@c4a/extract-ts 0.7.1 → 0.7.5

package/README.md

@@ -4,14 +4,13 @@
 
 `@c4a/extract-ts` turns TypeScript, JavaScript, TSX, and JSX structure into deterministic code
 evidence for Context knowledge production. It implements the
-`ExtractionPlugin` protocol from `@c4a/extract` and is the default plugin used
-by the SDK `extractTs({ source, collection: "codegraph" })` phase for npm-style
-packages.
+`ExtractionPlugin` protocol from `@c4a/extract` and is the community Code
+Indexer Provider's ECMAScript parser for npm-style packages.
 
 It extracts code facts; it does not decide product meaning, write approved
 Markdown, or choose the user's source boundary. Knowledge-workspace users reach
-it through the Context Agent entry and a confirmed extraction phase. The direct
-APIs below are for reusable structural analysis and project-owned adapters.
+it through the selected Code Indexer Provider. The direct APIs below are for
+Provider authors and reusable structural analysis.
 
 ## Package Role
 
@@ -24,12 +23,12 @@ confirmed TypeScript/JavaScript boundary
           ↓
 entry detection + export tracing + AST facts
           ↓
-raw code snapshot → review candidates → approved knowledge
+Indexer facts → reader-oriented Candidates → approved knowledge
 ```
 
 ## React Router structural facts
 
-Projects that use `extractCustom()` can reuse `extractReactRouterRoutes()` to
+Indexer Providers can reuse `extractReactRouterRoutes()` to
 index JSX `<Route>` declarations and route-object arrays. It reports paths,
 components, redirects, conditions, import sources, notes, and source locations
 without classifying product meaning:
@@ -83,10 +82,9 @@ const moduleFacts = extractEcmaScriptModuleExports(source, "src/index.cjs");
 - package version propagation into `ExtractionResult.package.version`
 - authored `.js`, `.jsx`, `.mjs`, and `.cjs` entry files without remapping them to TypeScript
 
-The Context project may override auto-detection with source-relative
-`extractTs.entries`, or use `mode: "scan"` to make every `include`-matched file
-an extraction root. These settings live in the Context workspace; consumers do
-not need to modify the analyzed package solely to declare extraction entries.
+The Code Indexer Provider may override auto-detection with source-relative
+entries, or use scan mode to make every include-matched file an extraction
+root. Consumers do not modify the analyzed package solely to configure parsing.
 
 Entry files are returned as module-relative paths. The repository runner later prefixes them to repo-relative paths in raw snapshots.
 
@@ -160,9 +158,8 @@ identities, per-file owner/disposition, explicit denominators, diagnostics, and
 the ordered parser receipt through `context.indexer.evidence-adapter-result/v1`.
 `ExtractionResult` remains available for the existing raw snapshot runner.
 
-The Context CLI consumes those rows during `context run extract:<source>:codegraph`
-to build review candidates for package/category/symbol knowledge Nodes. The
-important projection inputs are:
+The Code Indexer Provider consumes those facts while producing the current
+Indexer result. Important authoring inputs include:
 
 - stable package names and versions
 - stable exported symbol names
@@ -187,33 +184,16 @@ const registry = new ExtractionPluginRegistry();
 registry.register(new TypeScriptPlugin());
 ```
 
-Workspace usage normally goes through a declared project phase:
-
-```ts
-import { defineProject, extractTs, reviewValidity, source } from "@c4a/context";
-
-const componentLib = source("component-lib");
-
-export default defineProject({
-  sources: [componentLib],
-  phases: [
-    extractTs({ source: componentLib, collection: "codegraph" }),
-    reviewValidity({ collection: "codegraph" }),
-  ],
-  packages: [],
-});
-```
-
-The source boundary is registered first, then the phase is previewed or run:
+Workspace usage registers the source boundary, then selects the Code Indexer
+Provider through `src/indexers.yaml`:
 
 ```bash
 context source add repo component-lib --local ../packages/component-lib
-context run extract:component-lib:codegraph --dry-run
-context run extract:component-lib:codegraph
+context status --format json
 ```
 
-Agents should not hand-build runner input or raw snapshots for normal Context
-workspace operation.
+Agents should not hand-build parser input, raw snapshots, or Candidate files
+for normal Context workspace operation.
 
 ## Development
 

package/README.zh-CN.md

@@ -3,12 +3,12 @@
 [English](./README.md)
 
 `@c4a/extract-ts` 将 TypeScript、JavaScript、TSX 和 JSX 结构转化为 Context 知识生产可使用的确定性
-代码证据。它实现 `@c4a/extract` 的 `ExtractionPlugin` 协议，也是 npm-style
-package 使用 `extractTs({ source, collection: "codegraph" })` 阶段时的默认插件。
+代码证据。它实现 `@c4a/extract` 的 `ExtractionPlugin` 协议，也是社区 Code
+Indexer Provider 面向 npm-style package 使用的 ECMAScript 解析器。
 
 它只提取代码事实，不判断产品含义、不写正式 Markdown，也不替用户选择来源边界。
-知识工作区用户通过 Context Agent 入口和已确认的提取阶段使用它；直接 API 面向
-可复用结构分析和项目自有 Adapter。
+知识工作区用户通过选中的 Code Indexer Provider 使用它；直接 API 面向 Provider
+作者和可复用结构分析。
 
 ## 在知识生产链中的职责
 
@@ -17,7 +17,7 @@ package 使用 `extractTs({ source, collection: "codegraph" })` 阶段时的默
           ↓
 入口检测 + 导出追踪 + AST 事实
           ↓
-原始代码快照 → 审核候选 → 正式知识
+Indexer 事实 → 面向读者的 Candidate → 正式知识
 ```
 
 `@c4a/extract-ts` 负责 npm package 入口检测和 ECMAScript-family AST 提取。它不直接写工作区；
@@ -29,7 +29,7 @@ package 使用 `extractTs({ source, collection: "codegraph" })` 阶段时的默
 
 ### React Router 结构事实
 
-使用 `extractCustom()` 的项目可以复用 `extractReactRouterRoutes()`，索引 JSX
+Indexer Provider 可以复用 `extractReactRouterRoutes()`，索引 JSX
 `<Route>` 声明和 route-object 数组：
 
 ```ts
@@ -83,9 +83,8 @@ const moduleFacts = extractEcmaScriptModuleExports(source, "src/index.cjs");
 - 将 package version 写入 `ExtractionResult.package.version`。
 - 直接识别源码目录中的 `.js`、`.jsx`、`.mjs` 和 `.cjs`，不会误映射为 TypeScript。
 
-Context 项目可以使用 source-relative `extractTs.entries` 覆盖自动检测，或用
-`mode: "scan"` 把所有 `include` 命中的文件作为提取根。这些设置属于知识工作区，
-无需修改被分析 package。
+Code Indexer Provider 可以使用 source-relative entries 覆盖自动检测，或用 scan
+模式把所有命中 include 的文件作为解析根；无需修改被分析 package。
 
 入口文件以 module-relative 路径返回，Repository Runner 再为原始快照补充
 repo-relative 前缀。
@@ -162,25 +161,9 @@ const registry = new ExtractionPluginRegistry();
 registry.register(new TypeScriptPlugin());
 ```
 
-正常工作区在 `src/index.ts` 声明项目阶段：
-
-```ts
-import { defineProject, extractTs, reviewValidity, source } from "@c4a/context";
-
-const componentLib = source("component-lib");
-
-export default defineProject({
-  sources: [componentLib],
-  phases: [
-    extractTs({ source: componentLib, collection: "codegraph" }),
-    reviewValidity({ collection: "codegraph" }),
-  ],
-  packages: [],
-});
-```
-
-实际提取由当前 Context Route 驱动。Agent 不应在正常知识生产中手工构造 Runner
-输入或原始快照。
+正常工作区先登记来源，再由 `src/indexers.yaml` 选择 Code Indexer Provider。
+实际解析由当前 Context Route 驱动。Agent 不应在正常知识生产中手工构造 Runner
+输入、原始快照或 Candidate 文件。
 
 ## 开发
 

package/index.js

@@ -11404,6 +11404,8 @@ function keyTokens(key) {
   return key.replace(/([a-z0-9])([A-Z])/gu, "$1-$2").replace(/[^A-Za-z0-9]+/gu, "-").toLowerCase().split("-").filter(Boolean);
 }
 function sensitiveKey(key, value) {
+  if (value === INDEXER_OUTPUT_REDACTION_MARKER)
+    return false;
   const tokens = keyTokens(key);
   if (tokens.length === 0)
     return false;
@@ -11445,12 +11447,12 @@ function replaceWithCount(value, pattern, replacement, count) {
 function redactKnownText(value, count) {
   let output = value;
   output = replaceWithCount(output, /-----BEGIN (?:RSA |EC |OPENSSH )?PRIVATE KEY-----[\s\S]*?-----END (?:RSA |EC |OPENSSH )?PRIVATE KEY-----/gu, INDEXER_OUTPUT_REDACTION_MARKER, count);
-  output = replaceWithCount(output, /(\bauthorization\s*:\s*(?:bearer|basic)\s+)[^\s,;]+/giu, (_match, prefix) => `${prefix}${INDEXER_OUTPUT_REDACTION_MARKER}`, count);
+  output = replaceWithCount(output, new RegExp(`(\\bauthorization\\s*:\\s*(?:bearer|basic)\\s+)` + `(?!${escapeRegExp(INDEXER_OUTPUT_REDACTION_MARKER)})[^\\s,;]+`, "giu"), (_match, prefix) => `${prefix}${INDEXER_OUTPUT_REDACTION_MARKER}`, count);
   output = replaceWithCount(output, /([a-z][a-z0-9+.-]*:\/\/)[^\s/@:]+:[^\s/@]+@/giu, (_match, prefix) => `${prefix}${INDEXER_OUTPUT_REDACTION_MARKER}@`, count);
-  output = replaceWithCount(output, /([?&](?:access_token|refresh_token|api_key|password|secret)=)[^&#\s]+/giu, (_match, prefix) => `${prefix}${INDEXER_OUTPUT_REDACTION_MARKER}`, count);
+  output = replaceWithCount(output, new RegExp(`([?&](?:access_token|refresh_token|api_key|password|secret)=)` + `(?!${escapeRegExp(INDEXER_OUTPUT_REDACTION_MARKER)})[^&#\\s]+`, "giu"), (_match, prefix) => `${prefix}${INDEXER_OUTPUT_REDACTION_MARKER}`, count);
   const key = "(?:[A-Za-z0-9_.-]*(?:password|passwd|pwd|secret|token|credential|cookie)[A-Za-z0-9_.-]*|api[-_]?key|access[-_]?(?:key|token)|private[-_]?key|client[-_]?secret|authorization)";
   const assignment = `(?:=\\s*|:\\s+(?=\\S)|:\\s*(?=["']))`;
-  output = replaceWithCount(output, new RegExp(`((?:["']?${key}["']?)\\s*${assignment})(?:"(?:\\\\.|[^"])*"|'(?:\\\\.|[^'])*'|[^\\s,;}\\]]+)`, "giu"), (_match, prefix) => `${prefix}"${INDEXER_OUTPUT_REDACTION_MARKER}"`, count);
+  output = replaceWithCount(output, new RegExp(`((?:["']?${key}["']?)\\s*${assignment})` + `(?!["']?${escapeRegExp(INDEXER_OUTPUT_REDACTION_MARKER)})` + `(?:"(?:\\\\.|[^"])*"|'(?:\\\\.|[^'])*'|[^\\s,;}\\]]+)`, "giu"), (_match, prefix) => `${prefix}"${INDEXER_OUTPUT_REDACTION_MARKER}"`, count);
   return output;
 }
 function redactBlockedText(value, blocked, count) {
@@ -12312,6 +12314,7 @@ var buildToSourcePaths = (withoutExtension) => {
   }
   const fullRest = rest.join("/");
   results.push(`src/${fullRest}`);
+  results.push(fullRest);
   return unique(results);
 };
 var createCandidatePaths = (value, options = {}) => {
@@ -12463,7 +12466,7 @@ var expandWorkspacePattern = async (rootDir, pattern, fs) => {
   if (pattern.endsWith("/*")) {
     const prefix = pattern.slice(0, -2);
     const baseDir = prefix ? `${rootDir}/${prefix}` : rootDir;
-    const entries = await fs.readdir(baseDir);
+    const entries = (await fs.readdir(baseDir)).sort();
     return entries.map((entry) => `${baseDir}/${entry}`.replace(/^\.\//, ""));
   }
   return [`${rootDir}/${pattern}`.replace(/^\.\//, "")];
@@ -12814,6 +12817,10 @@ var codeExtractRunnerInputSchema = exports_external.object({
 });
 // ../extract/src/evidenceAdapter.ts
 function fact2(input) {
+  const payload = redactIndexerOutput({
+    channel: "success-payload",
+    value: input.payload
+  }).value;
   return createIndexerEvidenceAdapterFact({
     source_ref: input.sourceRef,
     module_ref: input.moduleRef,
@@ -12821,10 +12828,21 @@ function fact2(input) {
     qualified_item_path: input.qualifiedItemPath,
     kind: input.kind,
     signature: input.signature,
-    payload: input.payload,
+    payload,
     denominator: input.denominator
   });
 }
+function uniqueFacts(facts) {
+  const byRef = new Map;
+  for (const value of facts) {
+    const previous = byRef.get(value.fact_ref);
+    if (previous !== undefined && indexerEvidenceAdapterProtocolDigest(previous) !== indexerEvidenceAdapterProtocolDigest(value)) {
+      throw new TypeError(`ExtractionResult emits conflicting facts for ${value.fact_ref}`);
+    }
+    byRef.set(value.fact_ref, previous ?? value);
+  }
+  return [...byRef.values()];
+}
 function semanticExtractionPayload(extraction) {
   return {
     version: extraction.version,
@@ -12854,6 +12872,13 @@ function relationSourceFile(relation, symbols, filePaths) {
   }
   return null;
 }
+function relationQualifiedItemPath(relation) {
+  const identityDigest = indexerEvidenceAdapterProtocolDigest({
+    from: relation.from,
+    to: relation.to
+  });
+  return `relation:${relation.type}@${relation.line ?? 0}#${identityDigest}`;
+}
 function diagnosticPayload(diagnostic) {
   return {
     code: diagnostic.code,
@@ -12916,6 +12941,10 @@ function extractionResultToEvidenceAdapterResult(extraction, invocation) {
     }
     const facts = [];
     if (coverageFile.disposition === "analyzed" && fileInfo) {
+      const source = invocation.source_files?.[normalizedPath2];
+      if (invocation.source_files !== undefined && source === undefined) {
+        throw new TypeError(`Analyzed file ${normalizedPath2} has no scoped source text`);
+      }
       facts.push(fact2({
         sourceRef: invocation.authorized_scope.source_ref,
         moduleRef: invocation.module_ref,
@@ -12923,7 +12952,11 @@ function extractionResultToEvidenceAdapterResult(extraction, invocation) {
         qualifiedItemPath: "file",
         kind: "source-file",
         signature: { path: normalizedPath2, language: fileInfo.language },
-        payload: fileInfo,
+        payload: source === undefined ? fileInfo : {
+          ...fileInfo,
+          line: 1,
+          endLine: source.split(/\r\n|\r|\n/u).length
+        },
         denominator: ownsDenominators ? "eligible-file" : "none"
       }));
       facts.push(fact2({
@@ -12960,7 +12993,7 @@ function extractionResultToEvidenceAdapterResult(extraction, invocation) {
           sourceRef: invocation.authorized_scope.source_ref,
           moduleRef: invocation.module_ref,
           normalizedPath: normalizedPath2,
-          qualifiedItemPath: `relation:${relation.type}:${relation.from}->${relation.to}@${relation.line ?? 0}`,
+          qualifiedItemPath: relationQualifiedItemPath(relation),
           kind: "code-relation",
           signature: relation,
           payload: relation,
@@ -12976,7 +13009,7 @@ function extractionResultToEvidenceAdapterResult(extraction, invocation) {
       role,
       coverage_tier: coverage.tier,
       disposition: coverageFile.disposition,
-      facts
+      facts: uniqueFacts(facts)
     };
   });
   const diagnostics = [
@@ -13017,6 +13050,184 @@ function extractionResultToEvidenceAdapterResult(extraction, invocation) {
     }]
   });
 }
+// ../../node_modules/.bun/jsonc-parser@3.3.1/node_modules/jsonc-parser/lib/esm/impl/scanner.js
+var CharacterCodes;
+(function(CharacterCodes2) {
+  CharacterCodes2[CharacterCodes2["lineFeed"] = 10] = "lineFeed";
+  CharacterCodes2[CharacterCodes2["carriageReturn"] = 13] = "carriageReturn";
+  CharacterCodes2[CharacterCodes2["space"] = 32] = "space";
+  CharacterCodes2[CharacterCodes2["_0"] = 48] = "_0";
+  CharacterCodes2[CharacterCodes2["_1"] = 49] = "_1";
+  CharacterCodes2[CharacterCodes2["_2"] = 50] = "_2";
+  CharacterCodes2[CharacterCodes2["_3"] = 51] = "_3";
+  CharacterCodes2[CharacterCodes2["_4"] = 52] = "_4";
+  CharacterCodes2[CharacterCodes2["_5"] = 53] = "_5";
+  CharacterCodes2[CharacterCodes2["_6"] = 54] = "_6";
+  CharacterCodes2[CharacterCodes2["_7"] = 55] = "_7";
+  CharacterCodes2[CharacterCodes2["_8"] = 56] = "_8";
+  CharacterCodes2[CharacterCodes2["_9"] = 57] = "_9";
+  CharacterCodes2[CharacterCodes2["a"] = 97] = "a";
+  CharacterCodes2[CharacterCodes2["b"] = 98] = "b";
+  CharacterCodes2[CharacterCodes2["c"] = 99] = "c";
+  CharacterCodes2[CharacterCodes2["d"] = 100] = "d";
+  CharacterCodes2[CharacterCodes2["e"] = 101] = "e";
+  CharacterCodes2[CharacterCodes2["f"] = 102] = "f";
+  CharacterCodes2[CharacterCodes2["g"] = 103] = "g";
+  CharacterCodes2[CharacterCodes2["h"] = 104] = "h";
+  CharacterCodes2[CharacterCodes2["i"] = 105] = "i";
+  CharacterCodes2[CharacterCodes2["j"] = 106] = "j";
+  CharacterCodes2[CharacterCodes2["k"] = 107] = "k";
+  CharacterCodes2[CharacterCodes2["l"] = 108] = "l";
+  CharacterCodes2[CharacterCodes2["m"] = 109] = "m";
+  CharacterCodes2[CharacterCodes2["n"] = 110] = "n";
+  CharacterCodes2[CharacterCodes2["o"] = 111] = "o";
+  CharacterCodes2[CharacterCodes2["p"] = 112] = "p";
+  CharacterCodes2[CharacterCodes2["q"] = 113] = "q";
+  CharacterCodes2[CharacterCodes2["r"] = 114] = "r";
+  CharacterCodes2[CharacterCodes2["s"] = 115] = "s";
+  CharacterCodes2[CharacterCodes2["t"] = 116] = "t";
+  CharacterCodes2[CharacterCodes2["u"] = 117] = "u";
+  CharacterCodes2[CharacterCodes2["v"] = 118] = "v";
+  CharacterCodes2[CharacterCodes2["w"] = 119] = "w";
+  CharacterCodes2[CharacterCodes2["x"] = 120] = "x";
+  CharacterCodes2[CharacterCodes2["y"] = 121] = "y";
+  CharacterCodes2[CharacterCodes2["z"] = 122] = "z";
+  CharacterCodes2[CharacterCodes2["A"] = 65] = "A";
+  CharacterCodes2[CharacterCodes2["B"] = 66] = "B";
+  CharacterCodes2[CharacterCodes2["C"] = 67] = "C";
+  CharacterCodes2[CharacterCodes2["D"] = 68] = "D";
+  CharacterCodes2[CharacterCodes2["E"] = 69] = "E";
+  CharacterCodes2[CharacterCodes2["F"] = 70] = "F";
+  CharacterCodes2[CharacterCodes2["G"] = 71] = "G";
+  CharacterCodes2[CharacterCodes2["H"] = 72] = "H";
+  CharacterCodes2[CharacterCodes2["I"] = 73] = "I";
+  CharacterCodes2[CharacterCodes2["J"] = 74] = "J";
+  CharacterCodes2[CharacterCodes2["K"] = 75] = "K";
+  CharacterCodes2[CharacterCodes2["L"] = 76] = "L";
+  CharacterCodes2[CharacterCodes2["M"] = 77] = "M";
+  CharacterCodes2[CharacterCodes2["N"] = 78] = "N";
+  CharacterCodes2[CharacterCodes2["O"] = 79] = "O";
+  CharacterCodes2[CharacterCodes2["P"] = 80] = "P";
+  CharacterCodes2[CharacterCodes2["Q"] = 81] = "Q";
+  CharacterCodes2[CharacterCodes2["R"] = 82] = "R";
+  CharacterCodes2[CharacterCodes2["S"] = 83] = "S";
+  CharacterCodes2[CharacterCodes2["T"] = 84] = "T";
+  CharacterCodes2[CharacterCodes2["U"] = 85] = "U";
+  CharacterCodes2[CharacterCodes2["V"] = 86] = "V";
+  CharacterCodes2[CharacterCodes2["W"] = 87] = "W";
+  CharacterCodes2[CharacterCodes2["X"] = 88] = "X";
+  CharacterCodes2[CharacterCodes2["Y"] = 89] = "Y";
+  CharacterCodes2[CharacterCodes2["Z"] = 90] = "Z";
+  CharacterCodes2[CharacterCodes2["asterisk"] = 42] = "asterisk";
+  CharacterCodes2[CharacterCodes2["backslash"] = 92] = "backslash";
+  CharacterCodes2[CharacterCodes2["closeBrace"] = 125] = "closeBrace";
+  CharacterCodes2[CharacterCodes2["closeBracket"] = 93] = "closeBracket";
+  CharacterCodes2[CharacterCodes2["colon"] = 58] = "colon";
+  CharacterCodes2[CharacterCodes2["comma"] = 44] = "comma";
+  CharacterCodes2[CharacterCodes2["dot"] = 46] = "dot";
+  CharacterCodes2[CharacterCodes2["doubleQuote"] = 34] = "doubleQuote";
+  CharacterCodes2[CharacterCodes2["minus"] = 45] = "minus";
+  CharacterCodes2[CharacterCodes2["openBrace"] = 123] = "openBrace";
+  CharacterCodes2[CharacterCodes2["openBracket"] = 91] = "openBracket";
+  CharacterCodes2[CharacterCodes2["plus"] = 43] = "plus";
+  CharacterCodes2[CharacterCodes2["slash"] = 47] = "slash";
+  CharacterCodes2[CharacterCodes2["formFeed"] = 12] = "formFeed";
+  CharacterCodes2[CharacterCodes2["tab"] = 9] = "tab";
+})(CharacterCodes || (CharacterCodes = {}));
+
+// ../../node_modules/.bun/jsonc-parser@3.3.1/node_modules/jsonc-parser/lib/esm/impl/string-intern.js
+var cachedSpaces = new Array(20).fill(0).map((_, index) => {
+  return " ".repeat(index);
+});
+var maxCachedValues = 200;
+var cachedBreakLinesWithSpaces = {
+  " ": {
+    "\n": new Array(maxCachedValues).fill(0).map((_, index) => {
+      return `
+` + " ".repeat(index);
+    }),
+    "\r": new Array(maxCachedValues).fill(0).map((_, index) => {
+      return "\r" + " ".repeat(index);
+    }),
+    "\r\n": new Array(maxCachedValues).fill(0).map((_, index) => {
+      return `\r
+` + " ".repeat(index);
+    })
+  },
+  "\t": {
+    "\n": new Array(maxCachedValues).fill(0).map((_, index) => {
+      return `
+` + "\t".repeat(index);
+    }),
+    "\r": new Array(maxCachedValues).fill(0).map((_, index) => {
+      return "\r" + "\t".repeat(index);
+    }),
+    "\r\n": new Array(maxCachedValues).fill(0).map((_, index) => {
+      return `\r
+` + "\t".repeat(index);
+    })
+  }
+};
+
+// ../../node_modules/.bun/jsonc-parser@3.3.1/node_modules/jsonc-parser/lib/esm/impl/parser.js
+var ParseOptions;
+(function(ParseOptions2) {
+  ParseOptions2.DEFAULT = {
+    allowTrailingComma: false
+  };
+})(ParseOptions || (ParseOptions = {}));
+
+// ../../node_modules/.bun/jsonc-parser@3.3.1/node_modules/jsonc-parser/lib/esm/main.js
+var ScanError;
+(function(ScanError2) {
+  ScanError2[ScanError2["None"] = 0] = "None";
+  ScanError2[ScanError2["UnexpectedEndOfComment"] = 1] = "UnexpectedEndOfComment";
+  ScanError2[ScanError2["UnexpectedEndOfString"] = 2] = "UnexpectedEndOfString";
+  ScanError2[ScanError2["UnexpectedEndOfNumber"] = 3] = "UnexpectedEndOfNumber";
+  ScanError2[ScanError2["InvalidUnicode"] = 4] = "InvalidUnicode";
+  ScanError2[ScanError2["InvalidEscapeCharacter"] = 5] = "InvalidEscapeCharacter";
+  ScanError2[ScanError2["InvalidCharacter"] = 6] = "InvalidCharacter";
+})(ScanError || (ScanError = {}));
+var SyntaxKind;
+(function(SyntaxKind2) {
+  SyntaxKind2[SyntaxKind2["OpenBraceToken"] = 1] = "OpenBraceToken";
+  SyntaxKind2[SyntaxKind2["CloseBraceToken"] = 2] = "CloseBraceToken";
+  SyntaxKind2[SyntaxKind2["OpenBracketToken"] = 3] = "OpenBracketToken";
+  SyntaxKind2[SyntaxKind2["CloseBracketToken"] = 4] = "CloseBracketToken";
+  SyntaxKind2[SyntaxKind2["CommaToken"] = 5] = "CommaToken";
+  SyntaxKind2[SyntaxKind2["ColonToken"] = 6] = "ColonToken";
+  SyntaxKind2[SyntaxKind2["NullKeyword"] = 7] = "NullKeyword";
+  SyntaxKind2[SyntaxKind2["TrueKeyword"] = 8] = "TrueKeyword";
+  SyntaxKind2[SyntaxKind2["FalseKeyword"] = 9] = "FalseKeyword";
+  SyntaxKind2[SyntaxKind2["StringLiteral"] = 10] = "StringLiteral";
+  SyntaxKind2[SyntaxKind2["NumericLiteral"] = 11] = "NumericLiteral";
+  SyntaxKind2[SyntaxKind2["LineCommentTrivia"] = 12] = "LineCommentTrivia";
+  SyntaxKind2[SyntaxKind2["BlockCommentTrivia"] = 13] = "BlockCommentTrivia";
+  SyntaxKind2[SyntaxKind2["LineBreakTrivia"] = 14] = "LineBreakTrivia";
+  SyntaxKind2[SyntaxKind2["Trivia"] = 15] = "Trivia";
+  SyntaxKind2[SyntaxKind2["Unknown"] = 16] = "Unknown";
+  SyntaxKind2[SyntaxKind2["EOF"] = 17] = "EOF";
+})(SyntaxKind || (SyntaxKind = {}));
+var ParseErrorCode;
+(function(ParseErrorCode2) {
+  ParseErrorCode2[ParseErrorCode2["InvalidSymbol"] = 1] = "InvalidSymbol";
+  ParseErrorCode2[ParseErrorCode2["InvalidNumberFormat"] = 2] = "InvalidNumberFormat";
+  ParseErrorCode2[ParseErrorCode2["PropertyNameExpected"] = 3] = "PropertyNameExpected";
+  ParseErrorCode2[ParseErrorCode2["ValueExpected"] = 4] = "ValueExpected";
+  ParseErrorCode2[ParseErrorCode2["ColonExpected"] = 5] = "ColonExpected";
+  ParseErrorCode2[ParseErrorCode2["CommaExpected"] = 6] = "CommaExpected";
+  ParseErrorCode2[ParseErrorCode2["CloseBraceExpected"] = 7] = "CloseBraceExpected";
+  ParseErrorCode2[ParseErrorCode2["CloseBracketExpected"] = 8] = "CloseBracketExpected";
+  ParseErrorCode2[ParseErrorCode2["EndOfFileExpected"] = 9] = "EndOfFileExpected";
+  ParseErrorCode2[ParseErrorCode2["InvalidCommentToken"] = 10] = "InvalidCommentToken";
+  ParseErrorCode2[ParseErrorCode2["UnexpectedEndOfComment"] = 11] = "UnexpectedEndOfComment";
+  ParseErrorCode2[ParseErrorCode2["UnexpectedEndOfString"] = 12] = "UnexpectedEndOfString";
+  ParseErrorCode2[ParseErrorCode2["UnexpectedEndOfNumber"] = 13] = "UnexpectedEndOfNumber";
+  ParseErrorCode2[ParseErrorCode2["InvalidUnicode"] = 14] = "InvalidUnicode";
+  ParseErrorCode2[ParseErrorCode2["InvalidEscapeCharacter"] = 15] = "InvalidEscapeCharacter";
+  ParseErrorCode2[ParseErrorCode2["InvalidCharacter"] = 16] = "InvalidCharacter";
+})(ParseErrorCode || (ParseErrorCode = {}));
+
 // ../../node_modules/.bun/eslint-visitor-keys@5.0.1/node_modules/eslint-visitor-keys/lib/visitor-keys.js
 var KEYS = {
   ArrayExpression: ["elements"],
@@ -13505,6 +13716,24 @@ var scriptKind = (filePath) => {
   return ts.ScriptKind.TS;
 };
 var createEcmaScriptSourceFile = (source, filePath) => ts.createSourceFile(filePath, source, ts.ScriptTarget.Latest, true, scriptKind(filePath));
+var treeSitterCompatibleJsxSource = (sourceFile, source) => {
+  const characters = source.split("");
+  let changed = false;
+  const visit2 = (node2) => {
+    const jsxPresentationLiteral = ts.isJsxText(node2) || ts.isStringLiteral(node2) && ts.isJsxAttribute(node2.parent);
+    if (jsxPresentationLiteral) {
+      for (let index = node2.getStart(sourceFile);index < node2.getEnd(); index += 1) {
+        if (characters[index] === "&") {
+          characters[index] = " ";
+          changed = true;
+        }
+      }
+    }
+    ts.forEachChild(node2, visit2);
+  };
+  visit2(sourceFile);
+  return changed ? characters.join("") : source;
+};
 var syntaxDiagnostics = (sourceFile) => {
   const diagnostics = sourceFile.parseDiagnostics ?? [];
   return diagnostics.map((diagnostic) => {
@@ -13688,7 +13917,7 @@ var collectBindings = (sourceFile) => {
 };
 var collectDynamicDiagnostics = (sourceFile) => {
   const diagnostics = [];
-  const visit = (node2) => {
+  const visit2 = (node2) => {
     if (ts2.isCallExpression(node2) && ts2.isIdentifier(node2.expression) && node2.expression.text === "require") {
       if (staticStringValue(node2.arguments[0]) === null) {
         const location = nodeLocation(sourceFile, node2);
@@ -13726,9 +13955,9 @@ var collectDynamicDiagnostics = (sourceFile) => {
         }
       }
     }
-    ts2.forEachChild(node2, visit);
+    ts2.forEachChild(node2, visit2);
   };
-  visit(sourceFile);
+  visit2(sourceFile);
   return diagnostics;
 };
 var analyzeCommonJsModule = (source, filePath) => {
@@ -13965,7 +14194,9 @@ var traceFile = async (filePath, fs, state) => {
     const commonJs = analyzeCommonJsModule(source, filePath);
     if (commonJs.diagnostics.length > 0)
       return [];
-    const tree = await parseFile(source, isJsxLikePath(filePath));
+    const jsxLike = isJsxLikePath(filePath);
+    const parserSource = jsxLike ? treeSitterCompatibleJsxSource(createEcmaScriptSourceFile(source, filePath), source) : source;
+    const tree = await parseFile(parserSource, jsxLike);
     if (!tree)
       return [];
     const root = tree.rootNode;
@@ -14410,7 +14641,7 @@ var declarationName = (node2, sourceFile) => {
 var collectStaticCallRelations = (source, filePath, importBindings) => {
   const sourceFile = createEcmaScriptSourceFile(source, filePath);
   const relations = [];
-  const visit = (node2, owner) => {
+  const visit2 = (node2, owner) => {
     const namedOwner = declarationName(node2, sourceFile) ?? owner;
     if (ts3.isCallExpression(node2) || ts3.isNewExpression(node2)) {
       const target = callTarget(node2.expression, sourceFile);
@@ -14420,9 +14651,9 @@ var collectStaticCallRelations = (source, filePath, importBindings) => {
         relations.push(createRelation("calls" /* Calls */, namedOwner, target, isExternal, nodeLocation(sourceFile, node2).line));
       }
     }
-    ts3.forEachChild(node2, (child) => visit(child, namedOwner));
+    ts3.forEachChild(node2, (child) => visit2(child, namedOwner));
   };
-  visit(sourceFile, filePath);
+  visit2(sourceFile, filePath);
   const seen = new Set;
   return relations.sort((left, right) => (left.line ?? 0) - (right.line ?? 0) || left.from.localeCompare(right.from) || left.to.localeCompare(right.to)).filter((relation) => {
     const key = `${relation.from}\x00${relation.to}\x00${relation.line ?? 0}`;
@@ -14724,7 +14955,9 @@ var analyzeFile = async (filePath, fs, resolver = { mappings: [] }) => {
   const source = await fs.readFile(filePath);
   const sourceFile = createEcmaScriptSourceFile(source, filePath);
   const commonJs = analyzeCommonJsModule(source, filePath);
-  const tree = await parseFile(source, isJsxLikePath(filePath));
+  const jsxLike = isJsxLikePath(filePath);
+  const parserSource = jsxLike ? treeSitterCompatibleJsxSource(sourceFile, source) : source;
+  const tree = await parseFile(parserSource, jsxLike);
   if (!tree) {
     const diagnostics = syntaxDiagnostics(sourceFile);
     if (diagnostics.length === 0) {
@@ -14789,6 +15022,26 @@ var analyzeFile = async (filePath, fs, resolver = { mappings: [] }) => {
 };
 
 // src/symbolExtractor.ts
+var relationIdentity = (relation) => JSON.stringify([
+  relation.type,
+  relation.from,
+  relation.to,
+  relation.isExternal,
+  relation.grounding,
+  relation.confidence,
+  relation.source,
+  relation.line ?? null
+]);
+var uniqueRelations = (relations) => {
+  const seen = new Set;
+  return relations.filter((relation) => {
+    const identity = relationIdentity(relation);
+    if (seen.has(identity))
+      return false;
+    seen.add(identity);
+    return true;
+  });
+};
 var extractSymbols = async (entries, fs, options) => {
   const resolver = await loadTsConfigPathResolver(fs);
   const exportedSymbols = [];
@@ -14797,15 +15050,19 @@ var extractSymbols = async (entries, fs, options) => {
   const filePaths = new Set;
   const exportedDeclarationKeys = new Set;
   const analyses = new Map;
+  const analysisScope = options.analysisPaths === undefined ? null : new Set(options.analysisPaths);
+  const isInScope = (path2) => analysisScope === null || analysisScope.has(path2);
   for (const entry of entries) {
     const traced = await traceExports(entry.path, fs, resolver);
-    traced.files.forEach((filePath) => filePaths.add(filePath));
-    for (const filePath of traced.files) {
+    traced.files.filter(isInScope).forEach((filePath) => filePaths.add(filePath));
+    for (const filePath of traced.files.filter(isInScope)) {
       if (!analyses.has(filePath)) {
         analyses.set(filePath, await analyzeFile(filePath, fs, resolver));
       }
     }
     for (const tracedExport of traced.exports) {
+      if (!isInScope(tracedExport.declarationFile))
+        continue;
       const analysis = analyses.get(tracedExport.declarationFile);
       const declaration = analysis?.declarations.get(tracedExport.localName);
       if (!declaration)
@@ -14823,6 +15080,12 @@ var extractSymbols = async (entries, fs, options) => {
       });
     }
   }
+  for (const filePath of [...analysisScope ?? []].sort()) {
+    filePaths.add(filePath);
+    if (!analyses.has(filePath)) {
+      analyses.set(filePath, await analyzeFile(filePath, fs, resolver));
+    }
+  }
   for (const [filePath, analysis] of analyses) {
     relations.push(...analysis.relations);
     for (const [name, declaration] of analysis.declarations) {
@@ -14863,6 +15126,7 @@ var extractSymbols = async (entries, fs, options) => {
       relations.push(createRelation("of_type" /* OfType */, sym.name, propsTypeName, false, sym.line));
     }
   }
+  const deduplicatedRelations = uniqueRelations(relations);
   const files = [...filePaths].sort().map((filePath) => ({
     path: filePath,
     language: ecmaScriptLanguage(filePath),
@@ -14890,7 +15154,7 @@ var extractSymbols = async (entries, fs, options) => {
     package: options.packageInfo,
     files,
     symbols,
-    relations,
+    relations: deduplicatedRelations,
     coverage: {
       tier: EXTRACT_TS_COVERAGE_TIER,
       capabilities: [...EXTRACT_TS_CAPABILITIES],
@@ -14902,7 +15166,7 @@ var extractSymbols = async (entries, fs, options) => {
       lines: files.reduce((sum, file) => sum + file.lines, 0),
       exportedSymbols: exportedSymbols.length,
       internalSymbols: internalSymbols.length,
-      relations: relations.length
+      relations: deduplicatedRelations.length
     }
   };
 };
@@ -14929,12 +15193,34 @@ class TypeScriptPlugin {
     if (!packageInfo) {
       throw new Error("TypeScriptPlugin.extractSymbols requires detectEntries to run first");
     }
-    const result = await extractSymbols(entries, fs, {
-      packageInfo,
-      pluginId: this.id
-    });
-    this.#lastDetection = null;
-    return result;
+    try {
+      return await extractSymbols(entries, fs, {
+        packageInfo,
+        pluginId: this.id
+      });
+    } finally {
+      this.#lastDetection = null;
+    }
+  }
+  async extractSymbolsInScope(entries, analysisPaths, fs) {
+    const detectedPackage = this.#lastDetection?.package;
+    if (!detectedPackage && entries.length > 0) {
+      throw new Error("TypeScriptPlugin.extractSymbolsInScope requires detectEntries before using package entries");
+    }
+    const packageInfo = detectedPackage ?? {
+      name: "unknown-package",
+      kind: "lib" /* Lib */,
+      language: packageLanguage(analysisPaths)
+    };
+    try {
+      return await extractSymbols(entries, fs, {
+        packageInfo,
+        pluginId: this.id,
+        analysisPaths
+      });
+    } finally {
+      this.#lastDetection = null;
+    }
   }
 }
 // src/reactRouter.ts
@@ -14957,16 +15243,16 @@ function scalarValue(node2) {
 }
 function findDynamicImport(node2) {
   let result;
-  const visit = (child) => {
+  const visit2 = (child) => {
     if (result)
       return;
     if (ts4.isCallExpression(child) && child.expression.kind === ts4.SyntaxKind.ImportKeyword && child.arguments[0] && ts4.isStringLiteralLike(child.arguments[0])) {
       result = child.arguments[0].text;
       return;
     }
-    ts4.forEachChild(child, visit);
+    ts4.forEachChild(child, visit2);
   };
-  visit(node2);
+  visit2(node2);
   return result;
 }
 function parseSource(source, filePath) {
@@ -15054,7 +15340,7 @@ function jsxScalar(attribute, constants2) {
 }
 function descendantTags(node2) {
   const tags = new Set;
-  const visit = (child) => {
+  const visit2 = (child) => {
     if (ts4.isJsxElement(child) || ts4.isJsxSelfClosingElement(child)) {
       const opening = ts4.isJsxElement(child) ? child.openingElement : child;
       const tag = opening.tagName.getText();
@@ -15063,14 +15349,14 @@ function descendantTags(node2) {
       if (!["Route", "Routes", "Suspense", "Fragment", "React.Fragment", "Navigate"].includes(tag))
         tags.add(tag);
     }
-    ts4.forEachChild(child, visit);
+    ts4.forEachChild(child, visit2);
   };
-  visit(node2);
+  visit2(node2);
   return [...tags];
 }
 function navigateTarget(node2, sourceFile) {
   let target;
-  const visit = (child) => {
+  const visit2 = (child) => {
     if (target)
       return;
     if (ts4.isJsxElement(child) || ts4.isJsxSelfClosingElement(child)) {
@@ -15089,9 +15375,9 @@ function navigateTarget(node2, sourceFile) {
         }
       }
     }
-    ts4.forEachChild(child, visit);
+    ts4.forEachChild(child, visit2);
   };
-  visit(node2);
+  visit2(node2);
   return target;
 }
 function leadingNote(node2, sourceFile) {
@@ -15173,7 +15459,7 @@ function extractReactRouterRoutes(source, filePath, options = {}) {
       });
     }
   };
-  const visit = (node2, parentPath) => {
+  const visit2 = (node2, parentPath) => {
     if (ts4.isJsxElement(node2) || ts4.isJsxSelfClosingElement(node2)) {
       const opening = ts4.isJsxElement(node2) ? node2.openingElement : node2;
       if (opening.tagName.getText() === "Route") {
@@ -15190,7 +15476,7 @@ function extractReactRouterRoutes(source, filePath, options = {}) {
         pushRoute({ node: node2, parentPath, path: typeof pathValue === "string" ? pathValue : "", index, ...component ? { component } : {}, candidates, ...redirectTo ? { redirectTo } : {} });
         if (ts4.isJsxElement(node2))
           for (const child of node2.children)
-            visit(child, fullPath);
+            visit2(child, fullPath);
         return;
       }
     }
@@ -15200,9 +15486,9 @@ function extractReactRouterRoutes(source, filePath, options = {}) {
         visitRouteObjects(node2.arguments[0], mountPath);
       }
     }
-    ts4.forEachChild(node2, (child) => visit(child, parentPath));
+    ts4.forEachChild(node2, (child) => visit2(child, parentPath));
   };
-  visit(parsed.sourceFile, mountPath);
+  visit2(parsed.sourceFile, mountPath);
   return routes.sort((left, right) => left.fullPath.localeCompare(right.fullPath) || left.location.startLine - right.location.startLine || left.location.startColumn - right.location.startColumn);
 }
 // src/moduleExports.ts

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@c4a/extract-ts",
-  "version": "0.7.1",
+  "version": "0.7.5",
   "type": "module",
   "description": "TypeScript and JavaScript extraction plugin for the Context ExtractionResult v2 contract",
   "license": "MIT",