sigmap 8.32.0 → 8.32.1

package/CHANGELOG.md

@@ -10,6 +10,18 @@ Format: [Semantic Versioning](https://semver.org/)
 
 ---
 
+## [8.32.1] — 2026-09-13
+
+
+### Fixed
+- **Retraction of a v8.32.0 claim.** That release stated "every extractor now discloses what a ceiling dropped". It did not. `vue.js` was registered but unreachable (`.vue` dispatches to `vue_sfc`), so disclosure was added to dead code while the live `.vue` handler kept truncating silently; `.tsx` (every React component), `.properties`, `.toml` and `.md` still cut output with a bare `slice()`; and `r.js` called `capWithNotice` but eight inner caps stopped collection at the ceiling so it never fired — and when forced to, reported `+1 more` where 50 signatures were hidden. All fixed, ceilings unchanged, with a test that fails against the pre-fix extractors. The root cause was that three of those languages have no test fixture (#582, #583, #584, #588)
+- The walk depth hid most code in JVM package layouts: `maxDepth: 6` indexed **6 of 47** Java files on spring-petclinic, because Java puts one directory per package segment. #561 had already raised the *graph* walk to 12, so extraction was the shallower half of an inconsistent pair. Depth now rises to 12 only for JVM layouts; an explicit `maxDepth` always wins. Gated JVM corpus 16.4% → 23.0% (#590)
+
+### Changed
+- Published retrieval hit@5 moves 81.1% → 78.9%. One of 18 repos accounts for it: spring-petclinic 100% → 60%, previously measured against an index holding 6 of 47 Java files. The leak-free `mined` corpus is flat and the leak-free `jvm` corpus improves, so the prior figure was inflated by under-indexing rather than this being a ranking regression. The two affected tasks are tracked as a ranking weakness the missing files were concealing (#592)
+
+---
+
 ## [8.32.0] — 2026-09-12
 
 ### Added
@@ -18,6 +30,7 @@ Format: [Semantic Versioning](https://semver.org/)
 
 ### Fixed
 - Every extractor now **discloses what a ceiling dropped** instead of truncating silently. 20 extractors gained `… +N more` markers on their member and per-file caps (23 now carry them in total). The ceilings themselves are unchanged — raising them is a separate, measured decision (#578, #576)
+  > **Correction (2026-09-13):** "every extractor" was wrong as published. `vue.js` was dead code — `.vue` dispatches to `vue_sfc` — so one of the 20 was unreachable; four reachable extractors (`.tsx`, `.properties`, `.toml`, `.md`) still truncated silently; and `r.js`'s disclosure was defeated by eight inner caps. Corrected in #589 (#582, #583, #584).
 - The retrieval gate reused a gitignored index, so a "regression" could be pure staleness — this produced three separate false investigations, including one re-baseline. It now regenerates every index it scores, including one per JVM repo driven from `benchmarks/config-overrides.json` rather than a hand-written config (#579)
 - The `sigmap lines` CLI example nested a fenced block inside a text fence; the inner fence closed early and a brace expression became a Vue interpolation. This failed the Pages build *after* v8.31.0 was tagged and published (#574)
 

package/README.md

@@ -59,12 +59,12 @@ That map is exactly what agentic grep is worst at: reproducible, auditable conte
 
 **Proof it pays off** (full benchmark below):
 <!--SM:whyMetrics-->
-- **81.1% hit@5** — right file in top 5 results (vs 44.0% single-shot grep baseline — 1.73× lift)
+- **78.9% hit@5** — right file in top 5 results (vs 44.0% single-shot grep baseline — 1.73× lift)
 - **96.8% token reduction** — average across 21 real repos
-- **64.8% task-success proxy** — modeled from retrieval tiers, not measured LLM sessions
-- **1.54 prompts per task** — down from 2.84 (45.7% fewer retries, modeled)
+- **62.9% task-success proxy** — modeled from retrieval tiers, not measured LLM sessions
+- **1.58 prompts per task** — down from 2.84 (44.4% fewer retries, modeled)
 <!--/SM:whyMetrics-->
-- **<!--SM:languages-->33<!--/SM:languages--> languages supported** — TypeScript, Python, Go, Rust, Java, R, and more
+- **<!--SM:languages-->32<!--/SM:languages--> languages supported** — TypeScript, Python, Go, Rust, Java, R, and more
 - **No vendor lock-in** — works with any AI assistant or local LLM
 - **No API costs** — use local models (Ollama, llama.cpp, vLLM) with zero token fees
 - **Full privacy** — keep your code and context on your machine
@@ -98,7 +98,7 @@ sigmap verify answer.md --report        # standalone red/amber/green HTML report
 | Without SigMap | With SigMap |
 |---|---|
 | ❌ Non-reproducible agent guesses | ✅ Deterministic map — same input, same output, every time |
-| ❌ "Trust me" AI answers | ✅ Grounded — right file in context <!--SM:hitWhole-->81%<!--/SM:hitWhole--> of the time, every symbol on a real line anchor |
+| ❌ "Trust me" AI answers | ✅ Grounded — right file in context <!--SM:hitWhole-->79%<!--/SM:hitWhole--> of the time, every symbol on a real line anchor |
 | ❌ Embeddings / vector DB required | ✅ Zero deps, no infra, fully offline |
 
 ---
@@ -125,10 +125,10 @@ Ask → Rank → Context → Validate → Judge → Learn
 Benchmark : sigmap-v8.32-main (21 repositories, including R language)
 Date      : 2026-09-12
 
-Hit@5          : 81.1%   (grep-agent baseline 44.0%  — 1.73× lift)
+Hit@5          : 78.9%   (grep-agent baseline 44.0%  — 1.73× lift)
 Token reduction: 96.8%   (across 21 repos)
-Prompt reduction : 45.7% (2.84 → 1.54 prompts per task, modeled)
-Task success   : 64.8%   (proxy — modeled from retrieval tiers)
+Prompt reduction : 44.4% (2.84 → 1.58 prompts per task, modeled)
+Task success   : 62.9%   (proxy — modeled from retrieval tiers)
 Repos tested   : 21 (JavaScript, Python, Go, Rust, Java, R, C++, C#, Dart, Swift, Ruby, PHP, Scala, Kotlin, and more)
 ```
 <!--/SM:benchmarkBlock-->
@@ -332,7 +332,7 @@ sigmap --health
 | Benchmark methodology | [benchmark.html](https://sigmap.io/guide/benchmark.html) |
 | Config reference | [config.html](https://sigmap.io/guide/config.html) |
 | Roadmap | [roadmap.html](https://sigmap.io/guide/roadmap.html) |
-| <!--SM:languages-->33<!--/SM:languages--> languages | [generalization.html](https://sigmap.io/guide/generalization.html) |
+| <!--SM:languages-->32<!--/SM:languages--> languages | [generalization.html](https://sigmap.io/guide/generalization.html) |
 
 ---
 
@@ -389,7 +389,7 @@ See [.github/PULL_REQUEST_TEMPLATE.md](.github/PULL_REQUEST_TEMPLATE.md) for the
 
 ---
 
-## <!--SM:languages-->33<!--/SM:languages--> languages
+## <!--SM:languages-->32<!--/SM:languages--> languages
 
 TypeScript · JavaScript · Python · Java · Kotlin · Go · Rust · C# · C/C++ · Ruby · PHP · Swift · Dart · Scala · Vue · Svelte · HTML · CSS/SCSS · YAML · Shell · SQL · GraphQL · Terraform · Protobuf · Dockerfile · TOML · XML · Properties · Markdown · R · GDScript
 

package/gen-context.js

@@ -1790,6 +1790,55 @@ __factories["./src/config/loader"] = function(module, exports) {
     });
   }
 
+  /**
+   * Directory depth needed to reach source under a JVM package layout.
+   *
+   * `maxDepth: 6` is right for the JS/Python-shaped trees it was tuned on, but a
+   * JVM project puts one directory per package segment — real code in
+   * `src/main/java/com/company/project/module/Class.java` sits 8-10 levels down.
+   * At depth 6 only the top-level package was indexed: on spring-petclinic, 6 of
+   * 47 Java files (#590).
+   *
+   * #561 already raised the dependency-graph walk to 12 for exactly this reason,
+   * so extraction was the shallower half of an inconsistent pair.
+   *
+   * Deepening globally is not free — it adds candidates to every repo and cost
+   * 2.2pp on the (filename-leaky) 105-task matrix corpus while the unbiased
+   * `mined` corpus stayed flat. So the depth is raised only where the layout
+   * demands it, leaving non-JVM repos byte-identical.
+   *
+   * @param {string} cwd
+   * @returns {boolean} true when the repo looks like a Maven/Gradle/sbt project
+   */
+  function _isJvmLayout(cwd) {
+    const MARKERS = ['pom.xml', 'build.gradle', 'build.gradle.kts', 'build.sbt', 'settings.gradle', 'settings.gradle.kts'];
+    for (const m of MARKERS) {
+      try { if (fs.existsSync(path.join(cwd, m))) return true; } catch { /* unreadable cwd */ }
+    }
+    for (const d of ['src/main/java', 'src/main/kotlin', 'src/main/scala']) {
+      try { if (fs.existsSync(path.join(cwd, d))) return true; } catch { /* ignore */ }
+    }
+    return false;
+  }
+
+  /** Walk depth for a JVM package layout — matches the graph walk from #561. */
+  const JVM_MAX_DEPTH = 12;
+
+  /**
+   * Raise `maxDepth` to the JVM depth when the layout needs it (#590).
+   * An explicit user value always wins, including a deliberately shallow one.
+   * @param {object} cfg   resolved config (mutated and returned)
+   * @param {string} cwd
+   * @param {boolean} userSetDepth
+   * @returns {object} cfg
+   */
+  function _applyJvmDepth(cfg, cwd, userSetDepth) {
+    if (!userSetDepth && cfg.maxDepth < JVM_MAX_DEPTH && _isJvmLayout(cwd)) {
+      cfg.maxDepth = JVM_MAX_DEPTH;
+    }
+    return cfg;
+  }
+
   /**
    * Load and merge configuration for a given working directory.
    *
@@ -1802,7 +1851,7 @@ __factories["./src/config/loader"] = function(module, exports) {
       const cfg = deepClone(DEFAULTS);
       const detected = detectAutoSrcDirs(cwd, cfg.exclude);
       if (detected.length > 0) cfg.srcDirs = detected;
-      return cfg;
+      return _applyJvmDepth(cfg, cwd, false);
     }
 
     let userConfig;
@@ -1814,7 +1863,7 @@ __factories["./src/config/loader"] = function(module, exports) {
       const cfg = deepClone(DEFAULTS);
       const detected = detectAutoSrcDirs(cwd, cfg.exclude);
       if (detected.length > 0) cfg.srcDirs = detected;
-      return cfg;
+      return _applyJvmDepth(cfg, cwd, false);
     }
 
     // Warn on unknown keys (helps catch typos)
@@ -1866,7 +1915,8 @@ __factories["./src/config/loader"] = function(module, exports) {
     } else if (Array.isArray(merged.adapters) && !userConfig.outputs) {
       merged.outputs = merged.adapters.filter((a) => ['copilot','claude','cursor','windsurf'].includes(a));
     }
-    return merged;
+
+    return _applyJvmDepth(merged, cwd, userConfig.maxDepth !== undefined);
   }
 
   function deepClone(obj) {
@@ -4118,7 +4168,6 @@ __factories["./src/eval/analyzer"] = function(module, exports) {
     '.scala': 'scala',   '.sc': 'scala',
     '.gd': 'gdscript',
     '.r': 'r',           '.R': 'r',
-    '.vue': 'vue',
     '.svelte': 'svelte',
     '.html': 'html',     '.htm': 'html',
     '.css': 'css',       '.scss': 'css', '.sass': 'css', '.less': 'css',
@@ -5922,7 +5971,6 @@ __factories["./src/extractors/dispatch"] = function(module, exports) {
     scala: __require('./src/extractors/scala'),
     gdscript: __require('./src/extractors/gdscript'),
     r: __require('./src/extractors/r'),
-    vue: __require('./src/extractors/vue'),
     vue_sfc: __require('./src/extractors/vue_sfc'),
     svelte: __require('./src/extractors/svelte'),
     html: __require('./src/extractors/html'),
@@ -6986,6 +7034,11 @@ __factories["./src/extractors/line-anchor"] = function(module, exports) {
 // ── ./src/extractors/markdown ──
 __factories["./src/extractors/markdown"] = function(module, exports) {
   
+  const { capWithNotice } = __require('./src/util/truncate');
+
+  // Ceiling discloses what it drops rather than truncating silently (#583).
+  const PER_FILE_LIMIT = 40;
+
   /**
    * Lightweight markdown technical indexer.
    * Captures headings and fenced code block language hints only.
@@ -7010,7 +7063,7 @@ __factories["./src/extractors/markdown"] = function(module, exports) {
       sigs.push(`code-fence ${lang}`);
     }
 
-    return Array.from(new Set(sigs)).slice(0, 40);
+    return capWithNotice(Array.from(new Set(sigs)), PER_FILE_LIMIT, 'headings');
   }
 
   module.exports = { extract };
@@ -7341,6 +7394,11 @@ __factories["./src/extractors/prdiff"] = function(module, exports) {
 // ── ./src/extractors/properties ──
 __factories["./src/extractors/properties"] = function(module, exports) {
   
+  const { capWithNotice } = __require('./src/util/truncate');
+
+  // Ceiling discloses what it drops rather than truncating silently (#583).
+  const PER_FILE_LIMIT = 50;
+
   /**
    * Extract signatures from .properties configuration files.
    * Captures key names, grouped by prefixes where possible.
@@ -7372,7 +7430,7 @@ __factories["./src/extractors/properties"] = function(module, exports) {
       sigs.push(`key ${key}`);
     }
 
-    return Array.from(new Set(sigs)).slice(0, 50);
+    return capWithNotice(Array.from(new Set(sigs)), PER_FILE_LIMIT, 'keys');
   }
 
   module.exports = { extract };
@@ -7845,7 +7903,7 @@ __factories["./src/extractors/r"] = function(module, exports) {
     //   ClassName <- R6::R6Class(...)
     const r6Re = /([\w.]+)\s*(?:<<-|<-|=)\s*(?:R6::)?R6Class\s*\(/g;
     let m;
-    while ((m = r6Re.exec(stripped)) !== null && sigs.length < 30) {
+    while ((m = r6Re.exec(stripped)) !== null) {
       const name = m[1];
       if (name.startsWith('.')) continue;
       const openIdx = r6Re.lastIndex - 1;
@@ -7856,7 +7914,6 @@ __factories["./src/extractors/r"] = function(module, exports) {
       sigs.push(`${name} <- R6Class("${classNameLit}")` + applyHint(docHints, name));
       for (const memberSig of extractListMethods(body, 8)) {
         sigs.push('  ' + memberSig);
-        if (sigs.length >= 30) break;
       }
       consumedRanges.push([m.index, closeIdx]);
       r6Re.lastIndex = closeIdx;
@@ -7866,7 +7923,7 @@ __factories["./src/extractors/r"] = function(module, exports) {
     //   ClassName <- new_class("ClassName", properties = list(...))
     const s7Classes = new Set();
     const s7Re = /([\w.]+)\s*(?:<<-|<-|=)\s*(?:S7::)?new_class\s*\(/g;
-    while ((m = s7Re.exec(stripped)) !== null && sigs.length < 30) {
+    while ((m = s7Re.exec(stripped)) !== null) {
       const name = m[1];
       if (name.startsWith('.')) continue;
       const openIdx = s7Re.lastIndex - 1;
@@ -7883,7 +7940,7 @@ __factories["./src/extractors/r"] = function(module, exports) {
 
     // S7 method dispatch: `method(generic, ClassName) <- function(args)`
     const s7MethodRe = /^[ \t]*method\s*\(\s*([\w.]+)\s*,\s*([\w.]+)\s*\)\s*(?:<<-|<-|=)\s*function\s*\(/gm;
-    while ((m = s7MethodRe.exec(stripped)) !== null && sigs.length < 30) {
+    while ((m = s7MethodRe.exec(stripped)) !== null) {
       if (!s7Classes.has(m[2])) continue;
       const argsStart = s7MethodRe.lastIndex - 1;
       const args = readBalancedParens(stripped, argsStart);
@@ -7896,7 +7953,7 @@ __factories["./src/extractors/r"] = function(module, exports) {
     // Skip matches whose position falls inside an R6/S7 class body — those have
     // already been emitted as indented members.
     const funcRe = /^(?:[ \t]*)([\w.]+)\s*(?:<<-|<-|=)\s*function\s*\(/gm;
-    while ((m = funcRe.exec(stripped)) !== null && sigs.length < 30) {
+    while ((m = funcRe.exec(stripped)) !== null) {
       const name = m[1];
       if (name.startsWith('.')) continue;
       if (inAnyRange(m.index, consumedRanges)) continue;
@@ -7908,15 +7965,12 @@ __factories["./src/extractors/r"] = function(module, exports) {
 
     // ── S4 ────────────────────────────────────────────────────────────────────
     for (const sm of stripped.matchAll(/^[ \t]*setGeneric\s*\(\s*["']([\w.]+)["']/gm)) {
-      if (sigs.length >= 30) break;
       sigs.push(`setGeneric("${sm[1]}")`);
     }
     for (const sm of stripped.matchAll(/^[ \t]*setMethod\s*\(\s*["']([\w.]+)["']\s*,\s*["']([\w.]+)["']/gm)) {
-      if (sigs.length >= 30) break;
       sigs.push(`setMethod("${sm[1]}", "${sm[2]}")`);
     }
     for (const sm of stripped.matchAll(/^[ \t]*setClass\s*\(\s*["']([\w.]+)["']/gm)) {
-      if (sigs.length >= 30) break;
       sigs.push(`setClass("${sm[1]}")`);
     }
 
@@ -8925,6 +8979,11 @@ __factories["./src/extractors/todos"] = function(module, exports) {
 // ── ./src/extractors/toml ──
 __factories["./src/extractors/toml"] = function(module, exports) {
   
+  const { capWithNotice } = __require('./src/util/truncate');
+
+  // Ceiling discloses what it drops rather than truncating silently (#583).
+  const PER_FILE_LIMIT = 40;
+
   /**
    * Extract signatures from TOML configuration files.
    * Focuses on section/table names and high-value keys.
@@ -8961,7 +9020,7 @@ __factories["./src/extractors/toml"] = function(module, exports) {
       }
     }
 
-    return Array.from(new Set(sigs)).slice(0, 40);
+    return capWithNotice(Array.from(new Set(sigs)), PER_FILE_LIMIT, 'entries');
   }
 
   module.exports = { extract };
@@ -9301,6 +9360,11 @@ __factories["./src/extractors/typescript"] = function(module, exports) {
 // ── ./src/extractors/typescript_react ──
 __factories["./src/extractors/typescript_react"] = function(module, exports) {
   
+  const { capWithNotice } = __require('./src/util/truncate');
+
+  // Ceiling discloses what it drops rather than truncating silently (#583).
+  const PER_FILE_LIMIT = 50;
+
   /**
    * Extract React component signatures from .tsx files.
    * Captures component props interfaces, hooks usage, and exports.
@@ -9355,97 +9419,7 @@ __factories["./src/extractors/typescript_react"] = function(module, exports) {
       sigs.push(`handler on${h}`);
     }
 
-    return Array.from(new Set(sigs)).slice(0, 50);
-  }
-
-  module.exports = { extract };
-  
-};
-
-// ── ./src/extractors/vue ──
-__factories["./src/extractors/vue"] = function(module, exports) {
-  
-  const { capWithNotice } = __require('./src/util/truncate');
-
-  // Ceiling sits above the default `maxSigsPerFile` so the configured budget
-  // governs output rather than a literal buried here, and omissions are disclosed (#576).
-  const PER_FILE_LIMIT = 25;
-
-  /**
-   * Extract signatures from Vue single-file components.
-   * @param {string} src - Raw file content
-   * @returns {string[]} Array of signature strings
-   */
-  function extract(src) {
-    if (!src || typeof src !== 'string') return [];
-    const sigs = [];
-
-    // Extract component name from filename hint if present or defineComponent
-    const nameMatch = src.match(/name\s*:\s*['"](\w+)['"]/);
-    if (nameMatch) sigs.push(`component ${nameMatch[1]}`);
-
-    // Extract <script> block
-    const scriptMatch = src.match(/<script(?:\s[^>]*)?>(?:\s*)([\s\S]*?)<\/script>/i);
-    if (!scriptMatch) return sigs;
-
-    const script = scriptMatch[1]
-      .replace(/\/\/.*$/gm, '')
-      .replace(/\/\*[\s\S]*?\*\//g, '');
-
-    // Props
-    const propsMatch = script.match(/props\s*:\s*(\{[\s\S]*?\})/);
-    if (propsMatch) {
-      const propNames = [];
-      for (const m of propsMatch[1].matchAll(/^\s+(\w+)\s*:/gm)) {
-        propNames.push(m[1]);
-      }
-      if (propNames.length > 0) sigs.push(`props: [${propNames.join(', ')}]`);
-    }
-
-    // Methods in options API
-    const methodsMatch = script.match(/methods\s*:\s*\{([\s\S]*?)\},?\s*(?:computed|watch|mounted|created|data|\})/);
-    if (methodsMatch) {
-      for (const m of methodsMatch[1].matchAll(/^\s+(?:async\s+)?(\w+)\s*\(([^)]*)\)(?:\s*:\s*([^{=\n]+))?/gm)) {
-        if (m[1].startsWith('_')) continue;
-        const asyncKw = m[0].includes('async') ? 'async ' : '';
-        const retStr = m[3] ? ` → ${normalizeType(m[3])}` : '';
-        sigs.push(`  ${asyncKw}${m[1]}(${normalizeParams(m[2])})${retStr}`);
-      }
-    }
-
-    // Top-level functions in <script> (e.g., composition API helpers)
-    for (const m of script.matchAll(/^(?:export\s+)?(?:async\s+)?function\s+(\w+)\s*\(([^)]*)\)(?:\s*:\s*([^{=\n]+))?/gm)) {
-      if (m[1].startsWith('_')) continue;
-      const asyncKw = m[0].includes('async') ? 'async ' : '';
-      const retStr = m[3] ? ` → ${normalizeType(m[3])}` : '';
-      sigs.push(`${asyncKw}function ${m[1]}(${normalizeParams(m[2])})${retStr}`);
-    }
-
-    // defineProps (Composition API)
-    const definePropsMatch = script.match(/defineProps(?:<[^>]*>)?\s*\(\s*(\{[\s\S]*?\})\s*\)/);
-    if (definePropsMatch) {
-      const propNames = [];
-      for (const m of definePropsMatch[1].matchAll(/^\s+(\w+)\s*:/gm)) {
-        propNames.push(m[1]);
-      }
-      if (propNames.length > 0) sigs.push(`defineProps: [${propNames.join(', ')}]`);
-    }
-
-    // Emits
-    const emitsMatch = script.match(/(?:defineEmits|emits)\s*(?::\s*|\(\s*)(\[[\s\S]*?\])/);
-    if (emitsMatch) sigs.push(`emits: ${emitsMatch[1].replace(/\s+/g, ' ')}`);
-
-    return capWithNotice(sigs, PER_FILE_LIMIT, 'signatures');
-  }
-
-  function normalizeParams(params) {
-    if (!params) return '';
-    return params.trim().replace(/\s+/g, ' ');
-  }
-
-  function normalizeType(type) {
-    if (!type) return '';
-    return type.trim().replace(/[;\s]+$/g, '').replace(/\s+/g, ' ').slice(0, 25);
+    return capWithNotice(Array.from(new Set(sigs)), PER_FILE_LIMIT, 'signatures');
   }
 
   module.exports = { extract };
@@ -15814,7 +15788,7 @@ __factories["./src/mcp/server"] = function(module, exports) {
 
   const SERVER_INFO = {
     name: 'sigmap',
-    version: '8.32.0',
+    version: '8.32.1',
     description: 'SigMap MCP server — code signatures on demand',
   };
 
@@ -22108,7 +22082,7 @@ function __tryGit(args, opts = {}) {
   catch (_) { return ''; }
 }
 
-const VERSION = '8.32.0';
+const VERSION = '8.32.1';
 const MARKER = '\n\n## Auto-generated signatures\n<!-- Updated by gen-context.js -->\n';
 
 function requireSourceOrBundled(key) {
@@ -26881,7 +26855,7 @@ function main() {
         '.cs': 'csharp', '.cpp': 'cpp', '.rb': 'ruby', '.php': 'php',
         '.swift': 'swift', '.dart': 'dart', '.scala': 'scala',
         '.r': 'r', '.R': 'r',
-        '.vue': 'vue', '.svelte': 'svelte', '.html': 'html',
+        '.vue': 'vue_sfc', '.svelte': 'svelte', '.html': 'html',
         '.css': 'css', '.yml': 'yaml', '.sh': 'shell',
       };
       const SPECIAL = { 'Dockerfile': 'dockerfile' };

package/llms-full.txt

@@ -11,7 +11,7 @@ ranking keeps the relevant context in scope (cutting tokens ~97% as a side
 effect), with no LLM calls, embeddings, or vector database. Works with Claude,
 Cursor, GitHub Copilot, Aider, Windsurf, local LLMs, and MCP.
 
-# Version: 8.32.0 | Benchmark: sigmap-v8.32-main (2026-09-12)
+# Version: 8.32.1 | Benchmark: sigmap-v8.32-main (2026-09-12)
 # Source: auto-generated from package.json, version.json, benchmarks/latest.json, src/mcp/tools.js, src/config/defaults.js
 # Regenerate: npm run generate:llms   |   Validate: npm run validate:llms
 
@@ -21,11 +21,11 @@ Cursor, GitHub Copilot, Aider, Windsurf, local LLMs, and MCP.
 
 | Metric | Without SigMap | With SigMap |
 |--------|----------------|-------------|
-| Retrieval hit@5 | 44.0% (single-shot grep) | 81.1% (1.73× lift) |
+| Retrieval hit@5 | 44.0% (single-shot grep) | 78.9% (1.73× lift) |
 | Token reduction | — | 96.8% average |
-| Task-success proxy (modeled) | — | 64.8% |
-| Prompts per task | 2.84 | 1.54 (45.7% fewer) |
-| Supported languages | — | 33 |
+| Task-success proxy (modeled) | — | 62.9% |
+| Prompts per task | 2.84 | 1.58 (44.4% fewer) |
+| Supported languages | — | 32 |
 | MCP tools | — | 21 |
 | npm runtime dependencies | — | 0 |
 
@@ -370,9 +370,9 @@ impact = {"depth":3,"includeSigs":true}
 
 ---
 
-## Supported languages (33 extractors)
+## Supported languages (32 extractors)
 
-cpp, csharp, css, dart, dockerfile, gdscript, go, graphql, html, java, javascript, kotlin, markdown, php, properties, protobuf, python, r, ruby, rust, scala, shell, sql, svelte, swift, terraform, toml, typescript, typescript_react, vue, vue_sfc, xml, yaml
+cpp, csharp, css, dart, dockerfile, gdscript, go, graphql, html, java, javascript, kotlin, markdown, php, properties, protobuf, python, r, ruby, rust, scala, shell, sql, svelte, swift, terraform, toml, typescript, typescript_react, vue_sfc, xml, yaml
 
 ---
 

package/llms.txt

@@ -11,7 +11,7 @@ ranking keeps the relevant context in scope (cutting tokens ~97% as a side
 effect), with no LLM calls, embeddings, or vector database. Works with Claude,
 Cursor, GitHub Copilot, Aider, Windsurf, local LLMs, and MCP.
 
-# Version: 8.32.0 | Benchmark: sigmap-v8.32-main (2026-09-12)
+# Version: 8.32.1 | Benchmark: sigmap-v8.32-main (2026-09-12)
 # Source: auto-generated from package.json, version.json, benchmarks/latest.json, src/mcp/tools.js, src/config/defaults.js
 # Regenerate: npm run generate:llms   |   Validate: npm run validate:llms
 
@@ -25,11 +25,11 @@ Cursor, GitHub Copilot, Aider, Windsurf, local LLMs, and MCP.
 
 ## Core metrics (benchmark: sigmap-v8.32-main, 2026-09-12)
 
-- hit@5 retrieval: 81.1% vs 44.0% single-shot grep baseline (1.73× lift)
+- hit@5 retrieval: 78.9% vs 44.0% single-shot grep baseline (1.73× lift)
 - Token reduction: 96.8% average across benchmark repos
-- Task-success proxy: 64.8% (modeled from retrieval tiers, not measured LLM sessions)
-- Prompts per task: 1.54 vs 2.84 baseline (45.7% fewer, modeled)
-- Languages: 33 supported · MCP tools: 21
+- Task-success proxy: 62.9% (modeled from retrieval tiers, not measured LLM sessions)
+- Prompts per task: 1.58 vs 2.84 baseline (44.4% fewer, modeled)
+- Languages: 32 supported · MCP tools: 21
 - Dependencies: zero npm runtime dependencies · fully offline
 
 ## Quick start

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "sigmap",
-  "version": "8.32.0",
+  "version": "8.32.1",
   "description": "The deterministic, verifiable grounding layer for AI code work — a zero-dependency signature-and-evidence map that grounds Claude, Cursor, Copilot, Aider, Windsurf, local LLMs & MCP agents against your real code (repo + installed libraries) so they stop hallucinating files, imports & APIs. Runs offline via npx; byte-stable output; ~97% token reduction as proof.",
   "main": "packages/core/index.js",
   "exports": {

package/packages/cli/package.json

@@ -1,6 +1,6 @@
 {
   "name": "sigmap-cli",
-  "version": "8.32.0",
+  "version": "8.32.1",
   "description": "SigMap CLI wrapper — thin adapter for programmatic CLI invocation",
   "main": "index.js",
   "keywords": [

package/packages/core/package.json

@@ -1,6 +1,6 @@
 {
   "name": "sigmap-core",
-  "version": "8.32.0",
+  "version": "8.32.1",
   "description": "SigMap core library — zero-dependency code signature extraction, retrieval, and security scanning",
   "main": "index.js",
   "keywords": [

package/src/config/loader.js

@@ -227,6 +227,55 @@ function _legacyDetectAutoSrcDirs(cwd, excludeList) {
   });
 }
 
+/**
+ * Directory depth needed to reach source under a JVM package layout.
+ *
+ * `maxDepth: 6` is right for the JS/Python-shaped trees it was tuned on, but a
+ * JVM project puts one directory per package segment — real code in
+ * `src/main/java/com/company/project/module/Class.java` sits 8-10 levels down.
+ * At depth 6 only the top-level package was indexed: on spring-petclinic, 6 of
+ * 47 Java files (#590).
+ *
+ * #561 already raised the dependency-graph walk to 12 for exactly this reason,
+ * so extraction was the shallower half of an inconsistent pair.
+ *
+ * Deepening globally is not free — it adds candidates to every repo and cost
+ * 2.2pp on the (filename-leaky) 105-task matrix corpus while the unbiased
+ * `mined` corpus stayed flat. So the depth is raised only where the layout
+ * demands it, leaving non-JVM repos byte-identical.
+ *
+ * @param {string} cwd
+ * @returns {boolean} true when the repo looks like a Maven/Gradle/sbt project
+ */
+function _isJvmLayout(cwd) {
+  const MARKERS = ['pom.xml', 'build.gradle', 'build.gradle.kts', 'build.sbt', 'settings.gradle', 'settings.gradle.kts'];
+  for (const m of MARKERS) {
+    try { if (fs.existsSync(path.join(cwd, m))) return true; } catch { /* unreadable cwd */ }
+  }
+  for (const d of ['src/main/java', 'src/main/kotlin', 'src/main/scala']) {
+    try { if (fs.existsSync(path.join(cwd, d))) return true; } catch { /* ignore */ }
+  }
+  return false;
+}
+
+/** Walk depth for a JVM package layout — matches the graph walk from #561. */
+const JVM_MAX_DEPTH = 12;
+
+/**
+ * Raise `maxDepth` to the JVM depth when the layout needs it (#590).
+ * An explicit user value always wins, including a deliberately shallow one.
+ * @param {object} cfg   resolved config (mutated and returned)
+ * @param {string} cwd
+ * @param {boolean} userSetDepth
+ * @returns {object} cfg
+ */
+function _applyJvmDepth(cfg, cwd, userSetDepth) {
+  if (!userSetDepth && cfg.maxDepth < JVM_MAX_DEPTH && _isJvmLayout(cwd)) {
+    cfg.maxDepth = JVM_MAX_DEPTH;
+  }
+  return cfg;
+}
+
 /**
  * Load and merge configuration for a given working directory.
  *
@@ -239,7 +288,7 @@ function loadConfig(cwd) {
     const cfg = deepClone(DEFAULTS);
     const detected = detectAutoSrcDirs(cwd, cfg.exclude);
     if (detected.length > 0) cfg.srcDirs = detected;
-    return cfg;
+    return _applyJvmDepth(cfg, cwd, false);
   }
 
   let userConfig;
@@ -251,7 +300,7 @@ function loadConfig(cwd) {
     const cfg = deepClone(DEFAULTS);
     const detected = detectAutoSrcDirs(cwd, cfg.exclude);
     if (detected.length > 0) cfg.srcDirs = detected;
-    return cfg;
+    return _applyJvmDepth(cfg, cwd, false);
   }
 
   // Warn on unknown keys (helps catch typos)
@@ -303,7 +352,8 @@ function loadConfig(cwd) {
   } else if (Array.isArray(merged.adapters) && !userConfig.outputs) {
     merged.outputs = merged.adapters.filter((a) => ['copilot','claude','cursor','windsurf'].includes(a));
   }
-  return merged;
+
+  return _applyJvmDepth(merged, cwd, userConfig.maxDepth !== undefined);
 }
 
 function deepClone(obj) {

package/src/eval/analyzer.js

@@ -31,7 +31,6 @@ const EXT_MAP = {
   '.scala': 'scala',   '.sc': 'scala',
   '.gd': 'gdscript',
   '.r': 'r',           '.R': 'r',
-  '.vue': 'vue',
   '.svelte': 'svelte',
   '.html': 'html',     '.htm': 'html',
   '.css': 'css',       '.scss': 'css', '.sass': 'css', '.less': 'css',

package/src/extractors/dispatch.js

@@ -30,7 +30,6 @@ const EXTRACTORS = {
   scala: require('./scala'),
   gdscript: require('./gdscript'),
   r: require('./r'),
-  vue: require('./vue'),
   vue_sfc: require('./vue_sfc'),
   svelte: require('./svelte'),
   html: require('./html'),

package/src/extractors/markdown.js

@@ -1,5 +1,10 @@
 'use strict';
 
+const { capWithNotice } = require('../util/truncate');
+
+// Ceiling discloses what it drops rather than truncating silently (#583).
+const PER_FILE_LIMIT = 40;
+
 /**
  * Lightweight markdown technical indexer.
  * Captures headings and fenced code block language hints only.
@@ -24,7 +29,7 @@ function extract(src) {
     sigs.push(`code-fence ${lang}`);
   }
 
-  return Array.from(new Set(sigs)).slice(0, 40);
+  return capWithNotice(Array.from(new Set(sigs)), PER_FILE_LIMIT, 'headings');
 }
 
 module.exports = { extract };

package/src/extractors/properties.js

@@ -1,5 +1,10 @@
 'use strict';
 
+const { capWithNotice } = require('../util/truncate');
+
+// Ceiling discloses what it drops rather than truncating silently (#583).
+const PER_FILE_LIMIT = 50;
+
 /**
  * Extract signatures from .properties configuration files.
  * Captures key names, grouped by prefixes where possible.
@@ -31,7 +36,7 @@ function extract(src) {
     sigs.push(`key ${key}`);
   }
 
-  return Array.from(new Set(sigs)).slice(0, 50);
+  return capWithNotice(Array.from(new Set(sigs)), PER_FILE_LIMIT, 'keys');
 }
 
 module.exports = { extract };

package/src/extractors/r.js

@@ -44,7 +44,7 @@ function extract(src) {
   //   ClassName <- R6::R6Class(...)
   const r6Re = /([\w.]+)\s*(?:<<-|<-|=)\s*(?:R6::)?R6Class\s*\(/g;
   let m;
-  while ((m = r6Re.exec(stripped)) !== null && sigs.length < 30) {
+  while ((m = r6Re.exec(stripped)) !== null) {
     const name = m[1];
     if (name.startsWith('.')) continue;
     const openIdx = r6Re.lastIndex - 1;
@@ -55,7 +55,6 @@ function extract(src) {
     sigs.push(`${name} <- R6Class("${classNameLit}")` + applyHint(docHints, name));
     for (const memberSig of extractListMethods(body, 8)) {
       sigs.push('  ' + memberSig);
-      if (sigs.length >= 30) break;
     }
     consumedRanges.push([m.index, closeIdx]);
     r6Re.lastIndex = closeIdx;
@@ -65,7 +64,7 @@ function extract(src) {
   //   ClassName <- new_class("ClassName", properties = list(...))
   const s7Classes = new Set();
   const s7Re = /([\w.]+)\s*(?:<<-|<-|=)\s*(?:S7::)?new_class\s*\(/g;
-  while ((m = s7Re.exec(stripped)) !== null && sigs.length < 30) {
+  while ((m = s7Re.exec(stripped)) !== null) {
     const name = m[1];
     if (name.startsWith('.')) continue;
     const openIdx = s7Re.lastIndex - 1;
@@ -82,7 +81,7 @@ function extract(src) {
 
   // S7 method dispatch: `method(generic, ClassName) <- function(args)`
   const s7MethodRe = /^[ \t]*method\s*\(\s*([\w.]+)\s*,\s*([\w.]+)\s*\)\s*(?:<<-|<-|=)\s*function\s*\(/gm;
-  while ((m = s7MethodRe.exec(stripped)) !== null && sigs.length < 30) {
+  while ((m = s7MethodRe.exec(stripped)) !== null) {
     if (!s7Classes.has(m[2])) continue;
     const argsStart = s7MethodRe.lastIndex - 1;
     const args = readBalancedParens(stripped, argsStart);
@@ -95,7 +94,7 @@ function extract(src) {
   // Skip matches whose position falls inside an R6/S7 class body — those have
   // already been emitted as indented members.
   const funcRe = /^(?:[ \t]*)([\w.]+)\s*(?:<<-|<-|=)\s*function\s*\(/gm;
-  while ((m = funcRe.exec(stripped)) !== null && sigs.length < 30) {
+  while ((m = funcRe.exec(stripped)) !== null) {
     const name = m[1];
     if (name.startsWith('.')) continue;
     if (inAnyRange(m.index, consumedRanges)) continue;
@@ -107,15 +106,12 @@ function extract(src) {
 
   // ── S4 ────────────────────────────────────────────────────────────────────
   for (const sm of stripped.matchAll(/^[ \t]*setGeneric\s*\(\s*["']([\w.]+)["']/gm)) {
-    if (sigs.length >= 30) break;
     sigs.push(`setGeneric("${sm[1]}")`);
   }
   for (const sm of stripped.matchAll(/^[ \t]*setMethod\s*\(\s*["']([\w.]+)["']\s*,\s*["']([\w.]+)["']/gm)) {
-    if (sigs.length >= 30) break;
     sigs.push(`setMethod("${sm[1]}", "${sm[2]}")`);
   }
   for (const sm of stripped.matchAll(/^[ \t]*setClass\s*\(\s*["']([\w.]+)["']/gm)) {
-    if (sigs.length >= 30) break;
     sigs.push(`setClass("${sm[1]}")`);
   }
 

package/src/extractors/toml.js

@@ -1,5 +1,10 @@
 'use strict';
 
+const { capWithNotice } = require('../util/truncate');
+
+// Ceiling discloses what it drops rather than truncating silently (#583).
+const PER_FILE_LIMIT = 40;
+
 /**
  * Extract signatures from TOML configuration files.
  * Focuses on section/table names and high-value keys.
@@ -36,7 +41,7 @@ function extract(src) {
     }
   }
 
-  return Array.from(new Set(sigs)).slice(0, 40);
+  return capWithNotice(Array.from(new Set(sigs)), PER_FILE_LIMIT, 'entries');
 }
 
 module.exports = { extract };

package/src/extractors/typescript_react.js

@@ -1,5 +1,10 @@
 'use strict';
 
+const { capWithNotice } = require('../util/truncate');
+
+// Ceiling discloses what it drops rather than truncating silently (#583).
+const PER_FILE_LIMIT = 50;
+
 /**
  * Extract React component signatures from .tsx files.
  * Captures component props interfaces, hooks usage, and exports.
@@ -54,7 +59,7 @@ function extract(src) {
     sigs.push(`handler on${h}`);
   }
 
-  return Array.from(new Set(sigs)).slice(0, 50);
+  return capWithNotice(Array.from(new Set(sigs)), PER_FILE_LIMIT, 'signatures');
 }
 
 module.exports = { extract };

package/src/mcp/server.js

@@ -18,7 +18,7 @@ const { readContext, searchSignatures, getMap, createCheckpoint, getRouting, exp
 
 const SERVER_INFO = {
   name: 'sigmap',
-  version: '8.32.0',
+  version: '8.32.1',
   description: 'SigMap MCP server — code signatures on demand',
 };
 

package/src/extractors/vue.js

@@ -1,86 +0,0 @@
-'use strict';
-
-const { capWithNotice } = require('../util/truncate');
-
-// Ceiling sits above the default `maxSigsPerFile` so the configured budget
-// governs output rather than a literal buried here, and omissions are disclosed (#576).
-const PER_FILE_LIMIT = 25;
-
-/**
- * Extract signatures from Vue single-file components.
- * @param {string} src - Raw file content
- * @returns {string[]} Array of signature strings
- */
-function extract(src) {
-  if (!src || typeof src !== 'string') return [];
-  const sigs = [];
-
-  // Extract component name from filename hint if present or defineComponent
-  const nameMatch = src.match(/name\s*:\s*['"](\w+)['"]/);
-  if (nameMatch) sigs.push(`component ${nameMatch[1]}`);
-
-  // Extract <script> block
-  const scriptMatch = src.match(/<script(?:\s[^>]*)?>(?:\s*)([\s\S]*?)<\/script>/i);
-  if (!scriptMatch) return sigs;
-
-  const script = scriptMatch[1]
-    .replace(/\/\/.*$/gm, '')
-    .replace(/\/\*[\s\S]*?\*\//g, '');
-
-  // Props
-  const propsMatch = script.match(/props\s*:\s*(\{[\s\S]*?\})/);
-  if (propsMatch) {
-    const propNames = [];
-    for (const m of propsMatch[1].matchAll(/^\s+(\w+)\s*:/gm)) {
-      propNames.push(m[1]);
-    }
-    if (propNames.length > 0) sigs.push(`props: [${propNames.join(', ')}]`);
-  }
-
-  // Methods in options API
-  const methodsMatch = script.match(/methods\s*:\s*\{([\s\S]*?)\},?\s*(?:computed|watch|mounted|created|data|\})/);
-  if (methodsMatch) {
-    for (const m of methodsMatch[1].matchAll(/^\s+(?:async\s+)?(\w+)\s*\(([^)]*)\)(?:\s*:\s*([^{=\n]+))?/gm)) {
-      if (m[1].startsWith('_')) continue;
-      const asyncKw = m[0].includes('async') ? 'async ' : '';
-      const retStr = m[3] ? ` → ${normalizeType(m[3])}` : '';
-      sigs.push(`  ${asyncKw}${m[1]}(${normalizeParams(m[2])})${retStr}`);
-    }
-  }
-
-  // Top-level functions in <script> (e.g., composition API helpers)
-  for (const m of script.matchAll(/^(?:export\s+)?(?:async\s+)?function\s+(\w+)\s*\(([^)]*)\)(?:\s*:\s*([^{=\n]+))?/gm)) {
-    if (m[1].startsWith('_')) continue;
-    const asyncKw = m[0].includes('async') ? 'async ' : '';
-    const retStr = m[3] ? ` → ${normalizeType(m[3])}` : '';
-    sigs.push(`${asyncKw}function ${m[1]}(${normalizeParams(m[2])})${retStr}`);
-  }
-
-  // defineProps (Composition API)
-  const definePropsMatch = script.match(/defineProps(?:<[^>]*>)?\s*\(\s*(\{[\s\S]*?\})\s*\)/);
-  if (definePropsMatch) {
-    const propNames = [];
-    for (const m of definePropsMatch[1].matchAll(/^\s+(\w+)\s*:/gm)) {
-      propNames.push(m[1]);
-    }
-    if (propNames.length > 0) sigs.push(`defineProps: [${propNames.join(', ')}]`);
-  }
-
-  // Emits
-  const emitsMatch = script.match(/(?:defineEmits|emits)\s*(?::\s*|\(\s*)(\[[\s\S]*?\])/);
-  if (emitsMatch) sigs.push(`emits: ${emitsMatch[1].replace(/\s+/g, ' ')}`);
-
-  return capWithNotice(sigs, PER_FILE_LIMIT, 'signatures');
-}
-
-function normalizeParams(params) {
-  if (!params) return '';
-  return params.trim().replace(/\s+/g, ' ');
-}
-
-function normalizeType(type) {
-  if (!type) return '';
-  return type.trim().replace(/[;\s]+$/g, '').replace(/\s+/g, ' ').slice(0, 25);
-}
-
-module.exports = { extract };