@opencodehub/ingestion 0.1.0

package/dist/pipeline/ownership-helpers/git-blame-batcher.js

@@ -0,0 +1,178 @@
+/**
+ * Batch-blame helper for the ownership phase.
+ *
+ * `git` exposes no native multi-file blame command, so we parallelise per-file
+ * `git blame --porcelain` invocations behind a bounded concurrency pool. Two
+ * one-time set-up steps precede the fleet to minimise per-file cost:
+ *   1. `git commit-graph write --reachable --changed-paths` materialises Bloom
+ *      filters the blame engine uses to skip unrelated commits (2-6× speedup).
+ *   2. We resolve parallelism to `max(2, availableParallelism - 2)` capped at
+ *      32 — past that, I/O contention dominates.
+ *
+ * The porcelain format is deterministic line-by-line:
+ *   <commit-sha> <orig-line> <final-line> [<group-size>]
+ *   author <name>
+ *   author-mail <<email>>
+ *   ... other headers, terminated by a `\t<literal source line>` marker.
+ *
+ * A commit-sha repeats its headers only on the first line of each group, so a
+ * cache of sha → author metadata short-circuits repeat headers. Output is a
+ * deterministic `Map<relPath, LineOwner[]>`; the map preserves caller-provided
+ * insertion order but each `LineOwner[]` is ordered by line number.
+ */
+import { execFile } from "node:child_process";
+import { availableParallelism } from "node:os";
+import { promisify } from "node:util";
+const execFileAsync = promisify(execFile);
+const GIT_MAX_BUFFER = 256 * 1024 * 1024; // 256 MB; large files still fit.
+const BLAME_POOL_CAP = 32;
+/**
+ * Run `git blame --porcelain` against each `relPaths` entry in parallel and
+ * return the aggregated attribution map.
+ *
+ * Determinism: outputs are inserted into the map in sorted `relPath` order
+ * regardless of completion order, and each per-file `LineOwner[]` is sorted
+ * by line number.
+ */
+export async function batchBlame(repoPath, relPaths, opts = {}) {
+    if (relPaths.length === 0) {
+        return { byFile: new Map(), subprocessCount: 0, skippedCount: 0 };
+    }
+    if (opts.warmCommitGraph !== false) {
+        try {
+            await execFileAsync("git", ["commit-graph", "write", "--reachable", "--changed-paths"], {
+                cwd: repoPath,
+                maxBuffer: GIT_MAX_BUFFER,
+            });
+        }
+        catch {
+            // Shallow clones, no commits, or missing bloom-filter support all
+            // surface as non-zero exit here. Blame still works without the warm-up.
+        }
+    }
+    const poolSize = resolvePoolSize(opts.concurrency);
+    const sortedPaths = [...relPaths].sort();
+    const results = new Map();
+    let subprocessCount = 0;
+    let skippedCount = 0;
+    const queue = [...sortedPaths];
+    const workers = [];
+    for (let i = 0; i < Math.min(poolSize, queue.length); i += 1) {
+        workers.push((async () => {
+            for (;;) {
+                const relPath = queue.shift();
+                if (relPath === undefined)
+                    return;
+                try {
+                    const { stdout } = await execFileAsync("git", ["blame", "--porcelain", "--", relPath], {
+                        cwd: repoPath,
+                        maxBuffer: GIT_MAX_BUFFER,
+                        encoding: "utf8",
+                    });
+                    subprocessCount += 1;
+                    const lines = parsePorcelainBlame(stdout);
+                    if (lines.length > 0) {
+                        results.set(relPath, lines);
+                    }
+                    else {
+                        skippedCount += 1;
+                    }
+                }
+                catch (err) {
+                    subprocessCount += 1;
+                    skippedCount += 1;
+                    opts.onWarn?.(relPath, err.message);
+                }
+            }
+        })());
+    }
+    await Promise.all(workers);
+    // Re-materialise the map in sorted path order so downstream iteration is
+    // deterministic even if worker completion order was not.
+    const ordered = new Map();
+    for (const relPath of sortedPaths) {
+        const entry = results.get(relPath);
+        if (entry !== undefined)
+            ordered.set(relPath, entry);
+    }
+    return { byFile: ordered, subprocessCount, skippedCount };
+}
+function resolvePoolSize(overrideMaybe) {
+    if (overrideMaybe !== undefined) {
+        const clamped = Math.floor(Math.max(1, Math.min(overrideMaybe, BLAME_POOL_CAP)));
+        return clamped;
+    }
+    const cpus = availableParallelism();
+    const target = Math.max(2, cpus - 2);
+    return Math.min(target, BLAME_POOL_CAP);
+}
+/**
+ * Parse `git blame --porcelain` output into per-line owners.
+ *
+ * The format groups consecutive lines attributed to the same commit; only the
+ * first line of each group includes the `author` / `author-mail` headers. We
+ * cache per-sha author metadata across groups so later groups can reuse the
+ * first group's headers.
+ */
+export function parsePorcelainBlame(stdout) {
+    if (stdout.length === 0)
+        return [];
+    const out = [];
+    const authorBySha = new Map();
+    const lines = stdout.split("\n");
+    let i = 0;
+    while (i < lines.length) {
+        const header = lines[i] ?? "";
+        i += 1;
+        // Header line: `<sha> <origLine> <finalLine> [<groupSize>]`.
+        const headerMatch = /^([0-9a-f]{7,40}) (\d+) (\d+)(?: (\d+))?$/.exec(header);
+        if (headerMatch === null)
+            continue;
+        const sha = headerMatch[1] ?? "";
+        // Position 3 is the final (post-move) line number — that's the blame
+        // output we care about. Position 2 is the commit-time origin line.
+        const finalLine = Number(headerMatch[3] ?? "0");
+        if (!Number.isFinite(finalLine) || finalLine <= 0)
+            continue;
+        let currentName = authorBySha.get(sha)?.name ?? "";
+        let currentEmail = authorBySha.get(sha)?.email ?? "";
+        while (i < lines.length) {
+            const line = lines[i] ?? "";
+            if (line.startsWith("\t")) {
+                // Tab-prefixed line is the source text of the current record.
+                i += 1;
+                break;
+            }
+            if (line.startsWith("author ")) {
+                currentName = line.slice("author ".length);
+            }
+            else if (line.startsWith("author-mail ")) {
+                const raw = line.slice("author-mail ".length).trim();
+                currentEmail = normaliseEmail(raw);
+            }
+            i += 1;
+        }
+        if (sha.length > 0 && currentEmail.length > 0) {
+            const cached = authorBySha.get(sha);
+            if (cached === undefined) {
+                authorBySha.set(sha, { name: currentName, email: currentEmail });
+            }
+            out.push({
+                line: finalLine,
+                email: currentEmail,
+                authorName: currentName,
+                sha,
+            });
+        }
+    }
+    out.sort((a, b) => a.line - b.line);
+    return out;
+}
+function normaliseEmail(rawWithBrackets) {
+    const trimmed = rawWithBrackets.trim();
+    if (trimmed.startsWith("<") && trimmed.endsWith(">")) {
+        return trimmed.slice(1, -1).trim().toLowerCase();
+    }
+    return trimmed.toLowerCase();
+}
+//# sourceMappingURL=git-blame-batcher.js.map

package/dist/pipeline/ownership-helpers/git-blame-batcher.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"git-blame-batcher.js","sourceRoot":"","sources":["../../../src/pipeline/ownership-helpers/git-blame-batcher.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;GAqBG;AAEH,OAAO,EAAE,QAAQ,EAAE,MAAM,oBAAoB,CAAC;AAC9C,OAAO,EAAE,oBAAoB,EAAE,MAAM,SAAS,CAAC;AAC/C,OAAO,EAAE,SAAS,EAAE,MAAM,WAAW,CAAC;AAEtC,MAAM,aAAa,GAAG,SAAS,CAAC,QAAQ,CAAC,CAAC;AAE1C,MAAM,cAAc,GAAG,GAAG,GAAG,IAAI,GAAG,IAAI,CAAC,CAAC,iCAAiC;AAC3E,MAAM,cAAc,GAAG,EAAE,CAAC;AAmC1B;;;;;;;GAOG;AACH,MAAM,CAAC,KAAK,UAAU,UAAU,CAC9B,QAAgB,EAChB,QAA2B,EAC3B,OAA0B,EAAE;IAE5B,IAAI,QAAQ,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;QAC1B,OAAO,EAAE,MAAM,EAAE,IAAI,GAAG,EAAE,EAAE,eAAe,EAAE,CAAC,EAAE,YAAY,EAAE,CAAC,EAAE,CAAC;IACpE,CAAC;IACD,IAAI,IAAI,CAAC,eAAe,KAAK,KAAK,EAAE,CAAC;QACnC,IAAI,CAAC;YACH,MAAM,aAAa,CAAC,KAAK,EAAE,CAAC,cAAc,EAAE,OAAO,EAAE,aAAa,EAAE,iBAAiB,CAAC,EAAE;gBACtF,GAAG,EAAE,QAAQ;gBACb,SAAS,EAAE,cAAc;aAC1B,CAAC,CAAC;QACL,CAAC;QAAC,MAAM,CAAC;YACP,kEAAkE;YAClE,wEAAwE;QAC1E,CAAC;IACH,CAAC;IAED,MAAM,QAAQ,GAAG,eAAe,CAAC,IAAI,CAAC,WAAW,CAAC,CAAC;IACnD,MAAM,WAAW,GAAG,CAAC,GAAG,QAAQ,CAAC,CAAC,IAAI,EAAE,CAAC;IACzC,MAAM,OAAO,GAAG,IAAI,GAAG,EAAuB,CAAC;IAC/C,IAAI,eAAe,GAAG,CAAC,CAAC;IACxB,IAAI,YAAY,GAAG,CAAC,CAAC;IAErB,MAAM,KAAK,GAAG,CAAC,GAAG,WAAW,CAAC,CAAC;IAC/B,MAAM,OAAO,GAAyB,EAAE,CAAC;IACzC,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC,GAAG,CAAC,QAAQ,EAAE,KAAK,CAAC,MAAM,CAAC,EAAE,CAAC,IAAI,CAAC,EAAE,CAAC;QAC7D,OAAO,CAAC,IAAI,CACV,CAAC,KAAK,IAAI,EAAE;YACV,SAAS,CAAC;gBACR,MAAM,OAAO,GAAG,KAAK,CAAC,KAAK,EAAE,CAAC;gBAC9B,IAAI,OAAO,KAAK,SAAS;oBAAE,OAAO;gBAClC,IAAI,CAAC;oBACH,MAAM,EAAE,MAAM,EAAE,GAAG,MAAM,aAAa,CAAC,KAAK,EAAE,CAAC,OAAO,EAAE,aAAa,EAAE,IAAI,EAAE,OAAO,CAAC,EAAE;wBACrF,GAAG,EAAE,QAAQ;wBACb,SAAS,EAAE,cAAc;wBACzB,QAAQ,EAAE,MAAM;qBACjB,CAAC,CAAC;oBACH,eAAe,IAAI,CAAC,CAAC;oBACrB,MAAM,KAAK,GAAG,mBAAmB,CAAC,MAAM,CAAC,CAAC;oBAC1C,IAAI,KAAK,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;wBACrB,OAAO,CAAC,GAAG,CAAC,OAAO,EAAE,KAAK,CAAC,CAAC;oBAC9B,CAAC;yBAAM,CAAC;wBACN,YAAY,IAAI,CAAC,CAAC;oBACpB,CAAC;gBACH,CAAC;gBAAC,OAAO,GAAG,EAAE,CAAC;oBACb,eAAe,IAAI,CAAC,CAAC;oBACrB,YAAY,IAAI,CAAC,CAAC;oBAClB,IAAI,CAAC,MAAM,EAAE,CAAC,OAAO,EAAG,GAAa,CAAC,OAAO,CAAC,CAAC;gBACjD,CAAC;YACH,CAAC;QACH,CAAC,CAAC,EAAE,CACL,CAAC;IACJ,CAAC;IACD,MAAM,OAAO,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC;IAE3B,yEAAyE;IACzE,yDAAyD;IACzD,MAAM,OAAO,GAAG,IAAI,GAAG,EAAuB,CAAC;IAC/C,KAAK,MAAM,OAAO,IAAI,WAAW,EAAE,CAAC;QAClC,MAAM,KAAK,GAAG,OAAO,CAAC,GAAG,CAAC,OAAO,CAAC,CAAC;QACnC,IAAI,KAAK,KAAK,SAAS;YAAE,OAAO,CAAC,GAAG,CAAC,OAAO,EAAE,KAAK,CAAC,CAAC;IACvD,CAAC;IACD,OAAO,EAAE,MAAM,EAAE,OAAO,EAAE,eAAe,EAAE,YAAY,EAAE,CAAC;AAC5D,CAAC;AAED,SAAS,eAAe,CAAC,aAAiC;IACxD,IAAI,aAAa,KAAK,SAAS,EAAE,CAAC;QAChC,MAAM,OAAO,GAAG,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,GAAG,CAAC,aAAa,EAAE,cAAc,CAAC,CAAC,CAAC,CAAC;QACjF,OAAO,OAAO,CAAC;IACjB,CAAC;IACD,MAAM,IAAI,GAAG,oBAAoB,EAAE,CAAC;IACpC,MAAM,MAAM,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,GAAG,CAAC,CAAC,CAAC;IACrC,OAAO,IAAI,CAAC,GAAG,CAAC,MAAM,EAAE,cAAc,CAAC,CAAC;AAC1C,CAAC;AAED;;;;;;;GAOG;AACH,MAAM,UAAU,mBAAmB,CAAC,MAAc;IAChD,IAAI,MAAM,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,EAAE,CAAC;IACnC,MAAM,GAAG,GAAgB,EAAE,CAAC;IAC5B,MAAM,WAAW,GAAG,IAAI,GAAG,EAA2C,CAAC;IACvE,MAAM,KAAK,GAAG,MAAM,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;IACjC,IAAI,CAAC,GAAG,CAAC,CAAC;IACV,OAAO,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC;QACxB,MAAM,MAAM,GAAG,KAAK,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;QAC9B,CAAC,IAAI,CAAC,CAAC;QACP,6DAA6D;QAC7D,MAAM,WAAW,GAAG,2CAA2C,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;QAC7E,IAAI,WAAW,KAAK,IAAI;YAAE,SAAS;QACnC,MAAM,GAAG,GAAG,WAAW,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;QACjC,qEAAqE;QACrE,mEAAmE;QACnE,MAAM,SAAS,GAAG,MAAM,CAAC,WAAW,CAAC,CAAC,CAAC,IAAI,GAAG,CAAC,CAAC;QAChD,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,SAAS,CAAC,IAAI,SAAS,IAAI,CAAC;YAAE,SAAS;QAC5D,IAAI,WAAW,GAAG,WAAW,CAAC,GAAG,CAAC,GAAG,CAAC,EAAE,IAAI,IAAI,EAAE,CAAC;QACnD,IAAI,YAAY,GAAG,WAAW,CAAC,GAAG,CAAC,GAAG,CAAC,EAAE,KAAK,IAAI,EAAE,CAAC;QACrD,OAAO,CAAC,GAAG,KAAK,CAAC,MAAM,EAAE,CAAC;YACxB,MAAM,IAAI,GAAG,KAAK,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;YAC5B,IAAI,IAAI,CAAC,UAAU,CAAC,IAAI,CAAC,EAAE,CAAC;gBAC1B,8DAA8D;gBAC9D,CAAC,IAAI,CAAC,CAAC;gBACP,MAAM;YACR,CAAC;YACD,IAAI,IAAI,CAAC,UAAU,CAAC,SAAS,CAAC,EAAE,CAAC;gBAC/B,WAAW,GAAG,IAAI,CAAC,KAAK,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC;YAC7C,CAAC;iBAAM,IAAI,IAAI,CAAC,UAAU,CAAC,cAAc,CAAC,EAAE,CAAC;gBAC3C,MAAM,GAAG,GAAG,IAAI,CAAC,KAAK,CAAC,cAAc,CAAC,MAAM,CAAC,CAAC,IAAI,EAAE,CAAC;gBACrD,YAAY,GAAG,cAAc,CAAC,GAAG,CAAC,CAAC;YACrC,CAAC;YACD,CAAC,IAAI,CAAC,CAAC;QACT,CAAC;QACD,IAAI,GAAG,CAAC,MAAM,GAAG,CAAC,IAAI,YAAY,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;YAC9C,MAAM,MAAM,GAAG,WAAW,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;YACpC,IAAI,MAAM,KAAK,SAAS,EAAE,CAAC;gBACzB,WAAW,CAAC,GAAG,CAAC,GAAG,EAAE,EAAE,IAAI,EAAE,WAAW,EAAE,KAAK,EAAE,YAAY,EAAE,CAAC,CAAC;YACnE,CAAC;YACD,GAAG,CAAC,IAAI,CAAC;gBACP,IAAI,EAAE,SAAS;gBACf,KAAK,EAAE,YAAY;gBACnB,UAAU,EAAE,WAAW;gBACvB,GAAG;aACJ,CAAC,CAAC;QACL,CAAC;IACH,CAAC;IACD,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,IAAI,GAAG,CAAC,CAAC,IAAI,CAAC,CAAC;IACpC,OAAO,GAAG,CAAC;AACb,CAAC;AAED,SAAS,cAAc,CAAC,eAAuB;IAC7C,MAAM,OAAO,GAAG,eAAe,CAAC,IAAI,EAAE,CAAC;IACvC,IAAI,OAAO,CAAC,UAAU,CAAC,GAAG,CAAC,IAAI,OAAO,CAAC,QAAQ,CAAC,GAAG,CAAC,EAAE,CAAC;QACrD,OAAO,OAAO,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,WAAW,EAAE,CAAC;IACnD,CAAC;IACD,OAAO,OAAO,CAAC,WAAW,EAAE,CAAC;AAC/B,CAAC"}

package/dist/pipeline/ownership-helpers/line-overlap.d.ts

@@ -0,0 +1,35 @@
+/**
+ * Given a symbol's line range and a file's per-line blame attribution,
+ * compute each contributor's normalised line share.
+ *
+ * The formula is deliberately simple: count how many of the symbol's lines
+ * each contributor claims, divide by the symbol's line count, and emit the
+ * result sorted by email hash for determinism.
+ *
+ * File-level shares reuse this helper by passing `startLine=1` and
+ * `endLine=fileLineCount`.
+ */
+import type { LineOwner } from "./git-blame-batcher.js";
+export interface ContributorWeight {
+    readonly email: string;
+    readonly lines: number;
+    /** Share in `[0, 1]` of the symbol's lines attributed to this contributor. */
+    readonly weight: number;
+}
+export interface AttributionOptions {
+    /** Minimum share to keep; defaults to 0 (no filter). */
+    readonly minWeight?: number;
+}
+/**
+ * Build a map from email → {lines, weight} for every contributor with at
+ * least one line in the `[startLine, endLine]` range.
+ *
+ * `fileBlame` is expected to be ordered by line ascending; the caller obtains
+ * it via `batchBlame`. Lines outside the range are ignored; lines inside
+ * the range with no blame entry (a rare parse-skip case) contribute to the
+ * denominator but not to any contributor's numerator.
+ */
+export declare function attributeSymbolOwnership(startLine: number, endLine: number, fileBlame: readonly LineOwner[], opts?: AttributionOptions): readonly ContributorWeight[];
+/** Sum of lines attributed to each contributor across the whole file. */
+export declare function attributeFileOwnership(fileBlame: readonly LineOwner[]): readonly ContributorWeight[];
+//# sourceMappingURL=line-overlap.d.ts.map

package/dist/pipeline/ownership-helpers/line-overlap.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"line-overlap.d.ts","sourceRoot":"","sources":["../../../src/pipeline/ownership-helpers/line-overlap.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;GAUG;AAEH,OAAO,KAAK,EAAE,SAAS,EAAE,MAAM,wBAAwB,CAAC;AAExD,MAAM,WAAW,iBAAiB;IAChC,QAAQ,CAAC,KAAK,EAAE,MAAM,CAAC;IACvB,QAAQ,CAAC,KAAK,EAAE,MAAM,CAAC;IACvB,8EAA8E;IAC9E,QAAQ,CAAC,MAAM,EAAE,MAAM,CAAC;CACzB;AAED,MAAM,WAAW,kBAAkB;IACjC,wDAAwD;IACxD,QAAQ,CAAC,SAAS,CAAC,EAAE,MAAM,CAAC;CAC7B;AAED;;;;;;;;GAQG;AACH,wBAAgB,wBAAwB,CACtC,SAAS,EAAE,MAAM,EACjB,OAAO,EAAE,MAAM,EACf,SAAS,EAAE,SAAS,SAAS,EAAE,EAC/B,IAAI,GAAE,kBAAuB,GAC5B,SAAS,iBAAiB,EAAE,CAuB9B;AAED,yEAAyE;AACzE,wBAAgB,sBAAsB,CACpC,SAAS,EAAE,SAAS,SAAS,EAAE,GAC9B,SAAS,iBAAiB,EAAE,CAO9B"}

package/dist/pipeline/ownership-helpers/line-overlap.js

@@ -0,0 +1,62 @@
+/**
+ * Given a symbol's line range and a file's per-line blame attribution,
+ * compute each contributor's normalised line share.
+ *
+ * The formula is deliberately simple: count how many of the symbol's lines
+ * each contributor claims, divide by the symbol's line count, and emit the
+ * result sorted by email hash for determinism.
+ *
+ * File-level shares reuse this helper by passing `startLine=1` and
+ * `endLine=fileLineCount`.
+ */
+/**
+ * Build a map from email → {lines, weight} for every contributor with at
+ * least one line in the `[startLine, endLine]` range.
+ *
+ * `fileBlame` is expected to be ordered by line ascending; the caller obtains
+ * it via `batchBlame`. Lines outside the range are ignored; lines inside
+ * the range with no blame entry (a rare parse-skip case) contribute to the
+ * denominator but not to any contributor's numerator.
+ */
+export function attributeSymbolOwnership(startLine, endLine, fileBlame, opts = {}) {
+    if (endLine < startLine)
+        return [];
+    const span = endLine - startLine + 1;
+    if (span <= 0)
+        return [];
+    const counts = new Map();
+    for (const owner of fileBlame) {
+        if (owner.line < startLine || owner.line > endLine)
+            continue;
+        counts.set(owner.email, (counts.get(owner.email) ?? 0) + 1);
+    }
+    if (counts.size === 0)
+        return [];
+    const minWeight = opts.minWeight ?? 0;
+    const entries = [];
+    for (const [email, lines] of counts) {
+        const weight = lines / span;
+        if (weight < minWeight)
+            continue;
+        entries.push({ email, lines, weight });
+    }
+    // Sort by descending weight, then ascending email for stable tiebreaks.
+    entries.sort((a, b) => {
+        if (a.weight !== b.weight)
+            return b.weight - a.weight;
+        return a.email < b.email ? -1 : a.email > b.email ? 1 : 0;
+    });
+    return entries;
+}
+/** Sum of lines attributed to each contributor across the whole file. */
+export function attributeFileOwnership(fileBlame) {
+    if (fileBlame.length === 0)
+        return [];
+    let maxLine = 0;
+    for (const owner of fileBlame) {
+        if (owner.line > maxLine)
+            maxLine = owner.line;
+    }
+    return attributeSymbolOwnership(1, maxLine, fileBlame);
+}
+//# sourceMappingURL=line-overlap.js.map

package/dist/pipeline/ownership-helpers/line-overlap.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"line-overlap.js","sourceRoot":"","sources":["../../../src/pipeline/ownership-helpers/line-overlap.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;GAUG;AAgBH;;;;;;;;GAQG;AACH,MAAM,UAAU,wBAAwB,CACtC,SAAiB,EACjB,OAAe,EACf,SAA+B,EAC/B,OAA2B,EAAE;IAE7B,IAAI,OAAO,GAAG,SAAS;QAAE,OAAO,EAAE,CAAC;IACnC,MAAM,IAAI,GAAG,OAAO,GAAG,SAAS,GAAG,CAAC,CAAC;IACrC,IAAI,IAAI,IAAI,CAAC;QAAE,OAAO,EAAE,CAAC;IACzB,MAAM,MAAM,GAAG,IAAI,GAAG,EAAkB,CAAC;IACzC,KAAK,MAAM,KAAK,IAAI,SAAS,EAAE,CAAC;QAC9B,IAAI,KAAK,CAAC,IAAI,GAAG,SAAS,IAAI,KAAK,CAAC,IAAI,GAAG,OAAO;YAAE,SAAS;QAC7D,MAAM,CAAC,GAAG,CAAC,KAAK,CAAC,KAAK,EAAE,CAAC,MAAM,CAAC,GAAG,CAAC,KAAK,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;IAC9D,CAAC;IACD,IAAI,MAAM,CAAC,IAAI,KAAK,CAAC;QAAE,OAAO,EAAE,CAAC;IACjC,MAAM,SAAS,GAAG,IAAI,CAAC,SAAS,IAAI,CAAC,CAAC;IACtC,MAAM,OAAO,GAAwB,EAAE,CAAC;IACxC,KAAK,MAAM,CAAC,KAAK,EAAE,KAAK,CAAC,IAAI,MAAM,EAAE,CAAC;QACpC,MAAM,MAAM,GAAG,KAAK,GAAG,IAAI,CAAC;QAC5B,IAAI,MAAM,GAAG,SAAS;YAAE,SAAS;QACjC,OAAO,CAAC,IAAI,CAAC,EAAE,KAAK,EAAE,KAAK,EAAE,MAAM,EAAE,CAAC,CAAC;IACzC,CAAC;IACD,wEAAwE;IACxE,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE;QACpB,IAAI,CAAC,CAAC,MAAM,KAAK,CAAC,CAAC,MAAM;YAAE,OAAO,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC,MAAM,CAAC;QACtD,OAAO,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IAC5D,CAAC,CAAC,CAAC;IACH,OAAO,OAAO,CAAC;AACjB,CAAC;AAED,yEAAyE;AACzE,MAAM,UAAU,sBAAsB,CACpC,SAA+B;IAE/B,IAAI,SAAS,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,EAAE,CAAC;IACtC,IAAI,OAAO,GAAG,CAAC,CAAC;IAChB,KAAK,MAAM,KAAK,IAAI,SAAS,EAAE,CAAC;QAC9B,IAAI,KAAK,CAAC,IAAI,GAAG,OAAO;YAAE,OAAO,GAAG,KAAK,CAAC,IAAI,CAAC;IACjD,CAAC;IACD,OAAO,wBAAwB,CAAC,CAAC,EAAE,OAAO,EAAE,SAAS,CAAC,CAAC;AACzD,CAAC"}

package/dist/pipeline/ownership-helpers/orphan.d.ts

@@ -0,0 +1,73 @@
+/**
+ * Orphan detection.
+ *
+ * Classifies each File into one of four lifecycle buckets, driven by
+ * three inputs per file:
+ *
+ *   - `topContributorLastSeenDays` from temporal signal 11.
+ *   - `coauthors365d`: count of distinct Co-authored-by: emails across the
+ *     past 365 days of commits touching this file.
+ *   - `decayedChurn` from temporal signal 6.
+ *
+ * Grades:
+ *   - `active`     — none of the lifecycle conditions fire.
+ *   - `orphaned`   — top contributor inactive for >180 days, no recent
+ *                    coauthors, but the file still shows non-trivial churn.
+ *   - `abandoned`  — top contributor inactive for >365 days with non-trivial
+ *                    churn remaining.
+ *   - `fossilized` — top contributor inactive for >730 days and no churn
+ *                    above epsilon; old code that nobody touches.
+ *
+ * `epsilon` is 1% of the repo-median `decayedChurn` so the threshold scales
+ * with repo activity — a high-churn monorepo demands a higher bar than a
+ * handful of scripts.
+ */
+export type OrphanGrade = "active" | "orphaned" | "abandoned" | "fossilized";
+export interface OrphanFileInput {
+    /** Days since the file's top contributor last appeared anywhere in the repo. */
+    readonly topContributorLastSeenDays: number | undefined;
+    /** Distinct Co-authored-by: emails in the past 365 days. */
+    readonly coauthors365d: number;
+    /** Sum of exponential-decay churn weights for the file. */
+    readonly decayedChurn: number;
+}
+export interface OrphanClassificationInput {
+    /** Optional override of the epsilon threshold; default = 1% of repo median. */
+    readonly epsilonOverride?: number;
+    /** Whether enough history exists; when `false`, all files return `"active"`. */
+    readonly hasEnoughHistory: boolean;
+}
+/**
+ * Compute the orphan epsilon threshold from a repo's decayed-churn vector.
+ * Zero-length input yields 0 (everything passes the churn filter vacuously).
+ */
+export declare function computeOrphanEpsilon(decayedChurnValues: readonly number[], percentOfMedian?: number): number;
+/**
+ * Classify a single file. Rule precedence (first match wins):
+ *  1. `fossilized` — oldest top contributor, churn at-or-below epsilon.
+ *  2. `abandoned` — top contributor inactive >365d, churn above epsilon.
+ *  3. `orphaned` — top contributor inactive >180d, zero coauthors in the
+ *     past year, churn above epsilon.
+ *  4. `active` — fallthrough.
+ */
+export declare function classifyOrphan(file: OrphanFileInput, ctx: {
+    readonly epsilon: number;
+    readonly hasEnoughHistory: boolean;
+}): OrphanGrade;
+/**
+ * Batch-classify every file in one pass. Returns a Map keyed by the original
+ * file key (caller chooses semantics).
+ */
+export declare function classifyOrphans<K>(files: ReadonlyMap<K, OrphanFileInput>, ctxIn: OrphanClassificationInput): Map<K, OrphanGrade>;
+/**
+ * Impact multiplier per grade.
+ *
+ * Scale (active / orphaned / abandoned / fossilized) = (1.0 / 1.3 / 1.6 / 1.6).
+ * Fossilized ties with abandoned at the top because dormant-old code often
+ * carries the most hidden complexity (the top contributor is long gone and
+ * no one has touched it, so nobody has warmed the file into working memory
+ * recently). Orphaned sits in the middle: the code still churns, just
+ * without the original author.
+ */
+export declare function orphanImpactMultiplier(grade: OrphanGrade | undefined): number;
+//# sourceMappingURL=orphan.d.ts.map

package/dist/pipeline/ownership-helpers/orphan.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"orphan.d.ts","sourceRoot":"","sources":["../../../src/pipeline/ownership-helpers/orphan.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;GAuBG;AAEH,MAAM,MAAM,WAAW,GAAG,QAAQ,GAAG,UAAU,GAAG,WAAW,GAAG,YAAY,CAAC;AAE7E,MAAM,WAAW,eAAe;IAC9B,gFAAgF;IAChF,QAAQ,CAAC,0BAA0B,EAAE,MAAM,GAAG,SAAS,CAAC;IACxD,4DAA4D;IAC5D,QAAQ,CAAC,aAAa,EAAE,MAAM,CAAC;IAC/B,2DAA2D;IAC3D,QAAQ,CAAC,YAAY,EAAE,MAAM,CAAC;CAC/B;AAED,MAAM,WAAW,yBAAyB;IACxC,+EAA+E;IAC/E,QAAQ,CAAC,eAAe,CAAC,EAAE,MAAM,CAAC;IAClC,gFAAgF;IAChF,QAAQ,CAAC,gBAAgB,EAAE,OAAO,CAAC;CACpC;AAMD;;;GAGG;AACH,wBAAgB,oBAAoB,CAClC,kBAAkB,EAAE,SAAS,MAAM,EAAE,EACrC,eAAe,SAAO,GACrB,MAAM,CASR;AAED;;;;;;;GAOG;AACH,wBAAgB,cAAc,CAC5B,IAAI,EAAE,eAAe,EACrB,GAAG,EAAE;IAAE,QAAQ,CAAC,OAAO,EAAE,MAAM,CAAC;IAAC,QAAQ,CAAC,gBAAgB,EAAE,OAAO,CAAA;CAAE,GACpE,WAAW,CAYb;AAED;;;GAGG;AACH,wBAAgB,eAAe,CAAC,CAAC,EAC/B,KAAK,EAAE,WAAW,CAAC,CAAC,EAAE,eAAe,CAAC,EACtC,KAAK,EAAE,yBAAyB,GAC/B,GAAG,CAAC,CAAC,EAAE,WAAW,CAAC,CAarB;AAED;;;;;;;;;GASG;AACH,wBAAgB,sBAAsB,CAAC,KAAK,EAAE,WAAW,GAAG,SAAS,GAAG,MAAM,CAmB7E"}

package/dist/pipeline/ownership-helpers/orphan.js

@@ -0,0 +1,117 @@
+/**
+ * Orphan detection.
+ *
+ * Classifies each File into one of four lifecycle buckets, driven by
+ * three inputs per file:
+ *
+ *   - `topContributorLastSeenDays` from temporal signal 11.
+ *   - `coauthors365d`: count of distinct Co-authored-by: emails across the
+ *     past 365 days of commits touching this file.
+ *   - `decayedChurn` from temporal signal 6.
+ *
+ * Grades:
+ *   - `active`     — none of the lifecycle conditions fire.
+ *   - `orphaned`   — top contributor inactive for >180 days, no recent
+ *                    coauthors, but the file still shows non-trivial churn.
+ *   - `abandoned`  — top contributor inactive for >365 days with non-trivial
+ *                    churn remaining.
+ *   - `fossilized` — top contributor inactive for >730 days and no churn
+ *                    above epsilon; old code that nobody touches.
+ *
+ * `epsilon` is 1% of the repo-median `decayedChurn` so the threshold scales
+ * with repo activity — a high-churn monorepo demands a higher bar than a
+ * handful of scripts.
+ */
+const ORPHANED_DAYS = 180;
+const ABANDONED_DAYS = 365;
+const FOSSILIZED_DAYS = 730;
+/**
+ * Compute the orphan epsilon threshold from a repo's decayed-churn vector.
+ * Zero-length input yields 0 (everything passes the churn filter vacuously).
+ */
+export function computeOrphanEpsilon(decayedChurnValues, percentOfMedian = 0.01) {
+    if (decayedChurnValues.length === 0)
+        return 0;
+    const sorted = [...decayedChurnValues].sort((a, b) => a - b);
+    const mid = Math.floor(sorted.length / 2);
+    const median = sorted.length % 2 === 1
+        ? (sorted[mid] ?? 0)
+        : ((sorted[mid - 1] ?? 0) + (sorted[mid] ?? 0)) / 2;
+    return median * percentOfMedian;
+}
+/**
+ * Classify a single file. Rule precedence (first match wins):
+ *  1. `fossilized` — oldest top contributor, churn at-or-below epsilon.
+ *  2. `abandoned` — top contributor inactive >365d, churn above epsilon.
+ *  3. `orphaned` — top contributor inactive >180d, zero coauthors in the
+ *     past year, churn above epsilon.
+ *  4. `active` — fallthrough.
+ */
+export function classifyOrphan(file, ctx) {
+    if (!ctx.hasEnoughHistory)
+        return "active";
+    const lastSeen = file.topContributorLastSeenDays;
+    if (lastSeen === undefined)
+        return "active";
+    const epsilon = ctx.epsilon;
+    const hasMeaningfulChurn = file.decayedChurn > epsilon;
+    if (lastSeen > FOSSILIZED_DAYS && !hasMeaningfulChurn)
+        return "fossilized";
+    if (lastSeen > ABANDONED_DAYS && hasMeaningfulChurn)
+        return "abandoned";
+    if (lastSeen > ORPHANED_DAYS && hasMeaningfulChurn && file.coauthors365d === 0) {
+        return "orphaned";
+    }
+    return "active";
+}
+/**
+ * Batch-classify every file in one pass. Returns a Map keyed by the original
+ * file key (caller chooses semantics).
+ */
+export function classifyOrphans(files, ctxIn) {
+    const out = new Map();
+    if (!ctxIn.hasEnoughHistory) {
+        for (const key of files.keys())
+            out.set(key, "active");
+        return out;
+    }
+    const decayedChurnValues = [];
+    for (const f of files.values())
+        decayedChurnValues.push(f.decayedChurn);
+    const epsilon = ctxIn.epsilonOverride ?? computeOrphanEpsilon(decayedChurnValues);
+    for (const [key, f] of files) {
+        out.set(key, classifyOrphan(f, { epsilon, hasEnoughHistory: true }));
+    }
+    return out;
+}
+/**
+ * Impact multiplier per grade.
+ *
+ * Scale (active / orphaned / abandoned / fossilized) = (1.0 / 1.3 / 1.6 / 1.6).
+ * Fossilized ties with abandoned at the top because dormant-old code often
+ * carries the most hidden complexity (the top contributor is long gone and
+ * no one has touched it, so nobody has warmed the file into working memory
+ * recently). Orphaned sits in the middle: the code still churns, just
+ * without the original author.
+ */
+export function orphanImpactMultiplier(grade) {
+    switch (grade) {
+        case "orphaned":
+            return 1.3;
+        case "abandoned":
+            return 1.6;
+        case "fossilized":
+            return 1.6;
+        case "active":
+        case undefined:
+            return 1.0;
+        default: {
+            // Exhaustiveness — new grades added to the union without updating this
+            // switch will fail TypeScript's checks at build time.
+            const _exhaust = grade;
+            void _exhaust;
+            return 1.0;
+        }
+    }
+}
+//# sourceMappingURL=orphan.js.map

package/dist/pipeline/ownership-helpers/orphan.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"orphan.js","sourceRoot":"","sources":["../../../src/pipeline/ownership-helpers/orphan.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;GAuBG;AAoBH,MAAM,aAAa,GAAG,GAAG,CAAC;AAC1B,MAAM,cAAc,GAAG,GAAG,CAAC;AAC3B,MAAM,eAAe,GAAG,GAAG,CAAC;AAE5B;;;GAGG;AACH,MAAM,UAAU,oBAAoB,CAClC,kBAAqC,EACrC,eAAe,GAAG,IAAI;IAEtB,IAAI,kBAAkB,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,CAAC,CAAC;IAC9C,MAAM,MAAM,GAAG,CAAC,GAAG,kBAAkB,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC;IAC7D,MAAM,GAAG,GAAG,IAAI,CAAC,KAAK,CAAC,MAAM,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;IAC1C,MAAM,MAAM,GACV,MAAM,CAAC,MAAM,GAAG,CAAC,KAAK,CAAC;QACrB,CAAC,CAAC,CAAC,MAAM,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC;QACpB,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC,GAAG,GAAG,CAAC,CAAC,IAAI,CAAC,CAAC,GAAG,CAAC,MAAM,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,GAAG,CAAC,CAAC;IACxD,OAAO,MAAM,GAAG,eAAe,CAAC;AAClC,CAAC;AAED;;;;;;;GAOG;AACH,MAAM,UAAU,cAAc,CAC5B,IAAqB,EACrB,GAAqE;IAErE,IAAI,CAAC,GAAG,CAAC,gBAAgB;QAAE,OAAO,QAAQ,CAAC;IAC3C,MAAM,QAAQ,GAAG,IAAI,CAAC,0BAA0B,CAAC;IACjD,IAAI,QAAQ,KAAK,SAAS;QAAE,OAAO,QAAQ,CAAC;IAC5C,MAAM,OAAO,GAAG,GAAG,CAAC,OAAO,CAAC;IAC5B,MAAM,kBAAkB,GAAG,IAAI,CAAC,YAAY,GAAG,OAAO,CAAC;IACvD,IAAI,QAAQ,GAAG,eAAe,IAAI,CAAC,kBAAkB;QAAE,OAAO,YAAY,CAAC;IAC3E,IAAI,QAAQ,GAAG,cAAc,IAAI,kBAAkB;QAAE,OAAO,WAAW,CAAC;IACxE,IAAI,QAAQ,GAAG,aAAa,IAAI,kBAAkB,IAAI,IAAI,CAAC,aAAa,KAAK,CAAC,EAAE,CAAC;QAC/E,OAAO,UAAU,CAAC;IACpB,CAAC;IACD,OAAO,QAAQ,CAAC;AAClB,CAAC;AAED;;;GAGG;AACH,MAAM,UAAU,eAAe,CAC7B,KAAsC,EACtC,KAAgC;IAEhC,MAAM,GAAG,GAAG,IAAI,GAAG,EAAkB,CAAC;IACtC,IAAI,CAAC,KAAK,CAAC,gBAAgB,EAAE,CAAC;QAC5B,KAAK,MAAM,GAAG,IAAI,KAAK,CAAC,IAAI,EAAE;YAAE,GAAG,CAAC,GAAG,CAAC,GAAG,EAAE,QAAQ,CAAC,CAAC;QACvD,OAAO,GAAG,CAAC;IACb,CAAC;IACD,MAAM,kBAAkB,GAAa,EAAE,CAAC;IACxC,KAAK,MAAM,CAAC,IAAI,KAAK,CAAC,MAAM,EAAE;QAAE,kBAAkB,CAAC,IAAI,CAAC,CAAC,CAAC,YAAY,CAAC,CAAC;IACxE,MAAM,OAAO,GAAG,KAAK,CAAC,eAAe,IAAI,oBAAoB,CAAC,kBAAkB,CAAC,CAAC;IAClF,KAAK,MAAM,CAAC,GAAG,EAAE,CAAC,CAAC,IAAI,KAAK,EAAE,CAAC;QAC7B,GAAG,CAAC,GAAG,CAAC,GAAG,EAAE,cAAc,CAAC,CAAC,EAAE,EAAE,OAAO,EAAE,gBAAgB,EAAE,IAAI,EAAE,CAAC,CAAC,CAAC;IACvE,CAAC;IACD,OAAO,GAAG,CAAC;AACb,CAAC;AAED;;;;;;;;;GASG;AACH,MAAM,UAAU,sBAAsB,CAAC,KAA8B;IACnE,QAAQ,KAAK,EAAE,CAAC;QACd,KAAK,UAAU;YACb,OAAO,GAAG,CAAC;QACb,KAAK,WAAW;YACd,OAAO,GAAG,CAAC;QACb,KAAK,YAAY;YACf,OAAO,GAAG,CAAC;QACb,KAAK,QAAQ,CAAC;QACd,KAAK,SAAS;YACZ,OAAO,GAAG,CAAC;QACb,OAAO,CAAC,CAAC,CAAC;YACR,uEAAuE;YACvE,sDAAsD;YACtD,MAAM,QAAQ,GAAU,KAAK,CAAC;YAC9B,KAAK,QAAQ,CAAC;YACd,OAAO,GAAG,CAAC;QACb,CAAC;IACH,CAAC;AACH,CAAC"}

package/dist/pipeline/phases/accesses.d.ts

@@ -0,0 +1,44 @@
+/**
+ * Accesses phase — emits ACCESSES edges from Function/Method/Constructor
+ * nodes to Property nodes, tagged `read` or `write`.
+ *
+ * Scope (v2): providers that opt in (TS, TSX, JS, Python) supply an
+ * {@link LanguageProvider.extractPropertyAccesses} hook; everyone else is a
+ * silent no-op. The hook receives per-file `definitions`, `captures`,
+ * `sourceText`, and `filePath` (mirroring the parse-phase hand-off) and
+ * returns {@link PropertyAccess} records keyed by the enclosing symbol's
+ * `NodeId`.
+ *
+ * Resolution: property names are FLAT. For every access we look up a matching
+ * `Property` node in order:
+ *   1. Same-file property with identical `name`. Matches by name only; a
+ *      future pass will tighten to (owner, name) pairs once TS classes
+ *      reliably surface field members.
+ *   2. Any other `Property` node with the same name (deterministically the
+ *      one with the smallest NodeId).
+ *   3. If neither hits, we synthesise a `Property:unresolved:<name>` stub
+ *      node. The stub serves as a stable anchor so a later re-run that
+ *      introduces the real field produces a CLEAN graph-hash delta instead
+ *      of orphaned accesses being silently dropped.
+ *
+ * Determinism:
+ *   - Files are iterated in sorted order.
+ *   - Providers sort their accesses before returning.
+ *   - The phase re-sorts by (fromId, toId, startLine) before calling
+ *     `graph.addEdge` so the final emit order is independent of provider
+ *     output order.
+ *   - A 50,000 edges-per-file cap guards against pathological generated
+ *     files; overage is dropped with a `warn` event.
+ */
+import type { PipelinePhase } from "../types.js";
+export declare const ACCESSES_PHASE_NAME = "accesses";
+export interface AccessesOutput {
+    /** Count of ACCESSES edges actually emitted across all files. */
+    readonly edgeCount: number;
+    /** Files whose accesses were truncated at {@link MAX_ACCESSES_PER_FILE}. */
+    readonly truncatedFiles: readonly string[];
+    /** Synthesised `Property:unresolved:<name>` anchors added to the graph. */
+    readonly unresolvedCount: number;
+}
+export declare const accessesPhase: PipelinePhase<AccessesOutput>;
+//# sourceMappingURL=accesses.d.ts.map

package/dist/pipeline/phases/accesses.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"accesses.d.ts","sourceRoot":"","sources":["../../../src/pipeline/phases/accesses.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GA+BG;AAKH,OAAO,KAAK,EAAmB,aAAa,EAAE,MAAM,aAAa,CAAC;AAIlE,eAAO,MAAM,mBAAmB,aAAa,CAAC;AAK9C,MAAM,WAAW,cAAc;IAC7B,iEAAiE;IACjE,QAAQ,CAAC,SAAS,EAAE,MAAM,CAAC;IAC3B,4EAA4E;IAC5E,QAAQ,CAAC,cAAc,EAAE,SAAS,MAAM,EAAE,CAAC;IAC3C,2EAA2E;IAC3E,QAAQ,CAAC,eAAe,EAAE,MAAM,CAAC;CAClC;AAED,eAAO,MAAM,aAAa,EAAE,aAAa,CAAC,cAAc,CAUvD,CAAC"}