@jigyasudham/veto 2.9.0 → 3.0.0

package/dist/transcripts/toc.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"toc.js","sourceRoot":"","sources":["../../src/transcripts/toc.ts"],"names":[],"mappings":"AAAA,8EAA8E;AAC9E,EAAE;AACF,8EAA8E;AAC9E,6EAA6E;AAC7E,gFAAgF;AAChF,8EAA8E;AAC9E,6EAA6E;AAC7E,+DAA+D;AAE/D,OAAO,EAAE,gBAAgB,EAAE,MAAM,YAAY,CAAC;AAC9C,OAAO,EAAE,UAAU,EAAE,OAAO,EAAE,QAAQ,EAAE,MAAM,cAAc,CAAC;AAE7D,MAAM,SAAS,GAAG,EAAE,CAAC;AAmBrB,SAAS,OAAO,CAAC,IAAY;IAC3B,MAAM,CAAC,GAAG,IAAI,CAAC,OAAO,CAAC,MAAM,EAAE,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC;IAC3C,OAAO,CAAC,CAAC,MAAM,GAAG,SAAS,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,EAAE,SAAS,GAAG,CAAC,CAAC,GAAG,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;AACpE,CAAC;AAED,MAAM,UAAU,QAAQ,CAAC,SAAiB;IACxC,MAAM,EAAE,GAAG,gBAAgB,EAAE,CAAC;IAC9B,MAAM,IAAI,GAAG,EAAE,CAAC,OAAO,CACrB,yFAAyF,CAC1F,CAAC,GAAG,CAAC,SAAS,CAAS,CAAC;IACzB,IAAI,IAAI,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,EAAE,CAAC;IAEjC,MAAM,QAAQ,GAAc,EAAE,CAAC;IAC/B,IAAI,GAAG,GAAoE,IAAI,CAAC;IAEhF,MAAM,IAAI,GAAG,CAAC,KAAS,EAAE,KAAa,EAAE,EAAE;QACxC,GAAG,GAAG;YACJ,KAAK,EAAE,QAAQ,CAAC,MAAM,EAAE,OAAO,EAAE,KAAK,CAAC,GAAG,EAAE,KAAK,EAAE,KAAK,CAAC,GAAG;YAC5D,OAAO,EAAE,KAAK,CAAC,MAAM,EAAE,MAAM,EAAE,KAAK,CAAC,MAAM,EAAE,KAAK;YAClD,YAAY,EAAE,CAAC,EAAE,iBAAiB,EAAE,CAAC,EAAE,SAAS,EAAE,CAAC,EAAE,KAAK,EAAE,EAAE,EAAE,KAAK,EAAE,EAAE,EAAE,QAAQ,EAAE,KAAK;YAC1F,MAAM,EAAE,IAAI,GAAG,EAAU,EAAE,MAAM,EAAE,IAAI,GAAG,EAAU;SACrD,CAAC;QACF,QAAQ,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;IACrB,CAAC,CAAC;IACF,MAAM,KAAK,GAAG,GAAG,EAAE;QACjB,IAAI,GAAG,EAAE,CAAC;YAAC,GAAG,CAAC,KAAK,GAAG,CAAC,GAAG,GAAG,CAAC,MAAM,CAAC,CAAC;YAAC,GAAG,CAAC,KAAK,GAAG,CAAC,GAAG,GAAG,CAAC,MAAM,CAAC,CAAC;QAAC,CAAC;IACxE,CAAC,CAAC;IAEF,KAAK,MAAM,CAAC,IAAI,IAAI,EAAE,CAAC;QACrB,IAAI,CAAC,CAAC,IAAI,KAAK,cAAc,EAAE,CAAC;YAC9B,KAAK,EAAE,CAAC;YACR,IAAI,CAAC,CAAC,EAAE,OAAO,CAAC,CAAC,CAAC,IAAI,IAAI,EAAE,CAAC,IAAI,WAAW,QAAQ,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC,CAAC;QACrE,CAAC;aAAM,IAAI,CAAC,GAAG,EAAE,CAAC;YAChB,kDAAkD;YAClD,IAAI,CAAC,CAAC,EAAE,iBAAiB,CAAC,CAAC;QAC7B,CAAC;QACD,MAAM,CAAC,GAAG,GAAI,CAAC;QACf,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC,GAAG,CAAC;QAChB,IAAI,CAAC,CAAC,MAAM,EAAE,CAAC;YAAC,IAAI,CAAC,CAAC,CAAC,OAAO,IAAI,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC,OAAO;gBAAE,CAAC,CAAC,OAAO,GAAG,CAAC,CAAC,MAAM,CAAC;YAAC,IAAI,CAAC,CAAC,CAAC,MAAM,IAAI,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC,MAAM;gBAAE,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC,MAAM,CAAC;QAAC,CAAC;QAC1I,QAAQ,CAAC,CAAC,IAAI,EAAE,CAAC;YACf,KAAK,cAAc;gBAAE,CAAC,CAAC,YAAY,EAAE,CAAC;gBAAC,MAAM;YAC7C,KAAK,mBAAmB;gBAAE,CAAC,CAAC,iBAAiB,EAAE,CAAC;gBAAC,MAAM;YACvD,KAAK,WAAW,CAAC,CAAC,CAAC;gBACjB,CAAC,CAAC,SAAS,EAAE,CAAC;gBACd,MAAM,IAAI,GAAG,CAAC,CAAC,SAAS,IAAI,MAAM,CAAC;gBACnC,CAAC,CAAC,MAAM,CAAC,GAAG,CAAC,IAAI,CAAC,CAAC;gBACnB,IAAI,CAAC,CAAC,IAAI,IAAI,UAAU,CAAC,GAAG,CAAC,IAAI,CAAC,EAAE,CAAC;oBAAC,MAAM,CAAC,GAAG,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC;oBAAC,IAAI,CAAC;wBAAE,CAAC,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;gBAAC,CAAC;gBACnG,MAAM;YACR,CAAC;YACD,KAAK,aAAa;gBAAE,IAAI,CAAC,CAAC,IAAI,IAAI,QAAQ,CAAC,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC;oBAAE,CAAC,CAAC,QAAQ,GAAG,IAAI,CAAC;gBAAC,MAAM;YAClF,OAAO,CAAC,CAAC,MAAM;QACjB,CAAC;IACH,CAAC;IACD,KAAK,EAAE,CAAC;IACR,OAAO,QAAQ,CAAC;AAClB,CAAC;AAED,mFAAmF;AACnF,MAAM,UAAU,SAAS,CAAC,QAAmB;IAC3C,OAAO,QAAQ,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE;QACtB,MAAM,IAAI,GAAG,CAAC,IAAI,CAAC,CAAC,KAAK,KAAK,CAAC,CAAC,OAAO,IAAI,CAAC,CAAC,KAAK,GAAG,CAAC,CAAC;QACvD,IAAI,CAAC,CAAC,QAAQ;YAAE,IAAI,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;QAClC,IAAI,CAAC,CAAC,KAAK,CAAC,MAAM;YAAE,IAAI,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;QACxE,IAAI,CAAC,CAAC,KAAK,CAAC,MAAM;YAAE,IAAI,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,CAAC;QACxE,OAAO,GAAG,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,KAAK,CAAC,CAAC,KAAK,EAAE,CAAC;IACzC,CAAC,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;AAChB,CAAC"}

package/dist/transcripts/tokenize.d.ts

@@ -0,0 +1,14 @@
+/** Sub-token expansion on by default; the ingest benchmark measures both ways. */
+export declare const SUBTOKENS_DEFAULT = true;
+/**
+ * Tokenize text for indexing or querying — always the same function.
+ *
+ * 1. Split on anything that is not a letter, digit, or `_ - . / @ +`.
+ * 2. Emit each token lowercased.
+ * 3. When it looks compound, also emit its lowercased sub-tokens:
+ *    `normalizeProjectDir` → normalizeprojectdir, normalize, project, dir;
+ *    `src/transcripts/search.ts` → the full path, src, transcripts, search, ts.
+ * 4. Drop blobs (see isBlob) and tokens with no letter or digit at all.
+ */
+export declare function tokenize(text: string, subtokens?: boolean): string[];
+//# sourceMappingURL=tokenize.d.ts.map

package/dist/transcripts/tokenize.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"tokenize.d.ts","sourceRoot":"","sources":["../../src/transcripts/tokenize.ts"],"names":[],"mappings":"AAgCA,kFAAkF;AAClF,eAAO,MAAM,iBAAiB,OAAO,CAAC;AAEtC;;;;;;;;;GASG;AACH,wBAAgB,QAAQ,CAAC,IAAI,EAAE,MAAM,EAAE,SAAS,GAAE,OAA2B,GAAG,MAAM,EAAE,CAsBvF"}

package/dist/transcripts/tokenize.js

@@ -0,0 +1,70 @@
+// Shared tokenizer — the single source of truth for the portable index.
+//
+// The one correctness property the whole search layer hangs on: index-time and
+// query-time tokenization are the SAME function. The FTS5 design broke this
+// subtly (toFtsQuery kept `foo/bar.ts` whole, FTS5's unicode61 re-split it
+// inside the quotes), which is why this file exists instead of two ad-hoc
+// paths. Property-tested in tests/transcripts/tokenize.test.ts: every token
+// emitted at index time finds its document when fed back as a query.
+//
+// No stemming — this corpus is identifiers, paths, and error strings, where
+// porter folding costs more than it buys (`E404` must survive intact).
+// Sub-token expansion (step 3 below) buys the recall stemming was buying.
+// Keep letters, digits, and the identifier/path glue chars; split on the rest.
+const SPLIT = /[^\p{L}\p{N}_\-./@+]+/u;
+// Sub-token boundaries inside a kept token: the glue chars…
+const SEP = /[_\-./@+]+/;
+// …and camelCase humps, detected before lowercasing.
+const CAMEL = /(?<=[\p{Ll}\p{N}])(?=\p{Lu})|(?<=\p{Lu})(?=\p{Lu}\p{Ll})/u;
+// Masked-secret residue and inline blobs: long pure-hex, or long base64-ish
+// runs that actually look like data (digits + mixed case or padding) rather
+// than a long camelCase identifier. They bloat postings and are never searched.
+const HEXY = /^[0-9a-fA-F]{32,}$/;
+const B64_CHARS = /^[A-Za-z0-9+/=]{32,}$/;
+function isBlob(tok) {
+    if (tok.length > 64)
+        return true;
+    if (HEXY.test(tok))
+        return true;
+    return B64_CHARS.test(tok) && /\d/.test(tok) && (/[+/=]/.test(tok) || (/[a-z]/.test(tok) && /[A-Z]/.test(tok)));
+}
+/** Sub-token expansion on by default; the ingest benchmark measures both ways. */
+export const SUBTOKENS_DEFAULT = true;
+/**
+ * Tokenize text for indexing or querying — always the same function.
+ *
+ * 1. Split on anything that is not a letter, digit, or `_ - . / @ +`.
+ * 2. Emit each token lowercased.
+ * 3. When it looks compound, also emit its lowercased sub-tokens:
+ *    `normalizeProjectDir` → normalizeprojectdir, normalize, project, dir;
+ *    `src/transcripts/search.ts` → the full path, src, transcripts, search, ts.
+ * 4. Drop blobs (see isBlob) and tokens with no letter or digit at all.
+ */
+export function tokenize(text, subtokens = SUBTOKENS_DEFAULT) {
+    const out = [];
+    for (const raw of text.split(SPLIT)) {
+        if (!raw || !/[\p{L}\p{N}]/u.test(raw) || isBlob(raw))
+            continue;
+        const whole = raw.toLowerCase();
+        out.push(whole);
+        if (!subtokens)
+            continue;
+        // Two granularities, both emitted: separator segments AND their camel
+        // parts. Needed for symmetry — a lowercased emitted token re-splits by
+        // separators only (its camel humps are gone), so those segment-level
+        // tokens must exist in the index too.
+        for (const seg of raw.split(SEP)) {
+            const s = seg.toLowerCase();
+            const ok = (t) => t && /[\p{L}\p{N}]/u.test(t) && !isBlob(t);
+            if (s !== whole && ok(s))
+                out.push(s);
+            for (const part of seg.split(CAMEL)) {
+                const p = part.toLowerCase();
+                if (p !== s && p !== whole && ok(p))
+                    out.push(p);
+            }
+        }
+    }
+    return out;
+}
+//# sourceMappingURL=tokenize.js.map

package/dist/transcripts/tokenize.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"tokenize.js","sourceRoot":"","sources":["../../src/transcripts/tokenize.ts"],"names":[],"mappings":"AAAA,wEAAwE;AACxE,EAAE;AACF,+EAA+E;AAC/E,4EAA4E;AAC5E,2EAA2E;AAC3E,0EAA0E;AAC1E,4EAA4E;AAC5E,qEAAqE;AACrE,EAAE;AACF,4EAA4E;AAC5E,uEAAuE;AACvE,0EAA0E;AAE1E,+EAA+E;AAC/E,MAAM,KAAK,GAAG,wBAAwB,CAAC;AACvC,4DAA4D;AAC5D,MAAM,GAAG,GAAG,YAAY,CAAC;AACzB,qDAAqD;AACrD,MAAM,KAAK,GAAG,2DAA2D,CAAC;AAE1E,4EAA4E;AAC5E,4EAA4E;AAC5E,gFAAgF;AAChF,MAAM,IAAI,GAAG,oBAAoB,CAAC;AAClC,MAAM,SAAS,GAAG,uBAAuB,CAAC;AAE1C,SAAS,MAAM,CAAC,GAAW;IACzB,IAAI,GAAG,CAAC,MAAM,GAAG,EAAE;QAAE,OAAO,IAAI,CAAC;IACjC,IAAI,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC;QAAE,OAAO,IAAI,CAAC;IAChC,OAAO,SAAS,CAAC,IAAI,CAAC,GAAG,CAAC,IAAI,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,OAAO,CAAC,IAAI,CAAC,GAAG,CAAC,IAAI,CAAC,OAAO,CAAC,IAAI,CAAC,GAAG,CAAC,IAAI,OAAO,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC;AAClH,CAAC;AAED,kFAAkF;AAClF,MAAM,CAAC,MAAM,iBAAiB,GAAG,IAAI,CAAC;AAEtC;;;;;;;;;GASG;AACH,MAAM,UAAU,QAAQ,CAAC,IAAY,EAAE,YAAqB,iBAAiB;IAC3E,MAAM,GAAG,GAAa,EAAE,CAAC;IACzB,KAAK,MAAM,GAAG,IAAI,IAAI,CAAC,KAAK,CAAC,KAAK,CAAC,EAAE,CAAC;QACpC,IAAI,CAAC,GAAG,IAAI,CAAC,eAAe,CAAC,IAAI,CAAC,GAAG,CAAC,IAAI,MAAM,CAAC,GAAG,CAAC;YAAE,SAAS;QAChE,MAAM,KAAK,GAAG,GAAG,CAAC,WAAW,EAAE,CAAC;QAChC,GAAG,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QAChB,IAAI,CAAC,SAAS;YAAE,SAAS;QACzB,sEAAsE;QACtE,uEAAuE;QACvE,qEAAqE;QACrE,sCAAsC;QACtC,KAAK,MAAM,GAAG,IAAI,GAAG,CAAC,KAAK,CAAC,GAAG,CAAC,EAAE,CAAC;YACjC,MAAM,CAAC,GAAG,GAAG,CAAC,WAAW,EAAE,CAAC;YAC5B,MAAM,EAAE,GAAG,CAAC,CAAS,EAAE,EAAE,CAAC,CAAC,IAAI,eAAe,CAAC,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC;YACrE,IAAI,CAAC,KAAK,KAAK,IAAI,EAAE,CAAC,CAAC,CAAC;gBAAE,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;YACtC,KAAK,MAAM,IAAI,IAAI,GAAG,CAAC,KAAK,CAAC,KAAK,CAAC,EAAE,CAAC;gBACpC,MAAM,CAAC,GAAG,IAAI,CAAC,WAAW,EAAE,CAAC;gBAC7B,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,KAAK,KAAK,IAAI,EAAE,CAAC,CAAC,CAAC;oBAAE,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;YACnD,CAAC;QACH,CAAC;IACH,CAAC;IACD,OAAO,GAAG,CAAC;AACb,CAAC"}

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@jigyasudham/veto",
   "mcpName": "io.github.jigyasudham/veto",
-  "version": "2.9.0",
+  "version": "3.0.0",
   "description": "93 agentic tools. 49 specialists. Every major AI CLI. Self-learning. Zero extra cost on subscriptions.",
   "keywords": [
     "mcp",
@@ -26,7 +26,7 @@
   "license": "MIT",
   "type": "module",
   "engines": {
-    "node": ">=22.5.0"
+    "node": ">=22.13.0 <23.0.0 || >=23.4.0"
   },
   "bin": {
     "veto": "dist/cli.js",

package/server.json

@@ -8,13 +8,13 @@
     "source": "github"
   },
   "websiteUrl": "https://github.com/jigyasudham/veto#readme",
-  "version": "2.9.0",
+  "version": "3.0.0",
   "packages": [
     {
       "registryType": "npm",
       "registryBaseUrl": "https://registry.npmjs.org",
       "identifier": "@jigyasudham/veto",
-      "version": "2.9.0",
+      "version": "3.0.0",
       "transport": {
         "type": "stdio"
       }