org-qmd 0.1.0

package/bin/qmd

@@ -0,0 +1,32 @@
+#!/bin/sh
+# Resolve symlinks so global installs (npm link / npm install -g) can find the
+# actual package directory instead of the global bin directory.
+SOURCE="$0"
+while [ -L "$SOURCE" ]; do
+  SOURCE_DIR="$(cd -P "$(dirname "$SOURCE")" && pwd)"
+  TARGET="$(readlink "$SOURCE")"
+  case "$TARGET" in
+    /*) SOURCE="$TARGET" ;;
+    *) SOURCE="$SOURCE_DIR/$TARGET" ;;
+  esac
+done
+
+# Detect the runtime used to install this package and use the matching one
+# to avoid native module ABI mismatches (e.g., better-sqlite3 compiled for bun vs node)
+DIR="$(cd -P "$(dirname "$SOURCE")/.." && pwd)"
+
+# Detect the package manager that installed dependencies by checking lockfiles.
+# $BUN_INSTALL is intentionally NOT checked — it only indicates that bun exists
+# on the system, not that it was used to install this package (see #361).
+#
+# package-lock.json takes priority: if it exists, npm installed the native
+# modules for Node.  The repo ships bun.lock, so without this check, source
+# builds that use npm would be incorrectly routed to bun, causing ABI
+# mismatches with better-sqlite3 / sqlite-vec (see #381).
+if [ -f "$DIR/package-lock.json" ]; then
+  exec node "$DIR/dist/cli/qmd.js" "$@"
+elif [ -f "$DIR/bun.lock" ] || [ -f "$DIR/bun.lockb" ]; then
+  exec bun "$DIR/dist/cli/qmd.js" "$@"
+else
+  exec node "$DIR/dist/cli/qmd.js" "$@"
+fi

package/dist/ast.d.ts

@@ -0,0 +1,64 @@
+/**
+ * AST-aware chunking support via web-tree-sitter.
+ *
+ * Provides language detection, AST break point extraction for supported
+ * code file types, and a stub for future symbol extraction.
+ *
+ * All functions degrade gracefully: parse failures or unsupported languages
+ * return empty arrays, falling back to regex-only chunking.
+ *
+ * ## Dependency Note
+ *
+ * Grammar packages (tree-sitter-typescript, etc.) are listed as
+ * optionalDependencies with pinned versions. They ship native prebuilds
+ * and source files (~72 MB total) but QMD only uses the .wasm files
+ * (~5 MB). If install size becomes a concern, the .wasm files can be
+ * bundled directly in the repo (e.g. assets/grammars/) and resolved
+ * via import.meta.url instead of require.resolve(), eliminating the
+ * grammar packages entirely.
+ */
+import type { BreakPoint } from "./store.js";
+export type SupportedLanguage = "typescript" | "tsx" | "javascript" | "python" | "go" | "rust";
+/**
+ * Detect language from file path extension.
+ * Returns null for unsupported or unknown extensions (including .md).
+ */
+export declare function detectLanguage(filepath: string): SupportedLanguage | null;
+/**
+ * Parse a source file and return break points at AST node boundaries.
+ *
+ * Returns an empty array for unsupported languages, parse failures,
+ * or grammar loading failures. Never throws.
+ *
+ * @param content - The file content to parse.
+ * @param filepath - The file path (used for language detection).
+ * @returns Array of BreakPoint objects suitable for merging with regex break points.
+ */
+export declare function getASTBreakPoints(content: string, filepath: string): Promise<BreakPoint[]>;
+/**
+ * Check which tree-sitter grammars are available.
+ * Returns a status object for each supported language.
+ */
+export declare function getASTStatus(): Promise<{
+    available: boolean;
+    languages: {
+        language: SupportedLanguage;
+        available: boolean;
+        error?: string;
+    }[];
+}>;
+/**
+ * Metadata about a code symbol within a chunk.
+ * Stubbed for Phase 2 — always returns empty array in Phase 1.
+ */
+export interface SymbolInfo {
+    name: string;
+    kind: string;
+    signature?: string;
+    line: number;
+}
+/**
+ * Extract symbol metadata for code within a byte range.
+ * Stubbed for Phase 2 — returns empty array.
+ */
+export declare function extractSymbols(_content: string, _language: string, _startPos: number, _endPos: number): SymbolInfo[];

package/dist/ast.js

@@ -0,0 +1,324 @@
+/**
+ * AST-aware chunking support via web-tree-sitter.
+ *
+ * Provides language detection, AST break point extraction for supported
+ * code file types, and a stub for future symbol extraction.
+ *
+ * All functions degrade gracefully: parse failures or unsupported languages
+ * return empty arrays, falling back to regex-only chunking.
+ *
+ * ## Dependency Note
+ *
+ * Grammar packages (tree-sitter-typescript, etc.) are listed as
+ * optionalDependencies with pinned versions. They ship native prebuilds
+ * and source files (~72 MB total) but QMD only uses the .wasm files
+ * (~5 MB). If install size becomes a concern, the .wasm files can be
+ * bundled directly in the repo (e.g. assets/grammars/) and resolved
+ * via import.meta.url instead of require.resolve(), eliminating the
+ * grammar packages entirely.
+ */
+import { createRequire } from "node:module";
+import { extname } from "node:path";
+const EXTENSION_MAP = {
+    ".ts": "typescript",
+    ".tsx": "tsx",
+    ".js": "javascript",
+    ".jsx": "tsx",
+    ".mts": "typescript",
+    ".cts": "typescript",
+    ".mjs": "javascript",
+    ".cjs": "javascript",
+    ".py": "python",
+    ".go": "go",
+    ".rs": "rust",
+};
+/**
+ * Detect language from file path extension.
+ * Returns null for unsupported or unknown extensions (including .md).
+ */
+export function detectLanguage(filepath) {
+    const ext = extname(filepath).toLowerCase();
+    return EXTENSION_MAP[ext] ?? null;
+}
+// =============================================================================
+// Grammar Resolution
+// =============================================================================
+/**
+ * Maps language to the npm package and wasm filename for the grammar.
+ */
+const GRAMMAR_MAP = {
+    typescript: { pkg: "tree-sitter-typescript", wasm: "tree-sitter-typescript.wasm" },
+    tsx: { pkg: "tree-sitter-typescript", wasm: "tree-sitter-tsx.wasm" },
+    javascript: { pkg: "tree-sitter-typescript", wasm: "tree-sitter-typescript.wasm" },
+    python: { pkg: "tree-sitter-python", wasm: "tree-sitter-python.wasm" },
+    go: { pkg: "tree-sitter-go", wasm: "tree-sitter-go.wasm" },
+    rust: { pkg: "tree-sitter-rust", wasm: "tree-sitter-rust.wasm" },
+};
+// =============================================================================
+// Per-Language Query Definitions
+// =============================================================================
+/**
+ * Tree-sitter S-expression queries for each language.
+ * Each capture name maps to a break point score via SCORE_MAP.
+ *
+ * For TypeScript/JavaScript, we match export_statement wrappers to get the
+ * correct start position (before `export`), plus bare declarations for
+ * non-exported code.
+ */
+const LANGUAGE_QUERIES = {
+    typescript: `
+    (export_statement) @export
+    (class_declaration) @class
+    (function_declaration) @func
+    (method_definition) @method
+    (interface_declaration) @iface
+    (type_alias_declaration) @type
+    (enum_declaration) @enum
+    (import_statement) @import
+    (lexical_declaration (variable_declarator value: (arrow_function))) @func
+    (lexical_declaration (variable_declarator value: (function_expression))) @func
+  `,
+    tsx: `
+    (export_statement) @export
+    (class_declaration) @class
+    (function_declaration) @func
+    (method_definition) @method
+    (interface_declaration) @iface
+    (type_alias_declaration) @type
+    (enum_declaration) @enum
+    (import_statement) @import
+    (lexical_declaration (variable_declarator value: (arrow_function))) @func
+    (lexical_declaration (variable_declarator value: (function_expression))) @func
+  `,
+    javascript: `
+    (export_statement) @export
+    (class_declaration) @class
+    (function_declaration) @func
+    (method_definition) @method
+    (import_statement) @import
+    (lexical_declaration (variable_declarator value: (arrow_function))) @func
+    (lexical_declaration (variable_declarator value: (function_expression))) @func
+  `,
+    python: `
+    (class_definition) @class
+    (function_definition) @func
+    (decorated_definition) @decorated
+    (import_statement) @import
+    (import_from_statement) @import
+  `,
+    go: `
+    (type_declaration) @type
+    (function_declaration) @func
+    (method_declaration) @method
+    (import_declaration) @import
+  `,
+    rust: `
+    (struct_item) @struct
+    (impl_item) @impl
+    (function_item) @func
+    (trait_item) @trait
+    (enum_item) @enum
+    (use_declaration) @import
+    (type_item) @type
+    (mod_item) @mod
+  `,
+};
+/**
+ * Score mapping from capture names to break point scores.
+ * Aligned with the markdown BREAK_PATTERNS scale (h1=100, h2=90, etc.)
+ * so findBestCutoff() decay works unchanged.
+ */
+const SCORE_MAP = {
+    class: 100,
+    iface: 100,
+    struct: 100,
+    trait: 100,
+    impl: 100,
+    mod: 100,
+    export: 90,
+    func: 90,
+    method: 90,
+    decorated: 90,
+    type: 80,
+    enum: 80,
+    import: 60,
+};
+// =============================================================================
+// Parser Caching & Initialization
+// =============================================================================
+let ParserClass = null;
+let LanguageClass = null;
+let QueryClass = null;
+let initPromise = null;
+/** Languages that have already failed to load — warn only once per process. */
+const failedLanguages = new Set();
+/** Cached grammar load promises. */
+const grammarCache = new Map();
+/** Cached compiled queries per language. */
+const queryCache = new Map();
+/**
+ * Initialize web-tree-sitter. Called once and cached.
+ */
+async function ensureInit() {
+    if (!initPromise) {
+        initPromise = (async () => {
+            const mod = await import("web-tree-sitter");
+            ParserClass = mod.Parser;
+            LanguageClass = mod.Language;
+            QueryClass = mod.Query;
+            await ParserClass.init();
+        })();
+    }
+    return initPromise;
+}
+/**
+ * Resolve the filesystem path to a grammar .wasm file.
+ * Uses createRequire to resolve from installed dependency packages.
+ */
+function resolveGrammarPath(language) {
+    const { pkg, wasm } = GRAMMAR_MAP[language];
+    const require = createRequire(import.meta.url);
+    return require.resolve(`${pkg}/${wasm}`);
+}
+/**
+ * Load and cache a grammar for the given language.
+ * Returns null on failure (logs once per language).
+ */
+async function loadGrammar(language) {
+    if (failedLanguages.has(language))
+        return null;
+    const wasmKey = GRAMMAR_MAP[language].wasm;
+    if (!grammarCache.has(wasmKey)) {
+        grammarCache.set(wasmKey, (async () => {
+            const path = resolveGrammarPath(language);
+            return LanguageClass.load(path);
+        })());
+    }
+    try {
+        return await grammarCache.get(wasmKey);
+    }
+    catch (err) {
+        failedLanguages.add(language);
+        grammarCache.delete(wasmKey);
+        console.warn(`[qmd] Failed to load tree-sitter grammar for ${language}: ${err}`);
+        return null;
+    }
+}
+/**
+ * Get or create a compiled query for the given language.
+ */
+function getQuery(language, grammar) {
+    if (!queryCache.has(language)) {
+        const source = LANGUAGE_QUERIES[language];
+        const query = new QueryClass(grammar, source);
+        queryCache.set(language, query);
+    }
+    return queryCache.get(language);
+}
+// =============================================================================
+// AST Break Point Extraction
+// =============================================================================
+/**
+ * Parse a source file and return break points at AST node boundaries.
+ *
+ * Returns an empty array for unsupported languages, parse failures,
+ * or grammar loading failures. Never throws.
+ *
+ * @param content - The file content to parse.
+ * @param filepath - The file path (used for language detection).
+ * @returns Array of BreakPoint objects suitable for merging with regex break points.
+ */
+export async function getASTBreakPoints(content, filepath) {
+    const language = detectLanguage(filepath);
+    if (!language)
+        return [];
+    try {
+        await ensureInit();
+        const grammar = await loadGrammar(language);
+        if (!grammar)
+            return [];
+        const parser = new ParserClass();
+        parser.setLanguage(grammar);
+        const tree = parser.parse(content);
+        if (!tree) {
+            parser.delete();
+            return [];
+        }
+        const query = getQuery(language, grammar);
+        const captures = query.captures(tree.rootNode);
+        // Deduplicate: at each byte position, keep the highest-scoring capture.
+        // This handles cases like export_statement wrapping a class_declaration
+        // at different offsets — we want the outermost (earliest) position.
+        const seen = new Map();
+        for (const cap of captures) {
+            const pos = cap.node.startIndex;
+            const score = SCORE_MAP[cap.name] ?? 20;
+            const type = `ast:${cap.name}`;
+            const existing = seen.get(pos);
+            if (!existing || score > existing.score) {
+                seen.set(pos, { pos, score, type });
+            }
+        }
+        tree.delete();
+        parser.delete();
+        return Array.from(seen.values()).sort((a, b) => a.pos - b.pos);
+    }
+    catch (err) {
+        console.warn(`[qmd] AST parse failed for ${filepath}, falling back to regex: ${err instanceof Error ? err.message : err}`);
+        return [];
+    }
+}
+// =============================================================================
+// Health / Status
+// =============================================================================
+/**
+ * Check which tree-sitter grammars are available.
+ * Returns a status object for each supported language.
+ */
+export async function getASTStatus() {
+    const languages = [];
+    try {
+        await ensureInit();
+    }
+    catch (err) {
+        return {
+            available: false,
+            languages: Object.keys(GRAMMAR_MAP).map(lang => ({
+                language: lang,
+                available: false,
+                error: `web-tree-sitter init failed: ${err instanceof Error ? err.message : err}`,
+            })),
+        };
+    }
+    for (const lang of Object.keys(GRAMMAR_MAP)) {
+        try {
+            const grammar = await loadGrammar(lang);
+            if (grammar) {
+                // Also verify the query compiles
+                getQuery(lang, grammar);
+                languages.push({ language: lang, available: true });
+            }
+            else {
+                languages.push({ language: lang, available: false, error: "grammar failed to load" });
+            }
+        }
+        catch (err) {
+            languages.push({
+                language: lang,
+                available: false,
+                error: err instanceof Error ? err.message : String(err),
+            });
+        }
+    }
+    return {
+        available: languages.some(l => l.available),
+        languages,
+    };
+}
+/**
+ * Extract symbol metadata for code within a byte range.
+ * Stubbed for Phase 2 — returns empty array.
+ */
+export function extractSymbols(_content, _language, _startPos, _endPos) {
+    return [];
+}

package/dist/cli/formatter.d.ts

@@ -0,0 +1,120 @@
+/**
+ * formatter.ts - Output formatting utilities for QMD
+ *
+ * Provides methods to format search results and documents into various output formats:
+ * JSON, CSV, XML, Markdown, files list, and CLI (colored terminal output).
+ */
+import type { SearchResult, MultiGetResult, DocumentResult } from "../store.js";
+export type { SearchResult, MultiGetResult, DocumentResult };
+export type MultiGetFile = {
+    filepath: string;
+    displayPath: string;
+    title: string;
+    body: string;
+    context?: string | null;
+    skipped: false;
+} | {
+    filepath: string;
+    displayPath: string;
+    title: string;
+    body: string;
+    context?: string | null;
+    skipped: true;
+    skipReason: string;
+};
+export type OutputFormat = "cli" | "csv" | "md" | "xml" | "files" | "json";
+export type FormatOptions = {
+    full?: boolean;
+    query?: string;
+    useColor?: boolean;
+    lineNumbers?: boolean;
+    intent?: string;
+};
+/**
+ * Add line numbers to text content.
+ * Each line becomes: "{lineNum}: {content}"
+ * @param text The text to add line numbers to
+ * @param startLine Optional starting line number (default: 1)
+ */
+export declare function addLineNumbers(text: string, startLine?: number): string;
+/**
+ * Extract short docid from a full hash (first 6 characters).
+ */
+export declare function getDocid(hash: string): string;
+export declare function escapeCSV(value: string | null | number): string;
+export declare function escapeXml(str: string): string;
+/**
+ * Format search results as JSON
+ */
+export declare function searchResultsToJson(results: SearchResult[], opts?: FormatOptions): string;
+/**
+ * Format search results as CSV
+ */
+export declare function searchResultsToCsv(results: SearchResult[], opts?: FormatOptions): string;
+/**
+ * Format search results as simple files list (docid,score,filepath,context)
+ */
+export declare function searchResultsToFiles(results: SearchResult[]): string;
+/**
+ * Format search results as Markdown
+ */
+export declare function searchResultsToMarkdown(results: SearchResult[], opts?: FormatOptions): string;
+/**
+ * Format search results as XML
+ */
+export declare function searchResultsToXml(results: SearchResult[], opts?: FormatOptions): string;
+/**
+ * Format search results for MCP (simpler CSV format with pre-extracted snippets)
+ */
+export declare function searchResultsToMcpCsv(results: {
+    docid: string;
+    file: string;
+    title: string;
+    score: number;
+    context: string | null;
+    snippet: string;
+}[]): string;
+/**
+ * Format documents as JSON
+ */
+export declare function documentsToJson(results: MultiGetFile[]): string;
+/**
+ * Format documents as CSV
+ */
+export declare function documentsToCsv(results: MultiGetFile[]): string;
+/**
+ * Format documents as files list
+ */
+export declare function documentsToFiles(results: MultiGetFile[]): string;
+/**
+ * Format documents as Markdown
+ */
+export declare function documentsToMarkdown(results: MultiGetFile[]): string;
+/**
+ * Format documents as XML
+ */
+export declare function documentsToXml(results: MultiGetFile[]): string;
+/**
+ * Format a single DocumentResult as JSON
+ */
+export declare function documentToJson(doc: DocumentResult): string;
+/**
+ * Format a single DocumentResult as Markdown
+ */
+export declare function documentToMarkdown(doc: DocumentResult): string;
+/**
+ * Format a single DocumentResult as XML
+ */
+export declare function documentToXml(doc: DocumentResult): string;
+/**
+ * Format a single document to the specified format
+ */
+export declare function formatDocument(doc: DocumentResult, format: OutputFormat): string;
+/**
+ * Format search results to the specified output format
+ */
+export declare function formatSearchResults(results: SearchResult[], format: OutputFormat, opts?: FormatOptions): string;
+/**
+ * Format documents to the specified output format
+ */
+export declare function formatDocuments(results: MultiGetFile[], format: OutputFormat): string;