@opencodehub/ingestion 0.1.0

package/dist/pipeline/dep-parsers/types.d.ts

@@ -0,0 +1,63 @@
+/**
+ * Shared types for the per-ecosystem manifest / lockfile parsers that the
+ * `dependencies` pipeline phase invokes.
+ *
+ * Each parser is a pure async function: given a repository-relative
+ * manifest or lockfile path and the absolute filesystem path, it returns
+ * a de-duplicated list of parsed dependencies. Parsers never throw on
+ * malformed input — they log via `onWarn` and return `[]` instead.
+ *
+ * License detection: each parser populates `ParsedDependency.license`
+ * directly from the manifest field when one is present (SPDX identifier
+ * or a free-form string — the `dependencies` phase normalises through
+ * `spdx-correct` before storage). The value is `undefined` when the
+ * manifest does not carry license metadata; the phase converts that to
+ * the canonical `"UNKNOWN"` sentinel on the DependencyNode.
+ */
+import type { DependencyNode } from "@opencodehub/core-types";
+export type Ecosystem = DependencyNode["ecosystem"];
+/**
+ * Minimal parsed dependency record. The `dependencies` phase promotes
+ * these into full `DependencyNode` entries once all parsers have run.
+ */
+export interface ParsedDependency {
+    readonly ecosystem: Ecosystem;
+    readonly name: string;
+    readonly version: string;
+    /** Repo-relative POSIX path to the manifest / lockfile. */
+    readonly lockfileSource: string;
+    /**
+     * License declaration as it appears in the manifest (SPDX id or a
+     * free-form expression). `undefined` when the manifest entry did not
+     * carry a license; the `dependencies` phase maps that to the
+     * `"UNKNOWN"` sentinel on the DependencyNode.
+     */
+    readonly license?: string;
+}
+/** Warning sink threaded through every parser for best-effort logging. */
+export type WarnFn = (message: string) => void;
+/** Input shape accepted by every parser. */
+export interface ParseDepsInput {
+    /** Absolute filesystem path to the manifest / lockfile. */
+    readonly absPath: string;
+    /** Repo-relative POSIX path (what goes into `lockfileSource`). */
+    readonly relPath: string;
+    /** Absolute repo root (needed by parsers that must locate sibling files). */
+    readonly repoRoot: string;
+    /** Warning hook. Errors must not propagate out of the parser. */
+    readonly onWarn: WarnFn;
+}
+export type ParseDepsFn = (input: ParseDepsInput) => Promise<readonly ParsedDependency[]>;
+/** Canonical comparison for stable `ParsedDependency[]` ordering. */
+export declare function compareParsedDependency(a: ParsedDependency, b: ParsedDependency): number;
+/**
+ * De-duplicate a list of `ParsedDependency` on
+ * `(ecosystem, name, version, lockfileSource)`. When two records collide,
+ * we keep the first occurrence but prefer a defined `license` over an
+ * undefined one — a lockfile with no license + a sibling manifest that
+ * carries one should surface as "has license" after dedup. Callers rely
+ * on the returned array being sorted canonically so graph output is
+ * byte-stable.
+ */
+export declare function dedupAndSort(deps: readonly ParsedDependency[]): readonly ParsedDependency[];
+//# sourceMappingURL=types.d.ts.map

package/dist/pipeline/dep-parsers/types.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"types.d.ts","sourceRoot":"","sources":["../../../src/pipeline/dep-parsers/types.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;GAeG;AAEH,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,yBAAyB,CAAC;AAE9D,MAAM,MAAM,SAAS,GAAG,cAAc,CAAC,WAAW,CAAC,CAAC;AAEpD;;;GAGG;AACH,MAAM,WAAW,gBAAgB;IAC/B,QAAQ,CAAC,SAAS,EAAE,SAAS,CAAC;IAC9B,QAAQ,CAAC,IAAI,EAAE,MAAM,CAAC;IACtB,QAAQ,CAAC,OAAO,EAAE,MAAM,CAAC;IACzB,2DAA2D;IAC3D,QAAQ,CAAC,cAAc,EAAE,MAAM,CAAC;IAChC;;;;;OAKG;IACH,QAAQ,CAAC,OAAO,CAAC,EAAE,MAAM,CAAC;CAC3B;AAED,0EAA0E;AAC1E,MAAM,MAAM,MAAM,GAAG,CAAC,OAAO,EAAE,MAAM,KAAK,IAAI,CAAC;AAE/C,4CAA4C;AAC5C,MAAM,WAAW,cAAc;IAC7B,2DAA2D;IAC3D,QAAQ,CAAC,OAAO,EAAE,MAAM,CAAC;IACzB,kEAAkE;IAClE,QAAQ,CAAC,OAAO,EAAE,MAAM,CAAC;IACzB,6EAA6E;IAC7E,QAAQ,CAAC,QAAQ,EAAE,MAAM,CAAC;IAC1B,iEAAiE;IACjE,QAAQ,CAAC,MAAM,EAAE,MAAM,CAAC;CACzB;AAED,MAAM,MAAM,WAAW,GAAG,CAAC,KAAK,EAAE,cAAc,KAAK,OAAO,CAAC,SAAS,gBAAgB,EAAE,CAAC,CAAC;AAE1F,qEAAqE;AACrE,wBAAgB,uBAAuB,CAAC,CAAC,EAAE,gBAAgB,EAAE,CAAC,EAAE,gBAAgB,GAAG,MAAM,CAMxF;AAED;;;;;;;;GAQG;AACH,wBAAgB,YAAY,CAAC,IAAI,EAAE,SAAS,gBAAgB,EAAE,GAAG,SAAS,gBAAgB,EAAE,CAiB3F"}

package/dist/pipeline/dep-parsers/types.js

@@ -0,0 +1,56 @@
+/**
+ * Shared types for the per-ecosystem manifest / lockfile parsers that the
+ * `dependencies` pipeline phase invokes.
+ *
+ * Each parser is a pure async function: given a repository-relative
+ * manifest or lockfile path and the absolute filesystem path, it returns
+ * a de-duplicated list of parsed dependencies. Parsers never throw on
+ * malformed input — they log via `onWarn` and return `[]` instead.
+ *
+ * License detection: each parser populates `ParsedDependency.license`
+ * directly from the manifest field when one is present (SPDX identifier
+ * or a free-form string — the `dependencies` phase normalises through
+ * `spdx-correct` before storage). The value is `undefined` when the
+ * manifest does not carry license metadata; the phase converts that to
+ * the canonical `"UNKNOWN"` sentinel on the DependencyNode.
+ */
+/** Canonical comparison for stable `ParsedDependency[]` ordering. */
+export function compareParsedDependency(a, b) {
+    if (a.ecosystem !== b.ecosystem)
+        return a.ecosystem < b.ecosystem ? -1 : 1;
+    if (a.name !== b.name)
+        return a.name < b.name ? -1 : 1;
+    if (a.version !== b.version)
+        return a.version < b.version ? -1 : 1;
+    if (a.lockfileSource !== b.lockfileSource)
+        return a.lockfileSource < b.lockfileSource ? -1 : 1;
+    return 0;
+}
+/**
+ * De-duplicate a list of `ParsedDependency` on
+ * `(ecosystem, name, version, lockfileSource)`. When two records collide,
+ * we keep the first occurrence but prefer a defined `license` over an
+ * undefined one — a lockfile with no license + a sibling manifest that
+ * carries one should surface as "has license" after dedup. Callers rely
+ * on the returned array being sorted canonically so graph output is
+ * byte-stable.
+ */
+export function dedupAndSort(deps) {
+    const byKey = new Map();
+    for (const d of deps) {
+        const key = `${d.ecosystem}\x00${d.name}\x00${d.version}\x00${d.lockfileSource}`;
+        const existing = byKey.get(key);
+        if (existing === undefined) {
+            byKey.set(key, d);
+            continue;
+        }
+        // Prefer an entry carrying a license over one that doesn't.
+        if (existing.license === undefined && d.license !== undefined) {
+            byKey.set(key, d);
+        }
+    }
+    const out = [...byKey.values()];
+    out.sort(compareParsedDependency);
+    return out;
+}
+//# sourceMappingURL=types.js.map

package/dist/pipeline/dep-parsers/types.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"types.js","sourceRoot":"","sources":["../../../src/pipeline/dep-parsers/types.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;GAeG;AA0CH,qEAAqE;AACrE,MAAM,UAAU,uBAAuB,CAAC,CAAmB,EAAE,CAAmB;IAC9E,IAAI,CAAC,CAAC,SAAS,KAAK,CAAC,CAAC,SAAS;QAAE,OAAO,CAAC,CAAC,SAAS,GAAG,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IAC3E,IAAI,CAAC,CAAC,IAAI,KAAK,CAAC,CAAC,IAAI;QAAE,OAAO,CAAC,CAAC,IAAI,GAAG,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IACvD,IAAI,CAAC,CAAC,OAAO,KAAK,CAAC,CAAC,OAAO;QAAE,OAAO,CAAC,CAAC,OAAO,GAAG,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IACnE,IAAI,CAAC,CAAC,cAAc,KAAK,CAAC,CAAC,cAAc;QAAE,OAAO,CAAC,CAAC,cAAc,GAAG,CAAC,CAAC,cAAc,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IAC/F,OAAO,CAAC,CAAC;AACX,CAAC;AAED;;;;;;;;GAQG;AACH,MAAM,UAAU,YAAY,CAAC,IAAiC;IAC5D,MAAM,KAAK,GAAG,IAAI,GAAG,EAA4B,CAAC;IAClD,KAAK,MAAM,CAAC,IAAI,IAAI,EAAE,CAAC;QACrB,MAAM,GAAG,GAAG,GAAG,CAAC,CAAC,SAAS,OAAO,CAAC,CAAC,IAAI,OAAO,CAAC,CAAC,OAAO,OAAO,CAAC,CAAC,cAAc,EAAE,CAAC;QACjF,MAAM,QAAQ,GAAG,KAAK,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;QAChC,IAAI,QAAQ,KAAK,SAAS,EAAE,CAAC;YAC3B,KAAK,CAAC,GAAG,CAAC,GAAG,EAAE,CAAC,CAAC,CAAC;YAClB,SAAS;QACX,CAAC;QACD,4DAA4D;QAC5D,IAAI,QAAQ,CAAC,OAAO,KAAK,SAAS,IAAI,CAAC,CAAC,OAAO,KAAK,SAAS,EAAE,CAAC;YAC9D,KAAK,CAAC,GAAG,CAAC,GAAG,EAAE,CAAC,CAAC,CAAC;QACpB,CAAC;IACH,CAAC;IACD,MAAM,GAAG,GAAG,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC,CAAC;IAChC,GAAG,CAAC,IAAI,CAAC,uBAAuB,CAAC,CAAC;IAClC,OAAO,GAAG,CAAC;AACb,CAAC"}

package/dist/pipeline/gitignore-stack.d.ts

@@ -0,0 +1,44 @@
+/**
+ * Gitignore rule stacking helper — encapsulates the "layered rule set"
+ * semantics git itself honours when nested `.gitignore` files are present.
+ *
+ * Git's rules (see `man gitignore`):
+ *   - A `.gitignore` in a subdirectory shadows and extends the rules from
+ *     its parents. Rules are evaluated from the repo root downward.
+ *   - A negation (`!pat`) at a deeper layer can re-include a path that a
+ *     parent layer excluded (and vice versa).
+ *   - For a given path, evaluation walks the chain top-down; later matches
+ *     override earlier ones. The final state (ignored / not-ignored) wins.
+ *
+ * This module does NOT scan the filesystem; it takes a pre-loaded
+ * {@link GitignoreChain} (directory-relative paths → parsed rules) and
+ * resolves the effective ignored state for a given path. The scan-time
+ * walker in `gitignore.ts` handles the actual filesystem recursion.
+ */
+import type { IgnoreRule } from "./gitignore.js";
+/**
+ * Map from repo-relative directory path (POSIX, no leading slash; empty
+ * string for the repo root) to that directory's parsed `.gitignore`
+ * rules. A directory with no `.gitignore` is simply absent from the map.
+ */
+export type GitignoreChain = ReadonlyMap<string, readonly IgnoreRule[]>;
+/**
+ * Walk every parent directory of `relPath` from the repo root downward,
+ * returning the sequence of layered rule sets that apply to it. Each
+ * returned entry carries the owning directory (so rule paths can be made
+ * relative to it) plus its rules.
+ */
+export declare function layersFor(relPath: string, chain: GitignoreChain): readonly {
+    readonly dir: string;
+    readonly rules: readonly IgnoreRule[];
+}[];
+/**
+ * Evaluate `relPath` against the layered chain. Rules at deeper layers
+ * override rules at shallower layers; within a layer, the last-matching
+ * rule wins (same as git). Returns `true` when the path should be
+ * ignored.
+ */
+export declare function shouldIgnoreLayered(relPath: string, chain: GitignoreChain, opts?: {
+    readonly isDirectory?: boolean;
+}): boolean;
+//# sourceMappingURL=gitignore-stack.d.ts.map

package/dist/pipeline/gitignore-stack.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"gitignore-stack.d.ts","sourceRoot":"","sources":["../../src/pipeline/gitignore-stack.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;GAgBG;AAEH,OAAO,KAAK,EAAE,UAAU,EAAE,MAAM,gBAAgB,CAAC;AAEjD;;;;GAIG;AACH,MAAM,MAAM,cAAc,GAAG,WAAW,CAAC,MAAM,EAAE,SAAS,UAAU,EAAE,CAAC,CAAC;AAExE;;;;;GAKG;AACH,wBAAgB,SAAS,CACvB,OAAO,EAAE,MAAM,EACf,KAAK,EAAE,cAAc,GACpB,SAAS;IAAE,QAAQ,CAAC,GAAG,EAAE,MAAM,CAAC;IAAC,QAAQ,CAAC,KAAK,EAAE,SAAS,UAAU,EAAE,CAAA;CAAE,EAAE,CAoB5E;AAED;;;;;GAKG;AACH,wBAAgB,mBAAmB,CACjC,OAAO,EAAE,MAAM,EACf,KAAK,EAAE,cAAc,EACrB,IAAI,GAAE;IAAE,QAAQ,CAAC,WAAW,CAAC,EAAE,OAAO,CAAA;CAAO,GAC5C,OAAO,CAgBT"}

package/dist/pipeline/gitignore-stack.js

@@ -0,0 +1,69 @@
+/**
+ * Gitignore rule stacking helper — encapsulates the "layered rule set"
+ * semantics git itself honours when nested `.gitignore` files are present.
+ *
+ * Git's rules (see `man gitignore`):
+ *   - A `.gitignore` in a subdirectory shadows and extends the rules from
+ *     its parents. Rules are evaluated from the repo root downward.
+ *   - A negation (`!pat`) at a deeper layer can re-include a path that a
+ *     parent layer excluded (and vice versa).
+ *   - For a given path, evaluation walks the chain top-down; later matches
+ *     override earlier ones. The final state (ignored / not-ignored) wins.
+ *
+ * This module does NOT scan the filesystem; it takes a pre-loaded
+ * {@link GitignoreChain} (directory-relative paths → parsed rules) and
+ * resolves the effective ignored state for a given path. The scan-time
+ * walker in `gitignore.ts` handles the actual filesystem recursion.
+ */
+/**
+ * Walk every parent directory of `relPath` from the repo root downward,
+ * returning the sequence of layered rule sets that apply to it. Each
+ * returned entry carries the owning directory (so rule paths can be made
+ * relative to it) plus its rules.
+ */
+export function layersFor(relPath, chain) {
+    // Build the sequence of ancestor directories, root first.
+    const out = [];
+    const rootRules = chain.get("");
+    if (rootRules !== undefined && rootRules.length > 0) {
+        out.push({ dir: "", rules: rootRules });
+    }
+    if (relPath === "" || !relPath.includes("/"))
+        return out;
+    const parts = relPath.split("/");
+    // Walk ancestors excluding the file's own name (parts[0..-1)).
+    let cur = "";
+    for (let i = 0; i < parts.length - 1; i += 1) {
+        cur = cur === "" ? parts[i] : `${cur}/${parts[i]}`;
+        const rules = chain.get(cur);
+        if (rules !== undefined && rules.length > 0) {
+            out.push({ dir: cur, rules });
+        }
+    }
+    return out;
+}
+/**
+ * Evaluate `relPath` against the layered chain. Rules at deeper layers
+ * override rules at shallower layers; within a layer, the last-matching
+ * rule wins (same as git). Returns `true` when the path should be
+ * ignored.
+ */
+export function shouldIgnoreLayered(relPath, chain, opts = {}) {
+    const isDir = opts.isDirectory === true;
+    let ignored = false;
+    for (const layer of layersFor(relPath, chain)) {
+        // Paths inside a `.gitignore` are interpreted relative to the
+        // directory that owns the file. Strip the ancestor prefix so the
+        // layer's rules match against the correct sub-path.
+        const pathUnderLayer = layer.dir === "" ? relPath : relPath.slice(layer.dir.length + 1);
+        for (const rule of layer.rules) {
+            if (rule.directoryOnly && !isDir)
+                continue;
+            if (rule.regex.test(pathUnderLayer)) {
+                ignored = !rule.negate;
+            }
+        }
+    }
+    return ignored;
+}
+//# sourceMappingURL=gitignore-stack.js.map

package/dist/pipeline/gitignore-stack.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"gitignore-stack.js","sourceRoot":"","sources":["../../src/pipeline/gitignore-stack.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;GAgBG;AAWH;;;;;GAKG;AACH,MAAM,UAAU,SAAS,CACvB,OAAe,EACf,KAAqB;IAErB,0DAA0D;IAC1D,MAAM,GAAG,GAAoD,EAAE,CAAC;IAChE,MAAM,SAAS,GAAG,KAAK,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;IAChC,IAAI,SAAS,KAAK,SAAS,IAAI,SAAS,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;QACpD,GAAG,CAAC,IAAI,CAAC,EAAE,GAAG,EAAE,EAAE,EAAE,KAAK,EAAE,SAAS,EAAE,CAAC,CAAC;IAC1C,CAAC;IAED,IAAI,OAAO,KAAK,EAAE,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,GAAG,CAAC;QAAE,OAAO,GAAG,CAAC;IACzD,MAAM,KAAK,GAAG,OAAO,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC;IACjC,+DAA+D;IAC/D,IAAI,GAAG,GAAG,EAAE,CAAC;IACb,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,KAAK,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC,IAAI,CAAC,EAAE,CAAC;QAC7C,GAAG,GAAG,GAAG,KAAK,EAAE,CAAC,CAAC,CAAE,KAAK,CAAC,CAAC,CAAY,CAAC,CAAC,CAAC,GAAG,GAAG,IAAI,KAAK,CAAC,CAAC,CAAC,EAAE,CAAC;QAC/D,MAAM,KAAK,GAAG,KAAK,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;QAC7B,IAAI,KAAK,KAAK,SAAS,IAAI,KAAK,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;YAC5C,GAAG,CAAC,IAAI,CAAC,EAAE,GAAG,EAAE,GAAG,EAAE,KAAK,EAAE,CAAC,CAAC;QAChC,CAAC;IACH,CAAC;IACD,OAAO,GAAG,CAAC;AACb,CAAC;AAED;;;;;GAKG;AACH,MAAM,UAAU,mBAAmB,CACjC,OAAe,EACf,KAAqB,EACrB,OAA2C,EAAE;IAE7C,MAAM,KAAK,GAAG,IAAI,CAAC,WAAW,KAAK,IAAI,CAAC;IACxC,IAAI,OAAO,GAAG,KAAK,CAAC;IACpB,KAAK,MAAM,KAAK,IAAI,SAAS,CAAC,OAAO,EAAE,KAAK,CAAC,EAAE,CAAC;QAC9C,8DAA8D;QAC9D,iEAAiE;QACjE,oDAAoD;QACpD,MAAM,cAAc,GAAG,KAAK,CAAC,GAAG,KAAK,EAAE,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,OAAO,CAAC,KAAK,CAAC,KAAK,CAAC,GAAG,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;QACxF,KAAK,MAAM,IAAI,IAAI,KAAK,CAAC,KAAK,EAAE,CAAC;YAC/B,IAAI,IAAI,CAAC,aAAa,IAAI,CAAC,KAAK;gBAAE,SAAS;YAC3C,IAAI,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,cAAc,CAAC,EAAE,CAAC;gBACpC,OAAO,GAAG,CAAC,IAAI,CAAC,MAAM,CAAC;YACzB,CAAC;QACH,CAAC;IACH,CAAC;IACD,OAAO,OAAO,CAAC;AACjB,CAAC"}

package/dist/pipeline/gitignore.d.ts

@@ -0,0 +1,67 @@
+/**
+ * Minimal gitignore evaluator.
+ *
+ * Supports the subset of `.gitignore` syntax used in practice:
+ *  - Comments (`#`) and blank lines.
+ *  - Trailing-slash directory-only matches.
+ *  - Leading-slash anchored-to-root matches.
+ *  - Negation (`!`) re-includes a previously excluded path.
+ *  - `*` (single segment), `?` (single char), `**` (any number of segments).
+ *  - Nested `.gitignore` files with layered negation. Rules stack from
+ *    repo root downward; deeper layers override shallower ones so
+ *    `docs/.gitignore` can negate rules set by the repo-root file.
+ *
+ * Not supported today: character classes (`[abc]`), escaped metacharacters
+ * (`\*`). We surface them as warnings when the operator enables verbose
+ * mode.
+ */
+import type { GitignoreChain } from "./gitignore-stack.js";
+export type { GitignoreChain } from "./gitignore-stack.js";
+export { shouldIgnoreLayered } from "./gitignore-stack.js";
+/** Parsed gitignore rule — order matters; later rules win. */
+export interface IgnoreRule {
+    /** Original source line, for debugging. */
+    readonly raw: string;
+    /** `true` if the rule re-includes (was prefixed with `!`). */
+    readonly negate: boolean;
+    /** `true` if the rule only matches directories (trailing `/`). */
+    readonly directoryOnly: boolean;
+    /** `true` if the pattern is anchored to the directory that declared it. */
+    readonly anchored: boolean;
+    /** RegExp derived from the glob pattern. */
+    readonly regex: RegExp;
+}
+/**
+ * Parse a `.gitignore` file's contents into a list of rules.
+ */
+export declare function parseGitignore(content: string): IgnoreRule[];
+/**
+ * Evaluate a path against a rule set or a layered chain. Two calling
+ * shapes are supported to preserve binary-compat with pre-P06 callers:
+ *
+ *   - Legacy: `shouldIgnore(relPath, rules[], opts)` treats the second
+ *     argument as a flat rule list. Later rules win.
+ *   - Layered: `shouldIgnore(relPath, chain, opts)` treats a `Map` or
+ *     {@link GitignoreChain}-style object as a layered rule stack.
+ *     Deeper layers override shallower ones (git's actual semantics).
+ *
+ * `relPath` must use forward slashes and be relative to the repo root.
+ */
+export declare function shouldIgnore(relPath: string, rulesOrChain: readonly IgnoreRule[] | GitignoreChain, opts?: {
+    readonly isDirectory?: boolean;
+}): boolean;
+/**
+ * Recursively load every `.gitignore` under `repoPath` into a single chain
+ * keyed by repo-relative directory path (POSIX, no leading slash; `""` is
+ * the repo root).
+ *
+ * The walker stops at hardcoded ignores (`node_modules`, `.git`, etc.)
+ * plus any directory whose own parent-layered rule set marks it as
+ * ignored — i.e. it does not recurse into a directory that the existing
+ * rules already ignore. This mirrors git's own behaviour and avoids
+ * expanding the ignore tree into `node_modules` subtrees.
+ */
+export declare function loadGitignoreChain(repoPath: string): Promise<Map<string, IgnoreRule[]>>;
+/** Hardcoded directory names we always skip, even absent a `.gitignore`. */
+export declare const HARDCODED_IGNORES: readonly string[];
+//# sourceMappingURL=gitignore.d.ts.map

package/dist/pipeline/gitignore.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"gitignore.d.ts","sourceRoot":"","sources":["../../src/pipeline/gitignore.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;GAgBG;AAKH,OAAO,KAAK,EAAE,cAAc,EAAE,MAAM,sBAAsB,CAAC;AAG3D,YAAY,EAAE,cAAc,EAAE,MAAM,sBAAsB,CAAC;AAC3D,OAAO,EAAE,mBAAmB,EAAE,MAAM,sBAAsB,CAAC;AAE3D,8DAA8D;AAC9D,MAAM,WAAW,UAAU;IACzB,2CAA2C;IAC3C,QAAQ,CAAC,GAAG,EAAE,MAAM,CAAC;IACrB,8DAA8D;IAC9D,QAAQ,CAAC,MAAM,EAAE,OAAO,CAAC;IACzB,kEAAkE;IAClE,QAAQ,CAAC,aAAa,EAAE,OAAO,CAAC;IAChC,2EAA2E;IAC3E,QAAQ,CAAC,QAAQ,EAAE,OAAO,CAAC;IAC3B,4CAA4C;IAC5C,QAAQ,CAAC,KAAK,EAAE,MAAM,CAAC;CACxB;AAED;;GAEG;AACH,wBAAgB,cAAc,CAAC,OAAO,EAAE,MAAM,GAAG,UAAU,EAAE,CAmC5D;AAuDD;;;;;;;;;;;GAWG;AACH,wBAAgB,YAAY,CAC1B,OAAO,EAAE,MAAM,EACf,YAAY,EAAE,SAAS,UAAU,EAAE,GAAG,cAAc,EACpD,IAAI,GAAE;IAAE,QAAQ,CAAC,WAAW,CAAC,EAAE,OAAO,CAAA;CAAO,GAC5C,OAAO,CAaT;AAED;;;;;;;;;;GAUG;AACH,wBAAsB,kBAAkB,CAAC,QAAQ,EAAE,MAAM,GAAG,OAAO,CAAC,GAAG,CAAC,MAAM,EAAE,UAAU,EAAE,CAAC,CAAC,CAQ7F;AAuCD,4EAA4E;AAC5E,eAAO,MAAM,iBAAiB,EAAE,SAAS,MAAM,EAe9C,CAAC"}

package/dist/pipeline/gitignore.js

@@ -0,0 +1,210 @@
+/**
+ * Minimal gitignore evaluator.
+ *
+ * Supports the subset of `.gitignore` syntax used in practice:
+ *  - Comments (`#`) and blank lines.
+ *  - Trailing-slash directory-only matches.
+ *  - Leading-slash anchored-to-root matches.
+ *  - Negation (`!`) re-includes a previously excluded path.
+ *  - `*` (single segment), `?` (single char), `**` (any number of segments).
+ *  - Nested `.gitignore` files with layered negation. Rules stack from
+ *    repo root downward; deeper layers override shallower ones so
+ *    `docs/.gitignore` can negate rules set by the repo-root file.
+ *
+ * Not supported today: character classes (`[abc]`), escaped metacharacters
+ * (`\*`). We surface them as warnings when the operator enables verbose
+ * mode.
+ */
+import { promises as fs } from "node:fs";
+import path from "node:path";
+import { META_DIR_NAME } from "@opencodehub/storage";
+import { shouldIgnoreLayered } from "./gitignore-stack.js";
+export { shouldIgnoreLayered } from "./gitignore-stack.js";
+/**
+ * Parse a `.gitignore` file's contents into a list of rules.
+ */
+export function parseGitignore(content) {
+    const rules = [];
+    const lines = content.split(/\r?\n/);
+    for (const raw of lines) {
+        // Strip trailing whitespace; gitignore treats trailing spaces as part
+        // of the pattern only when escaped. We don't support escaped spaces.
+        const trimmed = raw.replace(/\s+$/, "");
+        if (trimmed === "" || trimmed.startsWith("#"))
+            continue;
+        let pattern = trimmed;
+        let negate = false;
+        if (pattern.startsWith("!")) {
+            negate = true;
+            pattern = pattern.slice(1);
+        }
+        let directoryOnly = false;
+        if (pattern.endsWith("/")) {
+            directoryOnly = true;
+            pattern = pattern.slice(0, -1);
+        }
+        let anchored = false;
+        if (pattern.startsWith("/")) {
+            anchored = true;
+            pattern = pattern.slice(1);
+        }
+        else if (pattern.includes("/")) {
+            // Middle-slash patterns are implicitly anchored per git's rules.
+            anchored = true;
+        }
+        const regex = globToRegex(pattern, anchored);
+        rules.push({ raw, negate, directoryOnly, anchored, regex });
+    }
+    return rules;
+}
+/**
+ * Translate a gitignore glob into a RegExp matching the relative path
+ * below the declaring directory. Paths are normalized to forward slashes
+ * by the caller before matching.
+ */
+function globToRegex(glob, anchored) {
+    // Walk character by character so we can handle `**` as a special token.
+    let out = "";
+    let i = 0;
+    while (i < glob.length) {
+        const c = glob[i];
+        if (c === "*") {
+            if (glob[i + 1] === "*") {
+                // `**` — zero or more path segments.
+                // `**/x` -> match `x` at any depth; `x/**` -> match anything under x.
+                const next = glob[i + 2];
+                if (next === "/") {
+                    out += "(?:.*/)?";
+                    i += 3;
+                    continue;
+                }
+                // Trailing `**` — match any remaining suffix.
+                out += ".*";
+                i += 2;
+                continue;
+            }
+            // Single `*` — match within a single path segment.
+            out += "[^/]*";
+            i += 1;
+            continue;
+        }
+        if (c === "?") {
+            out += "[^/]";
+            i += 1;
+            continue;
+        }
+        // Escape regex metacharacters.
+        if (/[.+^${}()|[\]\\]/.test(c)) {
+            out += `\\${c}`;
+            i += 1;
+            continue;
+        }
+        out += c;
+        i += 1;
+    }
+    const anchorPrefix = anchored ? "^" : "(?:^|.*/)";
+    // Allow the pattern to match either the full path or a path prefix that
+    // ends on a `/` (so `build` matches `build/foo.ts`).
+    const suffix = "(?:/|$)";
+    return new RegExp(`${anchorPrefix}${out}${suffix}`);
+}
+/**
+ * Evaluate a path against a rule set or a layered chain. Two calling
+ * shapes are supported to preserve binary-compat with pre-P06 callers:
+ *
+ *   - Legacy: `shouldIgnore(relPath, rules[], opts)` treats the second
+ *     argument as a flat rule list. Later rules win.
+ *   - Layered: `shouldIgnore(relPath, chain, opts)` treats a `Map` or
+ *     {@link GitignoreChain}-style object as a layered rule stack.
+ *     Deeper layers override shallower ones (git's actual semantics).
+ *
+ * `relPath` must use forward slashes and be relative to the repo root.
+ */
+export function shouldIgnore(relPath, rulesOrChain, opts = {}) {
+    if (rulesOrChain instanceof Map) {
+        return shouldIgnoreLayered(relPath, rulesOrChain, opts);
+    }
+    const isDir = opts.isDirectory === true;
+    let ignored = false;
+    for (const rule of rulesOrChain) {
+        if (rule.directoryOnly && !isDir)
+            continue;
+        if (rule.regex.test(relPath)) {
+            ignored = !rule.negate;
+        }
+    }
+    return ignored;
+}
+/**
+ * Recursively load every `.gitignore` under `repoPath` into a single chain
+ * keyed by repo-relative directory path (POSIX, no leading slash; `""` is
+ * the repo root).
+ *
+ * The walker stops at hardcoded ignores (`node_modules`, `.git`, etc.)
+ * plus any directory whose own parent-layered rule set marks it as
+ * ignored — i.e. it does not recurse into a directory that the existing
+ * rules already ignore. This mirrors git's own behaviour and avoids
+ * expanding the ignore tree into `node_modules` subtrees.
+ */
+export async function loadGitignoreChain(repoPath) {
+    const chain = new Map();
+    const hardcoded = new Set(HARDCODED_IGNORES);
+    await loadDir(repoPath, "", chain, hardcoded);
+    // Always return a `""` entry — scan.ts and other callers rely on its
+    // presence (empty array is fine when no root .gitignore exists).
+    if (!chain.has(""))
+        chain.set("", []);
+    return chain;
+}
+async function loadDir(repoPath, relDir, chain, hardcoded) {
+    const absDir = path.join(repoPath, relDir);
+    const ignoreFile = path.join(absDir, ".gitignore");
+    try {
+        const content = await fs.readFile(ignoreFile, "utf8");
+        chain.set(relDir, parseGitignore(content));
+    }
+    catch {
+        // No .gitignore here — nothing to stack for this layer. Deeper
+        // layers may still contribute rules.
+    }
+    let entries;
+    try {
+        entries = await fs.readdir(absDir, { withFileTypes: true });
+    }
+    catch {
+        return;
+    }
+    for (const entry of entries) {
+        if (!entry.isDirectory())
+            continue;
+        if (hardcoded.has(entry.name))
+            continue;
+        const childRel = relDir === "" ? entry.name : `${relDir}/${entry.name}`;
+        // Respect the chain we've built up so far so we don't descend into a
+        // subtree the parent layer already excluded. A directory ignored at
+        // layer N is still ignored at layer N+1 unless a deeper .gitignore
+        // re-includes it via a negation — and in that case we need to walk
+        // in anyway. The conservative win here is to recurse: the extra
+        // .gitignore reads are cheap and we match git's behaviour more
+        // exactly.
+        await loadDir(repoPath, childRel, chain, hardcoded);
+    }
+}
+/** Hardcoded directory names we always skip, even absent a `.gitignore`. */
+export const HARDCODED_IGNORES = [
+    "node_modules",
+    ".git",
+    ".svn",
+    ".hg",
+    "dist",
+    "build",
+    "target",
+    META_DIR_NAME,
+    ".venv",
+    "__pycache__",
+    ".next",
+    ".nuxt",
+    ".turbo",
+    "coverage",
+];
+//# sourceMappingURL=gitignore.js.map

package/dist/pipeline/gitignore.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"gitignore.js","sourceRoot":"","sources":["../../src/pipeline/gitignore.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;GAgBG;AAEH,OAAO,EAAE,QAAQ,IAAI,EAAE,EAAE,MAAM,SAAS,CAAC;AACzC,OAAO,IAAI,MAAM,WAAW,CAAC;AAC7B,OAAO,EAAE,aAAa,EAAE,MAAM,sBAAsB,CAAC;AAErD,OAAO,EAAE,mBAAmB,EAAE,MAAM,sBAAsB,CAAC;AAG3D,OAAO,EAAE,mBAAmB,EAAE,MAAM,sBAAsB,CAAC;AAgB3D;;GAEG;AACH,MAAM,UAAU,cAAc,CAAC,OAAe;IAC5C,MAAM,KAAK,GAAiB,EAAE,CAAC;IAC/B,MAAM,KAAK,GAAG,OAAO,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC;IACrC,KAAK,MAAM,GAAG,IAAI,KAAK,EAAE,CAAC;QACxB,sEAAsE;QACtE,qEAAqE;QACrE,MAAM,OAAO,GAAG,GAAG,CAAC,OAAO,CAAC,MAAM,EAAE,EAAE,CAAC,CAAC;QACxC,IAAI,OAAO,KAAK,EAAE,IAAI,OAAO,CAAC,UAAU,CAAC,GAAG,CAAC;YAAE,SAAS;QAExD,IAAI,OAAO,GAAG,OAAO,CAAC;QACtB,IAAI,MAAM,GAAG,KAAK,CAAC;QACnB,IAAI,OAAO,CAAC,UAAU,CAAC,GAAG,CAAC,EAAE,CAAC;YAC5B,MAAM,GAAG,IAAI,CAAC;YACd,OAAO,GAAG,OAAO,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC;QAC7B,CAAC;QAED,IAAI,aAAa,GAAG,KAAK,CAAC;QAC1B,IAAI,OAAO,CAAC,QAAQ,CAAC,GAAG,CAAC,EAAE,CAAC;YAC1B,aAAa,GAAG,IAAI,CAAC;YACrB,OAAO,GAAG,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC;QACjC,CAAC;QAED,IAAI,QAAQ,GAAG,KAAK,CAAC;QACrB,IAAI,OAAO,CAAC,UAAU,CAAC,GAAG,CAAC,EAAE,CAAC;YAC5B,QAAQ,GAAG,IAAI,CAAC;YAChB,OAAO,GAAG,OAAO,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC;QAC7B,CAAC;aAAM,IAAI,OAAO,CAAC,QAAQ,CAAC,GAAG,CAAC,EAAE,CAAC;YACjC,iEAAiE;YACjE,QAAQ,GAAG,IAAI,CAAC;QAClB,CAAC;QAED,MAAM,KAAK,GAAG,WAAW,CAAC,OAAO,EAAE,QAAQ,CAAC,CAAC;QAC7C,KAAK,CAAC,IAAI,CAAC,EAAE,GAAG,EAAE,MAAM,EAAE,aAAa,EAAE,QAAQ,EAAE,KAAK,EAAE,CAAC,CAAC;IAC9D,CAAC;IACD,OAAO,KAAK,CAAC;AACf,CAAC;AAED;;;;GAIG;AACH,SAAS,WAAW,CAAC,IAAY,EAAE,QAAiB;IAClD,wEAAwE;IACxE,IAAI,GAAG,GAAG,EAAE,CAAC;IACb,IAAI,CAAC,GAAG,CAAC,CAAC;IACV,OAAO,CAAC,GAAG,IAAI,CAAC,MAAM,EAAE,CAAC;QACvB,MAAM,CAAC,GAAG,IAAI,CAAC,CAAC,CAAW,CAAC;QAC5B,IAAI,CAAC,KAAK,GAAG,EAAE,CAAC;YACd,IAAI,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,KAAK,GAAG,EAAE,CAAC;gBACxB,qCAAqC;gBACrC,sEAAsE;gBACtE,MAAM,IAAI,GAAG,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;gBACzB,IAAI,IAAI,KAAK,GAAG,EAAE,CAAC;oBACjB,GAAG,IAAI,UAAU,CAAC;oBAClB,CAAC,IAAI,CAAC,CAAC;oBACP,SAAS;gBACX,CAAC;gBACD,8CAA8C;gBAC9C,GAAG,IAAI,IAAI,CAAC;gBACZ,CAAC,IAAI,CAAC,CAAC;gBACP,SAAS;YACX,CAAC;YACD,mDAAmD;YACnD,GAAG,IAAI,OAAO,CAAC;YACf,CAAC,IAAI,CAAC,CAAC;YACP,SAAS;QACX,CAAC;QACD,IAAI,CAAC,KAAK,GAAG,EAAE,CAAC;YACd,GAAG,IAAI,MAAM,CAAC;YACd,CAAC,IAAI,CAAC,CAAC;YACP,SAAS;QACX,CAAC;QACD,+BAA+B;QAC/B,IAAI,kBAAkB,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC;YAC/B,GAAG,IAAI,KAAK,CAAC,EAAE,CAAC;YAChB,CAAC,IAAI,CAAC,CAAC;YACP,SAAS;QACX,CAAC;QACD,GAAG,IAAI,CAAC,CAAC;QACT,CAAC,IAAI,CAAC,CAAC;IACT,CAAC;IAED,MAAM,YAAY,GAAG,QAAQ,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,WAAW,CAAC;IAClD,wEAAwE;IACxE,qDAAqD;IACrD,MAAM,MAAM,GAAG,SAAS,CAAC;IACzB,OAAO,IAAI,MAAM,CAAC,GAAG,YAAY,GAAG,GAAG,GAAG,MAAM,EAAE,CAAC,CAAC;AACtD,CAAC;AAED;;;;;;;;;;;GAWG;AACH,MAAM,UAAU,YAAY,CAC1B,OAAe,EACf,YAAoD,EACpD,OAA2C,EAAE;IAE7C,IAAI,YAAY,YAAY,GAAG,EAAE,CAAC;QAChC,OAAO,mBAAmB,CAAC,OAAO,EAAE,YAAY,EAAE,IAAI,CAAC,CAAC;IAC1D,CAAC;IACD,MAAM,KAAK,GAAG,IAAI,CAAC,WAAW,KAAK,IAAI,CAAC;IACxC,IAAI,OAAO,GAAG,KAAK,CAAC;IACpB,KAAK,MAAM,IAAI,IAAI,YAAqC,EAAE,CAAC;QACzD,IAAI,IAAI,CAAC,aAAa,IAAI,CAAC,KAAK;YAAE,SAAS;QAC3C,IAAI,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,OAAO,CAAC,EAAE,CAAC;YAC7B,OAAO,GAAG,CAAC,IAAI,CAAC,MAAM,CAAC;QACzB,CAAC;IACH,CAAC;IACD,OAAO,OAAO,CAAC;AACjB,CAAC;AAED;;;;;;;;;;GAUG;AACH,MAAM,CAAC,KAAK,UAAU,kBAAkB,CAAC,QAAgB;IACvD,MAAM,KAAK,GAAG,IAAI,GAAG,EAAwB,CAAC;IAC9C,MAAM,SAAS,GAAG,IAAI,GAAG,CAAS,iBAAiB,CAAC,CAAC;IACrD,MAAM,OAAO,CAAC,QAAQ,EAAE,EAAE,EAAE,KAAK,EAAE,SAAS,CAAC,CAAC;IAC9C,qEAAqE;IACrE,iEAAiE;IACjE,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,EAAE,CAAC;QAAE,KAAK,CAAC,GAAG,CAAC,EAAE,EAAE,EAAE,CAAC,CAAC;IACtC,OAAO,KAAK,CAAC;AACf,CAAC;AAED,KAAK,UAAU,OAAO,CACpB,QAAgB,EAChB,MAAc,EACd,KAAgC,EAChC,SAA8B;IAE9B,MAAM,MAAM,GAAG,IAAI,CAAC,IAAI,CAAC,QAAQ,EAAE,MAAM,CAAC,CAAC;IAC3C,MAAM,UAAU,GAAG,IAAI,CAAC,IAAI,CAAC,MAAM,EAAE,YAAY,CAAC,CAAC;IACnD,IAAI,CAAC;QACH,MAAM,OAAO,GAAG,MAAM,EAAE,CAAC,QAAQ,CAAC,UAAU,EAAE,MAAM,CAAC,CAAC;QACtD,KAAK,CAAC,GAAG,CAAC,MAAM,EAAE,cAAc,CAAC,OAAO,CAAC,CAAC,CAAC;IAC7C,CAAC;IAAC,MAAM,CAAC;QACP,+DAA+D;QAC/D,qCAAqC;IACvC,CAAC;IAED,IAAI,OAAmC,CAAC;IACxC,IAAI,CAAC;QACH,OAAO,GAAG,MAAM,EAAE,CAAC,OAAO,CAAC,MAAM,EAAE,EAAE,aAAa,EAAE,IAAI,EAAE,CAAC,CAAC;IAC9D,CAAC;IAAC,MAAM,CAAC;QACP,OAAO;IACT,CAAC;IACD,KAAK,MAAM,KAAK,IAAI,OAAO,EAAE,CAAC;QAC5B,IAAI,CAAC,KAAK,CAAC,WAAW,EAAE;YAAE,SAAS;QACnC,IAAI,SAAS,CAAC,GAAG,CAAC,KAAK,CAAC,IAAI,CAAC;YAAE,SAAS;QACxC,MAAM,QAAQ,GAAG,MAAM,KAAK,EAAE,CAAC,CAAC,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC,GAAG,MAAM,IAAI,KAAK,CAAC,IAAI,EAAE,CAAC;QACxE,qEAAqE;QACrE,oEAAoE;QACpE,mEAAmE;QACnE,mEAAmE;QACnE,gEAAgE;QAChE,+DAA+D;QAC/D,WAAW;QACX,MAAM,OAAO,CAAC,QAAQ,EAAE,QAAQ,EAAE,KAAK,EAAE,SAAS,CAAC,CAAC;IACtD,CAAC;AACH,CAAC;AAED,4EAA4E;AAC5E,MAAM,CAAC,MAAM,iBAAiB,GAAsB;IAClD,cAAc;IACd,MAAM;IACN,MAAM;IACN,KAAK;IACL,MAAM;IACN,OAAO;IACP,QAAQ;IACR,aAAa;IACb,OAAO;IACP,aAAa;IACb,OAAO;IACP,OAAO;IACP,QAAQ;IACR,UAAU;CACX,CAAC"}

package/dist/pipeline/index.d.ts

@@ -0,0 +1,53 @@
+/**
+ * Barrel exports for the pipeline subsystem.
+ */
+export type { IgnoreRule } from "./gitignore.js";
+export { HARDCODED_IGNORES, loadGitignoreChain, parseGitignore, shouldIgnore, } from "./gitignore.js";
+export type { RunIngestionOptions, RunPipelineResult } from "./orchestrator.js";
+export { runIngestion } from "./orchestrator.js";
+export type { CochangeOptions, CochangeOutput } from "./phases/cochange.js";
+export { COCHANGE_PHASE_NAME, cochangePhase, DEFAULT_MAX_FILES_PER_COMMIT, } from "./phases/cochange.js";
+export type { ComplexityOutput } from "./phases/complexity.js";
+export { COMPLEXITY_PHASE_NAME, complexityPhase } from "./phases/complexity.js";
+export type { CachedCapture, CachedExtractions, CacheEntry, CacheKey, } from "./phases/content-cache.js";
+export { CACHE_VERSION, cacheFilePath, computeCacheSize, deriveCacheKey, readCacheEntry, writeCacheEntry, } from "./phases/content-cache.js";
+export { DEFAULT_PHASES } from "./phases/default-set.js";
+export type { EmbedderPhaseOutput, EmbeddingHashCacheAdapter } from "./phases/embeddings.js";
+export { EMBEDDER_PHASE_NAME, EMBEDDING_HASH_CACHE_OPTIONS_KEY, embeddingsPhase, } from "./phases/embeddings.js";
+export type { FetchesOutput } from "./phases/fetches.js";
+export { FETCHES_PHASE_NAME, fetchesPhase, UNRESOLVED_FETCH_TARGET_PREFIX, } from "./phases/fetches.js";
+export type { IncrementalScopeOutput } from "./phases/incremental-scope.js";
+export { INCREMENTAL_SCOPE_PHASE_NAME, incrementalScopePhase, } from "./phases/incremental-scope.js";
+export type { MarkdownOutput } from "./phases/markdown.js";
+export { MARKDOWN_PHASE_NAME, markdownPhase } from "./phases/markdown.js";
+export type { OpenApiOutput } from "./phases/openapi.js";
+export { OPENAPI_PHASE_NAME, openapiPhase } from "./phases/openapi.js";
+export type { OrmOutput } from "./phases/orm.js";
+export { ORM_EXTERNAL_PATH, ORM_PHASE_NAME, ormPhase } from "./phases/orm.js";
+export type { OwnershipOptions, OwnershipOutput } from "./phases/ownership.js";
+export { OWNERSHIP_PHASE_NAME, ownershipPhase } from "./phases/ownership.js";
+export type { ParseOutput } from "./phases/parse.js";
+export { PARSE_PHASE_NAME, parsePhase } from "./phases/parse.js";
+export type { ProfileOutput } from "./phases/profile.js";
+export { PROFILE_PHASE_NAME, profilePhase } from "./phases/profile.js";
+export type { GitProbe, RepoNodePhaseInput, RepoNodePhaseOutput, } from "./phases/repo-node.js";
+export { defaultGitProbe, deriveLanguageStats, deriveLocalRepoUri, deriveRepoUri, REPO_NODE_PHASE_NAME, repoNodePhase, runRepoNodePhase, } from "./phases/repo-node.js";
+export type { RiskSnapshotOptions, RiskSnapshotOutput } from "./phases/risk-snapshot.js";
+export { RISK_SNAPSHOT_PHASE_NAME, riskSnapshotPhase, } from "./phases/risk-snapshot.js";
+export type { RoutesOutput } from "./phases/routes.js";
+export { ROUTES_PHASE_NAME, routesPhase } from "./phases/routes.js";
+export type { SbomOutput } from "./phases/sbom.js";
+export { SBOM_PHASE_NAME, sbomPhase } from "./phases/sbom.js";
+export type { ScannedFile, ScanOutput } from "./phases/scan.js";
+export { SCAN_PHASE_NAME, scanPhase } from "./phases/scan.js";
+export type { StructureOutput } from "./phases/structure.js";
+export { STRUCTURE_PHASE_NAME, structurePhase } from "./phases/structure.js";
+export type { SummarizePhaseOutput, SummarizerAdapter, SummaryCacheAdapter, } from "./phases/summarize.js";
+export { __setSummarizePhaseTestHooks__, SUMMARIZE_PHASE_NAME, SUMMARY_CACHE_OPTIONS_KEY, summarizePhase, } from "./phases/summarize.js";
+export type { TemporalCommitManifest, TemporalOptions, TemporalOutput, } from "./phases/temporal.js";
+export { TEMPORAL_PHASE_NAME, temporalPhase } from "./phases/temporal.js";
+export type { ToolsOutput } from "./phases/tools.js";
+export { TOOLS_PHASE_NAME, toolsPhase } from "./phases/tools.js";
+export { PipelineGraphError, runPipeline, topologicalSort, validatePipeline, } from "./runner.js";
+export type { PhaseResult, PipelineContext, PipelineOptions, PipelinePhase, PreviousGraph, ProgressEvent, } from "./types.js";
+//# sourceMappingURL=index.d.ts.map

package/dist/pipeline/index.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../../src/pipeline/index.ts"],"names":[],"mappings":"AAAA;;GAEG;AAEH,YAAY,EAAE,UAAU,EAAE,MAAM,gBAAgB,CAAC;AACjD,OAAO,EACL,iBAAiB,EACjB,kBAAkB,EAClB,cAAc,EACd,YAAY,GACb,MAAM,gBAAgB,CAAC;AACxB,YAAY,EAAE,mBAAmB,EAAE,iBAAiB,EAAE,MAAM,mBAAmB,CAAC;AAChF,OAAO,EAAE,YAAY,EAAE,MAAM,mBAAmB,CAAC;AACjD,YAAY,EAAE,eAAe,EAAE,cAAc,EAAE,MAAM,sBAAsB,CAAC;AAC5E,OAAO,EACL,mBAAmB,EACnB,aAAa,EACb,4BAA4B,GAC7B,MAAM,sBAAsB,CAAC;AAC9B,YAAY,EAAE,gBAAgB,EAAE,MAAM,wBAAwB,CAAC;AAC/D,OAAO,EAAE,qBAAqB,EAAE,eAAe,EAAE,MAAM,wBAAwB,CAAC;AAChF,YAAY,EACV,aAAa,EACb,iBAAiB,EACjB,UAAU,EACV,QAAQ,GACT,MAAM,2BAA2B,CAAC;AACnC,OAAO,EACL,aAAa,EACb,aAAa,EACb,gBAAgB,EAChB,cAAc,EACd,cAAc,EACd,eAAe,GAChB,MAAM,2BAA2B,CAAC;AACnC,OAAO,EAAE,cAAc,EAAE,MAAM,yBAAyB,CAAC;AACzD,YAAY,EAAE,mBAAmB,EAAE,yBAAyB,EAAE,MAAM,wBAAwB,CAAC;AAC7F,OAAO,EACL,mBAAmB,EACnB,gCAAgC,EAChC,eAAe,GAChB,MAAM,wBAAwB,CAAC;AAChC,YAAY,EAAE,aAAa,EAAE,MAAM,qBAAqB,CAAC;AACzD,OAAO,EACL,kBAAkB,EAClB,YAAY,EACZ,8BAA8B,GAC/B,MAAM,qBAAqB,CAAC;AAC7B,YAAY,EAAE,sBAAsB,EAAE,MAAM,+BAA+B,CAAC;AAC5E,OAAO,EACL,4BAA4B,EAC5B,qBAAqB,GACtB,MAAM,+BAA+B,CAAC;AACvC,YAAY,EAAE,cAAc,EAAE,MAAM,sBAAsB,CAAC;AAC3D,OAAO,EAAE,mBAAmB,EAAE,aAAa,EAAE,MAAM,sBAAsB,CAAC;AAC1E,YAAY,EAAE,aAAa,EAAE,MAAM,qBAAqB,CAAC;AACzD,OAAO,EAAE,kBAAkB,EAAE,YAAY,EAAE,MAAM,qBAAqB,CAAC;AACvE,YAAY,EAAE,SAAS,EAAE,MAAM,iBAAiB,CAAC;AACjD,OAAO,EAAE,iBAAiB,EAAE,cAAc,EAAE,QAAQ,EAAE,MAAM,iBAAiB,CAAC;AAC9E,YAAY,EAAE,gBAAgB,EAAE,eAAe,EAAE,MAAM,uBAAuB,CAAC;AAC/E,OAAO,EAAE,oBAAoB,EAAE,cAAc,EAAE,MAAM,uBAAuB,CAAC;AAC7E,YAAY,EAAE,WAAW,EAAE,MAAM,mBAAmB,CAAC;AACrD,OAAO,EAAE,gBAAgB,EAAE,UAAU,EAAE,MAAM,mBAAmB,CAAC;AACjE,YAAY,EAAE,aAAa,EAAE,MAAM,qBAAqB,CAAC;AACzD,OAAO,EAAE,kBAAkB,EAAE,YAAY,EAAE,MAAM,qBAAqB,CAAC;AACvE,YAAY,EACV,QAAQ,EACR,kBAAkB,EAClB,mBAAmB,GACpB,MAAM,uBAAuB,CAAC;AAC/B,OAAO,EACL,eAAe,EACf,mBAAmB,EACnB,kBAAkB,EAClB,aAAa,EACb,oBAAoB,EACpB,aAAa,EACb,gBAAgB,GACjB,MAAM,uBAAuB,CAAC;AAC/B,YAAY,EAAE,mBAAmB,EAAE,kBAAkB,EAAE,MAAM,2BAA2B,CAAC;AACzF,OAAO,EACL,wBAAwB,EACxB,iBAAiB,GAClB,MAAM,2BAA2B,CAAC;AACnC,YAAY,EAAE,YAAY,EAAE,MAAM,oBAAoB,CAAC;AACvD,OAAO,EAAE,iBAAiB,EAAE,WAAW,EAAE,MAAM,oBAAoB,CAAC;AACpE,YAAY,EAAE,UAAU,EAAE,MAAM,kBAAkB,CAAC;AACnD,OAAO,EAAE,eAAe,EAAE,SAAS,EAAE,MAAM,kBAAkB,CAAC;AAC9D,YAAY,EAAE,WAAW,EAAE,UAAU,EAAE,MAAM,kBAAkB,CAAC;AAChE,OAAO,EAAE,eAAe,EAAE,SAAS,EAAE,MAAM,kBAAkB,CAAC;AAC9D,YAAY,EAAE,eAAe,EAAE,MAAM,uBAAuB,CAAC;AAC7D,OAAO,EAAE,oBAAoB,EAAE,cAAc,EAAE,MAAM,uBAAuB,CAAC;AAC7E,YAAY,EACV,oBAAoB,EACpB,iBAAiB,EACjB,mBAAmB,GACpB,MAAM,uBAAuB,CAAC;AAC/B,OAAO,EACL,8BAA8B,EAC9B,oBAAoB,EACpB,yBAAyB,EACzB,cAAc,GACf,MAAM,uBAAuB,CAAC;AAC/B,YAAY,EACV,sBAAsB,EACtB,eAAe,EACf,cAAc,GACf,MAAM,sBAAsB,CAAC;AAC9B,OAAO,EAAE,mBAAmB,EAAE,aAAa,EAAE,MAAM,sBAAsB,CAAC;AAC1E,YAAY,EAAE,WAAW,EAAE,MAAM,mBAAmB,CAAC;AACrD,OAAO,EAAE,gBAAgB,EAAE,UAAU,EAAE,MAAM,mBAAmB,CAAC;AACjE,OAAO,EACL,kBAAkB,EAClB,WAAW,EACX,eAAe,EACf,gBAAgB,GACjB,MAAM,aAAa,CAAC;AACrB,YAAY,EACV,WAAW,EACX,eAAe,EACf,eAAe,EACf,aAAa,EACb,aAAa,EACb,aAAa,GACd,MAAM,YAAY,CAAC"}

package/dist/pipeline/index.js

@@ -0,0 +1,29 @@
+/**
+ * Barrel exports for the pipeline subsystem.
+ */
+export { HARDCODED_IGNORES, loadGitignoreChain, parseGitignore, shouldIgnore, } from "./gitignore.js";
+export { runIngestion } from "./orchestrator.js";
+export { COCHANGE_PHASE_NAME, cochangePhase, DEFAULT_MAX_FILES_PER_COMMIT, } from "./phases/cochange.js";
+export { COMPLEXITY_PHASE_NAME, complexityPhase } from "./phases/complexity.js";
+export { CACHE_VERSION, cacheFilePath, computeCacheSize, deriveCacheKey, readCacheEntry, writeCacheEntry, } from "./phases/content-cache.js";
+export { DEFAULT_PHASES } from "./phases/default-set.js";
+export { EMBEDDER_PHASE_NAME, EMBEDDING_HASH_CACHE_OPTIONS_KEY, embeddingsPhase, } from "./phases/embeddings.js";
+export { FETCHES_PHASE_NAME, fetchesPhase, UNRESOLVED_FETCH_TARGET_PREFIX, } from "./phases/fetches.js";
+export { INCREMENTAL_SCOPE_PHASE_NAME, incrementalScopePhase, } from "./phases/incremental-scope.js";
+export { MARKDOWN_PHASE_NAME, markdownPhase } from "./phases/markdown.js";
+export { OPENAPI_PHASE_NAME, openapiPhase } from "./phases/openapi.js";
+export { ORM_EXTERNAL_PATH, ORM_PHASE_NAME, ormPhase } from "./phases/orm.js";
+export { OWNERSHIP_PHASE_NAME, ownershipPhase } from "./phases/ownership.js";
+export { PARSE_PHASE_NAME, parsePhase } from "./phases/parse.js";
+export { PROFILE_PHASE_NAME, profilePhase } from "./phases/profile.js";
+export { defaultGitProbe, deriveLanguageStats, deriveLocalRepoUri, deriveRepoUri, REPO_NODE_PHASE_NAME, repoNodePhase, runRepoNodePhase, } from "./phases/repo-node.js";
+export { RISK_SNAPSHOT_PHASE_NAME, riskSnapshotPhase, } from "./phases/risk-snapshot.js";
+export { ROUTES_PHASE_NAME, routesPhase } from "./phases/routes.js";
+export { SBOM_PHASE_NAME, sbomPhase } from "./phases/sbom.js";
+export { SCAN_PHASE_NAME, scanPhase } from "./phases/scan.js";
+export { STRUCTURE_PHASE_NAME, structurePhase } from "./phases/structure.js";
+export { __setSummarizePhaseTestHooks__, SUMMARIZE_PHASE_NAME, SUMMARY_CACHE_OPTIONS_KEY, summarizePhase, } from "./phases/summarize.js";
+export { TEMPORAL_PHASE_NAME, temporalPhase } from "./phases/temporal.js";
+export { TOOLS_PHASE_NAME, toolsPhase } from "./phases/tools.js";
+export { PipelineGraphError, runPipeline, topologicalSort, validatePipeline, } from "./runner.js";
+//# sourceMappingURL=index.js.map

package/dist/pipeline/index.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"index.js","sourceRoot":"","sources":["../../src/pipeline/index.ts"],"names":[],"mappings":"AAAA;;GAEG;AAGH,OAAO,EACL,iBAAiB,EACjB,kBAAkB,EAClB,cAAc,EACd,YAAY,GACb,MAAM,gBAAgB,CAAC;AAExB,OAAO,EAAE,YAAY,EAAE,MAAM,mBAAmB,CAAC;AAEjD,OAAO,EACL,mBAAmB,EACnB,aAAa,EACb,4BAA4B,GAC7B,MAAM,sBAAsB,CAAC;AAE9B,OAAO,EAAE,qBAAqB,EAAE,eAAe,EAAE,MAAM,wBAAwB,CAAC;AAOhF,OAAO,EACL,aAAa,EACb,aAAa,EACb,gBAAgB,EAChB,cAAc,EACd,cAAc,EACd,eAAe,GAChB,MAAM,2BAA2B,CAAC;AACnC,OAAO,EAAE,cAAc,EAAE,MAAM,yBAAyB,CAAC;AAEzD,OAAO,EACL,mBAAmB,EACnB,gCAAgC,EAChC,eAAe,GAChB,MAAM,wBAAwB,CAAC;AAEhC,OAAO,EACL,kBAAkB,EAClB,YAAY,EACZ,8BAA8B,GAC/B,MAAM,qBAAqB,CAAC;AAE7B,OAAO,EACL,4BAA4B,EAC5B,qBAAqB,GACtB,MAAM,+BAA+B,CAAC;AAEvC,OAAO,EAAE,mBAAmB,EAAE,aAAa,EAAE,MAAM,sBAAsB,CAAC;AAE1E,OAAO,EAAE,kBAAkB,EAAE,YAAY,EAAE,MAAM,qBAAqB,CAAC;AAEvE,OAAO,EAAE,iBAAiB,EAAE,cAAc,EAAE,QAAQ,EAAE,MAAM,iBAAiB,CAAC;AAE9E,OAAO,EAAE,oBAAoB,EAAE,cAAc,EAAE,MAAM,uBAAuB,CAAC;AAE7E,OAAO,EAAE,gBAAgB,EAAE,UAAU,EAAE,MAAM,mBAAmB,CAAC;AAEjE,OAAO,EAAE,kBAAkB,EAAE,YAAY,EAAE,MAAM,qBAAqB,CAAC;AAMvE,OAAO,EACL,eAAe,EACf,mBAAmB,EACnB,kBAAkB,EAClB,aAAa,EACb,oBAAoB,EACpB,aAAa,EACb,gBAAgB,GACjB,MAAM,uBAAuB,CAAC;AAE/B,OAAO,EACL,wBAAwB,EACxB,iBAAiB,GAClB,MAAM,2BAA2B,CAAC;AAEnC,OAAO,EAAE,iBAAiB,EAAE,WAAW,EAAE,MAAM,oBAAoB,CAAC;AAEpE,OAAO,EAAE,eAAe,EAAE,SAAS,EAAE,MAAM,kBAAkB,CAAC;AAE9D,OAAO,EAAE,eAAe,EAAE,SAAS,EAAE,MAAM,kBAAkB,CAAC;AAE9D,OAAO,EAAE,oBAAoB,EAAE,cAAc,EAAE,MAAM,uBAAuB,CAAC;AAM7E,OAAO,EACL,8BAA8B,EAC9B,oBAAoB,EACpB,yBAAyB,EACzB,cAAc,GACf,MAAM,uBAAuB,CAAC;AAM/B,OAAO,EAAE,mBAAmB,EAAE,aAAa,EAAE,MAAM,sBAAsB,CAAC;AAE1E,OAAO,EAAE,gBAAgB,EAAE,UAAU,EAAE,MAAM,mBAAmB,CAAC;AACjE,OAAO,EACL,kBAAkB,EAClB,WAAW,EACX,eAAe,EACf,gBAAgB,GACjB,MAAM,aAAa,CAAC"}