@omfalos/mokosh 0.2.0 → 0.3.1

package/dist/index.d.mts

@@ -357,8 +357,9 @@ interface FunctionCallInfo {
  * field matches `functionName`. Callees are found by looking at the defining
  * file's `callEdges` for edges whose `from` field matches `functionName`.
  *
- * Call edges are populated only for TypeScript/JavaScript files. Functions in
- * other language files will return empty `callers` and `callees` arrays.
+ * Call edges are populated only for TypeScript/JavaScript, Go, and Python files
+ * (see `CALL_EDGE_TYPES` in `../language-support`). Functions in other language
+ * files will return empty `callers` and `callees` arrays.
  *
  * @param {Graph} graph - The import graph that carries `callEdges` on each node.
  * @param {string} functionName - Exact name of the function to look up.
@@ -441,6 +442,56 @@ declare function saveChangeImpactCache(cache: ChangeImpactCache, cachePath: stri
  */
 declare function loadChangeImpactCache(cachePath: string): ChangeImpactCache | null;
 
+/**
+ * Sliding-window (shingle) hashing over a normalized token stream, plus chain-merging of
+ * consecutive matching windows into contiguous duplicate blocks. Shared by every language
+ * `tokenize()` supports — the shingling step itself has no language awareness at all.
+ */
+
+interface DuplicateOccurrence {
+    file: string;
+    startLine: number;
+    endLine: number;
+}
+interface DuplicateGroup {
+    /** Two locations sharing this duplicated block. */
+    occurrences: [DuplicateOccurrence, DuplicateOccurrence];
+    /** Line span covered by the block (measured on the first occurrence). */
+    lines: number;
+    /** Token-window length backing this block, after chain-merging adjacent windows. */
+    tokens: number;
+}
+
+interface FindDuplicatesOptions {
+    /** Minimum duplicated block size, in source lines, to report (default 6). */
+    minLines?: number | undefined;
+    /** Shingle window size, in tokens — the smallest duplicate the scan can detect (default 15). */
+    windowSize?: number | undefined;
+    /** When true (default), string/number literals are normalized too, so only structural shape
+     *  — not the specific values used — drives a match. Set false for stricter, Type-1-only matching. */
+    ignoreLiterals?: boolean | undefined;
+    /** Caps the number of duplicate blocks returned, largest-first (default 50). */
+    limit?: number | undefined;
+    /** Directory names to exclude, matched against any path segment (default `DEFAULT_IGNORE_DIRS`
+     *  — `node_modules`, `dist`, `.git`, `mokosh-cache`, `coverage`, etc.). Pass `[]` to disable. */
+    ignoreDirs?: readonly string[] | undefined;
+}
+/**
+ * @description Scans every file already present in `graph` for cross-file (and within-file)
+ *   duplicated code, using a language-agnostic token-shingling pipeline: comments are stripped
+ *   per `FileType`, identifiers (and, by default, literals) are normalized to placeholders so
+ *   renamed-variable copies still match, then a sliding token window is hashed and chain-merged
+ *   into contiguous blocks. Works uniformly across every language in `DEFAULT_EXTENSIONS`.
+ * @param graph - The graph to scan; its node paths (already ignore-rule-filtered) are the file
+ *   list, re-read from disk since duplication data isn't cached on `FileNode`.
+ * @param rootDir - Absolute project root that graph paths are relative to.
+ * @param options - `minLines`/`windowSize` tune sensitivity; `ignoreLiterals` toggles Type-2 vs
+ *   Type-1 matching; `ignoreDirs` excludes files under matching directory names; `limit` caps
+ *   results. Lock files are always excluded, independent of `ignoreDirs`.
+ * @returns Duplicate blocks, each a pair of occurrences, sorted largest-first.
+ */
+declare function findDuplicates(graph: Graph, rootDir: string, options?: FindDuplicatesOptions): Promise<DuplicateGroup[]>;
+
 /** Controls how aggressively `detectFeatures` promotes files to features. */
 interface FeatureDetectionOptions {
     /**
@@ -956,7 +1007,8 @@ interface FindComplexFunctionsOptions {
 }
 /**
  * @description Scans every file's per-function complexity breakdown and returns functions/methods
- *   at or above the given threshold, sorted worst-first. TypeScript/JavaScript only.
+ *   at or above the given threshold, sorted worst-first. Populated for TypeScript/JavaScript, Go,
+ *   and Python.
  * @param graph - The graph to scan.
  * @param options - `metric` picks which score to threshold/sort on (default `cognitiveComplexity`); `threshold` is the minimum score to include (default 10); `limit` caps the results (default 20).
  * @returns Matching functions, sorted worst-first.
@@ -1012,7 +1064,7 @@ declare function hasCoverageData(graph: Graph): boolean;
 
 /** Symbol-name lookup across the whole graph — generalizes queryCallGraph beyond TS/JS functions. */
 
-type SymbolPrecision = "call" | "import-symbol" | "file-level";
+type SymbolPrecision = "call" | "file-level";
 interface SymbolCaller {
     file: string;
     callerFunction: string;
@@ -1031,9 +1083,11 @@ interface SymbolMatch {
 /**
  * @description Finds every file that exports a symbol by name, with the best available
  *   usage info per match. Precision depends on what the defining file's language parser
- *   tracks: TS/JS gets function-level callers via call edges, Python gets named-import
- *   tracking, everything else falls back to whole-file dependents (import-level, not
- *   symbol-level — the file might not even use this specific export).
+ *   tracks: TS/JS, Go, and Python get function-level callers via call edges (`"call"`);
+ *   everything else falls back to whole-file dependents (`"file-level"` — import-level, not
+ *   symbol-level, since the file might not even use this specific export). Coverage still
+ *   differs across the `"call"` languages: TS/JS tracks any directly imported symbol, Go only
+ *   package-qualified calls, Python only bare calls to `from <module> import <name>` symbols.
  * @param graph - The graph to search.
  * @param name - Exact export name to look up.
  * @returns One entry per file that exports `name`; empty if no file does (including
@@ -1299,4 +1353,4 @@ declare function createWorkspaceGraph(rootDir: string, options?: {
  */
 declare function getAllProjectFiles(rootDir: string, options?: ScanOptions): string[];
 
-export { type ApiSurface, type ApplyTagsFileResult, type ApplyTagsResult, CALL_EDGE_TYPES, CallEdge, type CalleeEntry, type CallerEntry$1 as CallerEntry, type ChangeImpactCache, type ComplexFunctionEntry, DEFAULT_EXTENSIONS, DEFAULT_IGNORE_DIRS, type DependencyGraph, EXPORT_TRACKING_TYPES, type ExportKind, type FeatureDetectionOptions, type FeatureDomain, type FeatureGraph, type FeatureGraphOptions, type FeatureInfo, FileNode, FileType, type FindComplexFunctionsOptions, type FunctionCallInfo, type GetAffectedOptions, type GetCallersOptions, Graph, type CallerEntry as GraphCallerEntry, type GraphExporter, IMPORT_SYMBOL_TYPES, ImportEdge, type LanguageCoverage, MermaidExporter, type ModuleResponsibility, type ModuleRole, type MokoshConfig, type MonorepoDetector, type MonorepoLayout, NodeCategory, type NodeMeta, type NodeQuery, type ParallelParsingOption, type PathWithSymbols, type ProposeTagsOptions, type PublicExport, type ResponsibilityGraph, type ScanOptions, type SerializedGraph, type SerializedWorkspaceGraph, type SlimNode, type SlimSerializedGraph, StructuredTag, type SymbolCaller, type SymbolImporter, type SymbolMatch, type SymbolPrecision, SymbolTraversalContext, type TestNodeIdentifier, type TraversalOptions, type TraversalVisitor, type TypeEdge, type TypeGraph, type TypeKind, type TypeNode, type TypeQueryResult, WorkspaceGraph, type WorkspacePackage, type WorkspacePackageSummary, type WorkspacePackagesSummary, applyConfig, applyTags, buildApiSurface, buildChangeImpactCache, buildFeatureGraph, buildResponsibilityGraph, buildTypeGraph, computeGraphHash, configToGraphOptions, createImportMap, createWorkspaceGraph, detectAllEntryPoints, detectEntryPoint, detectFeatures, detectMonorepo, filterGraph, findComplexFunctions, findSymbol, getAffected, getAllProjectFiles, getCallers, getDependencies, getDependents, getLanguageCoverage, getNodeMeta, hasCoverageData, isChangeImpactCacheValid, loadChangeImpactCache, loadCoverageMap, loadMokoshConfig, parseQuery, proposeAffectedTests, proposeTags, queryCallGraph, queryChangeImpact, queryTypeGraph, registerConfigMatcher, registerMonorepoDetector, registerParser, registerTestLibrary, registerTestPattern, saveChangeImpactCache, slimSerialize, summarizeWorkspacePackages, toMermaid };
+export { type ApiSurface, type ApplyTagsFileResult, type ApplyTagsResult, CALL_EDGE_TYPES, CallEdge, type CalleeEntry, type CallerEntry$1 as CallerEntry, type ChangeImpactCache, type ComplexFunctionEntry, DEFAULT_EXTENSIONS, DEFAULT_IGNORE_DIRS, type DependencyGraph, type DuplicateGroup, type DuplicateOccurrence, EXPORT_TRACKING_TYPES, type ExportKind, type FeatureDetectionOptions, type FeatureDomain, type FeatureGraph, type FeatureGraphOptions, type FeatureInfo, FileNode, FileType, type FindComplexFunctionsOptions, type FindDuplicatesOptions, type FunctionCallInfo, type GetAffectedOptions, type GetCallersOptions, Graph, type CallerEntry as GraphCallerEntry, type GraphExporter, IMPORT_SYMBOL_TYPES, ImportEdge, type LanguageCoverage, MermaidExporter, type ModuleResponsibility, type ModuleRole, type MokoshConfig, type MonorepoDetector, type MonorepoLayout, NodeCategory, type NodeMeta, type NodeQuery, type ParallelParsingOption, type PathWithSymbols, type ProposeTagsOptions, type PublicExport, type ResponsibilityGraph, type ScanOptions, type SerializedGraph, type SerializedWorkspaceGraph, type SlimNode, type SlimSerializedGraph, StructuredTag, type SymbolCaller, type SymbolImporter, type SymbolMatch, type SymbolPrecision, SymbolTraversalContext, type TestNodeIdentifier, type TraversalOptions, type TraversalVisitor, type TypeEdge, type TypeGraph, type TypeKind, type TypeNode, type TypeQueryResult, WorkspaceGraph, type WorkspacePackage, type WorkspacePackageSummary, type WorkspacePackagesSummary, applyConfig, applyTags, buildApiSurface, buildChangeImpactCache, buildFeatureGraph, buildResponsibilityGraph, buildTypeGraph, computeGraphHash, configToGraphOptions, createImportMap, createWorkspaceGraph, detectAllEntryPoints, detectEntryPoint, detectFeatures, detectMonorepo, filterGraph, findComplexFunctions, findDuplicates, findSymbol, getAffected, getAllProjectFiles, getCallers, getDependencies, getDependents, getLanguageCoverage, getNodeMeta, hasCoverageData, isChangeImpactCacheValid, loadChangeImpactCache, loadCoverageMap, loadMokoshConfig, parseQuery, proposeAffectedTests, proposeTags, queryCallGraph, queryChangeImpact, queryTypeGraph, registerConfigMatcher, registerMonorepoDetector, registerParser, registerTestLibrary, registerTestPattern, saveChangeImpactCache, slimSerialize, summarizeWorkspacePackages, toMermaid };

package/dist/index.d.ts

@@ -357,8 +357,9 @@ interface FunctionCallInfo {
  * field matches `functionName`. Callees are found by looking at the defining
  * file's `callEdges` for edges whose `from` field matches `functionName`.
  *
- * Call edges are populated only for TypeScript/JavaScript files. Functions in
- * other language files will return empty `callers` and `callees` arrays.
+ * Call edges are populated only for TypeScript/JavaScript, Go, and Python files
+ * (see `CALL_EDGE_TYPES` in `../language-support`). Functions in other language
+ * files will return empty `callers` and `callees` arrays.
  *
  * @param {Graph} graph - The import graph that carries `callEdges` on each node.
  * @param {string} functionName - Exact name of the function to look up.
@@ -441,6 +442,56 @@ declare function saveChangeImpactCache(cache: ChangeImpactCache, cachePath: stri
  */
 declare function loadChangeImpactCache(cachePath: string): ChangeImpactCache | null;
 
+/**
+ * Sliding-window (shingle) hashing over a normalized token stream, plus chain-merging of
+ * consecutive matching windows into contiguous duplicate blocks. Shared by every language
+ * `tokenize()` supports — the shingling step itself has no language awareness at all.
+ */
+
+interface DuplicateOccurrence {
+    file: string;
+    startLine: number;
+    endLine: number;
+}
+interface DuplicateGroup {
+    /** Two locations sharing this duplicated block. */
+    occurrences: [DuplicateOccurrence, DuplicateOccurrence];
+    /** Line span covered by the block (measured on the first occurrence). */
+    lines: number;
+    /** Token-window length backing this block, after chain-merging adjacent windows. */
+    tokens: number;
+}
+
+interface FindDuplicatesOptions {
+    /** Minimum duplicated block size, in source lines, to report (default 6). */
+    minLines?: number | undefined;
+    /** Shingle window size, in tokens — the smallest duplicate the scan can detect (default 15). */
+    windowSize?: number | undefined;
+    /** When true (default), string/number literals are normalized too, so only structural shape
+     *  — not the specific values used — drives a match. Set false for stricter, Type-1-only matching. */
+    ignoreLiterals?: boolean | undefined;
+    /** Caps the number of duplicate blocks returned, largest-first (default 50). */
+    limit?: number | undefined;
+    /** Directory names to exclude, matched against any path segment (default `DEFAULT_IGNORE_DIRS`
+     *  — `node_modules`, `dist`, `.git`, `mokosh-cache`, `coverage`, etc.). Pass `[]` to disable. */
+    ignoreDirs?: readonly string[] | undefined;
+}
+/**
+ * @description Scans every file already present in `graph` for cross-file (and within-file)
+ *   duplicated code, using a language-agnostic token-shingling pipeline: comments are stripped
+ *   per `FileType`, identifiers (and, by default, literals) are normalized to placeholders so
+ *   renamed-variable copies still match, then a sliding token window is hashed and chain-merged
+ *   into contiguous blocks. Works uniformly across every language in `DEFAULT_EXTENSIONS`.
+ * @param graph - The graph to scan; its node paths (already ignore-rule-filtered) are the file
+ *   list, re-read from disk since duplication data isn't cached on `FileNode`.
+ * @param rootDir - Absolute project root that graph paths are relative to.
+ * @param options - `minLines`/`windowSize` tune sensitivity; `ignoreLiterals` toggles Type-2 vs
+ *   Type-1 matching; `ignoreDirs` excludes files under matching directory names; `limit` caps
+ *   results. Lock files are always excluded, independent of `ignoreDirs`.
+ * @returns Duplicate blocks, each a pair of occurrences, sorted largest-first.
+ */
+declare function findDuplicates(graph: Graph, rootDir: string, options?: FindDuplicatesOptions): Promise<DuplicateGroup[]>;
+
 /** Controls how aggressively `detectFeatures` promotes files to features. */
 interface FeatureDetectionOptions {
     /**
@@ -956,7 +1007,8 @@ interface FindComplexFunctionsOptions {
 }
 /**
  * @description Scans every file's per-function complexity breakdown and returns functions/methods
- *   at or above the given threshold, sorted worst-first. TypeScript/JavaScript only.
+ *   at or above the given threshold, sorted worst-first. Populated for TypeScript/JavaScript, Go,
+ *   and Python.
  * @param graph - The graph to scan.
  * @param options - `metric` picks which score to threshold/sort on (default `cognitiveComplexity`); `threshold` is the minimum score to include (default 10); `limit` caps the results (default 20).
  * @returns Matching functions, sorted worst-first.
@@ -1012,7 +1064,7 @@ declare function hasCoverageData(graph: Graph): boolean;
 
 /** Symbol-name lookup across the whole graph — generalizes queryCallGraph beyond TS/JS functions. */
 
-type SymbolPrecision = "call" | "import-symbol" | "file-level";
+type SymbolPrecision = "call" | "file-level";
 interface SymbolCaller {
     file: string;
     callerFunction: string;
@@ -1031,9 +1083,11 @@ interface SymbolMatch {
 /**
  * @description Finds every file that exports a symbol by name, with the best available
  *   usage info per match. Precision depends on what the defining file's language parser
- *   tracks: TS/JS gets function-level callers via call edges, Python gets named-import
- *   tracking, everything else falls back to whole-file dependents (import-level, not
- *   symbol-level — the file might not even use this specific export).
+ *   tracks: TS/JS, Go, and Python get function-level callers via call edges (`"call"`);
+ *   everything else falls back to whole-file dependents (`"file-level"` — import-level, not
+ *   symbol-level, since the file might not even use this specific export). Coverage still
+ *   differs across the `"call"` languages: TS/JS tracks any directly imported symbol, Go only
+ *   package-qualified calls, Python only bare calls to `from <module> import <name>` symbols.
  * @param graph - The graph to search.
  * @param name - Exact export name to look up.
  * @returns One entry per file that exports `name`; empty if no file does (including
@@ -1299,4 +1353,4 @@ declare function createWorkspaceGraph(rootDir: string, options?: {
  */
 declare function getAllProjectFiles(rootDir: string, options?: ScanOptions): string[];
 
-export { type ApiSurface, type ApplyTagsFileResult, type ApplyTagsResult, CALL_EDGE_TYPES, CallEdge, type CalleeEntry, type CallerEntry$1 as CallerEntry, type ChangeImpactCache, type ComplexFunctionEntry, DEFAULT_EXTENSIONS, DEFAULT_IGNORE_DIRS, type DependencyGraph, EXPORT_TRACKING_TYPES, type ExportKind, type FeatureDetectionOptions, type FeatureDomain, type FeatureGraph, type FeatureGraphOptions, type FeatureInfo, FileNode, FileType, type FindComplexFunctionsOptions, type FunctionCallInfo, type GetAffectedOptions, type GetCallersOptions, Graph, type CallerEntry as GraphCallerEntry, type GraphExporter, IMPORT_SYMBOL_TYPES, ImportEdge, type LanguageCoverage, MermaidExporter, type ModuleResponsibility, type ModuleRole, type MokoshConfig, type MonorepoDetector, type MonorepoLayout, NodeCategory, type NodeMeta, type NodeQuery, type ParallelParsingOption, type PathWithSymbols, type ProposeTagsOptions, type PublicExport, type ResponsibilityGraph, type ScanOptions, type SerializedGraph, type SerializedWorkspaceGraph, type SlimNode, type SlimSerializedGraph, StructuredTag, type SymbolCaller, type SymbolImporter, type SymbolMatch, type SymbolPrecision, SymbolTraversalContext, type TestNodeIdentifier, type TraversalOptions, type TraversalVisitor, type TypeEdge, type TypeGraph, type TypeKind, type TypeNode, type TypeQueryResult, WorkspaceGraph, type WorkspacePackage, type WorkspacePackageSummary, type WorkspacePackagesSummary, applyConfig, applyTags, buildApiSurface, buildChangeImpactCache, buildFeatureGraph, buildResponsibilityGraph, buildTypeGraph, computeGraphHash, configToGraphOptions, createImportMap, createWorkspaceGraph, detectAllEntryPoints, detectEntryPoint, detectFeatures, detectMonorepo, filterGraph, findComplexFunctions, findSymbol, getAffected, getAllProjectFiles, getCallers, getDependencies, getDependents, getLanguageCoverage, getNodeMeta, hasCoverageData, isChangeImpactCacheValid, loadChangeImpactCache, loadCoverageMap, loadMokoshConfig, parseQuery, proposeAffectedTests, proposeTags, queryCallGraph, queryChangeImpact, queryTypeGraph, registerConfigMatcher, registerMonorepoDetector, registerParser, registerTestLibrary, registerTestPattern, saveChangeImpactCache, slimSerialize, summarizeWorkspacePackages, toMermaid };
+export { type ApiSurface, type ApplyTagsFileResult, type ApplyTagsResult, CALL_EDGE_TYPES, CallEdge, type CalleeEntry, type CallerEntry$1 as CallerEntry, type ChangeImpactCache, type ComplexFunctionEntry, DEFAULT_EXTENSIONS, DEFAULT_IGNORE_DIRS, type DependencyGraph, type DuplicateGroup, type DuplicateOccurrence, EXPORT_TRACKING_TYPES, type ExportKind, type FeatureDetectionOptions, type FeatureDomain, type FeatureGraph, type FeatureGraphOptions, type FeatureInfo, FileNode, FileType, type FindComplexFunctionsOptions, type FindDuplicatesOptions, type FunctionCallInfo, type GetAffectedOptions, type GetCallersOptions, Graph, type CallerEntry as GraphCallerEntry, type GraphExporter, IMPORT_SYMBOL_TYPES, ImportEdge, type LanguageCoverage, MermaidExporter, type ModuleResponsibility, type ModuleRole, type MokoshConfig, type MonorepoDetector, type MonorepoLayout, NodeCategory, type NodeMeta, type NodeQuery, type ParallelParsingOption, type PathWithSymbols, type ProposeTagsOptions, type PublicExport, type ResponsibilityGraph, type ScanOptions, type SerializedGraph, type SerializedWorkspaceGraph, type SlimNode, type SlimSerializedGraph, StructuredTag, type SymbolCaller, type SymbolImporter, type SymbolMatch, type SymbolPrecision, SymbolTraversalContext, type TestNodeIdentifier, type TraversalOptions, type TraversalVisitor, type TypeEdge, type TypeGraph, type TypeKind, type TypeNode, type TypeQueryResult, WorkspaceGraph, type WorkspacePackage, type WorkspacePackageSummary, type WorkspacePackagesSummary, applyConfig, applyTags, buildApiSurface, buildChangeImpactCache, buildFeatureGraph, buildResponsibilityGraph, buildTypeGraph, computeGraphHash, configToGraphOptions, createImportMap, createWorkspaceGraph, detectAllEntryPoints, detectEntryPoint, detectFeatures, detectMonorepo, filterGraph, findComplexFunctions, findDuplicates, findSymbol, getAffected, getAllProjectFiles, getCallers, getDependencies, getDependents, getLanguageCoverage, getNodeMeta, hasCoverageData, isChangeImpactCacheValid, loadChangeImpactCache, loadCoverageMap, loadMokoshConfig, parseQuery, proposeAffectedTests, proposeTags, queryCallGraph, queryChangeImpact, queryTypeGraph, registerConfigMatcher, registerMonorepoDetector, registerParser, registerTestLibrary, registerTestPattern, saveChangeImpactCache, slimSerialize, summarizeWorkspacePackages, toMermaid };