Debug, PartialEq, Eq)]
pub enum FileOutcome {
    Processed,
    PartialParse { warning: String },
    Failed { error: String },
}

#[derive(Clone, Debug, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
pub struct SymbolRecord {
    pub name: String,
    pub kind: SymbolKind,
    pub depth: u32,
    pub sort_order: u32,
    pub byte_range: (u32, u32),
    pub line_range: (u32, u32),
    pub doc_byte_range: Option<(u32, u32)>,
    #[serde(default)]
    pub item_byte_range: Option<(u32, u32)>,
}

impl SymbolRecord {
    /// Returns the start byte of the full editable item when available.
    pub fn item_start(&self) -> u32 {
        self.item_byte_range.map_or_else(
            || {
                self.doc_byte_range
                    .map_or(self.byte_range.0, |(start, _)| start)
            },
            |(start, _)| start,
        )
    }

    /// Returns the end byte of the full editable item when available.
    pub fn item_end(&self) -> u32 {
        self.item_byte_range
            .map_or(self.byte_range.1, |(_, end)| end)
    }

    /// Returns the full editable item range when available.
    pub fn item_range(&self) -> (u32, u32) {
        (self.item_start(), self.item_end())
    }

    /// Returns the parser's core symbol node range.
    pub fn core_range(&self) -> (u32, u32) {
        self.byte_range
    }

    /// Returns the effective start byte, including doc comments if present.
    pub fn effective_start(&self) -> u32 {
        self.doc_byte_range
            .map_or(self.byte_range.0, |(start, _)| start)
    }
}

#[derive(Clone, Copy, Debug, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
pub enum SymbolKind {
    Function,
    Method,
    Class,
    Struct,
    Enum,
    Interface,
    Module,
    Constant,
    Variable,
    Type,
    Trait,
    Impl,
    Other,
    Key,
    Section,
}

impl fmt::Display for SymbolKind {
    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
        let prefix = match self {
            Self::Function => "fn",
            Self::Method => "fn",
            Self::Class => "class",
            Self::Struct => "struct",
            Self::Enum => "enum",
            Self::Interface => "interface",
            Self::Module => "mod",
            Self::Constant => "const",
            Self::Variable => "let",
            Self::Type => "type",
            Self::Trait => "trait",
            Self::Impl => "impl",
            Self::Other => "other",
            Self::Key => "key",
            Self::Section => "section",
        };
        write!(f, "{prefix}")
    }
}

/// A single cross-reference (call site, import, type usage, or macro use) extracted
/// from a source file. Part of the Phase 4 cross-reference pipeline.
#[derive(Clone, Debug, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
pub struct ReferenceRecord {
    /// Simple name at the reference site (e.g. "new", "process", "HashMap").
    pub name: String,
    /// Best-effort qualified name when available (e.g. "Vec::new", "fmt.Println").
    pub qualified_name: Option<String>,
    /// What kind of reference this is.
    pub kind: ReferenceKind,
    /// Byte range in the source file (start, end).
    pub byte_range: (u32, u32),
    /// Line range in the source file (start, end — zero-indexed).
    pub line_range: (u32, u32),
    /// Index into the file's symbol list for the innermost containing definition.
    /// `None` means the reference is at module/top level.
    pub enclosing_symbol_index: Option<u32>,
}

/// Discriminates the semantic role of a cross-reference.
#[derive(Clone, Copy, Debug, PartialEq, Eq, serde::Serialize, serde::Deserialize)]
pub enum ReferenceKind {
    /// A function or method call site.
    Call,
    /// An import/use/require statement.
    Import,
    /// A type annotation, generic parameter, or other type usage.
    TypeUsage,
    /// A macro invocation.
    MacroUse,
    /// A trait/interface implementation relationship.
    /// `name` holds the trait/interface name, `qualified_name` holds the implementing type.
    Implements,
    /// A read of a named `const`/`static` value at a value-expression site
    /// (e.g. iterated in a `for` loop, used as a receiver, or passed as a bare
    /// argument). Distinct from `Call` so callee/type analysis stays clean.
    ValueUse,
}

impl fmt::Display for ReferenceKind {
    fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
        let s = match self {
            Self::Call => "call",
            Self::Import => "import",
            Self::TypeUsage => "type_usage",
            Self::MacroUse => "macro_use",
            Self::Implements => "implements",
            Self::ValueUse => "value_use",
        };
        write!(f, "{s}")
    }
}

/// Returns the index of the innermost `SymbolRecord` whose `line_range` contains
/// `ref_line`, or `None` if the reference is at module level.
///
/// "Innermost" is defined as the symbol with the latest `line_range.0` (start line)
/// that still contains `ref_line`. This handles nested function definitions correctly.
pub fn find_enclosing_symbol(symbols: &[SymbolRecord], ref_line: u32) -> Option<u32> {
    let mut best: Option<(u32, u32)> = None; // (start_line, index)
    for (idx, sym) in symbols.iter().enumerate() {
        let (start, end) = sym.line_range;
        if ref_line >= start && ref_line <= end {
            match best {
                None => best = Some((start, idx as u32)),
                Some((best_start, _)) if start > best_start => best = Some((start, idx as u32)),
                _ => {}
            }
        }
    }
    best.map(|(_, idx)| idx)
}

/// Admission tier — whether a file is eligible for indexing/parsing at all.
/// Separate from NoiseClass (which is about ranking/filtering signal).
#[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
pub enum AdmissionTier {
    /// Tier 1: Fully indexed — parsed, symbols extracted, text searchable.
    Normal,
    /// Tier 2: Metadata only — path, size, classification stored. No parsing.
    MetadataOnly,
    /// Tier 3: Hard-skipped — counted in health, minimal registration.
    HardSkip,
}

/// Reason a file was placed in Tier 2 or Tier 3.
#[derive(Debug, Clone, Copy, PartialEq, Eq)]
pub enum SkipReason {
    SizeCeiling,
    DenylistedExtension,
    SizeThreshold,
    BinaryContent,
    /// File is a dependency lockfile (e.g. `package-lock.json`, `Cargo.lock`).
    /// Demoted to Tier-2 metadata-only so its machine-generated content does not
    /// flood the index with thousands of junk key/value symbols (a single
    /// `package-lock.json` can mint ~9k JSON-key symbols, dominating symbol counts
    /// and `conventions` complexity heuristics). The path stays searchable as
    /// metadata; only symbol extraction is skipped. See `LOCKFILE_BASENAMES`.
    DependencyLockfile,
    /// SF-009: file demoted to Tier-2 because it is not git-tracked, under the
    /// opt-in `SYMFORGE_EXCLUDE_UNTRACKED` policy (default OFF). Only minted when
    /// that env gate is explicitly enabled; the default admission path never
    /// produces this reason, so admission defaults are unchanged.
    Untracked,
    /// SF-004 / SF-012: file exists on disk and is small/non-binary, but its
    /// extension maps to no supported tree-sitter grammar (e.g. `.tcl`, `.sh`,
    /// `.m`, `.eex`, extensionless `LICENSE`/`Makefile`). It cannot be parsed, so
    /// it is admitted Tier-2 metadata-only instead of being stored with a
    /// contradictory Tier-1/Normal decision (which made it vanish from tier
    /// accounting and minted a false "File not found"). The path stays searchable
    /// as metadata; only symbol extraction is skipped.
    UnsupportedLanguage,
}

impl std::fmt::Display for SkipReason {
    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
        match self {
            SkipReason::SizeCeiling => write!(f, ">100MB"),
            SkipReason::DenylistedExtension => write!(f, "artifact"),
            SkipReason::SizeThreshold => write!(f, ">1MB"),
            SkipReason::BinaryContent => write!(f, "binary"),
            SkipReason::DependencyLockfile => write!(f, "lockfile"),
            SkipReason::Untracked => write!(f, "untracked"),
            SkipReason::UnsupportedLanguage => write!(f, "unsupported language"),
        }
    }
}

/// Structured result from the admission gate.
#[derive(Debug, Clone, Copy, PartialEq, Eq)]
pub struct AdmissionDecision {
    pub tier: AdmissionTier,
    pub reason: Option<SkipReason>,
}

impl AdmissionDecision {
    pub fn normal() -> Self {
        Self {
            tier: AdmissionTier::Normal,
            reason: None,
        }
    }
    pub fn skip(tier: AdmissionTier, reason: SkipReason) -> Self {
        Self {
            tier,
            reason: Some(reason),
        }
    }
}

/// Metadata record for a file that was not fully indexed (Tier 2 or Tier 3).
/// Stores the AdmissionDecision directly — no re-derivation downstream.
#[derive(Debug, Clone)]
pub struct SkippedFile {
    pub path: String,
    pub size: u64,
    pub extension: Option<String>,
    pub decision: AdmissionDecision,
}

impl SkippedFile {
    pub fn tier(&self) -> AdmissionTier {
        self.decision.tier
    }
    pub fn reason(&self) -> Option<SkipReason> {
        self.decision.reason
    }
}

pub const HARD_SKIP_BYTES: u64 = 100 * 1024 * 1024;
pub const METADATA_ONLY_BYTES: u64 = 1024 * 1024;
pub const BINARY_SNIFF_BYTES: usize = 8192;

const DENYLISTED_EXTENSIONS: &[&str] = &[
    // ML models
    "safetensors",
    "ckpt",
    "pt",
    "onnx",
    "gguf",
    "pth",
    // VM/disk images
    "vmdk",
    "iso",
    "img",
    "qcow2",
    // Archives
    "tar",
    "gz",
    "zip",
    "7z",
    "rar",
    "bz2",
    "xz",
    "zst",
    // Databases
    "db",
    "sqlite",
    "sqlite3",
    "mdb",
    // Media
    "mp3",
    "mp4",
    "wav",
    "avi",
    "mov",
    "mkv",
    "png",
    "jpg",
    "jpeg",
    "gif",
    "bmp",
    "ico",
    "woff",
    "woff2",
    "ttf",
    "eot",
    // Binary
    "bin",
    // Executables and libraries
    "exe",
    "dll",
    "so",
    "dylib",
    "class",
];

pub fn is_denylisted_extension(ext: &str) -> bool {
    DENYLISTED_EXTENSIONS.contains(&ext.to_lowercase().as_str())
}

/// Dependency lockfiles matched by exact basename (case-sensitive, as these
/// names are conventionally fixed). These are machine-generated manifests whose
/// content (resolved dependency trees) parses into thousands of structurally
/// meaningless symbols — e.g. a single `package-lock.json` can mint ~9k JSON-key
/// symbols, dwarfing real source and skewing `conventions` complexity stats.
/// Admission demotes them to Tier-2 (metadata only): the path stays searchable,
/// but no symbols are extracted.
///
/// Cross-reference: the search ranker keeps an OVERLAPPING but distinct list,
/// `CHORE_ANCHOR_FILENAMES` in `src/live_index/rank_signals.rs`, used to suppress
/// co-change promotion. That list is intentionally NOT shared: it also contains
/// non-lockfile chore anchors (`CHANGELOG.md`, `.release-please-manifest.json`)
/// and lives in the `live_index` layer, which `domain` must not depend on. The
/// two lists answer different questions (admission vs. ranking) and are kept
/// separate on purpose; update both when the lockfile ecosystem changes.
const LOCKFILE_BASENAMES: &[&str] = &[
    "package-lock.json",
    "npm-shrinkwrap.json",
    "yarn.lock",
    "pnpm-lock.yaml",
    "bun.lock",
    "Cargo.lock",
    "composer.lock",
    "Gemfile.lock",
    "poetry.lock",
    "uv.lock",
    "Pipfile.lock",
    "flake.lock",
    "packages.lock.json",
    "go.sum",
];

/// Returns `true` when `path`'s file name is an exact (case-sensitive) match for
/// a known dependency lockfile. See [`LOCKFILE_BASENAMES`].
pub fn is_dependency_lockfile(path: &std::path::Path) -> bool {
    path.file_name()
        .and_then(|n| n.to_str())
        .is_some_and(|name| LOCKFILE_BASENAMES.contains(&name))
}

/// Maximum number of NUL byte offsets [`scan_nul_bytes`] enumerates explicitly.
/// Beyond this we report only the count plus the first few offsets — the warning
/// exists to alert, not to exhaustively map every NUL.
pub const NUL_OFFSET_SAMPLE_LIMIT: usize = 5;

/// Result of scanning a byte slice for literal NUL (`0x00`) bytes.
///
/// A NUL is valid UTF-8, so it survives `String::from_utf8_lossy` intact and is
/// then rendered invisibly (most terminals show nothing or a space). Source code
/// that smuggles a NUL into a string/template literal (legal for Node, Python,
/// etc.) therefore reads back as text that does NOT byte-match the file — an
/// agent copying that rendered text for an edit produces a mismatch. This scan
/// powers an honest warning at the content-rendering boundary.
#[derive(Clone, Debug, Default, PartialEq, Eq)]
pub struct NulScan {
    /// Total count of NUL bytes in the scanned slice.
    pub count: usize,
    /// Byte offsets (relative to the scanned slice) of the first
    /// [`NUL_OFFSET_SAMPLE_LIMIT`] NUL bytes, in ascending order.
    pub first_offsets: Vec<usize>,
}

impl NulScan {
    /// `true` when the scanned slice contained at least one NUL byte.
    pub fn has_nul(&self) -> bool {
        self.count > 0
    }
}

/// Scan `content` for literal NUL (`0x00`) bytes, returning a count and the
/// first [`NUL_OFFSET_SAMPLE_LIMIT`] offsets. Pure; scans the WHOLE slice it is
/// given (callers pass the exact bytes whose rendering they are warning about).
pub fn scan_nul_bytes(content: &[u8]) -> NulScan {
    let mut count = 0usize;
    let mut first_offsets = Vec::new();
    for (offset, &byte) in content.iter().enumerate() {
        if byte == 0 {
            count += 1;
            if first_offsets.len() < NUL_OFFSET_SAMPLE_LIMIT {
                first_offsets.push(offset);
            }
        }
    }
    NulScan {
        count,
        first_offsets,
    }
}

/// Render a single-line warning describing NUL bytes found by [`scan_nul_bytes`],
/// or `None` when the slice is clean. The message is deliberately blunt: copied
/// rendered text will not byte-match the file, so edits must use byte-exact tools.
pub fn nul_byte_warning_line(scan: &NulScan) -> Option<String> {
    if !scan.has_nul() {
        return None;
    }
    let offsets = scan
        .first_offsets
        .iter()
        .map(|o| o.to_string())
        .collect::<Vec<_>>()
        .join(", ");
    let plural = if scan.count == 1 { "" } else { "s" };
    let offset_suffix = if scan.count > scan.first_offsets.len() {
        format!("{offsets}, ...")
    } else {
        offsets
    };
    Some(format!(
        "WARNING: this file contains {} NUL byte{plural} (0x00) at byte offset(s) {offset_suffix}. \
         They are rendered invisibly above; copied text will NOT byte-match the file. \
         Edit this file with byte-exact tools, not rendered text.",
        scan.count
    ))
}

#[cfg(test)]
mod tests {
    use super::*;

    #[test]
    fn test_symbol_kind_display_function() {
        assert_eq!(SymbolKind::Function.to_string(), "fn");
    }

    #[test]
    fn test_is_c_header_path() {
        // `.h` headers (any case, any directory separator) are flagged.
        assert!(LanguageId::is_c_header_path("foo.h"));
        assert!(LanguageId::is_c_header_path("src/runner/flutter_window.h"));
        assert!(LanguageId::is_c_header_path("windows\\runner\\utils.h"));
        assert!(LanguageId::is_c_header_path("API.H"));
        // C++ header extensions map to Cpp directly and are NOT `.h`.
        assert!(!LanguageId::is_c_header_path("foo.hpp"));
        assert!(!LanguageId::is_c_header_path("foo.hxx"));
        assert!(!LanguageId::is_c_header_path("foo.hh"));
        // Non-headers.
        assert!(!LanguageId::is_c_header_path("foo.c"));
        assert!(!LanguageId::is_c_header_path("foo.cpp"));
        assert!(!LanguageId::is_c_header_path("README"));
        // A `.h` in the directory name, not the file, must not match.
        assert!(!LanguageId::is_c_header_path("foo.h/bar.c"));
    }

    #[test]
    fn test_symbol_kind_display_method() {
        assert_eq!(SymbolKind::Method.to_string(), "fn");
    }

    #[test]
    fn test_symbol_kind_display_class() {
        assert_eq!(SymbolKind::Class.to_string(), "class");
    }

    #[test]
    fn test_symbol_kind_display_struct() {
        assert_eq!(SymbolKind::Struct.to_string(), "struct");
    }

    #[test]
    fn test_symbol_kind_display_enum() {
        assert_eq!(SymbolKind::Enum.to_string(), "enum");
    }

    #[test]
    fn test_symbol_kind_display_interface() {
        assert_eq!(SymbolKind::Interface.to_string(), "interface");
    }

    #[test]
    fn test_symbol_kind_display_module() {
        assert_eq!(SymbolKind::Module.to_string(), "mod");
    }

    #[test]
    fn test_symbol_kind_display_constant() {
        assert_eq!(SymbolKind::Constant.to_string(), "const");
    }

    #[test]
    fn test_symbol_kind_display_variable() {
        assert_eq!(SymbolKind::Variable.to_string(), "let");
    }

    #[test]
    fn test_symbol_kind_display_type() {
        assert_eq!(SymbolKind::Type.to_string(), "type");
    }

    #[test]
    fn test_symbol_kind_display_trait() {
        assert_eq!(SymbolKind::Trait.to_string(), "trait");
    }

    #[test]
    fn test_symbol_kind_display_impl() {
        assert_eq!(SymbolKind::Impl.to_string(), "impl");
    }

    #[test]
    fn test_symbol_kind_display_other() {
        assert_eq!(SymbolKind::Other.to_string(), "other");
    }

    // --- ReferenceRecord and ReferenceKind ---

    #[test]
    fn test_reference_kind_all_variants_constructible() {
        let _call = ReferenceKind::Call;
        let _import = ReferenceKind::Import;
        let _type_usage = ReferenceKind::TypeUsage;
        let _macro_use = ReferenceKind::MacroUse;
    }

    #[test]
    fn test_reference_kind_display_call() {
        assert_eq!(ReferenceKind::Call.to_string(), "call");
    }

    #[test]
    fn test_reference_kind_display_import() {
        assert_eq!(ReferenceKind::Import.to_string(), "import");
    }

    #[test]
    fn test_reference_kind_display_type_usage() {
        assert_eq!(ReferenceKind::TypeUsage.to_string(), "type_usage");
    }

    #[test]
    fn test_reference_kind_display_macro_use() {
        assert_eq!(ReferenceKind::MacroUse.to_string(), "macro_use");
    }

    #[test]
    fn test_reference_record_construction_with_all_fields() {
        let r = ReferenceRecord {
            name: "foo".to_string(),
            qualified_name: Some("Bar::foo".to_string()),
            kind: ReferenceKind::Call,
            byte_range: (10, 20),
            line_range: (1, 1),
            enclosing_symbol_index: Some(0),
        };
        assert_eq!(r.name, "foo");
        assert_eq!(r.qualified_name.as_deref(), Some("Bar::foo"));
        assert_eq!(r.kind, ReferenceKind::Call);
        assert_eq!(r.byte_range, (10, 20));
        assert_eq!(r.line_range, (1, 1));
        assert_eq!(r.enclosing_symbol_index, Some(0));
    }

    #[test]
    fn test_reference_record_without_optional_fields() {
        let r = ReferenceRecord {
            name: "baz".to_string(),
            qualified_name: None,
            kind: ReferenceKind::Import,
            byte_range: (0, 5),
            line_range: (0, 0),
            enclosing_symbol_index: None,
        };
        assert!(r.qualified_name.is_none());
        assert!(r.enclosing_symbol_index.is_none());
    }

    #[test]
    fn test_file_processing_result_backward_compat_with_empty_refs() {
        use std::collections::HashMap;
        let result = FileProcessingResult {
            relative_path: "test.rs".to_string(),
            language: LanguageId::Rust,
            classification: FileClassification::for_code_path("test.rs"),
            outcome: FileOutcome::Processed,
            parse_diagnostic: None,
            symbols: vec![],
            byte_len: 0,
            content_hash: "abc".to_string(),
            references: vec![],
            alias_map: HashMap::new(),
        };
        assert!(result.references.is_empty());
        assert!(result.alias_map.is_empty());
    }

    #[test]
    fn test_file_classification_for_code_path_marks_code_only_by_default() {
        let classification = FileClassification::for_code_path("src/lib.rs");

        assert!(classification.is_code());
        assert!(!classification.is_text());
        assert!(!classification.is_binary());
        assert!(!classification.is_generated);
        assert!(!classification.is_test);
        assert!(!classification.is_vendor);
    }

    #[test]
    fn test_file_classification_for_code_path_marks_noise_tags_from_path() {
        let generated = FileClassification::for_code_path("src/generated/client.pb.go");
        assert!(generated.is_generated);

        let test_file = FileClassification::for_code_path("tests/parser_spec.rs");
        assert!(test_file.is_test);

        let vendor_file = FileClassification::for_code_path("node_modules/pkg/index.js");
        assert!(vendor_file.is_vendor);
    }

    #[test]
    fn test_find_enclosing_symbol_innermost_for_nested() {
        // outer: line 0..10, inner: line 3..6
        let symbols = vec![
            SymbolRecord {
                name: "outer".to_string(),
                kind: SymbolKind::Function,
                depth: 0,
                sort_order: 0,
                byte_range: (0, 100),
                line_range: (0, 10),
                doc_byte_range: None,
                item_byte_range: None,
            },
            SymbolRecord {
                name: "inner".to_string(),
                kind: SymbolKind::Function,
                depth: 1,
                sort_order: 1,
                byte_range: (30, 60),
                line_range: (3, 6),
                doc_byte_range: None,
                item_byte_range: None,
            },
        ];
        // Reference at line 4 is inside both - should return inner (index 1)
        let idx = find_enclosing_symbol(&symbols, 4);
        assert_eq!(idx, Some(1), "should return innermost enclosing symbol");
    }

    #[test]
    fn test_find_enclosing_symbol_none_at_module_level() {
        let symbols = vec![SymbolRecord {
            name: "foo".to_string(),
            kind: SymbolKind::Function,
            depth: 0,
            sort_order: 0,
            byte_range: (50, 100),
            line_range: (5, 10),
            doc_byte_range: None,
            item_byte_range: None,
        }];
        // Reference at line 0 is not inside any symbol
        let idx = find_enclosing_symbol(&symbols, 0);
        assert_eq!(idx, None, "should return None when not inside any symbol");
    }

    #[test]
    fn test_symbol_record_item_helpers_fall_back_to_doc_range() {
        let symbol = SymbolRecord {
            name: "target".to_string(),
            kind: SymbolKind::Function,
            depth: 0,
            sort_order: 0,
            byte_range: (20, 40),
            line_range: (1, 3),
            doc_byte_range: Some((10, 20)),
            item_byte_range: None,
        };

        assert_eq!(symbol.item_start(), 10);
        assert_eq!(symbol.item_end(), 40);
        assert_eq!(symbol.item_range(), (10, 40));
        assert_eq!(symbol.core_range(), (20, 40));
    }

    #[test]
    fn test_symbol_record_item_helpers_prefer_item_range() {
        let symbol = SymbolRecord {
            name: "target".to_string(),
            kind: SymbolKind::Function,
            depth: 0,
            sort_order: 0,
            byte_range: (20, 40),
            line_range: (1, 3),
            doc_byte_range: Some((10, 20)),
            item_byte_range: Some((5, 44)),
        };

        assert_eq!(symbol.item_start(), 5);
        assert_eq!(symbol.item_end(), 44);
        assert_eq!(symbol.item_range(), (5, 44));
        assert_eq!(symbol.core_range(), (20, 40));
    }

    #[test]
    fn test_admission_tier_variants() {
        let t1 = AdmissionTier::Normal;
        let t2 = AdmissionTier::MetadataOnly;
        let t3 = AdmissionTier::HardSkip;
        assert_ne!(t1, t2);
        assert_ne!(t2, t3);
        assert_ne!(t1, t3);
    }

    #[test]
    fn test_extension_is_denylisted() {
        assert!(is_denylisted_extension("safetensors"));
        assert!(is_denylisted_extension("ckpt"));
        assert!(is_denylisted_extension("zip"));
        assert!(is_denylisted_extension("mp4"));
        assert!(is_denylisted_extension("woff2"));
        assert!(is_denylisted_extension("png"));
        assert!(is_denylisted_extension("bin"));
    }

    #[test]
    fn test_extension_not_denylisted() {
        assert!(!is_denylisted_extension("rs"));
        assert!(!is_denylisted_extension("ts"));
        assert!(!is_denylisted_extension("json"));
        assert!(!is_denylisted_extension("svg")); // SVG intentionally NOT denylisted
        assert!(!is_denylisted_extension("md"));
        assert!(!is_denylisted_extension("toml"));
    }

    #[test]
    fn test_new_executable_extensions_denylisted() {
        assert!(is_denylisted_extension("exe"));
        assert!(is_denylisted_extension("dll"));
        assert!(is_denylisted_extension("so"));
        assert!(is_denylisted_extension("dylib"));
        assert!(is_denylisted_extension("class"));
    }

    #[test]
    fn test_denylist_case_insensitive() {
        assert!(is_denylisted_extension("DLL"));
        assert!(is_denylisted_extension("So"));
        assert!(is_denylisted_extension("EXE"));
        assert!(is_denylisted_extension("Dylib"));
        assert!(is_denylisted_extension("CLASS"));
    }

    #[test]
    fn test_skipped_file_creation() {
        let decision =
            AdmissionDecision::skip(AdmissionTier::MetadataOnly, SkipReason::DenylistedExtension);
        let sf = SkippedFile {
            path: "model.safetensors".into(),
            size: 4_200_000_000,
            extension: Some("safetensors".into()),
            decision,
        };
        assert_eq!(sf.tier(), AdmissionTier::MetadataOnly);
        assert_eq!(sf.reason(), Some(SkipReason::DenylistedExtension));
    }

    #[test]
    fn test_scan_nul_bytes_clean_slice_reports_none() {
        let scan = scan_nul_bytes(b"fn main() {}\n");
        assert!(!scan.has_nul());
        assert_eq!(scan.count, 0);
        assert!(scan.first_offsets.is_empty());
        assert_eq!(nul_byte_warning_line(&scan), None);
    }

    #[test]
    fn test_scan_nul_bytes_empty_slice_reports_none() {
        let scan = scan_nul_bytes(b"");
        assert!(!scan.has_nul());
        assert_eq!(nul_byte_warning_line(&scan), None);
    }

    #[test]
    fn test_scan_nul_bytes_single_offset_and_singular_warning() {
        let scan = scan_nul_bytes(b"ab\0cd");
        assert_eq!(scan.count, 1);
        assert_eq!(scan.first_offsets, vec![2]);
        let warning = nul_byte_warning_line(&scan).expect("NUL present");
        assert!(warning.contains("1 NUL byte (0x00)"), "warning: {warning}");
        assert!(
            warning.contains("byte offset(s) 2."),
            "single offset, no ellipsis: {warning}"
        );
        assert!(warning.contains("byte-exact tools"));
    }

    #[test]
    fn test_scan_nul_bytes_plural_warning() {
        let scan = scan_nul_bytes(b"\0a\0b");
        assert_eq!(scan.count, 2);
        assert_eq!(scan.first_offsets, vec![0, 2]);
        let warning = nul_byte_warning_line(&scan).expect("NUL present");
        assert!(warning.contains("2 NUL bytes (0x00)"), "warning: {warning}");
        assert!(
            warning.contains("byte offset(s) 0, 2."),
            "warning: {warning}"
        );
    }

    #[test]
    fn test_scan_nul_bytes_caps_offsets_and_adds_ellipsis() {
        // 7 NUL bytes; only the first NUL_OFFSET_SAMPLE_LIMIT (5) are enumerated.
        let scan = scan_nul_bytes(b"\0\0\0\0\0\0\0");
        assert_eq!(scan.count, 7);
        assert_eq!(scan.first_offsets.len(), NUL_OFFSET_SAMPLE_LIMIT);
        assert_eq!(scan.first_offsets, vec![0, 1, 2, 3, 4]);
        let warning = nul_byte_warning_line(&scan).expect("NUL present");
        assert!(warning.contains("7 NUL bytes"), "warning: {warning}");
        assert!(
            warning.contains("0, 1, 2, 3, 4, ..."),
            "more NULs than sampled => ellipsis: {warning}"
        );
    }
}