gitnexus 1.6.6-rc.156 → 1.6.6-rc.158
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/dist/cli/analyze.js +16 -0
- package/dist/core/ingestion/call-processor.d.ts +2 -13
- package/dist/core/ingestion/call-processor.js +2 -79
- package/dist/core/ingestion/export-detection.d.ts +2 -1
- package/dist/core/ingestion/export-detection.js +2 -1
- package/dist/core/ingestion/field-extractors/configs/dart.d.ts +0 -6
- package/dist/core/ingestion/field-extractors/configs/dart.js +61 -40
- package/dist/core/ingestion/field-extractors/configs/jvm.js +43 -4
- package/dist/core/ingestion/field-extractors/configs/swift.d.ts +12 -1
- package/dist/core/ingestion/field-extractors/configs/swift.js +30 -5
- package/dist/core/ingestion/filesystem-walker.d.ts +0 -9
- package/dist/core/ingestion/filesystem-walker.js +0 -11
- package/dist/core/ingestion/languages/c/import-decomposer.d.ts +9 -0
- package/dist/core/ingestion/languages/c/import-decomposer.js +15 -0
- package/dist/core/ingestion/languages/cpp/import-decomposer.d.ts +7 -0
- package/dist/core/ingestion/languages/cpp/import-decomposer.js +12 -0
- package/dist/core/ingestion/languages/dart/query.js +40 -0
- package/dist/core/ingestion/languages/kotlin/captures.js +105 -0
- package/dist/core/ingestion/languages/kotlin/query.js +30 -0
- package/dist/core/ingestion/languages/python/captures.js +4 -3
- package/dist/core/ingestion/method-extractors/configs/jvm.js +25 -10
- package/dist/core/ingestion/method-extractors/configs/swift.js +11 -4
- package/dist/core/ingestion/pipeline-phases/parse-impl.js +0 -10
- package/dist/core/ingestion/scope-resolution/graph-bridge/ids.d.ts +1 -0
- package/dist/core/ingestion/scope-resolution/graph-bridge/ids.js +10 -0
- package/dist/core/ingestion/scope-resolution/graph-bridge/node-lookup.js +15 -0
- package/dist/core/ingestion/tree-sitter-queries.d.ts +4 -4
- package/dist/core/ingestion/tree-sitter-queries.js +125 -6
- package/dist/core/ingestion/utils/ast-helpers.d.ts +36 -0
- package/dist/core/ingestion/utils/ast-helpers.js +68 -0
- package/dist/core/ingestion/variable-extractors/configs/c-cpp.js +49 -0
- package/dist/core/ingestion/variable-extractors/configs/dart.js +76 -53
- package/dist/core/ingestion/variable-extractors/configs/jvm.js +35 -8
- package/dist/core/ingestion/workers/parse-worker.js +46 -4
- package/dist/core/ingestion/workers/worker-pool.js +25 -2
- package/dist/core/lbug/lbug-adapter.js +29 -10
- package/dist/core/lbug/pool-adapter.js +38 -7
- package/dist/core/lbug/query-result-utils.d.ts +19 -0
- package/dist/core/lbug/query-result-utils.js +28 -0
- package/dist/mcp/local/local-backend.js +14 -1
- package/dist/types/pipeline.d.ts +6 -3
- package/package.json +1 -2
- package/dist/core/ingestion/ast-cache.d.ts +0 -26
- package/dist/core/ingestion/ast-cache.js +0 -48
- package/dist/core/ingestion/pipeline-phases/orm-extraction.d.ts +0 -22
- package/dist/core/ingestion/pipeline-phases/orm-extraction.js +0 -92
package/dist/cli/analyze.js
CHANGED
|
@@ -41,6 +41,22 @@ const writeFatalToStderr = (label, err) => {
|
|
|
41
41
|
realStderrWrite(`\n ${label}: ${message}\n`);
|
|
42
42
|
if (isErr && err.stack)
|
|
43
43
|
realStderrWrite(`${err.stack}\n`);
|
|
44
|
+
// Walk and print the `cause` chain. The phase runner wraps the underlying
|
|
45
|
+
// failure as `new Error("Phase 'X' failed: …", { cause })`, so the original
|
|
46
|
+
// error (e.g. a WorkerPoolDispatchError carrying the worker-side stack from
|
|
47
|
+
// #2068) is only reachable via `.cause`. Without this the user sees the
|
|
48
|
+
// wrapper's main-thread stack and never the real frame. `cause.stack` already
|
|
49
|
+
// begins with the cause's message, so we print the stack alone (not message +
|
|
50
|
+
// stack) to avoid repeating it. Depth-bounded so a cyclic `cause` can't loop
|
|
51
|
+
// (the phase runner wraps one level; the bound leaves headroom for future
|
|
52
|
+
// nesting); uses realStderrWrite so the redirected console.error's ANSI
|
|
53
|
+
// clear-line wrapping can't erase it (#1169).
|
|
54
|
+
const MAX_CAUSE_DEPTH = 5;
|
|
55
|
+
let cause = isErr ? err.cause : undefined;
|
|
56
|
+
for (let depth = 0; depth < MAX_CAUSE_DEPTH && cause instanceof Error; depth++) {
|
|
57
|
+
realStderrWrite(`\n Caused by: ${cause.stack ?? cause.message}\n`);
|
|
58
|
+
cause = cause.cause;
|
|
59
|
+
}
|
|
44
60
|
};
|
|
45
61
|
let fatalHandlersInstalled = false;
|
|
46
62
|
/**
|
|
@@ -9,14 +9,12 @@
|
|
|
9
9
|
*
|
|
10
10
|
* - `processRoutesFromExtracted` — CALLS edges from framework routes
|
|
11
11
|
* (e.g. Laravel) to their controller methods.
|
|
12
|
-
* - `processNextjsFetchRoutes` / `
|
|
13
|
-
*
|
|
14
|
-
* Next.js Route nodes.
|
|
12
|
+
* - `processNextjsFetchRoutes` / `extractConsumerAccessedKeys` — FETCHES edges
|
|
13
|
+
* from `fetch()` calls to Next.js Route nodes.
|
|
15
14
|
* - `buildExportedTypeMapFromGraph` — exported symbol → return/declared type
|
|
16
15
|
* map, consumed by the cross-file enrichment pass.
|
|
17
16
|
*/
|
|
18
17
|
import { KnowledgeGraph } from '../graph/types.js';
|
|
19
|
-
import { ASTCache } from './ast-cache.js';
|
|
20
18
|
import type { SemanticModel, SymbolTableReader } from './model/index.js';
|
|
21
19
|
import type { ExtractedRoute, ExtractedFetchCall } from './workers/parse-worker.js';
|
|
22
20
|
/** Per-file resolved type bindings for exported symbols.
|
|
@@ -76,12 +74,3 @@ export declare const extractConsumerAccessedKeys: (content: string) => string[];
|
|
|
76
74
|
*/
|
|
77
75
|
export declare const processNextjsFetchRoutes: (graph: KnowledgeGraph, fetchCalls: ExtractedFetchCall[], routeRegistry: Map<string, string>, // routeURL → handlerFilePath
|
|
78
76
|
consumerContents?: Map<string, string>) => void;
|
|
79
|
-
/**
|
|
80
|
-
* Extract fetch() calls from source files (sequential path).
|
|
81
|
-
* Workers handle this via tree-sitter captures in parse-worker; this function
|
|
82
|
-
* provides the same extraction for the sequential fallback path.
|
|
83
|
-
*/
|
|
84
|
-
export declare const extractFetchCallsFromFiles: (files: {
|
|
85
|
-
path: string;
|
|
86
|
-
content: string;
|
|
87
|
-
}[], astCache: ASTCache) => Promise<ExtractedFetchCall[]>;
|
|
@@ -9,20 +9,13 @@
|
|
|
9
9
|
*
|
|
10
10
|
* - `processRoutesFromExtracted` — CALLS edges from framework routes
|
|
11
11
|
* (e.g. Laravel) to their controller methods.
|
|
12
|
-
* - `processNextjsFetchRoutes` / `
|
|
13
|
-
*
|
|
14
|
-
* Next.js Route nodes.
|
|
12
|
+
* - `processNextjsFetchRoutes` / `extractConsumerAccessedKeys` — FETCHES edges
|
|
13
|
+
* from `fetch()` calls to Next.js Route nodes.
|
|
15
14
|
* - `buildExportedTypeMapFromGraph` — exported symbol → return/declared type
|
|
16
15
|
* map, consumed by the cross-file enrichment pass.
|
|
17
16
|
*/
|
|
18
|
-
import Parser from 'tree-sitter';
|
|
19
|
-
import { isLanguageAvailable, loadParser, loadLanguage } from '../tree-sitter/parser-loader.js';
|
|
20
|
-
import { getProvider } from './languages/index.js';
|
|
21
17
|
import { generateId } from '../../lib/utils.js';
|
|
22
|
-
import { getLanguageFromFilename } from '../../_shared/index.js';
|
|
23
18
|
import { yieldToEventLoop } from './utils/event-loop.js';
|
|
24
|
-
import { parseSourceSafe } from '../tree-sitter/safe-parse.js';
|
|
25
|
-
import { getTreeSitterBufferSize } from './constants.js';
|
|
26
19
|
import { normalizeFetchURL, routeMatches } from './route-extractors/nextjs.js';
|
|
27
20
|
import { extractReturnTypeName } from './type-extractors/shared.js';
|
|
28
21
|
const MAX_EXPORTS_PER_FILE = 500;
|
|
@@ -408,73 +401,3 @@ consumerContents) => {
|
|
|
408
401
|
}
|
|
409
402
|
}
|
|
410
403
|
};
|
|
411
|
-
/**
|
|
412
|
-
* Extract fetch() calls from source files (sequential path).
|
|
413
|
-
* Workers handle this via tree-sitter captures in parse-worker; this function
|
|
414
|
-
* provides the same extraction for the sequential fallback path.
|
|
415
|
-
*/
|
|
416
|
-
export const extractFetchCallsFromFiles = async (files, astCache) => {
|
|
417
|
-
const parser = await loadParser();
|
|
418
|
-
const result = [];
|
|
419
|
-
for (const file of files) {
|
|
420
|
-
const language = getLanguageFromFilename(file.path);
|
|
421
|
-
if (!language)
|
|
422
|
-
continue;
|
|
423
|
-
if (!isLanguageAvailable(language))
|
|
424
|
-
continue;
|
|
425
|
-
const provider = getProvider(language);
|
|
426
|
-
const queryStr = provider.treeSitterQueries;
|
|
427
|
-
if (!queryStr)
|
|
428
|
-
continue;
|
|
429
|
-
await loadLanguage(language, file.path);
|
|
430
|
-
let tree = astCache.get(file.path);
|
|
431
|
-
if (!tree) {
|
|
432
|
-
const parseContent = provider.preprocessSource?.(file.content, file.path) ?? file.content;
|
|
433
|
-
try {
|
|
434
|
-
tree = parseSourceSafe(parser, parseContent, undefined, {
|
|
435
|
-
bufferSize: getTreeSitterBufferSize(parseContent),
|
|
436
|
-
});
|
|
437
|
-
}
|
|
438
|
-
catch {
|
|
439
|
-
continue;
|
|
440
|
-
}
|
|
441
|
-
astCache.set(file.path, tree);
|
|
442
|
-
}
|
|
443
|
-
let matches;
|
|
444
|
-
try {
|
|
445
|
-
const lang = parser.getLanguage();
|
|
446
|
-
const query = new Parser.Query(lang, queryStr);
|
|
447
|
-
matches = query.matches(tree.rootNode);
|
|
448
|
-
}
|
|
449
|
-
catch {
|
|
450
|
-
continue;
|
|
451
|
-
}
|
|
452
|
-
for (const match of matches) {
|
|
453
|
-
const captureMap = {};
|
|
454
|
-
match.captures.forEach((c) => (captureMap[c.name] = c.node));
|
|
455
|
-
if (captureMap['route.fetch']) {
|
|
456
|
-
const urlNode = captureMap['route.url'] ?? captureMap['route.template_url'];
|
|
457
|
-
if (urlNode) {
|
|
458
|
-
result.push({
|
|
459
|
-
filePath: file.path,
|
|
460
|
-
fetchURL: urlNode.text,
|
|
461
|
-
lineNumber: captureMap['route.fetch'].startPosition.row,
|
|
462
|
-
});
|
|
463
|
-
}
|
|
464
|
-
}
|
|
465
|
-
else if (captureMap['http_client'] && captureMap['http_client.url']) {
|
|
466
|
-
const method = captureMap['http_client.method']?.text;
|
|
467
|
-
const url = captureMap['http_client.url'].text;
|
|
468
|
-
const HTTP_CLIENT_ONLY = new Set(['head', 'options', 'request', 'ajax']);
|
|
469
|
-
if (method && HTTP_CLIENT_ONLY.has(method) && url.startsWith('/')) {
|
|
470
|
-
result.push({
|
|
471
|
-
filePath: file.path,
|
|
472
|
-
fetchURL: url,
|
|
473
|
-
lineNumber: captureMap['http_client'].startPosition.row,
|
|
474
|
-
});
|
|
475
|
-
}
|
|
476
|
-
}
|
|
477
|
-
}
|
|
478
|
-
}
|
|
479
|
-
return result;
|
|
480
|
-
};
|
|
@@ -4,7 +4,8 @@
|
|
|
4
4
|
* Determines whether a symbol (function, class, etc.) is exported/public
|
|
5
5
|
* in its language. This is a pure function — safe for use in worker threads.
|
|
6
6
|
*
|
|
7
|
-
*
|
|
7
|
+
* Used by the language providers during worker parsing (parse-worker.ts) — the
|
|
8
|
+
* sole parse path. (Sequential parsing was removed.)
|
|
8
9
|
*/
|
|
9
10
|
import { type SyntaxNode } from './utils/ast-helpers.js';
|
|
10
11
|
/** Handler type: given a node and symbol name, return true if the symbol is exported/public. */
|
|
@@ -4,7 +4,8 @@
|
|
|
4
4
|
* Determines whether a symbol (function, class, etc.) is exported/public
|
|
5
5
|
* in its language. This is a pure function — safe for use in worker threads.
|
|
6
6
|
*
|
|
7
|
-
*
|
|
7
|
+
* Used by the language providers during worker parsing (parse-worker.ts) — the
|
|
8
|
+
* sole parse path. (Sequential parsing was removed.)
|
|
8
9
|
*/
|
|
9
10
|
import { findSiblingChild } from './utils/ast-helpers.js';
|
|
10
11
|
// ============================================================================
|
|
@@ -1,8 +1,2 @@
|
|
|
1
1
|
import type { FieldExtractionConfig } from '../generic.js';
|
|
2
|
-
/**
|
|
3
|
-
* Dart field extraction config.
|
|
4
|
-
*
|
|
5
|
-
* Dart class fields appear as declaration nodes inside class_body.
|
|
6
|
-
* Visibility is convention-based: underscore prefix = private.
|
|
7
|
-
*/
|
|
8
2
|
export declare const dartConfig: FieldExtractionConfig;
|
|
@@ -5,41 +5,68 @@ import { extractSimpleTypeName } from '../../type-extractors/shared.js';
|
|
|
5
5
|
/**
|
|
6
6
|
* Dart field extraction config.
|
|
7
7
|
*
|
|
8
|
-
* Dart class fields appear as declaration nodes inside class_body
|
|
8
|
+
* Dart class fields appear as `declaration` nodes inside `class_body`.
|
|
9
|
+
* Two shapes carry the field name(s):
|
|
10
|
+
* - instance / plain fields → `initialized_identifier_list`
|
|
11
|
+
* (`int z = 0;`, `int a = 1, b = 2;`)
|
|
12
|
+
* - `static const` / `static final` / `const` fields → `static_final_declaration_list`
|
|
13
|
+
* (`static const a = 1;`, `static final String b = 'x', c = 'y';`)
|
|
14
|
+
* Both shapes may declare SEVERAL fields in one declaration, so name extraction
|
|
15
|
+
* is multi-name (`extractNames`). The structure query (`DART_QUERIES`) emits one
|
|
16
|
+
* `@definition.property` per name for both shapes; this config enriches each.
|
|
17
|
+
*
|
|
9
18
|
* Visibility is convention-based: underscore prefix = private.
|
|
10
19
|
*/
|
|
20
|
+
/** All field names declared by a `declaration` node, across both Dart shapes. */
|
|
21
|
+
function extractDartFieldNames(node) {
|
|
22
|
+
const names = [];
|
|
23
|
+
for (let i = 0; i < node.namedChildCount; i++) {
|
|
24
|
+
const child = node.namedChild(i);
|
|
25
|
+
if (!child)
|
|
26
|
+
continue;
|
|
27
|
+
// instance / plain fields: initialized_identifier_list > initialized_identifier > identifier
|
|
28
|
+
if (child.type === 'initialized_identifier_list') {
|
|
29
|
+
for (let j = 0; j < child.namedChildCount; j++) {
|
|
30
|
+
const init = child.namedChild(j);
|
|
31
|
+
if (init?.type === 'initialized_identifier') {
|
|
32
|
+
const ident = init.firstNamedChild;
|
|
33
|
+
if (ident?.type === 'identifier')
|
|
34
|
+
names.push(ident.text);
|
|
35
|
+
}
|
|
36
|
+
}
|
|
37
|
+
}
|
|
38
|
+
// static const / final fields: static_final_declaration_list > static_final_declaration > identifier
|
|
39
|
+
if (child.type === 'static_final_declaration_list') {
|
|
40
|
+
for (let j = 0; j < child.namedChildCount; j++) {
|
|
41
|
+
const decl = child.namedChild(j);
|
|
42
|
+
if (decl?.type === 'static_final_declaration') {
|
|
43
|
+
const ident = decl.firstNamedChild;
|
|
44
|
+
if (ident?.type === 'identifier')
|
|
45
|
+
names.push(ident.text);
|
|
46
|
+
}
|
|
47
|
+
}
|
|
48
|
+
}
|
|
49
|
+
}
|
|
50
|
+
return names;
|
|
51
|
+
}
|
|
11
52
|
export const dartConfig = {
|
|
12
53
|
language: SupportedLanguages.Dart,
|
|
13
54
|
typeDeclarationNodes: ['class_definition'],
|
|
14
55
|
fieldNodeTypes: ['declaration'],
|
|
15
56
|
bodyNodeTypes: ['class_body'],
|
|
16
57
|
defaultVisibility: 'public',
|
|
58
|
+
// One AST `declaration` node may declare several fields (`int a, b;`,
|
|
59
|
+
// `static final String b = 'x', c = 'y';`), so use the multi-name path.
|
|
17
60
|
extractName(node) {
|
|
18
|
-
|
|
19
|
-
|
|
20
|
-
|
|
21
|
-
|
|
22
|
-
for (let j = 0; j < child.namedChildCount; j++) {
|
|
23
|
-
const init = child.namedChild(j);
|
|
24
|
-
if (init?.type === 'initialized_identifier') {
|
|
25
|
-
const ident = init.firstNamedChild;
|
|
26
|
-
if (ident?.type === 'identifier')
|
|
27
|
-
return ident.text;
|
|
28
|
-
}
|
|
29
|
-
}
|
|
30
|
-
}
|
|
31
|
-
if (child?.type === 'initialized_identifier') {
|
|
32
|
-
const ident = child.firstNamedChild;
|
|
33
|
-
if (ident?.type === 'identifier')
|
|
34
|
-
return ident.text;
|
|
35
|
-
}
|
|
36
|
-
}
|
|
37
|
-
// fallback: look for direct identifier
|
|
38
|
-
const name = node.childForFieldName('name');
|
|
39
|
-
return name?.text;
|
|
61
|
+
return extractDartFieldNames(node)[0];
|
|
62
|
+
},
|
|
63
|
+
extractNames(node) {
|
|
64
|
+
return extractDartFieldNames(node);
|
|
40
65
|
},
|
|
41
66
|
extractType(node) {
|
|
42
|
-
// declaration > type_identifier (
|
|
67
|
+
// declaration > type_identifier (the type annotation, present for both the
|
|
68
|
+
// instance-field shape and `static final String b = …`). `static const a = 1;`
|
|
69
|
+
// has no annotation → undefined (untyped).
|
|
43
70
|
for (let i = 0; i < node.namedChildCount; i++) {
|
|
44
71
|
const child = node.namedChild(i);
|
|
45
72
|
if (child && (child.type === 'type_identifier' || child.type === 'function_type')) {
|
|
@@ -48,28 +75,22 @@ export const dartConfig = {
|
|
|
48
75
|
}
|
|
49
76
|
return undefined;
|
|
50
77
|
},
|
|
78
|
+
// Per-name: Dart convention is underscore-prefixed = private. A single
|
|
79
|
+
// declaration can mix visibilities (`static const _p = 1, q = 2;`), so the
|
|
80
|
+
// decision is keyed on the individual field name.
|
|
81
|
+
extractVisibilityForName(_node, name) {
|
|
82
|
+
return name.startsWith('_') ? 'private' : 'public';
|
|
83
|
+
},
|
|
51
84
|
extractVisibility(node) {
|
|
52
|
-
|
|
53
|
-
|
|
54
|
-
for (let i = 0; i < node.namedChildCount; i++) {
|
|
55
|
-
const child = node.namedChild(i);
|
|
56
|
-
if (child?.type === 'initialized_identifier_list') {
|
|
57
|
-
for (let j = 0; j < child.namedChildCount; j++) {
|
|
58
|
-
const init = child.namedChild(j);
|
|
59
|
-
if (init?.type === 'initialized_identifier') {
|
|
60
|
-
const ident = init.firstNamedChild;
|
|
61
|
-
if (ident?.text?.startsWith('_'))
|
|
62
|
-
return 'private';
|
|
63
|
-
}
|
|
64
|
-
}
|
|
65
|
-
}
|
|
66
|
-
}
|
|
67
|
-
return 'public';
|
|
85
|
+
const first = extractDartFieldNames(node)[0];
|
|
86
|
+
return first?.startsWith('_') ? 'private' : 'public';
|
|
68
87
|
},
|
|
69
88
|
isStatic(node) {
|
|
70
89
|
return hasKeyword(node, 'static');
|
|
71
90
|
},
|
|
72
91
|
isReadonly(node) {
|
|
92
|
+
// `final` / `const` (both `final_builtin`/`const_builtin` nodes whose text
|
|
93
|
+
// is `final`/`const`) are read-only.
|
|
73
94
|
return hasKeyword(node, 'final') || hasKeyword(node, 'const');
|
|
74
95
|
},
|
|
75
96
|
};
|
|
@@ -60,12 +60,49 @@ export const javaConfig = {
|
|
|
60
60
|
// Kotlin
|
|
61
61
|
// ---------------------------------------------------------------------------
|
|
62
62
|
const KOTLIN_VIS = new Set(['public', 'private', 'protected', 'internal']);
|
|
63
|
+
/** A property_declaration is a companion-object member when its nearest
|
|
64
|
+
* class-body ancestor is the body of a companion_object (F52, issue #1919).
|
|
65
|
+
* Companion members are addressed statically through the enclosing class
|
|
66
|
+
* (`C.TAG`), so they are marked static. */
|
|
67
|
+
function isInsideKotlinCompanion(node) {
|
|
68
|
+
for (let cur = node.parent; cur !== null; cur = cur.parent) {
|
|
69
|
+
if (cur.type === 'class_body')
|
|
70
|
+
return cur.parent?.type === 'companion_object';
|
|
71
|
+
if (cur.type === 'companion_object')
|
|
72
|
+
return true;
|
|
73
|
+
}
|
|
74
|
+
return false;
|
|
75
|
+
}
|
|
63
76
|
export const kotlinConfig = {
|
|
64
77
|
language: SupportedLanguages.Kotlin,
|
|
65
|
-
|
|
78
|
+
// F52: include companion_object so a companion property's innermost
|
|
79
|
+
// class-container owner (findEnclosingClassNode returns the companion_object)
|
|
80
|
+
// is recognized as a type declaration and its nested class_body is walked.
|
|
81
|
+
// The structure query already creates the Property node and owns it on the
|
|
82
|
+
// ENCLOSING class for anonymous companions / on the named companion Class —
|
|
83
|
+
// this entry only drives field-metadata enrichment, so it does NOT change
|
|
84
|
+
// ownership or emit a second node (no double-count).
|
|
85
|
+
typeDeclarationNodes: ['class_declaration', 'object_declaration', 'companion_object'],
|
|
66
86
|
fieldNodeTypes: ['property_declaration'],
|
|
67
87
|
bodyNodeTypes: ['class_body'],
|
|
68
88
|
defaultVisibility: 'public',
|
|
89
|
+
// F52: an anonymous `companion object { ... }` has no name child, so the
|
|
90
|
+
// generic factory's `childForFieldName('name')` owner lookup is empty and
|
|
91
|
+
// `extract()` would bail before walking the body. Supply a stable owner
|
|
92
|
+
// name (the named companion's identifier, else "Companion") so the body IS
|
|
93
|
+
// walked; the resulting FieldInfo map is keyed by field NAME only, so the
|
|
94
|
+
// owner name does not affect which Property node gets enriched.
|
|
95
|
+
extractOwnerName(node) {
|
|
96
|
+
const typeIdentifierText = node.namedChildren.find((c) => c.type === 'type_identifier')?.text;
|
|
97
|
+
if (node.type === 'companion_object') {
|
|
98
|
+
// Anonymous companions have no type_identifier — fall back to "Companion".
|
|
99
|
+
return typeIdentifierText ?? 'Companion';
|
|
100
|
+
}
|
|
101
|
+
const name = node.childForFieldName('name');
|
|
102
|
+
if (name)
|
|
103
|
+
return name.text;
|
|
104
|
+
return typeIdentifierText;
|
|
105
|
+
},
|
|
69
106
|
extractName(node) {
|
|
70
107
|
// property_declaration > variable_declaration > simple_identifier
|
|
71
108
|
for (let i = 0; i < node.namedChildCount; i++) {
|
|
@@ -107,9 +144,11 @@ export const kotlinConfig = {
|
|
|
107
144
|
extractVisibility(node) {
|
|
108
145
|
return findVisibility(node, KOTLIN_VIS, 'public', 'modifiers');
|
|
109
146
|
},
|
|
110
|
-
isStatic(
|
|
111
|
-
// Kotlin
|
|
112
|
-
|
|
147
|
+
isStatic(node) {
|
|
148
|
+
// Kotlin has no `static`, but companion-object members are accessed
|
|
149
|
+
// statically through the enclosing class (`C.TAG`) — mark them static
|
|
150
|
+
// so the field metadata reflects that (F52).
|
|
151
|
+
return isInsideKotlinCompanion(node);
|
|
113
152
|
},
|
|
114
153
|
isReadonly(node) {
|
|
115
154
|
// 'val' = readonly, 'var' = mutable
|
|
@@ -2,7 +2,18 @@ import type { FieldExtractionConfig } from '../generic.js';
|
|
|
2
2
|
/**
|
|
3
3
|
* Swift field extraction config.
|
|
4
4
|
*
|
|
5
|
-
* Handles property_declaration inside class_body / protocol_body
|
|
5
|
+
* Handles property_declaration inside class_body / protocol_body and
|
|
6
|
+
* protocol_property_declaration inside protocol_body (F75 — protocol property
|
|
7
|
+
* requirements like "var title: String { get }").
|
|
8
|
+
*
|
|
6
9
|
* tree-sitter-swift uses property_declaration for stored/computed properties.
|
|
10
|
+
* A protocol property requirement parses to its own node type,
|
|
11
|
+
* protocol_property_declaration, whose name lives in a "name:" pattern field
|
|
12
|
+
* (pattern > value_binding_pattern + simple_identifier(bound_identifier)), its
|
|
13
|
+
* type in a sibling type_annotation, and its "{ get }" / "{ get set }" in a
|
|
14
|
+
* protocol_property_requirements child. Note: Swift reuses the "name:" field
|
|
15
|
+
* across many positions (func name, every parameter label, parameter/return
|
|
16
|
+
* type), so the name is synthesized from the simple_identifier inside the
|
|
17
|
+
* pattern rather than read blindly off "name:".
|
|
7
18
|
*/
|
|
8
19
|
export declare const swiftConfig: FieldExtractionConfig;
|
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
// gitnexus/src/core/ingestion/field-extractors/configs/swift.ts
|
|
2
2
|
import { SupportedLanguages } from '../../../../_shared/index.js';
|
|
3
|
-
import { hasKeyword, findVisibility } from './helpers.js';
|
|
3
|
+
import { hasKeyword, hasModifier, findVisibility } from './helpers.js';
|
|
4
4
|
import { extractSimpleTypeName } from '../../type-extractors/shared.js';
|
|
5
5
|
const SWIFT_VIS = new Set([
|
|
6
6
|
'public',
|
|
@@ -12,17 +12,32 @@ const SWIFT_VIS = new Set([
|
|
|
12
12
|
/**
|
|
13
13
|
* Swift field extraction config.
|
|
14
14
|
*
|
|
15
|
-
* Handles property_declaration inside class_body / protocol_body
|
|
15
|
+
* Handles property_declaration inside class_body / protocol_body and
|
|
16
|
+
* protocol_property_declaration inside protocol_body (F75 — protocol property
|
|
17
|
+
* requirements like "var title: String { get }").
|
|
18
|
+
*
|
|
16
19
|
* tree-sitter-swift uses property_declaration for stored/computed properties.
|
|
20
|
+
* A protocol property requirement parses to its own node type,
|
|
21
|
+
* protocol_property_declaration, whose name lives in a "name:" pattern field
|
|
22
|
+
* (pattern > value_binding_pattern + simple_identifier(bound_identifier)), its
|
|
23
|
+
* type in a sibling type_annotation, and its "{ get }" / "{ get set }" in a
|
|
24
|
+
* protocol_property_requirements child. Note: Swift reuses the "name:" field
|
|
25
|
+
* across many positions (func name, every parameter label, parameter/return
|
|
26
|
+
* type), so the name is synthesized from the simple_identifier inside the
|
|
27
|
+
* pattern rather than read blindly off "name:".
|
|
17
28
|
*/
|
|
18
29
|
export const swiftConfig = {
|
|
19
30
|
language: SupportedLanguages.Swift,
|
|
20
31
|
typeDeclarationNodes: ['class_declaration', 'protocol_declaration'],
|
|
21
|
-
fieldNodeTypes: ['property_declaration'],
|
|
32
|
+
fieldNodeTypes: ['property_declaration', 'protocol_property_declaration'],
|
|
22
33
|
bodyNodeTypes: ['class_body', 'protocol_body'],
|
|
23
34
|
defaultVisibility: 'internal',
|
|
24
35
|
extractName(node) {
|
|
25
|
-
// property_declaration > pattern > simple_identifier
|
|
36
|
+
// property_declaration > pattern > simple_identifier, and
|
|
37
|
+
// protocol_property_declaration > name: (pattern ... simple_identifier).
|
|
38
|
+
// For protocol_property_declaration the pattern wraps a leading
|
|
39
|
+
// value_binding_pattern ("var") plus the simple_identifier — the loop
|
|
40
|
+
// below skips the binding keyword and returns the identifier.
|
|
26
41
|
for (let i = 0; i < node.namedChildCount; i++) {
|
|
27
42
|
const child = node.namedChild(i);
|
|
28
43
|
if (child?.type === 'pattern') {
|
|
@@ -56,7 +71,17 @@ export const swiftConfig = {
|
|
|
56
71
|
return findVisibility(node, SWIFT_VIS, 'internal', 'modifiers');
|
|
57
72
|
},
|
|
58
73
|
isStatic(node) {
|
|
59
|
-
|
|
74
|
+
// `static`/`class` (type-level) modifiers live inside a `modifiers`
|
|
75
|
+
// wrapper for both property_declaration and protocol_property_declaration
|
|
76
|
+
// (e.g. `static var shared: P { get }`), so check the wrapper too.
|
|
77
|
+
// `hasKeyword` compares each direct child by `.text` equality: it matches a
|
|
78
|
+
// single-modifier wrapper (`modifiers.text === 'static'`) but fails for a
|
|
79
|
+
// multi-modifier wrapper (`private static` → `modifiers.text === 'private static'`),
|
|
80
|
+
// which `hasModifier` handles by descending into the wrapper's children.
|
|
81
|
+
return (hasKeyword(node, 'static') ||
|
|
82
|
+
hasKeyword(node, 'class') ||
|
|
83
|
+
hasModifier(node, 'modifiers', 'static') ||
|
|
84
|
+
hasModifier(node, 'modifiers', 'class'));
|
|
60
85
|
},
|
|
61
86
|
isReadonly(node) {
|
|
62
87
|
// 'let' = constant/readonly, 'var' = variable
|
|
@@ -1,7 +1,3 @@
|
|
|
1
|
-
export interface FileEntry {
|
|
2
|
-
path: string;
|
|
3
|
-
content: string;
|
|
4
|
-
}
|
|
5
1
|
/** Lightweight entry — path + size from stat, no content in memory */
|
|
6
2
|
export interface ScannedFile {
|
|
7
3
|
path: string;
|
|
@@ -21,8 +17,3 @@ export declare const walkRepositoryPaths: (repoPath: string, onProgress?: (curre
|
|
|
21
17
|
* Returns a Map for O(1) lookup. Silently skips files that fail to read.
|
|
22
18
|
*/
|
|
23
19
|
export declare const readFileContents: (repoPath: string, relativePaths: string[]) => Promise<Map<string, string>>;
|
|
24
|
-
/**
|
|
25
|
-
* Legacy API — scans and reads everything into memory.
|
|
26
|
-
* Used by sequential fallback path only.
|
|
27
|
-
*/
|
|
28
|
-
export declare const walkRepository: (repoPath: string, onProgress?: (current: number, total: number, filePath: string) => void) => Promise<FileEntry[]>;
|
|
@@ -111,14 +111,3 @@ export const readFileContents = async (repoPath, relativePaths) => {
|
|
|
111
111
|
}
|
|
112
112
|
return contents;
|
|
113
113
|
};
|
|
114
|
-
/**
|
|
115
|
-
* Legacy API — scans and reads everything into memory.
|
|
116
|
-
* Used by sequential fallback path only.
|
|
117
|
-
*/
|
|
118
|
-
export const walkRepository = async (repoPath, onProgress) => {
|
|
119
|
-
const scanned = await walkRepositoryPaths(repoPath, onProgress);
|
|
120
|
-
const contents = await readFileContents(repoPath, scanned.map((f) => f.path));
|
|
121
|
-
return scanned
|
|
122
|
-
.filter((f) => contents.has(f.path))
|
|
123
|
-
.map((f) => ({ path: f.path, content: contents.get(f.path) }));
|
|
124
|
-
};
|
|
@@ -4,5 +4,14 @@ import { type SyntaxNode } from '../../utils/ast-helpers.js';
|
|
|
4
4
|
* Decompose a `preproc_include` node into a CaptureMatch with structured
|
|
5
5
|
* import captures. C #include maps to a wildcard import (all symbols
|
|
6
6
|
* from the header are visible).
|
|
7
|
+
*
|
|
8
|
+
* Only literal include paths are emitted as import sources:
|
|
9
|
+
* #include <stdio.h> → system_lib_string
|
|
10
|
+
* #include "local.h" → string_literal
|
|
11
|
+
* A computed include like `#include HEADER_MACRO` carries an `identifier`
|
|
12
|
+
* path node (the macro name, not a header path). Emitting it as an import
|
|
13
|
+
* source produces a garbage literal edge, so we skip it entirely — matching
|
|
14
|
+
* the convention in interpretCImport, which drops imports with no resolvable
|
|
15
|
+
* source (issue #1919 F5).
|
|
7
16
|
*/
|
|
8
17
|
export declare function splitCInclude(node: SyntaxNode): CaptureMatch | null;
|
|
@@ -3,10 +3,20 @@ import { nodeToCapture, syntheticCapture } from '../../utils/ast-helpers.js';
|
|
|
3
3
|
* Decompose a `preproc_include` node into a CaptureMatch with structured
|
|
4
4
|
* import captures. C #include maps to a wildcard import (all symbols
|
|
5
5
|
* from the header are visible).
|
|
6
|
+
*
|
|
7
|
+
* Only literal include paths are emitted as import sources:
|
|
8
|
+
* #include <stdio.h> → system_lib_string
|
|
9
|
+
* #include "local.h" → string_literal
|
|
10
|
+
* A computed include like `#include HEADER_MACRO` carries an `identifier`
|
|
11
|
+
* path node (the macro name, not a header path). Emitting it as an import
|
|
12
|
+
* source produces a garbage literal edge, so we skip it entirely — matching
|
|
13
|
+
* the convention in interpretCImport, which drops imports with no resolvable
|
|
14
|
+
* source (issue #1919 F5).
|
|
6
15
|
*/
|
|
7
16
|
export function splitCInclude(node) {
|
|
8
17
|
// node.type === 'preproc_include'
|
|
9
18
|
// path field: (string_literal (string_content)) | (system_lib_string)
|
|
19
|
+
// | (identifier) ← computed macro include, NOT a header path
|
|
10
20
|
const pathNode = node.childForFieldName?.('path') ?? null;
|
|
11
21
|
if (pathNode === null) {
|
|
12
22
|
// Fallback: scan children
|
|
@@ -23,6 +33,11 @@ export function splitCInclude(node) {
|
|
|
23
33
|
return buildIncludeCapture(node, pathNode);
|
|
24
34
|
}
|
|
25
35
|
function buildIncludeCapture(node, pathNode) {
|
|
36
|
+
// Skip computed includes (`#include MACRO`) — the path is an `identifier`,
|
|
37
|
+
// not a literal header path. Emitting it would create a garbage import.
|
|
38
|
+
if (pathNode.type !== 'string_literal' && pathNode.type !== 'system_lib_string') {
|
|
39
|
+
return null;
|
|
40
|
+
}
|
|
26
41
|
let raw;
|
|
27
42
|
if (pathNode.type === 'string_literal') {
|
|
28
43
|
// string_literal has children: `"`, string_content, `"`
|
|
@@ -4,6 +4,13 @@ import { type SyntaxNode } from '../../utils/ast-helpers.js';
|
|
|
4
4
|
* Decompose a `preproc_include` node into a CaptureMatch with structured
|
|
5
5
|
* import captures. C++ #include maps to a wildcard import (all symbols
|
|
6
6
|
* from the header are visible). Identical to C's splitCInclude.
|
|
7
|
+
*
|
|
8
|
+
* Only literal include paths are emitted as import sources:
|
|
9
|
+
* #include <map> → system_lib_string
|
|
10
|
+
* #include "User.h" → string_literal
|
|
11
|
+
* A computed include like `#include HEADER_MACRO` carries an `identifier`
|
|
12
|
+
* path node (the macro name, not a header path); we skip it so it never
|
|
13
|
+
* becomes a garbage literal import source (issue #1919 F5).
|
|
7
14
|
*/
|
|
8
15
|
export declare function splitCppInclude(node: SyntaxNode): CaptureMatch | null;
|
|
9
16
|
/**
|
|
@@ -3,6 +3,13 @@ import { nodeToCapture, syntheticCapture } from '../../utils/ast-helpers.js';
|
|
|
3
3
|
* Decompose a `preproc_include` node into a CaptureMatch with structured
|
|
4
4
|
* import captures. C++ #include maps to a wildcard import (all symbols
|
|
5
5
|
* from the header are visible). Identical to C's splitCInclude.
|
|
6
|
+
*
|
|
7
|
+
* Only literal include paths are emitted as import sources:
|
|
8
|
+
* #include <map> → system_lib_string
|
|
9
|
+
* #include "User.h" → string_literal
|
|
10
|
+
* A computed include like `#include HEADER_MACRO` carries an `identifier`
|
|
11
|
+
* path node (the macro name, not a header path); we skip it so it never
|
|
12
|
+
* becomes a garbage literal import source (issue #1919 F5).
|
|
6
13
|
*/
|
|
7
14
|
export function splitCppInclude(node) {
|
|
8
15
|
const pathNode = node.childForFieldName?.('path') ?? null;
|
|
@@ -20,6 +27,11 @@ export function splitCppInclude(node) {
|
|
|
20
27
|
return buildIncludeCapture(node, pathNode);
|
|
21
28
|
}
|
|
22
29
|
function buildIncludeCapture(node, pathNode) {
|
|
30
|
+
// Skip computed includes (`#include MACRO`) — the path is an `identifier`,
|
|
31
|
+
// not a literal header path. Emitting it would create a garbage import.
|
|
32
|
+
if (pathNode.type !== 'string_literal' && pathNode.type !== 'system_lib_string') {
|
|
33
|
+
return null;
|
|
34
|
+
}
|
|
23
35
|
let raw;
|
|
24
36
|
if (pathNode.type === 'string_literal') {
|
|
25
37
|
const content = pathNode.namedChildren.find((c) => c.type === 'string_content');
|
|
@@ -37,6 +37,46 @@ const DART_SCOPE_QUERY = `
|
|
|
37
37
|
(extension_declaration name: (identifier) @declaration.name) @declaration.class
|
|
38
38
|
(enum_declaration name: (identifier) @declaration.name) @declaration.enum
|
|
39
39
|
|
|
40
|
+
; ── Declarations — type aliases (old-style + new-style function typedefs) ────
|
|
41
|
+
; Both forms parse as type_alias; the name position differs, and a generic
|
|
42
|
+
; <T> parameter list intervenes for the generic variants. Per #1919 review CF2,
|
|
43
|
+
; a generic type_parameters node sits between the name and the next anchor, so
|
|
44
|
+
; the non-generic adjacency patterns silently drop the generic forms. Four
|
|
45
|
+
; standalone patterns (NOT one alternation — the tree-sitter 0.21 hazard drops
|
|
46
|
+
; sibling branches) keep the name capture unambiguous and single-match per form:
|
|
47
|
+
; non-generic old-style typedef int Cmp(int a, int b);
|
|
48
|
+
; children: return-type, NAME, formal_parameter_list
|
|
49
|
+
; generic old-style typedef int Cmp<T>(T a, T b); (CF2)
|
|
50
|
+
; children: return-type, NAME, type_parameters, formal_parameter_list
|
|
51
|
+
; non-generic new-style typedef Pred = bool Function(int);
|
|
52
|
+
; children: NAME, "=", function_type
|
|
53
|
+
; generic new-style typedef Mapper<T> = T Function(T);
|
|
54
|
+
; children: NAME, type_parameters, "=", function_type
|
|
55
|
+
; The alias name is the type_identifier immediately before the param list (old)
|
|
56
|
+
; or before "=" (new); for the generic forms it is the one immediately before
|
|
57
|
+
; the intervening type_parameters. Mirrors Kotlin's @declaration.type_alias
|
|
58
|
+
; rule; the generic scope-extractor maps "type_alias" → TypeAlias.
|
|
59
|
+
(type_alias
|
|
60
|
+
(type_identifier) @declaration.name
|
|
61
|
+
.
|
|
62
|
+
(formal_parameter_list)) @declaration.type_alias
|
|
63
|
+
(type_alias
|
|
64
|
+
(type_identifier) @declaration.name
|
|
65
|
+
.
|
|
66
|
+
(type_parameters)
|
|
67
|
+
.
|
|
68
|
+
(formal_parameter_list)) @declaration.type_alias
|
|
69
|
+
(type_alias
|
|
70
|
+
(type_identifier) @declaration.name
|
|
71
|
+
.
|
|
72
|
+
"=") @declaration.type_alias
|
|
73
|
+
(type_alias
|
|
74
|
+
(type_identifier) @declaration.name
|
|
75
|
+
.
|
|
76
|
+
(type_parameters)
|
|
77
|
+
.
|
|
78
|
+
"=") @declaration.type_alias
|
|
79
|
+
|
|
40
80
|
; ── Declarations — top-level functions (parent is program, not method) ───────
|
|
41
81
|
(program
|
|
42
82
|
(function_signature
|