@opencodehub/ingestion 0.1.0

package/dist/pipeline/phases/content-cache.js

@@ -0,0 +1,323 @@
+/**
+ * Content-addressed parse cache.
+ *
+ * Per-file JSON sidecar under `<repo>/.codehub/parse-cache/`, keyed on
+ * `sha256(content) + grammarSha + pipelineVersion`. Grammar and pipeline
+ * version changes invalidate cleanly because they participate in the cache
+ * filename — readers that look up the fresh key just miss on old entries.
+ *
+ * Layout:
+ *
+ *   .codehub/parse-cache/<contentSha[0..2]>/<contentSha>-<grammarSha[0..6]>-<pipelineVersion>.json
+ *
+ * The 2-char shard prefix keeps each directory under ~2k entries at 100k
+ * file scale, which most filesystems handle well. Short grammarSha + full
+ * pipelineVersion in the filename keep the disk layout human-inspectable
+ * (you can `rm -rf .codehub/parse-cache/` at any time without corruption —
+ * the cache is purely a latency optimization, never source of truth).
+ *
+ * This module is a **utility**, not a pipeline phase:
+ *   - Scan writes `grammarSha` onto each `ScannedFile`.
+ *   - `parse.ts` consumes the utility: lookup → replay → fill. Hits bypass
+ *     the worker pool entirely; misses go through the worker pool and then
+ *     serialize their {@link CachedExtractions} back to disk for next time.
+ *
+ * Atomic writes go through `write-file-atomic` (reused helper pattern from
+ * `packages/cli/src/fs-atomic.ts`): temp file + fsync + rename. Reads
+ * tolerate missing or malformed files by returning `null` — a cache miss
+ * is normal behavior, never an error.
+ */
+import { promises as fs } from "node:fs";
+import path from "node:path";
+import { default as wfa } from "write-file-atomic";
+/** Current cache envelope schema. Bump in lockstep with breaking changes. */
+export const CACHE_VERSION = 2;
+/** Shard prefix length; 2 chars → 256 buckets, each <~400 entries at 100k. */
+const SHARD_PREFIX_LEN = 2;
+/** Grammar SHA bytes embedded in the filename (first 6 hex chars). */
+const GRAMMAR_SHA_IN_FILENAME_LEN = 6;
+/**
+ * Build a cache key from its three components. Pure; performs no IO.
+ * Inputs must already be content-addressed (sha256 hex) or semver strings —
+ * no normalization is applied here.
+ */
+export function deriveCacheKey(contentSha, grammarSha, pipelineVersion) {
+    return { contentSha, grammarSha, pipelineVersion };
+}
+/**
+ * Resolve the on-disk path for a cache key under `cacheDir`.
+ *
+ * Layout:
+ *   `<cacheDir>/<contentSha[0..2]>/<contentSha>-<grammarSha[0..6]>-<pipelineVersion>.json`
+ *
+ * The grammarSha/pipelineVersion suffix ensures two simultaneous entries
+ * for the same content (e.g. before/after a grammar bump) cannot clobber
+ * each other — older entries simply become unreachable and are reclaimed
+ * by the LRU sweep in {@link evictIfOverCap}, which runs after every
+ * `writeCacheEntry` when `CODEHUB_PARSE_CACHE_MAX_BYTES` (default `1GiB`)
+ * is non-zero.
+ */
+export function cacheFilePath(cacheDir, key) {
+    const shard = key.contentSha.slice(0, SHARD_PREFIX_LEN);
+    const shortGrammar = key.grammarSha.slice(0, GRAMMAR_SHA_IN_FILENAME_LEN);
+    const filename = `${key.contentSha}-${shortGrammar}-${key.pipelineVersion}.json`;
+    return path.join(cacheDir, shard, filename);
+}
+/**
+ * Read a cache entry. Returns `null` on any failure (missing file, read
+ * error, JSON parse error, shape mismatch) — a cache miss is never fatal.
+ *
+ * This deliberately does NOT validate the envelope deeply; downstream
+ * consumers treat the shape as advisory and will rebuild on any surprise.
+ */
+export async function readCacheEntry(cacheDir, key) {
+    const filePath = cacheFilePath(cacheDir, key);
+    let text;
+    try {
+        text = await fs.readFile(filePath, "utf8");
+    }
+    catch {
+        return null;
+    }
+    let parsed;
+    try {
+        parsed = JSON.parse(text);
+    }
+    catch {
+        return null;
+    }
+    if (!isCacheEntry(parsed)) {
+        return null;
+    }
+    // Double-check the envelope matches the composed key — protects against a
+    // filesystem races where a stale/misnamed file could shadow a fresh one.
+    if (parsed.grammarSha !== key.grammarSha || parsed.pipelineVersion !== key.pipelineVersion) {
+        return null;
+    }
+    return parsed;
+}
+/**
+ * Default cap when `CODEHUB_PARSE_CACHE_MAX_BYTES` is unset. 1 GiB keeps a
+ * generous headroom on a typical dev box while preventing the cache from
+ * growing without bound on long-lived analyzer hosts. Set the env var to
+ * `0` to disable eviction entirely (useful for ephemeral CI runners).
+ */
+const DEFAULT_CACHE_CAP = "1GiB";
+/**
+ * Write a cache entry atomically. Creates the shard directory if missing.
+ * Never throws on `mkdir EEXIST`; other IO failures propagate to the caller.
+ *
+ * After a successful write, runs {@link evictIfOverCap} against the cap
+ * sourced from `CODEHUB_PARSE_CACHE_MAX_BYTES` (default `1GiB`; `0`
+ * disables). Eviction errors are swallowed — a cache-eviction failure
+ * is never fatal to the pipeline.
+ */
+export async function writeCacheEntry(cacheDir, key, entry) {
+    const filePath = cacheFilePath(cacheDir, key);
+    const parentDir = path.dirname(filePath);
+    // recursive:true already tolerates EEXIST; kept explicit for readability.
+    await fs.mkdir(parentDir, { recursive: true });
+    const payload = `${JSON.stringify(entry, null, 2)}\n`;
+    await writeFileAtomicAsync(filePath, payload);
+    // Post-write LRU sweep — gated on env, errors swallowed.
+    const cap = parseHumanSizeBytes(process.env["CODEHUB_PARSE_CACHE_MAX_BYTES"] ?? DEFAULT_CACHE_CAP);
+    if (cap > 0) {
+        try {
+            await evictIfOverCap(cacheDir, cap);
+        }
+        catch {
+            // Cache-eviction failure is never fatal; caller still got their write.
+        }
+    }
+}
+/**
+ * Parse a human-readable size string (e.g. `"1GiB"`, `"500MB"`, `"0"`) into
+ * bytes. Numeric inputs pass through clamped to non-negative. Unknown
+ * units, malformed input, or negative numbers all yield `0` (which the
+ * eviction code treats as "disabled"). Both decimal (KB/MB/GB/TB) and
+ * binary (KiB/MiB/GiB/TiB) prefixes are supported.
+ */
+export function parseHumanSizeBytes(input) {
+    if (typeof input === "number")
+        return Number.isFinite(input) ? Math.max(0, Math.floor(input)) : 0;
+    const m = /^\s*(\d+(?:\.\d+)?)\s*([KMGT]i?B?|B)?\s*$/i.exec(input);
+    if (!m)
+        return 0;
+    const n = Number.parseFloat(m[1] ?? "0");
+    if (!Number.isFinite(n) || n < 0)
+        return 0;
+    const unit = (m[2] ?? "").toUpperCase();
+    const mult = {
+        "": 1,
+        B: 1,
+        KB: 1_000,
+        KIB: 1024,
+        MB: 1_000_000,
+        MIB: 1024 ** 2,
+        GB: 1_000_000_000,
+        GIB: 1024 ** 3,
+        TB: 1_000_000_000_000,
+        TIB: 1024 ** 4,
+    };
+    return Math.floor(n * (mult[unit] ?? 1));
+}
+/**
+ * LRU-evict cache entries until total on-disk bytes ≤ `0.9 × capBytes`.
+ *
+ * Walks the same shard layout as {@link computeCacheSize}: each top-level
+ * directory under `cacheDir` is treated as a shard, and every regular
+ * file inside it is a candidate. Entries are sorted by mtime ascending,
+ * then unlinked in oldest-first order until the running total reaches
+ * the 90 % water-mark — the headroom prevents thrash where each new
+ * write evicts exactly one older entry.
+ *
+ * Behavior:
+ *   - `capBytes <= 0` short-circuits (eviction disabled).
+ *   - Missing `cacheDir` is a no-op.
+ *   - Per-file errors during stat or unlink are swallowed (skipped).
+ *   - Total under cap → no work done.
+ *
+ * Cache layout reminder: `<cacheDir>/<shard:2>/<contentSha>-<grammar:6>-<pipelineVersion>.json`.
+ */
+export async function evictIfOverCap(cacheDir, capBytes) {
+    if (capBytes <= 0)
+        return;
+    const candidates = [];
+    let total = 0;
+    let shards;
+    try {
+        shards = await fs.readdir(cacheDir, { withFileTypes: true });
+    }
+    catch {
+        return; // Missing cache dir → nothing to evict.
+    }
+    // Deterministic shard order matches computeCacheSize.
+    shards.sort((a, b) => (a.name < b.name ? -1 : a.name > b.name ? 1 : 0));
+    for (const shard of shards) {
+        if (!shard.isDirectory())
+            continue;
+        const shardPath = path.join(cacheDir, shard.name);
+        let entries;
+        try {
+            entries = await fs.readdir(shardPath, { withFileTypes: true });
+        }
+        catch {
+            continue;
+        }
+        entries.sort((a, b) => (a.name < b.name ? -1 : a.name > b.name ? 1 : 0));
+        for (const e of entries) {
+            if (!e.isFile())
+                continue;
+            const entryPath = path.join(shardPath, e.name);
+            try {
+                const s = await fs.stat(entryPath);
+                candidates.push({ path: entryPath, size: s.size, mtimeMs: s.mtimeMs });
+                total += s.size;
+            }
+            catch {
+                // File vanished mid-traversal; skip.
+            }
+        }
+    }
+    if (total <= capBytes)
+        return;
+    const target = Math.floor(0.9 * capBytes);
+    // Oldest first → LRU eviction order.
+    candidates.sort((a, b) => a.mtimeMs - b.mtimeMs);
+    for (const c of candidates) {
+        if (total <= target)
+            break;
+        try {
+            await fs.unlink(c.path);
+            total -= c.size;
+        }
+        catch {
+            // Concurrent unlink, EACCES, etc. — keep going; over-cap is recoverable.
+        }
+    }
+}
+/**
+ * Walk `cacheDir` and report total file count + byte size. Used by the
+ * meta-sidecar cache-stats path. Returns zeros when the directory does
+ * not exist.
+ */
+export async function computeCacheSize(cacheDir) {
+    let fileCount = 0;
+    let bytes = 0;
+    try {
+        const shards = await fs.readdir(cacheDir, { withFileTypes: true });
+        // Sort for deterministic traversal — size totals are commutative but
+        // deterministic iteration makes error messages reproducible.
+        shards.sort((a, b) => (a.name < b.name ? -1 : a.name > b.name ? 1 : 0));
+        for (const shard of shards) {
+            if (!shard.isDirectory())
+                continue;
+            const shardPath = path.join(cacheDir, shard.name);
+            let entries;
+            try {
+                entries = await fs.readdir(shardPath, { withFileTypes: true });
+            }
+            catch {
+                continue;
+            }
+            entries.sort((a, b) => (a.name < b.name ? -1 : a.name > b.name ? 1 : 0));
+            for (const e of entries) {
+                if (!e.isFile())
+                    continue;
+                const entryPath = path.join(shardPath, e.name);
+                try {
+                    const s = await fs.stat(entryPath);
+                    fileCount += 1;
+                    bytes += s.size;
+                }
+                catch {
+                    // Skip files that disappeared during traversal.
+                }
+            }
+        }
+    }
+    catch {
+        // Missing cache dir → zeroes.
+    }
+    return { fileCount, bytes };
+}
+/** Promise wrapper around write-file-atomic's callback/promise dual API. */
+async function writeFileAtomicAsync(filePath, contents) {
+    await wfa(filePath, contents, { encoding: "utf8", fsync: true });
+}
+function isCacheEntry(v) {
+    if (v === null || typeof v !== "object")
+        return false;
+    const o = v;
+    if (o["cacheVersion"] !== CACHE_VERSION)
+        return false;
+    if (typeof o["grammarSha"] !== "string")
+        return false;
+    if (typeof o["pipelineVersion"] !== "string")
+        return false;
+    const ex = o["extractions"];
+    if (ex === null || typeof ex !== "object")
+        return false;
+    const e = ex;
+    if (!Array.isArray(e["definitions"]))
+        return false;
+    if (!Array.isArray(e["calls"]))
+        return false;
+    if (!Array.isArray(e["imports"]))
+        return false;
+    if (!Array.isArray(e["heritage"]))
+        return false;
+    // `captures` is optional at v2 — validate only if present.
+    if (o["captures"] !== undefined && !Array.isArray(o["captures"]))
+        return false;
+    const md = o["metadata"];
+    if (md === null || typeof md !== "object")
+        return false;
+    const m = md;
+    if (typeof m["language"] !== "string")
+        return false;
+    if (typeof m["byteSize"] !== "number")
+        return false;
+    return true;
+}
+//# sourceMappingURL=content-cache.js.map

package/dist/pipeline/phases/content-cache.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"content-cache.js","sourceRoot":"","sources":["../../../src/pipeline/phases/content-cache.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;GA4BG;AAEH,OAAO,EAAE,QAAQ,IAAI,EAAE,EAAE,MAAM,SAAS,CAAC;AACzC,OAAO,IAAI,MAAM,WAAW,CAAC;AAC7B,OAAO,EAAE,OAAO,IAAI,GAAG,EAAE,MAAM,mBAAmB,CAAC;AAQnD,6EAA6E;AAC7E,MAAM,CAAC,MAAM,aAAa,GAAG,CAAC,CAAC;AAC/B,8EAA8E;AAC9E,MAAM,gBAAgB,GAAG,CAAC,CAAC;AAC3B,sEAAsE;AACtE,MAAM,2BAA2B,GAAG,CAAC,CAAC;AAiEtC;;;;GAIG;AACH,MAAM,UAAU,cAAc,CAC5B,UAAkB,EAClB,UAAkB,EAClB,eAAuB;IAEvB,OAAO,EAAE,UAAU,EAAE,UAAU,EAAE,eAAe,EAAE,CAAC;AACrD,CAAC;AAED;;;;;;;;;;;;GAYG;AACH,MAAM,UAAU,aAAa,CAAC,QAAgB,EAAE,GAAa;IAC3D,MAAM,KAAK,GAAG,GAAG,CAAC,UAAU,CAAC,KAAK,CAAC,CAAC,EAAE,gBAAgB,CAAC,CAAC;IACxD,MAAM,YAAY,GAAG,GAAG,CAAC,UAAU,CAAC,KAAK,CAAC,CAAC,EAAE,2BAA2B,CAAC,CAAC;IAC1E,MAAM,QAAQ,GAAG,GAAG,GAAG,CAAC,UAAU,IAAI,YAAY,IAAI,GAAG,CAAC,eAAe,OAAO,CAAC;IACjF,OAAO,IAAI,CAAC,IAAI,CAAC,QAAQ,EAAE,KAAK,EAAE,QAAQ,CAAC,CAAC;AAC9C,CAAC;AAED;;;;;;GAMG;AACH,MAAM,CAAC,KAAK,UAAU,cAAc,CAAC,QAAgB,EAAE,GAAa;IAClE,MAAM,QAAQ,GAAG,aAAa,CAAC,QAAQ,EAAE,GAAG,CAAC,CAAC;IAC9C,IAAI,IAAY,CAAC;IACjB,IAAI,CAAC;QACH,IAAI,GAAG,MAAM,EAAE,CAAC,QAAQ,CAAC,QAAQ,EAAE,MAAM,CAAC,CAAC;IAC7C,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,IAAI,CAAC;IACd,CAAC;IACD,IAAI,MAAe,CAAC;IACpB,IAAI,CAAC;QACH,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;IAC5B,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,IAAI,CAAC;IACd,CAAC;IACD,IAAI,CAAC,YAAY,CAAC,MAAM,CAAC,EAAE,CAAC;QAC1B,OAAO,IAAI,CAAC;IACd,CAAC;IACD,0EAA0E;IAC1E,yEAAyE;IACzE,IAAI,MAAM,CAAC,UAAU,KAAK,GAAG,CAAC,UAAU,IAAI,MAAM,CAAC,eAAe,KAAK,GAAG,CAAC,eAAe,EAAE,CAAC;QAC3F,OAAO,IAAI,CAAC;IACd,CAAC;IACD,OAAO,MAAM,CAAC;AAChB,CAAC;AAED;;;;;GAKG;AACH,MAAM,iBAAiB,GAAG,MAAM,CAAC;AAEjC;;;;;;;;GAQG;AACH,MAAM,CAAC,KAAK,UAAU,eAAe,CACnC,QAAgB,EAChB,GAAa,EACb,KAAiB;IAEjB,MAAM,QAAQ,GAAG,aAAa,CAAC,QAAQ,EAAE,GAAG,CAAC,CAAC;IAC9C,MAAM,SAAS,GAAG,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC;IACzC,0EAA0E;IAC1E,MAAM,EAAE,CAAC,KAAK,CAAC,SAAS,EAAE,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC;IAC/C,MAAM,OAAO,GAAG,GAAG,IAAI,CAAC,SAAS,CAAC,KAAK,EAAE,IAAI,EAAE,CAAC,CAAC,IAAI,CAAC;IACtD,MAAM,oBAAoB,CAAC,QAAQ,EAAE,OAAO,CAAC,CAAC;IAC9C,yDAAyD;IACzD,MAAM,GAAG,GAAG,mBAAmB,CAC7B,OAAO,CAAC,GAAG,CAAC,+BAA+B,CAAC,IAAI,iBAAiB,CAClE,CAAC;IACF,IAAI,GAAG,GAAG,CAAC,EAAE,CAAC;QACZ,IAAI,CAAC;YACH,MAAM,cAAc,CAAC,QAAQ,EAAE,GAAG,CAAC,CAAC;QACtC,CAAC;QAAC,MAAM,CAAC;YACP,uEAAuE;QACzE,CAAC;IACH,CAAC;AACH,CAAC;AAED;;;;;;GAMG;AACH,MAAM,UAAU,mBAAmB,CAAC,KAAsB;IACxD,IAAI,OAAO,KAAK,KAAK,QAAQ;QAAE,OAAO,MAAM,CAAC,QAAQ,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IAClG,MAAM,CAAC,GAAG,4CAA4C,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;IACnE,IAAI,CAAC,CAAC;QAAE,OAAO,CAAC,CAAC;IACjB,MAAM,CAAC,GAAG,MAAM,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,GAAG,CAAC,CAAC;IACzC,IAAI,CAAC,MAAM,CAAC,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC;QAAE,OAAO,CAAC,CAAC;IAC3C,MAAM,IAAI,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,WAAW,EAAE,CAAC;IACxC,MAAM,IAAI,GAA2B;QACnC,EAAE,EAAE,CAAC;QACL,CAAC,EAAE,CAAC;QACJ,EAAE,EAAE,KAAK;QACT,GAAG,EAAE,IAAI;QACT,EAAE,EAAE,SAAS;QACb,GAAG,EAAE,IAAI,IAAI,CAAC;QACd,EAAE,EAAE,aAAa;QACjB,GAAG,EAAE,IAAI,IAAI,CAAC;QACd,EAAE,EAAE,iBAAiB;QACrB,GAAG,EAAE,IAAI,IAAI,CAAC;KACf,CAAC;IACF,OAAO,IAAI,CAAC,KAAK,CAAC,CAAC,GAAG,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;AAC3C,CAAC;AAED;;;;;;;;;;;;;;;;;GAiBG;AACH,MAAM,CAAC,KAAK,UAAU,cAAc,CAAC,QAAgB,EAAE,QAAgB;IACrE,IAAI,QAAQ,IAAI,CAAC;QAAE,OAAO;IAO1B,MAAM,UAAU,GAAgB,EAAE,CAAC;IACnC,IAAI,KAAK,GAAG,CAAC,CAAC;IAEd,IAAI,MAAkC,CAAC;IACvC,IAAI,CAAC;QACH,MAAM,GAAG,MAAM,EAAE,CAAC,OAAO,CAAC,QAAQ,EAAE,EAAE,aAAa,EAAE,IAAI,EAAE,CAAC,CAAC;IAC/D,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,CAAC,wCAAwC;IAClD,CAAC;IACD,sDAAsD;IACtD,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,IAAI,GAAG,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,GAAG,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;IACxE,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE,CAAC;QAC3B,IAAI,CAAC,KAAK,CAAC,WAAW,EAAE;YAAE,SAAS;QACnC,MAAM,SAAS,GAAG,IAAI,CAAC,IAAI,CAAC,QAAQ,EAAE,KAAK,CAAC,IAAI,CAAC,CAAC;QAClD,IAAI,OAAmC,CAAC;QACxC,IAAI,CAAC;YACH,OAAO,GAAG,MAAM,EAAE,CAAC,OAAO,CAAC,SAAS,EAAE,EAAE,aAAa,EAAE,IAAI,EAAE,CAAC,CAAC;QACjE,CAAC;QAAC,MAAM,CAAC;YACP,SAAS;QACX,CAAC;QACD,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,IAAI,GAAG,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,GAAG,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;QACzE,KAAK,MAAM,CAAC,IAAI,OAAO,EAAE,CAAC;YACxB,IAAI,CAAC,CAAC,CAAC,MAAM,EAAE;gBAAE,SAAS;YAC1B,MAAM,SAAS,GAAG,IAAI,CAAC,IAAI,CAAC,SAAS,EAAE,CAAC,CAAC,IAAI,CAAC,CAAC;YAC/C,IAAI,CAAC;gBACH,MAAM,CAAC,GAAG,MAAM,EAAE,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;gBACnC,UAAU,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,SAAS,EAAE,IAAI,EAAE,CAAC,CAAC,IAAI,EAAE,OAAO,EAAE,CAAC,CAAC,OAAO,EAAE,CAAC,CAAC;gBACvE,KAAK,IAAI,CAAC,CAAC,IAAI,CAAC;YAClB,CAAC;YAAC,MAAM,CAAC;gBACP,qCAAqC;YACvC,CAAC;QACH,CAAC;IACH,CAAC;IAED,IAAI,KAAK,IAAI,QAAQ;QAAE,OAAO;IAE9B,MAAM,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,GAAG,GAAG,QAAQ,CAAC,CAAC;IAC1C,qCAAqC;IACrC,UAAU,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,OAAO,GAAG,CAAC,CAAC,OAAO,CAAC,CAAC;IACjD,KAAK,MAAM,CAAC,IAAI,UAAU,EAAE,CAAC;QAC3B,IAAI,KAAK,IAAI,MAAM;YAAE,MAAM;QAC3B,IAAI,CAAC;YACH,MAAM,EAAE,CAAC,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC;YACxB,KAAK,IAAI,CAAC,CAAC,IAAI,CAAC;QAClB,CAAC;QAAC,MAAM,CAAC;YACP,yEAAyE;QAC3E,CAAC;IACH,CAAC;AACH,CAAC;AAED;;;;GAIG;AACH,MAAM,CAAC,KAAK,UAAU,gBAAgB,CACpC,QAAgB;IAEhB,IAAI,SAAS,GAAG,CAAC,CAAC;IAClB,IAAI,KAAK,GAAG,CAAC,CAAC;IACd,IAAI,CAAC;QACH,MAAM,MAAM,GAAG,MAAM,EAAE,CAAC,OAAO,CAAC,QAAQ,EAAE,EAAE,aAAa,EAAE,IAAI,EAAE,CAAC,CAAC;QACnE,qEAAqE;QACrE,6DAA6D;QAC7D,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,IAAI,GAAG,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,GAAG,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;QACxE,KAAK,MAAM,KAAK,IAAI,MAAM,EAAE,CAAC;YAC3B,IAAI,CAAC,KAAK,CAAC,WAAW,EAAE;gBAAE,SAAS;YACnC,MAAM,SAAS,GAAG,IAAI,CAAC,IAAI,CAAC,QAAQ,EAAE,KAAK,CAAC,IAAI,CAAC,CAAC;YAClD,IAAI,OAAmC,CAAC;YACxC,IAAI,CAAC;gBACH,OAAO,GAAG,MAAM,EAAE,CAAC,OAAO,CAAC,SAAS,EAAE,EAAE,aAAa,EAAE,IAAI,EAAE,CAAC,CAAC;YACjE,CAAC;YAAC,MAAM,CAAC;gBACP,SAAS;YACX,CAAC;YACD,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,EAAE,CAAC,CAAC,CAAC,CAAC,IAAI,GAAG,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,GAAG,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;YACzE,KAAK,MAAM,CAAC,IAAI,OAAO,EAAE,CAAC;gBACxB,IAAI,CAAC,CAAC,CAAC,MAAM,EAAE;oBAAE,SAAS;gBAC1B,MAAM,SAAS,GAAG,IAAI,CAAC,IAAI,CAAC,SAAS,EAAE,CAAC,CAAC,IAAI,CAAC,CAAC;gBAC/C,IAAI,CAAC;oBACH,MAAM,CAAC,GAAG,MAAM,EAAE,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;oBACnC,SAAS,IAAI,CAAC,CAAC;oBACf,KAAK,IAAI,CAAC,CAAC,IAAI,CAAC;gBAClB,CAAC;gBAAC,MAAM,CAAC;oBACP,gDAAgD;gBAClD,CAAC;YACH,CAAC;QACH,CAAC;IACH,CAAC;IAAC,MAAM,CAAC;QACP,8BAA8B;IAChC,CAAC;IACD,OAAO,EAAE,SAAS,EAAE,KAAK,EAAE,CAAC;AAC9B,CAAC;AAED,4EAA4E;AAC5E,KAAK,UAAU,oBAAoB,CAAC,QAAgB,EAAE,QAAgB;IACpE,MAAM,GAAG,CAAC,QAAQ,EAAE,QAAQ,EAAE,EAAE,QAAQ,EAAE,MAAM,EAAE,KAAK,EAAE,IAAI,EAAE,CAAC,CAAC;AACnE,CAAC;AAED,SAAS,YAAY,CAAC,CAAU;IAC9B,IAAI,CAAC,KAAK,IAAI,IAAI,OAAO,CAAC,KAAK,QAAQ;QAAE,OAAO,KAAK,CAAC;IACtD,MAAM,CAAC,GAAG,CAA4B,CAAC;IACvC,IAAI,CAAC,CAAC,cAAc,CAAC,KAAK,aAAa;QAAE,OAAO,KAAK,CAAC;IACtD,IAAI,OAAO,CAAC,CAAC,YAAY,CAAC,KAAK,QAAQ;QAAE,OAAO,KAAK,CAAC;IACtD,IAAI,OAAO,CAAC,CAAC,iBAAiB,CAAC,KAAK,QAAQ;QAAE,OAAO,KAAK,CAAC;IAC3D,MAAM,EAAE,GAAG,CAAC,CAAC,aAAa,CAAC,CAAC;IAC5B,IAAI,EAAE,KAAK,IAAI,IAAI,OAAO,EAAE,KAAK,QAAQ;QAAE,OAAO,KAAK,CAAC;IACxD,MAAM,CAAC,GAAG,EAA6B,CAAC;IACxC,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,aAAa,CAAC,CAAC;QAAE,OAAO,KAAK,CAAC;IACnD,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC;QAAE,OAAO,KAAK,CAAC;IAC7C,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC;QAAE,OAAO,KAAK,CAAC;IAC/C,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,UAAU,CAAC,CAAC;QAAE,OAAO,KAAK,CAAC;IAChD,2DAA2D;IAC3D,IAAI,CAAC,CAAC,UAAU,CAAC,KAAK,SAAS,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,UAAU,CAAC,CAAC;QAAE,OAAO,KAAK,CAAC;IAC/E,MAAM,EAAE,GAAG,CAAC,CAAC,UAAU,CAAC,CAAC;IACzB,IAAI,EAAE,KAAK,IAAI,IAAI,OAAO,EAAE,KAAK,QAAQ;QAAE,OAAO,KAAK,CAAC;IACxD,MAAM,CAAC,GAAG,EAA6B,CAAC;IACxC,IAAI,OAAO,CAAC,CAAC,UAAU,CAAC,KAAK,QAAQ;QAAE,OAAO,KAAK,CAAC;IACpD,IAAI,OAAO,CAAC,CAAC,UAAU,CAAC,KAAK,QAAQ;QAAE,OAAO,KAAK,CAAC;IACpD,OAAO,IAAI,CAAC;AACd,CAAC"}

package/dist/pipeline/phases/coverage-parsers/cobertura.d.ts

@@ -0,0 +1,25 @@
+/**
+ * Cobertura coverage.xml parser.
+ *
+ * Cobertura / Python coverage-xml reports nest per-file coverage under:
+ *   <coverage>
+ *     <packages>
+ *       <package>
+ *         <classes>
+ *           <class filename="src/foo.py">
+ *             <lines>
+ *               <line number="1" hits="2"/>
+ *               <line number="2" hits="0"/>
+ *             </lines>
+ *           </class>
+ *
+ * The phase uses `fast-xml-parser` (already a dep). We set
+ * `ignoreAttributes=false` so the `number`/`hits`/`filename` attributes are
+ * reachable via the attribute-prefix key (`@_`).
+ *
+ * Determinism: the parser returns a map, which the calling phase sorts
+ * before mutating the graph.
+ */
+import { type FileCoverage } from "./types.js";
+export declare function parseCobertura(raw: string, repoRoot: string): ReadonlyMap<string, FileCoverage>;
+//# sourceMappingURL=cobertura.d.ts.map

package/dist/pipeline/phases/coverage-parsers/cobertura.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"cobertura.d.ts","sourceRoot":"","sources":["../../../../src/pipeline/phases/coverage-parsers/cobertura.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;GAqBG;AAGH,OAAO,EAAc,KAAK,YAAY,EAAS,MAAM,YAAY,CAAC;AAOlE,wBAAgB,cAAc,CAAC,GAAG,EAAE,MAAM,EAAE,QAAQ,EAAE,MAAM,GAAG,WAAW,CAAC,MAAM,EAAE,YAAY,CAAC,CA6C/F"}

package/dist/pipeline/phases/coverage-parsers/cobertura.js

@@ -0,0 +1,139 @@
+/**
+ * Cobertura coverage.xml parser.
+ *
+ * Cobertura / Python coverage-xml reports nest per-file coverage under:
+ *   <coverage>
+ *     <packages>
+ *       <package>
+ *         <classes>
+ *           <class filename="src/foo.py">
+ *             <lines>
+ *               <line number="1" hits="2"/>
+ *               <line number="2" hits="0"/>
+ *             </lines>
+ *           </class>
+ *
+ * The phase uses `fast-xml-parser` (already a dep). We set
+ * `ignoreAttributes=false` so the `number`/`hits`/`filename` attributes are
+ * reachable via the attribute-prefix key (`@_`).
+ *
+ * Determinism: the parser returns a map, which the calling phase sorts
+ * before mutating the graph.
+ */
+import { XMLParser } from "fast-xml-parser";
+import { canonLines, ratio } from "./types.js";
+export function parseCobertura(raw, repoRoot) {
+    const parser = new XMLParser({
+        ignoreAttributes: false,
+        attributeNamePrefix: "@_",
+        // Force singletons into arrays where we walk over them.
+        isArray: (name) => name === "package" ||
+            name === "class" ||
+            name === "line" ||
+            name === "packages" ||
+            name === "classes" ||
+            name === "lines",
+    });
+    let parsed;
+    try {
+        parsed = parser.parse(raw);
+    }
+    catch {
+        return new Map();
+    }
+    const classes = collectClasses(parsed);
+    const out = new Map();
+    for (const c of classes) {
+        const filename = c.filename;
+        if (filename === undefined || filename.length === 0)
+            continue;
+        const linesIn = c.lines ?? [];
+        if (linesIn.length === 0)
+            continue;
+        let covered = 0;
+        const coveredLines = [];
+        for (const line of linesIn) {
+            if (line.hits > 0) {
+                covered += 1;
+                coveredLines.push(line.line);
+            }
+        }
+        const total = linesIn.length;
+        const relPath = normalisePath(filename, repoRoot);
+        out.set(relPath, {
+            filePath: relPath,
+            coveredLines: canonLines(coveredLines),
+            totalLines: total,
+            coveragePercent: ratio(covered, total),
+        });
+    }
+    return out;
+}
+function collectClasses(parsed) {
+    const out = [];
+    const root = parsed?.coverage;
+    const packagesContainer = root?.packages;
+    if (!Array.isArray(packagesContainer))
+        return out;
+    // Each entry inside `packages[]` is itself an object that *may* nest a
+    // further `package` array (repeat container). Walk one more level so both
+    // shapes work.
+    for (const container of packagesContainer) {
+        const packages = container?.package;
+        if (!Array.isArray(packages))
+            continue;
+        for (const pkg of packages) {
+            const classesContainer = pkg?.classes;
+            if (!Array.isArray(classesContainer))
+                continue;
+            for (const cContainer of classesContainer) {
+                const classes = cContainer?.class;
+                if (!Array.isArray(classes))
+                    continue;
+                for (const cls of classes) {
+                    const filename = cls["@_filename"];
+                    const linesContainer = cls?.lines;
+                    const linesRaw = Array.isArray(linesContainer) && linesContainer.length > 0
+                        ? (linesContainer[0]?.line ?? [])
+                        : [];
+                    const entries = [];
+                    if (Array.isArray(linesRaw)) {
+                        for (const lv of linesRaw) {
+                            const rec = lv;
+                            const lineNum = numberAttr(rec["@_number"]);
+                            const hits = numberAttr(rec["@_hits"]);
+                            if (lineNum === undefined || hits === undefined)
+                                continue;
+                            entries.push({ line: lineNum, hits });
+                        }
+                    }
+                    out.push({
+                        ...(typeof filename === "string" ? { filename } : {}),
+                        lines: entries,
+                    });
+                }
+            }
+        }
+    }
+    return out;
+}
+function numberAttr(v) {
+    if (typeof v === "number")
+        return Number.isFinite(v) ? v : undefined;
+    if (typeof v === "string") {
+        const n = Number.parseInt(v, 10);
+        return Number.isInteger(n) ? n : undefined;
+    }
+    return undefined;
+}
+function normalisePath(raw, repoRoot) {
+    const posix = raw.replace(/\\/g, "/");
+    if (!posix.startsWith("/"))
+        return posix;
+    const rootPosix = repoRoot.replace(/\\/g, "/");
+    const prefix = rootPosix.endsWith("/") ? rootPosix : `${rootPosix}/`;
+    if (posix.startsWith(prefix))
+        return posix.slice(prefix.length);
+    return posix;
+}
+//# sourceMappingURL=cobertura.js.map

package/dist/pipeline/phases/coverage-parsers/cobertura.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"cobertura.js","sourceRoot":"","sources":["../../../../src/pipeline/phases/coverage-parsers/cobertura.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;GAqBG;AAEH,OAAO,EAAE,SAAS,EAAE,MAAM,iBAAiB,CAAC;AAC5C,OAAO,EAAE,UAAU,EAAqB,KAAK,EAAE,MAAM,YAAY,CAAC;AAOlE,MAAM,UAAU,cAAc,CAAC,GAAW,EAAE,QAAgB;IAC1D,MAAM,MAAM,GAAG,IAAI,SAAS,CAAC;QAC3B,gBAAgB,EAAE,KAAK;QACvB,mBAAmB,EAAE,IAAI;QACzB,wDAAwD;QACxD,OAAO,EAAE,CAAC,IAAI,EAAE,EAAE,CAChB,IAAI,KAAK,SAAS;YAClB,IAAI,KAAK,OAAO;YAChB,IAAI,KAAK,MAAM;YACf,IAAI,KAAK,UAAU;YACnB,IAAI,KAAK,SAAS;YAClB,IAAI,KAAK,OAAO;KACnB,CAAC,CAAC;IACH,IAAI,MAAe,CAAC;IACpB,IAAI,CAAC;QACH,MAAM,GAAG,MAAM,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC;IAC7B,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,IAAI,GAAG,EAAE,CAAC;IACnB,CAAC;IAED,MAAM,OAAO,GAAG,cAAc,CAAC,MAAM,CAAC,CAAC;IACvC,MAAM,GAAG,GAAG,IAAI,GAAG,EAAwB,CAAC;IAC5C,KAAK,MAAM,CAAC,IAAI,OAAO,EAAE,CAAC;QACxB,MAAM,QAAQ,GAAG,CAAC,CAAC,QAAQ,CAAC;QAC5B,IAAI,QAAQ,KAAK,SAAS,IAAI,QAAQ,CAAC,MAAM,KAAK,CAAC;YAAE,SAAS;QAC9D,MAAM,OAAO,GAAG,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QAC9B,IAAI,OAAO,CAAC,MAAM,KAAK,CAAC;YAAE,SAAS;QACnC,IAAI,OAAO,GAAG,CAAC,CAAC;QAChB,MAAM,YAAY,GAAa,EAAE,CAAC;QAClC,KAAK,MAAM,IAAI,IAAI,OAAO,EAAE,CAAC;YAC3B,IAAI,IAAI,CAAC,IAAI,GAAG,CAAC,EAAE,CAAC;gBAClB,OAAO,IAAI,CAAC,CAAC;gBACb,YAAY,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;YAC/B,CAAC;QACH,CAAC;QACD,MAAM,KAAK,GAAG,OAAO,CAAC,MAAM,CAAC;QAC7B,MAAM,OAAO,GAAG,aAAa,CAAC,QAAQ,EAAE,QAAQ,CAAC,CAAC;QAClD,GAAG,CAAC,GAAG,CAAC,OAAO,EAAE;YACf,QAAQ,EAAE,OAAO;YACjB,YAAY,EAAE,UAAU,CAAC,YAAY,CAAC;YACtC,UAAU,EAAE,KAAK;YACjB,eAAe,EAAE,KAAK,CAAC,OAAO,EAAE,KAAK,CAAC;SACvC,CAAC,CAAC;IACL,CAAC;IACD,OAAO,GAAG,CAAC;AACb,CAAC;AAED,SAAS,cAAc,CAAC,MAAe;IACrC,MAAM,GAAG,GAAiB,EAAE,CAAC;IAC7B,MAAM,IAAI,GAAI,MAAiC,EAAE,QAAQ,CAAC;IAC1D,MAAM,iBAAiB,GAAI,IAAiC,EAAE,QAAQ,CAAC;IACvE,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,iBAAiB,CAAC;QAAE,OAAO,GAAG,CAAC;IAClD,uEAAuE;IACvE,0EAA0E;IAC1E,eAAe;IACf,KAAK,MAAM,SAAS,IAAI,iBAAiB,EAAE,CAAC;QAC1C,MAAM,QAAQ,GAAI,SAAqC,EAAE,OAAO,CAAC;QACjE,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,QAAQ,CAAC;YAAE,SAAS;QACvC,KAAK,MAAM,GAAG,IAAI,QAAQ,EAAE,CAAC;YAC3B,MAAM,gBAAgB,GAAI,GAA+B,EAAE,OAAO,CAAC;YACnE,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,gBAAgB,CAAC;gBAAE,SAAS;YAC/C,KAAK,MAAM,UAAU,IAAI,gBAAgB,EAAE,CAAC;gBAC1C,MAAM,OAAO,GAAI,UAAoC,EAAE,KAAK,CAAC;gBAC7D,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,OAAO,CAAC;oBAAE,SAAS;gBACtC,KAAK,MAAM,GAAG,IAAI,OAAO,EAAE,CAAC;oBAC1B,MAAM,QAAQ,GAAI,GAA+B,CAAC,YAAY,CAAC,CAAC;oBAChE,MAAM,cAAc,GAAI,GAA6B,EAAE,KAAK,CAAC;oBAC7D,MAAM,QAAQ,GACZ,KAAK,CAAC,OAAO,CAAC,cAAc,CAAC,IAAI,cAAc,CAAC,MAAM,GAAG,CAAC;wBACxD,CAAC,CAAC,CAAE,cAAc,CAAC,CAAC,CAA0B,EAAE,IAAI,IAAI,EAAE,CAAC;wBAC3D,CAAC,CAAC,EAAE,CAAC;oBACT,MAAM,OAAO,GAAqC,EAAE,CAAC;oBACrD,IAAI,KAAK,CAAC,OAAO,CAAC,QAAQ,CAAC,EAAE,CAAC;wBAC5B,KAAK,MAAM,EAAE,IAAI,QAAQ,EAAE,CAAC;4BAC1B,MAAM,GAAG,GAAG,EAA6B,CAAC;4BAC1C,MAAM,OAAO,GAAG,UAAU,CAAC,GAAG,CAAC,UAAU,CAAC,CAAC,CAAC;4BAC5C,MAAM,IAAI,GAAG,UAAU,CAAC,GAAG,CAAC,QAAQ,CAAC,CAAC,CAAC;4BACvC,IAAI,OAAO,KAAK,SAAS,IAAI,IAAI,KAAK,SAAS;gCAAE,SAAS;4BAC1D,OAAO,CAAC,IAAI,CAAC,EAAE,IAAI,EAAE,OAAO,EAAE,IAAI,EAAE,CAAC,CAAC;wBACxC,CAAC;oBACH,CAAC;oBACD,GAAG,CAAC,IAAI,CAAC;wBACP,GAAG,CAAC,OAAO,QAAQ,KAAK,QAAQ,CAAC,CAAC,CAAC,EAAE,QAAQ,EAAE,CAAC,CAAC,CAAC,EAAE,CAAC;wBACrD,KAAK,EAAE,OAAO;qBACf,CAAC,CAAC;gBACL,CAAC;YACH,CAAC;QACH,CAAC;IACH,CAAC;IACD,OAAO,GAAG,CAAC;AACb,CAAC;AAED,SAAS,UAAU,CAAC,CAAU;IAC5B,IAAI,OAAO,CAAC,KAAK,QAAQ;QAAE,OAAO,MAAM,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC;IACrE,IAAI,OAAO,CAAC,KAAK,QAAQ,EAAE,CAAC;QAC1B,MAAM,CAAC,GAAG,MAAM,CAAC,QAAQ,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;QACjC,OAAO,MAAM,CAAC,SAAS,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC;IAC7C,CAAC;IACD,OAAO,SAAS,CAAC;AACnB,CAAC;AAED,SAAS,aAAa,CAAC,GAAW,EAAE,QAAgB;IAClD,MAAM,KAAK,GAAG,GAAG,CAAC,OAAO,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC;IACtC,IAAI,CAAC,KAAK,CAAC,UAAU,CAAC,GAAG,CAAC;QAAE,OAAO,KAAK,CAAC;IACzC,MAAM,SAAS,GAAG,QAAQ,CAAC,OAAO,CAAC,KAAK,EAAE,GAAG,CAAC,CAAC;IAC/C,MAAM,MAAM,GAAG,SAAS,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,GAAG,SAAS,GAAG,CAAC;IACrE,IAAI,KAAK,CAAC,UAAU,CAAC,MAAM,CAAC;QAAE,OAAO,KAAK,CAAC,KAAK,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC;IAChE,OAAO,KAAK,CAAC;AACf,CAAC"}

package/dist/pipeline/phases/coverage-parsers/coverage-py.d.ts

@@ -0,0 +1,25 @@
+/**
+ * coverage.py JSON report parser.
+ *
+ * Shape emitted by `coverage json` (a.k.a. `coverage report --format=json`):
+ *   {
+ *     "files": {
+ *       "src/foo.py": {
+ *         "executed_lines": [1, 2, 3],
+ *         "missing_lines":  [7, 10],
+ *         "summary": { "covered_lines": 3, "num_statements": 5, ... }
+ *       }
+ *     }
+ *   }
+ *
+ * We key on `executed_lines` for the covered set and sum
+ * `executed_lines + missing_lines` to get the total. That matches coverage.py
+ * semantics more precisely than the `summary.num_statements` field, which
+ * excludes excluded-but-executable lines.
+ *
+ * Paths are already repo-relative POSIX in coverage.py output, so no
+ * rewriting is necessary.
+ */
+import { type FileCoverage } from "./types.js";
+export declare function parseCoveragePy(raw: string, _repoRoot: string): ReadonlyMap<string, FileCoverage>;
+//# sourceMappingURL=coverage-py.d.ts.map

package/dist/pipeline/phases/coverage-parsers/coverage-py.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"coverage-py.d.ts","sourceRoot":"","sources":["../../../../src/pipeline/phases/coverage-parsers/coverage-py.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;GAqBG;AAEH,OAAO,EAAc,KAAK,YAAY,EAAS,MAAM,YAAY,CAAC;AAWlE,wBAAgB,eAAe,CAAC,GAAG,EAAE,MAAM,EAAE,SAAS,EAAE,MAAM,GAAG,WAAW,CAAC,MAAM,EAAE,YAAY,CAAC,CAwBjG"}

package/dist/pipeline/phases/coverage-parsers/coverage-py.js

@@ -0,0 +1,51 @@
+/**
+ * coverage.py JSON report parser.
+ *
+ * Shape emitted by `coverage json` (a.k.a. `coverage report --format=json`):
+ *   {
+ *     "files": {
+ *       "src/foo.py": {
+ *         "executed_lines": [1, 2, 3],
+ *         "missing_lines":  [7, 10],
+ *         "summary": { "covered_lines": 3, "num_statements": 5, ... }
+ *       }
+ *     }
+ *   }
+ *
+ * We key on `executed_lines` for the covered set and sum
+ * `executed_lines + missing_lines` to get the total. That matches coverage.py
+ * semantics more precisely than the `summary.num_statements` field, which
+ * excludes excluded-but-executable lines.
+ *
+ * Paths are already repo-relative POSIX in coverage.py output, so no
+ * rewriting is necessary.
+ */
+import { canonLines, ratio } from "./types.js";
+export function parseCoveragePy(raw, _repoRoot) {
+    const out = new Map();
+    let parsed;
+    try {
+        parsed = JSON.parse(raw);
+    }
+    catch {
+        return out;
+    }
+    const files = parsed?.files;
+    if (files === undefined || typeof files !== "object")
+        return out;
+    for (const [path, rec] of Object.entries(files)) {
+        const executed = Array.isArray(rec?.executed_lines) ? rec.executed_lines : [];
+        const missing = Array.isArray(rec?.missing_lines) ? rec.missing_lines : [];
+        const total = executed.length + missing.length;
+        if (total === 0)
+            continue;
+        out.set(path, {
+            filePath: path,
+            coveredLines: canonLines(executed),
+            totalLines: total,
+            coveragePercent: ratio(executed.length, total),
+        });
+    }
+    return out;
+}
+//# sourceMappingURL=coverage-py.js.map

package/dist/pipeline/phases/coverage-parsers/coverage-py.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"coverage-py.js","sourceRoot":"","sources":["../../../../src/pipeline/phases/coverage-parsers/coverage-py.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;GAqBG;AAEH,OAAO,EAAE,UAAU,EAAqB,KAAK,EAAE,MAAM,YAAY,CAAC;AAWlE,MAAM,UAAU,eAAe,CAAC,GAAW,EAAE,SAAiB;IAC5D,MAAM,GAAG,GAAG,IAAI,GAAG,EAAwB,CAAC;IAC5C,IAAI,MAAe,CAAC;IACpB,IAAI,CAAC;QACH,MAAM,GAAG,IAAI,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC;IAC3B,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,GAAG,CAAC;IACb,CAAC;IACD,MAAM,KAAK,GAAI,MAAsB,EAAE,KAAK,CAAC;IAC7C,IAAI,KAAK,KAAK,SAAS,IAAI,OAAO,KAAK,KAAK,QAAQ;QAAE,OAAO,GAAG,CAAC;IAEjE,KAAK,MAAM,CAAC,IAAI,EAAE,GAAG,CAAC,IAAI,MAAM,CAAC,OAAO,CAAC,KAAK,CAAC,EAAE,CAAC;QAChD,MAAM,QAAQ,GAAG,KAAK,CAAC,OAAO,CAAC,GAAG,EAAE,cAAc,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,cAAc,CAAC,CAAC,CAAC,EAAE,CAAC;QAC9E,MAAM,OAAO,GAAG,KAAK,CAAC,OAAO,CAAC,GAAG,EAAE,aAAa,CAAC,CAAC,CAAC,CAAC,GAAG,CAAC,aAAa,CAAC,CAAC,CAAC,EAAE,CAAC;QAC3E,MAAM,KAAK,GAAG,QAAQ,CAAC,MAAM,GAAG,OAAO,CAAC,MAAM,CAAC;QAC/C,IAAI,KAAK,KAAK,CAAC;YAAE,SAAS;QAC1B,GAAG,CAAC,GAAG,CAAC,IAAI,EAAE;YACZ,QAAQ,EAAE,IAAI;YACd,YAAY,EAAE,UAAU,CAAC,QAAQ,CAAC;YAClC,UAAU,EAAE,KAAK;YACjB,eAAe,EAAE,KAAK,CAAC,QAAQ,CAAC,MAAM,EAAE,KAAK,CAAC;SAC/C,CAAC,CAAC;IACL,CAAC;IACD,OAAO,GAAG,CAAC;AACb,CAAC"}

package/dist/pipeline/phases/coverage-parsers/jacoco.d.ts

@@ -0,0 +1,32 @@
+/**
+ * JaCoCo XML parser.
+ *
+ * JaCoCo's XML report tree:
+ *   <report>
+ *     <package name="com/foo">
+ *       <sourcefile name="Bar.java">
+ *         <line nr="12" mi="0" ci="3" mb="0" cb="0"/>
+ *         ...
+ *       </sourcefile>
+ *
+ * Per the JaCoCo DTD, `mi` is missed-instructions, `ci` is covered-
+ * instructions. We treat a line as covered when `ci > 0`. Total is the
+ * count of `<line>` elements for the sourcefile.
+ *
+ * Path assembly: the JaCoCo `name` attribute on `<package>` is slash-joined
+ * (e.g. `com/foo`), and the `<sourcefile>` name is the basename. We join
+ * them into a single POSIX path for graph matching. Callers may additionally
+ * set a project-source-root prefix via `options.sourceRoot` so the returned
+ * path aligns with the scan output.
+ */
+import { type FileCoverage } from "./types.js";
+export interface JacocoOptions {
+    /**
+     * Optional repo-relative prefix to prepend when assembling the source path
+     * (e.g. `src/main/java`). Defaults to the empty string because JaCoCo's
+     * report layout varies by build tool.
+     */
+    readonly sourceRoot?: string;
+}
+export declare function parseJacoco(raw: string, _repoRoot: string, options?: JacocoOptions): ReadonlyMap<string, FileCoverage>;
+//# sourceMappingURL=jacoco.d.ts.map

package/dist/pipeline/phases/coverage-parsers/jacoco.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"jacoco.d.ts","sourceRoot":"","sources":["../../../../src/pipeline/phases/coverage-parsers/jacoco.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;GAoBG;AAGH,OAAO,EAAc,KAAK,YAAY,EAAS,MAAM,YAAY,CAAC;AAElE,MAAM,WAAW,aAAa;IAC5B;;;;OAIG;IACH,QAAQ,CAAC,UAAU,CAAC,EAAE,MAAM,CAAC;CAC9B;AAED,wBAAgB,WAAW,CACzB,GAAG,EAAE,MAAM,EACX,SAAS,EAAE,MAAM,EACjB,OAAO,GAAE,aAAkB,GAC1B,WAAW,CAAC,MAAM,EAAE,YAAY,CAAC,CAwDnC"}