@aroman22/codegraph-vba-darwin-arm64 1.5.2 → 1.6.0

package/lib/dist/db/schema.sql

@@ -68,7 +68,15 @@ CREATE TABLE IF NOT EXISTS files (
     errors TEXT -- JSON array
 );
 
--- Unresolved References: References that need resolution after full indexing
+-- Unresolved References: References that need resolution after full indexing.
+-- status lifecycle: rows are inserted 'pending' by extraction; a completed
+-- resolution pass either deletes a row (resolved) or marks it 'failed'
+-- (attempted, no match — kept so a later sync can retry it when a changed
+-- file introduces a symbol that could satisfy it, #1240). name_tail is the
+-- last segment of reference_name ('util.greet' → 'greet'), written when a
+-- row is marked failed, so the retry lookup matches new node names against
+-- dotted refs too. Rows follow their from_node via ON DELETE CASCADE, so
+-- re-extracting or deleting a file clears its stale rows in any status.
 CREATE TABLE IF NOT EXISTS unresolved_refs (
     id INTEGER PRIMARY KEY AUTOINCREMENT,
     from_node_id TEXT NOT NULL,
@@ -79,6 +87,8 @@ CREATE TABLE IF NOT EXISTS unresolved_refs (
     candidates TEXT, -- JSON array
     file_path TEXT NOT NULL DEFAULT '',
     language TEXT NOT NULL DEFAULT 'unknown',
+    status TEXT NOT NULL DEFAULT 'pending',
+    name_tail TEXT NOT NULL DEFAULT '',
     FOREIGN KEY (from_node_id) REFERENCES nodes(id) ON DELETE CASCADE
 );
 
@@ -124,6 +134,25 @@ CREATE TRIGGER IF NOT EXISTS nodes_au AFTER UPDATE ON nodes BEGIN
     VALUES (NEW.rowid, NEW.id, NEW.name, NEW.qualified_name, NEW.docstring, NEW.signature);
 END;
 
+-- Prose-word → symbol-name lookup for the prompt hook's graph-derived gate.
+-- One row per (segment, name): segment is a lowercased word of a symbol name
+-- ("OrderStateMachine" → order, state, machine — see identifier-segments.ts),
+-- which lets natural-language prompt words be verified against the graph in
+-- any language whose technical nouns are Latin script. File nodes are
+-- excluded — a file's basename duplicates the symbols inside it and skews the
+-- singleton-vs-cluster rarity statistics. FTS can't serve this lookup (its
+-- tokenizer keeps camelCase names as single tokens), so segments are
+-- materialized on the node write path.
+-- Deletions leave orphan rows ON PURPOSE: rows are PROPOSALS, always
+-- re-verified against nodes before being surfaced (CodeGraph.getSegmentMatches),
+-- and a full index clears the table at its start. Populated lazily on old
+-- databases (empty until the next index/sync heals it).
+CREATE TABLE IF NOT EXISTS name_segment_vocab (
+    segment TEXT NOT NULL,
+    name TEXT NOT NULL,
+    PRIMARY KEY (segment, name)
+) WITHOUT ROWID;
+
 -- Edge indexes.
 -- idx_edges_source / idx_edges_target are intentionally omitted —
 -- the (source, kind) and (target, kind) composites below cover the
@@ -134,6 +163,17 @@ CREATE INDEX IF NOT EXISTS idx_edges_kind ON edges(kind);
 CREATE INDEX IF NOT EXISTS idx_edges_source_kind ON edges(source, kind);
 CREATE INDEX IF NOT EXISTS idx_edges_target_kind ON edges(target, kind);
 
+-- Edge identity uniqueness. An edge IS uniquely (source, target, kind, line,
+-- col); insertEdge uses `INSERT OR IGNORE`, but without something UNIQUE to
+-- conflict on it behaved like a plain INSERT, so two passes emitting the same
+-- edge produced byte-identical duplicate rows that inflated counts and flowed
+-- into callers/impact (#1034). IFNULL folds the nullable line/col so
+-- coordinate-less edges (synthesized / file-level) dedup too — SQLite treats
+-- each NULL as distinct otherwise. Migration v6 dedups existing rows + adds
+-- this on older databases.
+CREATE UNIQUE INDEX IF NOT EXISTS idx_edges_identity
+  ON edges(source, target, kind, IFNULL(line, -1), IFNULL(col, -1));
+
 -- File indexes
 CREATE INDEX IF NOT EXISTS idx_files_language ON files(language);
 CREATE INDEX IF NOT EXISTS idx_files_modified_at ON files(modified_at);
@@ -143,6 +183,8 @@ CREATE INDEX IF NOT EXISTS idx_unresolved_from_node ON unresolved_refs(from_node
 CREATE INDEX IF NOT EXISTS idx_unresolved_name ON unresolved_refs(reference_name);
 CREATE INDEX IF NOT EXISTS idx_unresolved_file_path ON unresolved_refs(file_path);
 CREATE INDEX IF NOT EXISTS idx_unresolved_from_name ON unresolved_refs(from_node_id, reference_name);
+CREATE INDEX IF NOT EXISTS idx_unresolved_status ON unresolved_refs(status);
+CREATE INDEX IF NOT EXISTS idx_unresolved_failed_tail ON unresolved_refs(name_tail) WHERE status = 'failed';
 CREATE INDEX IF NOT EXISTS idx_edges_provenance ON edges(provenance);
 
 -- Project metadata for version/provenance tracking

package/lib/dist/db/wal-valve.d.ts

@@ -0,0 +1,106 @@
+/**
+ * WAL checkpoint valve — bounds WAL growth while auto-checkpointing is
+ * deferred during a bulk index (#1231).
+ *
+ * Why deferral: SQLite's default `wal_autocheckpoint` (1000 pages) re-writes
+ * hot B-tree/FTS pages into the main DB file over and over during a bulk
+ * index — measured at ~95% of ALL disk I/O, and the difference between 45s
+ * and 19+ minutes on HDD-class storage (150 random IOPS). Deferring
+ * checkpoints turns the store into pure sequential WAL appends; each backfill
+ * pass writes distinct pages once, in page order (≈ sequential).
+ *
+ * Why a valve: unbounded deferral is its own failure mode, both measured in
+ * the #1231 repro. The WAL duplicates hot pages per COMMIT, so it grows far
+ * faster than the DB (5.9GB WAL for a ~340MB DB on a 3.3k-file index) —
+ * filling the disk, and poisoning every subsequent read that must page
+ * through it (the first resolution-phase read blocked the main thread >60s
+ * and the #850 liveness watchdog killed the healthy index). The valve
+ * watches WAL growth on a timer and, past a soft threshold, backfills with
+ * `PRAGMA wal_checkpoint(PASSIVE)` on a worker-thread connection — PASSIVE
+ * never blocks the writer, and off-thread means the main thread (and the
+ * watchdog heartbeat) keep turning regardless of how long a backfill takes.
+ *
+ * The load-bearing subtlety: a WAL file's SIZE never shrinks. After a full
+ * backfill, the writer's next commit RESTARTS the WAL from the top and the
+ * frames recycle inside the same file — so raw size says nothing about the
+ * un-backfilled backlog, and a size-triggered valve degenerates into firing
+ * (and pausing the writer) forever once the file passes its threshold
+ * (measured: guava crawled at ~9min per 160 files). Instead the valve
+ * tracks `sizeAtLastFullBackfill` — refreshed whenever a checkpoint reports
+ * `log === checkpointed` (everything backfilled) — and triggers on GROWTH
+ * beyond that baseline, which only happens when genuinely un-backfilled
+ * frames push past the file's high-water mark.
+ *
+ * Backpressure: if the writer outruns the checkpointer past a hard cap of
+ * growth (2× soft), {@link backpressure} pauses the writer (at a safe,
+ * between-transactions boundary) until a FULL backfill lands. One in-flight
+ * pass is not enough: on a disk saturated by the writer, every concurrent
+ * PASSIVE pass is already stale by the time it finishes (the writer appended
+ * past its snapshot), so neither SQLite's WAL wrap nor the baseline ever
+ * trigger and the WAL grows without bound (measured: 5.9GB on guava at 150
+ * IOPS, then a >60s read stall and a watchdog kill). With the writer parked,
+ * the next pass covers everything, the WAL wraps on the following commit,
+ * and the pause is the disk's honest catch-up cost — the correct terminal
+ * mode when hardware genuinely can't keep up with the append rate.
+ */
+import type { DatabaseConnection } from './index';
+/**
+ * Resolve the valve's soft threshold from the `CODEGRAPH_WAL_VALVE_MB`
+ * override; non-numeric / non-positive values fall back to the default.
+ */
+export declare function resolveWalValveMb(envVal: string | undefined): number;
+export declare class WalCheckpointValve {
+    private readonly db;
+    private readonly intervalMs;
+    private readonly log;
+    private timer;
+    private inflight;
+    /** Writer pause in progress (hard cap breached): passes loop until a full backfill. */
+    private pause;
+    /**
+     * WAL file size observed when a checkpoint last reported the ENTIRE WAL
+     * backfilled. Growth is measured against this baseline — see the header
+     * comment for why absolute size cannot be used.
+     */
+    private sizeAtLastFullBackfill;
+    private readonly softBytes;
+    private readonly hardBytes;
+    constructor(db: DatabaseConnection, softMb?: number, intervalMs?: number, log?: (msg: string) => void);
+    private mb;
+    /** Un-backfilled growth estimate: bytes the WAL has grown past the last full backfill. */
+    private growthBytes;
+    /** Begin watching the WAL. Idempotent; the timer never holds the loop open. */
+    start(): void;
+    /** Stop watching. Any in-flight checkpoint keeps running — await drain(). */
+    stop(): void;
+    /** One poll: fire an off-thread passive checkpoint when growth passes the soft threshold. */
+    check(): void;
+    /**
+     * Writer-side backstop, called at a between-transactions boundary. Returns
+     * null (no wait) while growth is under the hard cap; past it, returns a
+     * promise that resolves only once a FULL backfill has landed — see the
+     * header comment for why a single pass is not enough on a saturated disk.
+     */
+    backpressure(): Promise<void> | null;
+    /** Await any in-flight checkpoint and writer pause. */
+    drain(): Promise<void>;
+    /**
+     * Phase-boundary fold: backfill the ENTIRE WAL now (off-thread, awaited).
+     * Called between bulk phases — e.g. after parsing, before resolution's
+     * first reads — so the next phase never pages a bulk-write-sized WAL on
+     * the main thread (the post-parse read against a multi-GB WAL is what
+     * blew the #850 watchdog's 60s window in the #1231 repro). The await
+     * keeps the event loop (and the watchdog heartbeat) turning.
+     */
+    foldNow(): Promise<void>;
+    /**
+     * With the writer parked on the returned promise, loop passive passes until
+     * one reports the entire WAL backfilled (typically the second: the first
+     * drains the pass that was already running against a stale snapshot). Gives
+     * up after a bounded number of passes — e.g. a reader pinning the WAL —
+     * because unbounded WAL growth degrades; a wedged writer never recovers.
+     */
+    private backfillFully;
+    private fire;
+}
+//# sourceMappingURL=wal-valve.d.ts.map

package/lib/dist/db/wal-valve.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"wal-valve.d.ts","sourceRoot":"","sources":["../../src/db/wal-valve.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GA4CG;AAEH,OAAO,KAAK,EAAE,kBAAkB,EAAE,MAAM,SAAS,CAAC;AAWlD;;;GAGG;AACH,wBAAgB,iBAAiB,CAAC,MAAM,EAAE,MAAM,GAAG,SAAS,GAAG,MAAM,CAMpE;AAED,qBAAa,kBAAkB;IAe3B,OAAO,CAAC,QAAQ,CAAC,EAAE;IAEnB,OAAO,CAAC,QAAQ,CAAC,UAAU;IAC3B,OAAO,CAAC,QAAQ,CAAC,GAAG;IAjBtB,OAAO,CAAC,KAAK,CAA+C;IAC5D,OAAO,CAAC,QAAQ,CAA8B;IAC9C,uFAAuF;IACvF,OAAO,CAAC,KAAK,CAA8B;IAC3C;;;;OAIG;IACH,OAAO,CAAC,sBAAsB,CAAK;IACnC,OAAO,CAAC,QAAQ,CAAC,SAAS,CAAS;IACnC,OAAO,CAAC,QAAQ,CAAC,SAAS,CAAS;gBAGhB,EAAE,EAAE,kBAAkB,EACvC,MAAM,GAAE,MAA8D,EACrD,UAAU,GAAE,MAA0B,EACtC,GAAG,GAAE,CAAC,GAAG,EAAE,MAAM,KAAK,IAAe;IAMxD,OAAO,CAAC,EAAE;IAIV,0FAA0F;IAC1F,OAAO,CAAC,WAAW;IAInB,+EAA+E;IAC/E,KAAK,IAAI,IAAI;IAMb,6EAA6E;IAC7E,IAAI,IAAI,IAAI;IAOZ,6FAA6F;IAC7F,KAAK,IAAI,IAAI;IAIb;;;;;OAKG;IACH,YAAY,IAAI,OAAO,CAAC,IAAI,CAAC,GAAG,IAAI;IAYpC,uDAAuD;IACjD,KAAK,IAAI,OAAO,CAAC,IAAI,CAAC;IAO5B;;;;;;;OAOG;IACG,OAAO,IAAI,OAAO,CAAC,IAAI,CAAC;IAQ9B;;;;;;OAMG;YACW,aAAa;IAc3B,OAAO,CAAC,IAAI;CAoBb"}

package/lib/dist/db/wal-valve.js

@@ -0,0 +1,208 @@
+"use strict";
+/**
+ * WAL checkpoint valve — bounds WAL growth while auto-checkpointing is
+ * deferred during a bulk index (#1231).
+ *
+ * Why deferral: SQLite's default `wal_autocheckpoint` (1000 pages) re-writes
+ * hot B-tree/FTS pages into the main DB file over and over during a bulk
+ * index — measured at ~95% of ALL disk I/O, and the difference between 45s
+ * and 19+ minutes on HDD-class storage (150 random IOPS). Deferring
+ * checkpoints turns the store into pure sequential WAL appends; each backfill
+ * pass writes distinct pages once, in page order (≈ sequential).
+ *
+ * Why a valve: unbounded deferral is its own failure mode, both measured in
+ * the #1231 repro. The WAL duplicates hot pages per COMMIT, so it grows far
+ * faster than the DB (5.9GB WAL for a ~340MB DB on a 3.3k-file index) —
+ * filling the disk, and poisoning every subsequent read that must page
+ * through it (the first resolution-phase read blocked the main thread >60s
+ * and the #850 liveness watchdog killed the healthy index). The valve
+ * watches WAL growth on a timer and, past a soft threshold, backfills with
+ * `PRAGMA wal_checkpoint(PASSIVE)` on a worker-thread connection — PASSIVE
+ * never blocks the writer, and off-thread means the main thread (and the
+ * watchdog heartbeat) keep turning regardless of how long a backfill takes.
+ *
+ * The load-bearing subtlety: a WAL file's SIZE never shrinks. After a full
+ * backfill, the writer's next commit RESTARTS the WAL from the top and the
+ * frames recycle inside the same file — so raw size says nothing about the
+ * un-backfilled backlog, and a size-triggered valve degenerates into firing
+ * (and pausing the writer) forever once the file passes its threshold
+ * (measured: guava crawled at ~9min per 160 files). Instead the valve
+ * tracks `sizeAtLastFullBackfill` — refreshed whenever a checkpoint reports
+ * `log === checkpointed` (everything backfilled) — and triggers on GROWTH
+ * beyond that baseline, which only happens when genuinely un-backfilled
+ * frames push past the file's high-water mark.
+ *
+ * Backpressure: if the writer outruns the checkpointer past a hard cap of
+ * growth (2× soft), {@link backpressure} pauses the writer (at a safe,
+ * between-transactions boundary) until a FULL backfill lands. One in-flight
+ * pass is not enough: on a disk saturated by the writer, every concurrent
+ * PASSIVE pass is already stale by the time it finishes (the writer appended
+ * past its snapshot), so neither SQLite's WAL wrap nor the baseline ever
+ * trigger and the WAL grows without bound (measured: 5.9GB on guava at 150
+ * IOPS, then a >60s read stall and a watchdog kill). With the writer parked,
+ * the next pass covers everything, the WAL wraps on the following commit,
+ * and the pause is the disk's honest catch-up cost — the correct terminal
+ * mode when hardware genuinely can't keep up with the append rate.
+ */
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.WalCheckpointValve = void 0;
+exports.resolveWalValveMb = resolveWalValveMb;
+/** Soft WAL-growth threshold (MB) that triggers an off-thread passive checkpoint. */
+const DEFAULT_WAL_VALVE_MB = 256;
+/** Hard cap = this × soft threshold; past it the writer pauses for a full backfill. */
+const HARD_CAP_MULTIPLIER = 2;
+/** Passes attempted per writer pause before giving up (a pinned reader could stall forever). */
+const MAX_PAUSED_BACKFILL_PASSES = 20;
+/** How often the timer looks at the WAL file size. */
+const CHECK_INTERVAL_MS = 2000;
+/**
+ * Resolve the valve's soft threshold from the `CODEGRAPH_WAL_VALVE_MB`
+ * override; non-numeric / non-positive values fall back to the default.
+ */
+function resolveWalValveMb(envVal) {
+    if (envVal !== undefined && envVal !== '') {
+        const n = Number(envVal);
+        if (Number.isFinite(n) && n > 0)
+            return Math.floor(n);
+    }
+    return DEFAULT_WAL_VALVE_MB;
+}
+class WalCheckpointValve {
+    db;
+    intervalMs;
+    log;
+    timer = null;
+    inflight = null;
+    /** Writer pause in progress (hard cap breached): passes loop until a full backfill. */
+    pause = null;
+    /**
+     * WAL file size observed when a checkpoint last reported the ENTIRE WAL
+     * backfilled. Growth is measured against this baseline — see the header
+     * comment for why absolute size cannot be used.
+     */
+    sizeAtLastFullBackfill = 0;
+    softBytes;
+    hardBytes;
+    constructor(db, softMb = resolveWalValveMb(process.env.CODEGRAPH_WAL_VALVE_MB), intervalMs = CHECK_INTERVAL_MS, log = () => { }) {
+        this.db = db;
+        this.intervalMs = intervalMs;
+        this.log = log;
+        this.softBytes = softMb * 1024 * 1024;
+        this.hardBytes = this.softBytes * HARD_CAP_MULTIPLIER;
+    }
+    mb(n) {
+        return `${Math.round(n / 1024 / 1024)}MB`;
+    }
+    /** Un-backfilled growth estimate: bytes the WAL has grown past the last full backfill. */
+    growthBytes() {
+        return this.db.getWalSizeBytes() - this.sizeAtLastFullBackfill;
+    }
+    /** Begin watching the WAL. Idempotent; the timer never holds the loop open. */
+    start() {
+        if (this.timer)
+            return;
+        this.timer = setInterval(() => this.check(), this.intervalMs);
+        this.timer.unref?.();
+    }
+    /** Stop watching. Any in-flight checkpoint keeps running — await drain(). */
+    stop() {
+        if (this.timer) {
+            clearInterval(this.timer);
+            this.timer = null;
+        }
+    }
+    /** One poll: fire an off-thread passive checkpoint when growth passes the soft threshold. */
+    check() {
+        if (!this.pause && !this.inflight && this.growthBytes() > this.softBytes)
+            this.fire();
+    }
+    /**
+     * Writer-side backstop, called at a between-transactions boundary. Returns
+     * null (no wait) while growth is under the hard cap; past it, returns a
+     * promise that resolves only once a FULL backfill has landed — see the
+     * header comment for why a single pass is not enough on a saturated disk.
+     */
+    backpressure() {
+        if (this.pause)
+            return this.pause;
+        if (this.growthBytes() <= this.hardBytes)
+            return null;
+        this.log(`backpressure: wal=${this.mb(this.db.getWalSizeBytes())} baseline=${this.mb(this.sizeAtLastFullBackfill)} — pausing writer for full backfill`);
+        const t0 = Date.now();
+        this.pause = this.backfillFully().finally(() => {
+            this.pause = null;
+            this.log(`backpressure released after ${Date.now() - t0}ms: wal=${this.mb(this.db.getWalSizeBytes())} baseline=${this.mb(this.sizeAtLastFullBackfill)}`);
+        });
+        return this.pause;
+    }
+    /** Await any in-flight checkpoint and writer pause. */
+    async drain() {
+        while (this.pause || this.inflight) {
+            if (this.pause)
+                await this.pause;
+            if (this.inflight)
+                await this.inflight;
+        }
+    }
+    /**
+     * Phase-boundary fold: backfill the ENTIRE WAL now (off-thread, awaited).
+     * Called between bulk phases — e.g. after parsing, before resolution's
+     * first reads — so the next phase never pages a bulk-write-sized WAL on
+     * the main thread (the post-parse read against a multi-GB WAL is what
+     * blew the #850 watchdog's 60s window in the #1231 repro). The await
+     * keeps the event loop (and the watchdog heartbeat) turning.
+     */
+    async foldNow() {
+        await this.drain();
+        if (this.growthBytes() <= 0)
+            return;
+        this.log(`foldNow: wal=${this.mb(this.db.getWalSizeBytes())} baseline=${this.mb(this.sizeAtLastFullBackfill)}`);
+        this.pause = this.backfillFully().finally(() => { this.pause = null; });
+        await this.pause;
+    }
+    /**
+     * With the writer parked on the returned promise, loop passive passes until
+     * one reports the entire WAL backfilled (typically the second: the first
+     * drains the pass that was already running against a stale snapshot). Gives
+     * up after a bounded number of passes — e.g. a reader pinning the WAL —
+     * because unbounded WAL growth degrades; a wedged writer never recovers.
+     */
+    async backfillFully() {
+        for (let i = 0; i < MAX_PAUSED_BACKFILL_PASSES; i++) {
+            if (this.inflight)
+                await this.inflight; // fold in the stale in-flight pass first
+            const res = await this.db.checkpointWalPassive();
+            if (!res)
+                return; // checkpoint machinery unavailable — don't spin
+            this.log(`backfill pass ${i + 1}: busy=${res.busy} log=${res.log} checkpointed=${res.checkpointed} wal=${this.mb(this.db.getWalSizeBytes())}`);
+            if (res.busy === 0 && res.log === res.checkpointed) {
+                this.sizeAtLastFullBackfill = this.db.getWalSizeBytes();
+                return;
+            }
+        }
+        this.log(`backfill gave up after ${MAX_PAUSED_BACKFILL_PASSES} passes — WAL stays unbounded this cycle`);
+    }
+    fire() {
+        const p = this.db
+            .checkpointWalPassive()
+            .then((res) => {
+            // Full backfill (busy 0, every log frame checkpointed) ⇒ the writer's
+            // next commit wraps the WAL; the file's current size becomes the new
+            // growth baseline. A partial pass (writer appended during it, or a
+            // read transaction pinned frames) leaves the baseline alone, so the
+            // next tick fires again and copies the remainder. In non-WAL mode
+            // SQLite reports log = checkpointed = -1, which is harmless here.
+            if (res && res.busy === 0 && res.log === res.checkpointed) {
+                this.sizeAtLastFullBackfill = this.db.getWalSizeBytes();
+            }
+        })
+            .catch(() => { })
+            .finally(() => {
+            if (this.inflight === p)
+                this.inflight = null;
+        });
+        this.inflight = p;
+    }
+}
+exports.WalCheckpointValve = WalCheckpointValve;
+//# sourceMappingURL=wal-valve.js.map

package/lib/dist/db/wal-valve.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"wal-valve.js","sourceRoot":"","sources":["../../src/db/wal-valve.ts"],"names":[],"mappings":";AAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;GA4CG;;;AAiBH,8CAMC;AAnBD,qFAAqF;AACrF,MAAM,oBAAoB,GAAG,GAAG,CAAC;AACjC,uFAAuF;AACvF,MAAM,mBAAmB,GAAG,CAAC,CAAC;AAC9B,gGAAgG;AAChG,MAAM,0BAA0B,GAAG,EAAE,CAAC;AACtC,sDAAsD;AACtD,MAAM,iBAAiB,GAAG,IAAI,CAAC;AAE/B;;;GAGG;AACH,SAAgB,iBAAiB,CAAC,MAA0B;IAC1D,IAAI,MAAM,KAAK,SAAS,IAAI,MAAM,KAAK,EAAE,EAAE,CAAC;QAC1C,MAAM,CAAC,GAAG,MAAM,CAAC,MAAM,CAAC,CAAC;QACzB,IAAI,MAAM,CAAC,QAAQ,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC;YAAE,OAAO,IAAI,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC;IACxD,CAAC;IACD,OAAO,oBAAoB,CAAC;AAC9B,CAAC;AAED,MAAa,kBAAkB;IAeV;IAEA;IACA;IAjBX,KAAK,GAA0C,IAAI,CAAC;IACpD,QAAQ,GAAyB,IAAI,CAAC;IAC9C,uFAAuF;IAC/E,KAAK,GAAyB,IAAI,CAAC;IAC3C;;;;OAIG;IACK,sBAAsB,GAAG,CAAC,CAAC;IAClB,SAAS,CAAS;IAClB,SAAS,CAAS;IAEnC,YACmB,EAAsB,EACvC,SAAiB,iBAAiB,CAAC,OAAO,CAAC,GAAG,CAAC,sBAAsB,CAAC,EACrD,aAAqB,iBAAiB,EACtC,MAA6B,GAAG,EAAE,GAAE,CAAC;QAHrC,OAAE,GAAF,EAAE,CAAoB;QAEtB,eAAU,GAAV,UAAU,CAA4B;QACtC,QAAG,GAAH,GAAG,CAAkC;QAEtD,IAAI,CAAC,SAAS,GAAG,MAAM,GAAG,IAAI,GAAG,IAAI,CAAC;QACtC,IAAI,CAAC,SAAS,GAAG,IAAI,CAAC,SAAS,GAAG,mBAAmB,CAAC;IACxD,CAAC;IAEO,EAAE,CAAC,CAAS;QAClB,OAAO,GAAG,IAAI,CAAC,KAAK,CAAC,CAAC,GAAG,IAAI,GAAG,IAAI,CAAC,IAAI,CAAC;IAC5C,CAAC;IAED,0FAA0F;IAClF,WAAW;QACjB,OAAO,IAAI,CAAC,EAAE,CAAC,eAAe,EAAE,GAAG,IAAI,CAAC,sBAAsB,CAAC;IACjE,CAAC;IAED,+EAA+E;IAC/E,KAAK;QACH,IAAI,IAAI,CAAC,KAAK;YAAE,OAAO;QACvB,IAAI,CAAC,KAAK,GAAG,WAAW,CAAC,GAAG,EAAE,CAAC,IAAI,CAAC,KAAK,EAAE,EAAE,IAAI,CAAC,UAAU,CAAC,CAAC;QAC9D,IAAI,CAAC,KAAK,CAAC,KAAK,EAAE,EAAE,CAAC;IACvB,CAAC;IAED,6EAA6E;IAC7E,IAAI;QACF,IAAI,IAAI,CAAC,KAAK,EAAE,CAAC;YACf,aAAa,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;YAC1B,IAAI,CAAC,KAAK,GAAG,IAAI,CAAC;QACpB,CAAC;IACH,CAAC;IAED,6FAA6F;IAC7F,KAAK;QACH,IAAI,CAAC,IAAI,CAAC,KAAK,IAAI,CAAC,IAAI,CAAC,QAAQ,IAAI,IAAI,CAAC,WAAW,EAAE,GAAG,IAAI,CAAC,SAAS;YAAE,IAAI,CAAC,IAAI,EAAE,CAAC;IACxF,CAAC;IAED;;;;;OAKG;IACH,YAAY;QACV,IAAI,IAAI,CAAC,KAAK;YAAE,OAAO,IAAI,CAAC,KAAK,CAAC;QAClC,IAAI,IAAI,CAAC,WAAW,EAAE,IAAI,IAAI,CAAC,SAAS;YAAE,OAAO,IAAI,CAAC;QACtD,IAAI,CAAC,GAAG,CAAC,qBAAqB,IAAI,CAAC,EAAE,CAAC,IAAI,CAAC,EAAE,CAAC,eAAe,EAAE,CAAC,aAAa,IAAI,CAAC,EAAE,CAAC,IAAI,CAAC,sBAAsB,CAAC,qCAAqC,CAAC,CAAC;QACxJ,MAAM,EAAE,GAAG,IAAI,CAAC,GAAG,EAAE,CAAC;QACtB,IAAI,CAAC,KAAK,GAAG,IAAI,CAAC,aAAa,EAAE,CAAC,OAAO,CAAC,GAAG,EAAE;YAC7C,IAAI,CAAC,KAAK,GAAG,IAAI,CAAC;YAClB,IAAI,CAAC,GAAG,CAAC,+BAA+B,IAAI,CAAC,GAAG,EAAE,GAAG,EAAE,WAAW,IAAI,CAAC,EAAE,CAAC,IAAI,CAAC,EAAE,CAAC,eAAe,EAAE,CAAC,aAAa,IAAI,CAAC,EAAE,CAAC,IAAI,CAAC,sBAAsB,CAAC,EAAE,CAAC,CAAC;QAC3J,CAAC,CAAC,CAAC;QACH,OAAO,IAAI,CAAC,KAAK,CAAC;IACpB,CAAC;IAED,uDAAuD;IACvD,KAAK,CAAC,KAAK;QACT,OAAO,IAAI,CAAC,KAAK,IAAI,IAAI,CAAC,QAAQ,EAAE,CAAC;YACnC,IAAI,IAAI,CAAC,KAAK;gBAAE,MAAM,IAAI,CAAC,KAAK,CAAC;YACjC,IAAI,IAAI,CAAC,QAAQ;gBAAE,MAAM,IAAI,CAAC,QAAQ,CAAC;QACzC,CAAC;IACH,CAAC;IAED;;;;;;;OAOG;IACH,KAAK,CAAC,OAAO;QACX,MAAM,IAAI,CAAC,KAAK,EAAE,CAAC;QACnB,IAAI,IAAI,CAAC,WAAW,EAAE,IAAI,CAAC;YAAE,OAAO;QACpC,IAAI,CAAC,GAAG,CAAC,gBAAgB,IAAI,CAAC,EAAE,CAAC,IAAI,CAAC,EAAE,CAAC,eAAe,EAAE,CAAC,aAAa,IAAI,CAAC,EAAE,CAAC,IAAI,CAAC,sBAAsB,CAAC,EAAE,CAAC,CAAC;QAChH,IAAI,CAAC,KAAK,GAAG,IAAI,CAAC,aAAa,EAAE,CAAC,OAAO,CAAC,GAAG,EAAE,GAAG,IAAI,CAAC,KAAK,GAAG,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC;QACxE,MAAM,IAAI,CAAC,KAAK,CAAC;IACnB,CAAC;IAED;;;;;;OAMG;IACK,KAAK,CAAC,aAAa;QACzB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,0BAA0B,EAAE,CAAC,EAAE,EAAE,CAAC;YACpD,IAAI,IAAI,CAAC,QAAQ;gBAAE,MAAM,IAAI,CAAC,QAAQ,CAAC,CAAC,yCAAyC;YACjF,MAAM,GAAG,GAAG,MAAM,IAAI,CAAC,EAAE,CAAC,oBAAoB,EAAE,CAAC;YACjD,IAAI,CAAC,GAAG;gBAAE,OAAO,CAAC,gDAAgD;YAClE,IAAI,CAAC,GAAG,CAAC,iBAAiB,CAAC,GAAG,CAAC,UAAU,GAAG,CAAC,IAAI,QAAQ,GAAG,CAAC,GAAG,iBAAiB,GAAG,CAAC,YAAY,QAAQ,IAAI,CAAC,EAAE,CAAC,IAAI,CAAC,EAAE,CAAC,eAAe,EAAE,CAAC,EAAE,CAAC,CAAC;YAC/I,IAAI,GAAG,CAAC,IAAI,KAAK,CAAC,IAAI,GAAG,CAAC,GAAG,KAAK,GAAG,CAAC,YAAY,EAAE,CAAC;gBACnD,IAAI,CAAC,sBAAsB,GAAG,IAAI,CAAC,EAAE,CAAC,eAAe,EAAE,CAAC;gBACxD,OAAO;YACT,CAAC;QACH,CAAC;QACD,IAAI,CAAC,GAAG,CAAC,0BAA0B,0BAA0B,0CAA0C,CAAC,CAAC;IAC3G,CAAC;IAEO,IAAI;QACV,MAAM,CAAC,GAAG,IAAI,CAAC,EAAE;aACd,oBAAoB,EAAE;aACtB,IAAI,CAAC,CAAC,GAAG,EAAE,EAAE;YACZ,sEAAsE;YACtE,qEAAqE;YACrE,mEAAmE;YACnE,oEAAoE;YACpE,kEAAkE;YAClE,kEAAkE;YAClE,IAAI,GAAG,IAAI,GAAG,CAAC,IAAI,KAAK,CAAC,IAAI,GAAG,CAAC,GAAG,KAAK,GAAG,CAAC,YAAY,EAAE,CAAC;gBAC1D,IAAI,CAAC,sBAAsB,GAAG,IAAI,CAAC,EAAE,CAAC,eAAe,EAAE,CAAC;YAC1D,CAAC;QACH,CAAC,CAAC;aACD,KAAK,CAAC,GAAG,EAAE,GAAqB,CAAC,CAAC;aAClC,OAAO,CAAC,GAAG,EAAE;YACZ,IAAI,IAAI,CAAC,QAAQ,KAAK,CAAC;gBAAE,IAAI,CAAC,QAAQ,GAAG,IAAI,CAAC;QAChD,CAAC,CAAC,CAAC;QACL,IAAI,CAAC,QAAQ,GAAG,CAAC,CAAC;IACpB,CAAC;CACF;AAxID,gDAwIC"}

package/lib/dist/directory.d.ts

@@ -86,10 +86,14 @@ export declare function findIndexedSubprojectRoots(root: string, opts?: {
     max?: number;
 }): string[];
 /**
- * Does `prompt` contain an explicit structural keyword (English or CJK)? A
- * keyword is a strong, self-contained signal, so the front-load hook fires on it
- * directly — no graph check needed. (A *code-token* match, by contrast, is only
- * a candidate the hook verifies against the graph first; see {@link extractCodeTokens}.)
+ * Does `prompt` contain an explicit structural keyword? A keyword is a strong,
+ * self-contained signal, so the front-load hook fires on it directly — no graph
+ * check needed. (A *code-token* match, by contrast, is only a candidate the
+ * hook verifies against the graph first; see {@link extractCodeTokens}.)
+ * Coverage is multilingual (#994, #1126): the ~29 languages with the largest
+ * developer populations, across Latin, Cyrillic, Greek, CJK, Hangul, Arabic,
+ * Hebrew, Thai, and Devanagari scripts. Languages beyond the keyword table
+ * still fire through the language-agnostic code-token path.
  */
 export declare function hasStructuralKeyword(prompt: string): boolean;
 /**
@@ -110,7 +114,7 @@ export declare function extractCodeTokens(prompt: string): string[];
 /**
  * Cheap, graph-free candidate gate for the front-load hook: could `prompt` be a
  * structural / flow / impact / "where-how" question worth front-loading context
- * for? True on an explicit keyword (English or CJK, issue #994) OR an
+ * for? True on an explicit keyword in any covered language (#994, #1126) OR an
  * identifier-shaped token. A keyword is sufficient to fire on its own; a
  * token-only match is only a candidate the hook then verifies against the graph
  * (a brand name like `JavaScript` is token-shaped but isn't a symbol). Every

package/lib/dist/directory.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"directory.d.ts","sourceRoot":"","sources":["../src/directory.ts"],"names":[],"mappings":"AAAA;;;;GAIG;AAWH;;;;;;;;;;;;;;;;;;GAkBG;AACH,wBAAgB,gBAAgB,IAAI,MAAM,CAqBzC;AAED;;;;;;GAMG;AACH,eAAO,MAAM,aAAa,QAAqB,CAAC;AAEhD;;;;;;GAMG;AACH,wBAAgB,kBAAkB,CAAC,IAAI,EAAE,MAAM,GAAG,OAAO,CAQxD;AAED;;GAEG;AACH,wBAAgB,eAAe,CAAC,WAAW,EAAE,MAAM,GAAG,MAAM,CAE3D;AAED;;;GAGG;AACH,wBAAgB,aAAa,CAAC,WAAW,EAAE,MAAM,GAAG,OAAO,CAQ1D;AAED;;;;;;;;GAQG;AACH;;;;;;;;;;;;;GAaG;AACH,wBAAgB,qBAAqB,CAAC,WAAW,EAAE,MAAM,GAAG,MAAM,GAAG,IAAI,CA8BxE;AAED,wBAAgB,wBAAwB,CAAC,SAAS,EAAE,MAAM,GAAG,MAAM,GAAG,IAAI,CAmBzE;AA2BD;;;;;;;GAOG;AACH,wBAAgB,0BAA0B,CACxC,IAAI,EAAE,MAAM,EACZ,IAAI,GAAE;IAAE,QAAQ,CAAC,EAAE,MAAM,CAAC;IAAC,GAAG,CAAC,EAAE,MAAM,CAAA;CAAO,GAC7C,MAAM,EAAE,CAmBV;AAyBD;;;;;GAKG;AACH,wBAAgB,oBAAoB,CAAC,MAAM,EAAE,MAAM,GAAG,OAAO,CAE5D;AAED;;;;;;;;;;;;;GAaG;AACH,wBAAgB,iBAAiB,CAAC,MAAM,EAAE,MAAM,GAAG,MAAM,EAAE,CAkB1D;AAED;;;;;;;;GAQG;AACH,wBAAgB,kBAAkB,CAAC,MAAM,EAAE,MAAM,GAAG,OAAO,CAE1D;AAED;;GAEG;AACH,MAAM,WAAW,aAAa;IAC5B;;yEAEqE;IACrE,WAAW,EAAE,MAAM,GAAG,IAAI,CAAC;IAC3B;;6EAEyE;IACzE,aAAa,EAAE,MAAM,EAAE,CAAC;IACxB;;+DAE2D;IAC3D,UAAU,EAAE,OAAO,CAAC;CACrB;AAED;;;;;;;;;;;;GAYG;AACH,wBAAgB,aAAa,CAAC,GAAG,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,GAAG,aAAa,CAwCxE;AAyDD;;;GAGG;AACH,wBAAgB,eAAe,CAAC,WAAW,EAAE,MAAM,GAAG,IAAI,CAkCzD;AAED;;GAEG;AACH,wBAAgB,eAAe,CAAC,WAAW,EAAE,MAAM,GAAG,IAAI,CAuBzD;AAED;;GAEG;AACH,wBAAgB,qBAAqB,CAAC,WAAW,EAAE,MAAM,GAAG,MAAM,EAAE,CA8BnE;AAED;;GAEG;AACH,wBAAgB,gBAAgB,CAAC,WAAW,EAAE,MAAM,GAAG,MAAM,CA+B5D;AAED;;GAEG;AACH,wBAAgB,kBAAkB,CAAC,WAAW,EAAE,MAAM,EAAE,UAAU,EAAE,MAAM,GAAG,MAAM,CAYlF;AAED;;GAEG;AACH,wBAAgB,iBAAiB,CAAC,WAAW,EAAE,MAAM,GAAG;IACtD,KAAK,EAAE,OAAO,CAAC;IACf,MAAM,EAAE,MAAM,EAAE,CAAC;CAClB,CA6BA"}
+{"version":3,"file":"directory.d.ts","sourceRoot":"","sources":["../src/directory.ts"],"names":[],"mappings":"AAAA;;;;GAIG;AAWH;;;;;;;;;;;;;;;;;;GAkBG;AACH,wBAAgB,gBAAgB,IAAI,MAAM,CAqBzC;AAED;;;;;;GAMG;AACH,eAAO,MAAM,aAAa,QAAqB,CAAC;AAEhD;;;;;;GAMG;AACH,wBAAgB,kBAAkB,CAAC,IAAI,EAAE,MAAM,GAAG,OAAO,CAQxD;AAED;;GAEG;AACH,wBAAgB,eAAe,CAAC,WAAW,EAAE,MAAM,GAAG,MAAM,CAE3D;AAED;;;GAGG;AACH,wBAAgB,aAAa,CAAC,WAAW,EAAE,MAAM,GAAG,OAAO,CAQ1D;AAED;;;;;;;;GAQG;AACH;;;;;;;;;;;;;GAaG;AACH,wBAAgB,qBAAqB,CAAC,WAAW,EAAE,MAAM,GAAG,MAAM,GAAG,IAAI,CA8BxE;AAED,wBAAgB,wBAAwB,CAAC,SAAS,EAAE,MAAM,GAAG,MAAM,GAAG,IAAI,CAmBzE;AA2BD;;;;;;;GAOG;AACH,wBAAgB,0BAA0B,CACxC,IAAI,EAAE,MAAM,EACZ,IAAI,GAAE;IAAE,QAAQ,CAAC,EAAE,MAAM,CAAC;IAAC,GAAG,CAAC,EAAE,MAAM,CAAA;CAAO,GAC7C,MAAM,EAAE,CAmBV;AAsND;;;;;;;;;GASG;AACH,wBAAgB,oBAAoB,CAAC,MAAM,EAAE,MAAM,GAAG,OAAO,CAK5D;AAED;;;;;;;;;;;;;GAaG;AACH,wBAAgB,iBAAiB,CAAC,MAAM,EAAE,MAAM,GAAG,MAAM,EAAE,CAkB1D;AAED;;;;;;;;GAQG;AACH,wBAAgB,kBAAkB,CAAC,MAAM,EAAE,MAAM,GAAG,OAAO,CAE1D;AAED;;GAEG;AACH,MAAM,WAAW,aAAa;IAC5B;;yEAEqE;IACrE,WAAW,EAAE,MAAM,GAAG,IAAI,CAAC;IAC3B;;6EAEyE;IACzE,aAAa,EAAE,MAAM,EAAE,CAAC;IACxB;;+DAE2D;IAC3D,UAAU,EAAE,OAAO,CAAC;CACrB;AAED;;;;;;;;;;;;GAYG;AACH,wBAAgB,aAAa,CAAC,GAAG,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,GAAG,aAAa,CAwCxE;AAyDD;;;GAGG;AACH,wBAAgB,eAAe,CAAC,WAAW,EAAE,MAAM,GAAG,IAAI,CAkCzD;AAED;;GAEG;AACH,wBAAgB,eAAe,CAAC,WAAW,EAAE,MAAM,GAAG,IAAI,CAuBzD;AAED;;GAEG;AACH,wBAAgB,qBAAqB,CAAC,WAAW,EAAE,MAAM,GAAG,MAAM,EAAE,CA8BnE;AAED;;GAEG;AACH,wBAAgB,gBAAgB,CAAC,WAAW,EAAE,MAAM,GAAG,MAAM,CA+B5D;AAED;;GAEG;AACH,wBAAgB,kBAAkB,CAAC,WAAW,EAAE,MAAM,EAAE,UAAU,EAAE,MAAM,GAAG,MAAM,CAYlF;AAED;;GAEG;AACH,wBAAgB,iBAAiB,CAAC,WAAW,EAAE,MAAM,GAAG;IACtD,KAAK,EAAE,OAAO,CAAC;IACf,MAAM,EAAE,MAAM,EAAE,CAAC;CAClB,CA6BA"}