@timmeck/brain 1.0.0 → 1.1.0

package/src/learning/rule-generator.ts

@@ -1,74 +1,74 @@
-import type { LearningConfig } from '../types/config.types.js';
-import type { RuleRepository } from '../db/repositories/rule.repository.js';
-import type { ErrorPattern } from './pattern-extractor.js';
-import { getLogger } from '../utils/logger.js';
-
-export interface GeneratedRule {
-  pattern: string;
-  action: string;
-  description: string;
-  confidence: number;
-  sourceErrorIds: number[];
-}
-
-/**
- * Generate prevention rules from extracted patterns.
- */
-export function generateRules(
-  patterns: ErrorPattern[],
-  config: LearningConfig,
-): GeneratedRule[] {
-  return patterns
-    .filter(p =>
-      p.occurrences >= config.minOccurrences &&
-      p.confidence >= config.minConfidence,
-    )
-    .map(pattern => ({
-      pattern: pattern.messageRegex,
-      action: pattern.confidence >= 0.90
-        ? `Auto-fix available for ${pattern.errorType}`
-        : `Suggestion: check ${pattern.errorType} pattern (${pattern.occurrences} occurrences)`,
-      description: `Auto-generated from ${pattern.occurrences} occurrences of ${pattern.errorType}`,
-      confidence: pattern.confidence,
-      sourceErrorIds: pattern.errorIds,
-    }));
-}
-
-/**
- * Persist generated rules to the database.
- */
-export function persistRules(
-  rules: GeneratedRule[],
-  ruleRepo: RuleRepository,
-  projectId?: number,
-): number {
-  const logger = getLogger();
-  let created = 0;
-
-  for (const rule of rules) {
-    // Check if similar rule already exists
-    const existing = ruleRepo.findByPattern(rule.pattern);
-    if (existing.length > 0) {
-      // Update confidence of existing rule
-      const best = existing[0]!;
-      if (rule.confidence > best.confidence) {
-        ruleRepo.update(best.id, { confidence: rule.confidence });
-      }
-      continue;
-    }
-
-    ruleRepo.create({
-      pattern: rule.pattern,
-      action: rule.action,
-      description: rule.description,
-      confidence: rule.confidence,
-      occurrences: 0,
-      active: 1,
-      project_id: projectId ?? null,
-    });
-    created++;
-    logger.info(`New rule generated: ${rule.pattern.substring(0, 50)}...`);
-  }
-
-  return created;
-}
+import type { LearningConfig } from '../types/config.types.js';
+import type { RuleRepository } from '../db/repositories/rule.repository.js';
+import type { ErrorPattern } from './pattern-extractor.js';
+import { getLogger } from '../utils/logger.js';
+
+export interface GeneratedRule {
+  pattern: string;
+  action: string;
+  description: string;
+  confidence: number;
+  sourceErrorIds: number[];
+}
+
+/**
+ * Generate prevention rules from extracted patterns.
+ */
+export function generateRules(
+  patterns: ErrorPattern[],
+  config: LearningConfig,
+): GeneratedRule[] {
+  return patterns
+    .filter(p =>
+      p.occurrences >= config.minOccurrences &&
+      p.confidence >= config.minConfidence,
+    )
+    .map(pattern => ({
+      pattern: pattern.messageRegex,
+      action: pattern.confidence >= 0.90
+        ? `Auto-fix available for ${pattern.errorType}`
+        : `Suggestion: check ${pattern.errorType} pattern (${pattern.occurrences} occurrences)`,
+      description: `Auto-generated from ${pattern.occurrences} occurrences of ${pattern.errorType}`,
+      confidence: pattern.confidence,
+      sourceErrorIds: pattern.errorIds,
+    }));
+}
+
+/**
+ * Persist generated rules to the database.
+ */
+export function persistRules(
+  rules: GeneratedRule[],
+  ruleRepo: RuleRepository,
+  projectId?: number,
+): number {
+  const logger = getLogger();
+  let created = 0;
+
+  for (const rule of rules) {
+    // Check if similar rule already exists
+    const existing = ruleRepo.findByPattern(rule.pattern);
+    if (existing.length > 0) {
+      // Update confidence of existing rule
+      const best = existing[0]!;
+      if (rule.confidence > best.confidence) {
+        ruleRepo.update(best.id, { confidence: rule.confidence });
+      }
+      continue;
+    }
+
+    ruleRepo.create({
+      pattern: rule.pattern,
+      action: rule.action,
+      description: rule.description,
+      confidence: rule.confidence,
+      occurrences: 0,
+      active: 1,
+      project_id: projectId ?? null,
+    });
+    created++;
+    logger.info(`New rule generated: ${rule.pattern.substring(0, 50)}...`);
+  }
+
+  return created;
+}

package/src/matching/error-matcher.ts

@@ -1,115 +1,115 @@
-import type { ErrorRecord } from '../types/error.types.js';
-import { tokenize } from './tokenizer.js';
-import { cosineSimilarity, jaccardSimilarity, levenshteinDistance } from './similarity.js';
-
-export interface SignalScore {
-  signal: string;
-  score: number;
-  weighted: number;
-}
-
-export interface MatchResult {
-  errorId: number;
-  score: number;
-  signals: SignalScore[];
-  isStrong: boolean;
-}
-
-interface MatchSignal {
-  name: string;
-  weight: number;
-  compute: (a: ErrorRecord, b: ErrorRecord) => number;
-}
-
-const SIGNALS: MatchSignal[] = [
-  { name: 'fingerprint', weight: 0.30, compute: fingerprintMatch },
-  { name: 'message_similarity', weight: 0.20, compute: messageSimilarity },
-  { name: 'type_match', weight: 0.15, compute: typeMatch },
-  { name: 'stack_similarity', weight: 0.15, compute: stackSimilarity },
-  { name: 'file_similarity', weight: 0.10, compute: fileSimilarity },
-  { name: 'context_similarity', weight: 0.10, compute: contextSimilarity },
-];
-
-const MATCH_THRESHOLD = 0.70;
-const STRONG_MATCH_THRESHOLD = 0.90;
-
-export function matchError(
-  incoming: ErrorRecord,
-  candidates: ErrorRecord[],
-): MatchResult[] {
-  return candidates
-    .map(candidate => {
-      const signals = SIGNALS.map(signal => {
-        const score = signal.compute(incoming, candidate);
-        return {
-          signal: signal.name,
-          score,
-          weighted: score * signal.weight,
-        };
-      });
-
-      const totalScore = signals.reduce((sum, s) => sum + s.weighted, 0);
-
-      return {
-        errorId: candidate.id,
-        score: totalScore,
-        signals,
-        isStrong: totalScore >= STRONG_MATCH_THRESHOLD,
-      };
-    })
-    .filter(result => result.score >= MATCH_THRESHOLD)
-    .sort((a, b) => b.score - a.score);
-}
-
-function fingerprintMatch(a: ErrorRecord, b: ErrorRecord): number {
-  return a.fingerprint === b.fingerprint ? 1.0 : 0.0;
-}
-
-function messageSimilarity(a: ErrorRecord, b: ErrorRecord): number {
-  const tokensA = tokenize(a.message);
-  const tokensB = tokenize(b.message);
-  return cosineSimilarity(tokensA, tokensB);
-}
-
-function typeMatch(a: ErrorRecord, b: ErrorRecord): number {
-  return a.type === b.type ? 1.0 : 0.0;
-}
-
-function stackSimilarity(a: ErrorRecord, b: ErrorRecord): number {
-  const rawA = a.raw_output ?? '';
-  const rawB = b.raw_output ?? '';
-
-  const frameRe = /at (?:(.+?) )?\(/g;
-  const extractFuncs = (raw: string) => {
-    const funcs: string[] = [];
-    let m: RegExpExecArray | null;
-    const re = new RegExp(frameRe.source, 'g');
-    while ((m = re.exec(raw)) !== null) {
-      if (m[1]) funcs.push(m[1]);
-    }
-    return funcs;
-  };
-
-  const funcsA = extractFuncs(rawA);
-  const funcsB = extractFuncs(rawB);
-
-  if (funcsA.length === 0 && funcsB.length === 0) return 0.5;
-  return jaccardSimilarity(funcsA, funcsB);
-}
-
-function fileSimilarity(a: ErrorRecord, b: ErrorRecord): number {
-  const pathA = a.file_path ?? '';
-  const pathB = b.file_path ?? '';
-  if (!pathA || !pathB) return 0.0;
-  if (pathA === pathB) return 1.0;
-  return levenshteinDistance(pathA, pathB);
-}
-
-function contextSimilarity(a: ErrorRecord, b: ErrorRecord): number {
-  const ctxA = a.context ?? '';
-  const ctxB = b.context ?? '';
-  if (!ctxA || !ctxB) return 0.0;
-  const tokensA = tokenize(ctxA);
-  const tokensB = tokenize(ctxB);
-  return cosineSimilarity(tokensA, tokensB);
-}
+import type { ErrorRecord } from '../types/error.types.js';
+import { tokenize } from './tokenizer.js';
+import { cosineSimilarity, jaccardSimilarity, levenshteinDistance } from './similarity.js';
+
+export interface SignalScore {
+  signal: string;
+  score: number;
+  weighted: number;
+}
+
+export interface MatchResult {
+  errorId: number;
+  score: number;
+  signals: SignalScore[];
+  isStrong: boolean;
+}
+
+interface MatchSignal {
+  name: string;
+  weight: number;
+  compute: (a: ErrorRecord, b: ErrorRecord) => number;
+}
+
+const SIGNALS: MatchSignal[] = [
+  { name: 'fingerprint', weight: 0.30, compute: fingerprintMatch },
+  { name: 'message_similarity', weight: 0.20, compute: messageSimilarity },
+  { name: 'type_match', weight: 0.15, compute: typeMatch },
+  { name: 'stack_similarity', weight: 0.15, compute: stackSimilarity },
+  { name: 'file_similarity', weight: 0.10, compute: fileSimilarity },
+  { name: 'context_similarity', weight: 0.10, compute: contextSimilarity },
+];
+
+const MATCH_THRESHOLD = 0.70;
+const STRONG_MATCH_THRESHOLD = 0.90;
+
+export function matchError(
+  incoming: ErrorRecord,
+  candidates: ErrorRecord[],
+): MatchResult[] {
+  return candidates
+    .map(candidate => {
+      const signals = SIGNALS.map(signal => {
+        const score = signal.compute(incoming, candidate);
+        return {
+          signal: signal.name,
+          score,
+          weighted: score * signal.weight,
+        };
+      });
+
+      const totalScore = signals.reduce((sum, s) => sum + s.weighted, 0);
+
+      return {
+        errorId: candidate.id,
+        score: totalScore,
+        signals,
+        isStrong: totalScore >= STRONG_MATCH_THRESHOLD,
+      };
+    })
+    .filter(result => result.score >= MATCH_THRESHOLD)
+    .sort((a, b) => b.score - a.score);
+}
+
+function fingerprintMatch(a: ErrorRecord, b: ErrorRecord): number {
+  return a.fingerprint === b.fingerprint ? 1.0 : 0.0;
+}
+
+function messageSimilarity(a: ErrorRecord, b: ErrorRecord): number {
+  const tokensA = tokenize(a.message);
+  const tokensB = tokenize(b.message);
+  return cosineSimilarity(tokensA, tokensB);
+}
+
+function typeMatch(a: ErrorRecord, b: ErrorRecord): number {
+  return a.type === b.type ? 1.0 : 0.0;
+}
+
+function stackSimilarity(a: ErrorRecord, b: ErrorRecord): number {
+  const rawA = a.raw_output ?? '';
+  const rawB = b.raw_output ?? '';
+
+  const frameRe = /at (?:(.+?) )?\(/g;
+  const extractFuncs = (raw: string) => {
+    const funcs: string[] = [];
+    let m: RegExpExecArray | null;
+    const re = new RegExp(frameRe.source, 'g');
+    while ((m = re.exec(raw)) !== null) {
+      if (m[1]) funcs.push(m[1]);
+    }
+    return funcs;
+  };
+
+  const funcsA = extractFuncs(rawA);
+  const funcsB = extractFuncs(rawB);
+
+  if (funcsA.length === 0 && funcsB.length === 0) return 0.5;
+  return jaccardSimilarity(funcsA, funcsB);
+}
+
+function fileSimilarity(a: ErrorRecord, b: ErrorRecord): number {
+  const pathA = a.file_path ?? '';
+  const pathB = b.file_path ?? '';
+  if (!pathA || !pathB) return 0.0;
+  if (pathA === pathB) return 1.0;
+  return levenshteinDistance(pathA, pathB);
+}
+
+function contextSimilarity(a: ErrorRecord, b: ErrorRecord): number {
+  const ctxA = a.context ?? '';
+  const ctxB = b.context ?? '';
+  if (!ctxA || !ctxB) return 0.0;
+  const tokensA = tokenize(ctxA);
+  const tokensB = tokenize(ctxB);
+  return cosineSimilarity(tokensA, tokensB);
+}

package/src/matching/fingerprint.ts

@@ -1,29 +1,29 @@
-import path from 'node:path';
-import { sha256 } from '../utils/hash.js';
-import type { StackFrame } from '../parsing/types.js';
-
-export function templateMessage(msg: string): string {
-  return msg
-    .replace(/[A-Z]:\\[\w\-.\\ ]+\.\w+/g, '<PATH>')
-    .replace(/\/[\w\-./ ]+\.\w+/g, '<PATH>')
-    .replace(/:(\d+):(\d+)/g, ':<LINE>:<COL>')
-    .replace(/line \d+/gi, 'line <LINE>')
-    .replace(/0x[0-9a-fA-F]+/g, '<ADDR>')
-    .replace(/[0-9a-f]{8}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{12}/gi, '<UUID>')
-    .replace(/https?:\/\/[^\s]+/g, '<URL>')
-    .replace(/\d{4}-\d{2}-\d{2}[T ]\d{2}:\d{2}:\d{2}/g, '<TIMESTAMP>');
-}
-
-export function generateFingerprint(
-  errorType: string,
-  message: string,
-  frames: StackFrame[],
-): string {
-  const template = templateMessage(message);
-  const topFrames = frames
-    .slice(0, 3)
-    .map(f => `${f.function_name || '<anon>'}@${path.basename(f.file_path || '<unknown>')}`)
-    .join('|');
-  const input = `${errorType}::${template}::${topFrames}`;
-  return sha256(input);
-}
+import path from 'node:path';
+import { sha256 } from '../utils/hash.js';
+import type { StackFrame } from '../parsing/types.js';
+
+export function templateMessage(msg: string): string {
+  return msg
+    .replace(/[A-Z]:\\[\w\-.\\ ]+\.\w+/g, '<PATH>')
+    .replace(/\/[\w\-./ ]+\.\w+/g, '<PATH>')
+    .replace(/:(\d+):(\d+)/g, ':<LINE>:<COL>')
+    .replace(/line \d+/gi, 'line <LINE>')
+    .replace(/0x[0-9a-fA-F]+/g, '<ADDR>')
+    .replace(/[0-9a-f]{8}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{12}/gi, '<UUID>')
+    .replace(/https?:\/\/[^\s]+/g, '<URL>')
+    .replace(/\d{4}-\d{2}-\d{2}[T ]\d{2}:\d{2}:\d{2}/g, '<TIMESTAMP>');
+}
+
+export function generateFingerprint(
+  errorType: string,
+  message: string,
+  frames: StackFrame[],
+): string {
+  const template = templateMessage(message);
+  const topFrames = frames
+    .slice(0, 3)
+    .map(f => `${f.function_name || '<anon>'}@${path.basename(f.file_path || '<unknown>')}`)
+    .join('|');
+  const input = `${errorType}::${template}::${topFrames}`;
+  return sha256(input);
+}

package/src/matching/similarity.ts

@@ -1,61 +1,61 @@
-export function levenshteinDistance(a: string, b: string): number {
-  if (a === b) return 1.0;
-  if (a.length === 0 || b.length === 0) return 0.0;
-
-  const dp: number[][] = Array(b.length + 1)
-    .fill(0)
-    .map(() => Array(a.length + 1).fill(0) as number[]);
-
-  for (let i = 0; i <= a.length; i++) dp[0]![i] = i;
-  for (let j = 0; j <= b.length; j++) dp[j]![0] = j;
-
-  for (let i = 1; i <= b.length; i++) {
-    for (let j = 1; j <= a.length; j++) {
-      const cost = a[j - 1] === b[i - 1] ? 0 : 1;
-      dp[i]![j] = Math.min(
-        dp[i - 1]![j]! + 1,
-        dp[i]![j - 1]! + 1,
-        dp[i - 1]![j - 1]! + cost,
-      );
-    }
-  }
-
-  return 1 - dp[b.length]![a.length]! / Math.max(a.length, b.length);
-}
-
-export function cosineSimilarity(tokensA: string[], tokensB: string[]): number {
-  if (tokensA.length === 0 || tokensB.length === 0) return 0.0;
-
-  const vocab = new Set([...tokensA, ...tokensB]);
-  const vecA = new Map<string, number>();
-  const vecB = new Map<string, number>();
-
-  for (const t of tokensA) vecA.set(t, (vecA.get(t) ?? 0) + 1);
-  for (const t of tokensB) vecB.set(t, (vecB.get(t) ?? 0) + 1);
-
-  let dot = 0;
-  let magA = 0;
-  let magB = 0;
-
-  for (const word of vocab) {
-    const a = vecA.get(word) ?? 0;
-    const b = vecB.get(word) ?? 0;
-    dot += a * b;
-    magA += a * a;
-    magB += b * b;
-  }
-
-  const denom = Math.sqrt(magA) * Math.sqrt(magB);
-  return denom === 0 ? 0 : dot / denom;
-}
-
-export function jaccardSimilarity(tokensA: string[], tokensB: string[]): number {
-  if (tokensA.length === 0 && tokensB.length === 0) return 0.0;
-
-  const setA = new Set(tokensA);
-  const setB = new Set(tokensB);
-  const intersection = new Set([...setA].filter(x => setB.has(x)));
-  const union = new Set([...setA, ...setB]);
-
-  return union.size === 0 ? 0 : intersection.size / union.size;
-}
+export function levenshteinDistance(a: string, b: string): number {
+  if (a === b) return 1.0;
+  if (a.length === 0 || b.length === 0) return 0.0;
+
+  const dp: number[][] = Array(b.length + 1)
+    .fill(0)
+    .map(() => Array(a.length + 1).fill(0) as number[]);
+
+  for (let i = 0; i <= a.length; i++) dp[0]![i] = i;
+  for (let j = 0; j <= b.length; j++) dp[j]![0] = j;
+
+  for (let i = 1; i <= b.length; i++) {
+    for (let j = 1; j <= a.length; j++) {
+      const cost = a[j - 1] === b[i - 1] ? 0 : 1;
+      dp[i]![j] = Math.min(
+        dp[i - 1]![j]! + 1,
+        dp[i]![j - 1]! + 1,
+        dp[i - 1]![j - 1]! + cost,
+      );
+    }
+  }
+
+  return 1 - dp[b.length]![a.length]! / Math.max(a.length, b.length);
+}
+
+export function cosineSimilarity(tokensA: string[], tokensB: string[]): number {
+  if (tokensA.length === 0 || tokensB.length === 0) return 0.0;
+
+  const vocab = new Set([...tokensA, ...tokensB]);
+  const vecA = new Map<string, number>();
+  const vecB = new Map<string, number>();
+
+  for (const t of tokensA) vecA.set(t, (vecA.get(t) ?? 0) + 1);
+  for (const t of tokensB) vecB.set(t, (vecB.get(t) ?? 0) + 1);
+
+  let dot = 0;
+  let magA = 0;
+  let magB = 0;
+
+  for (const word of vocab) {
+    const a = vecA.get(word) ?? 0;
+    const b = vecB.get(word) ?? 0;
+    dot += a * b;
+    magA += a * a;
+    magB += b * b;
+  }
+
+  const denom = Math.sqrt(magA) * Math.sqrt(magB);
+  return denom === 0 ? 0 : dot / denom;
+}
+
+export function jaccardSimilarity(tokensA: string[], tokensB: string[]): number {
+  if (tokensA.length === 0 && tokensB.length === 0) return 0.0;
+
+  const setA = new Set(tokensA);
+  const setB = new Set(tokensB);
+  const intersection = new Set([...setA].filter(x => setB.has(x)));
+  const union = new Set([...setA, ...setB]);
+
+  return union.size === 0 ? 0 : intersection.size / union.size;
+}