@rigour-labs/core 2.0.0 → 2.2.0

package/src/services/context-engine.ts

@@ -0,0 +1,104 @@
+import { FileScanner } from '../utils/scanner.js';
+import { Config } from '../types/index.js';
+import path from 'path';
+import fs from 'fs-extra';
+
+export interface ProjectAnchor {
+    id: string;
+    type: 'env' | 'naming' | 'import';
+    pattern: string;
+    confidence: number;
+    occurrences: number;
+}
+
+export interface GoldenRecord {
+    anchors: ProjectAnchor[];
+    metadata: {
+        scannedFiles: number;
+        detectedCasing: 'camelCase' | 'snake_case' | 'PascalCase' | 'unknown';
+    };
+}
+
+export class ContextEngine {
+    constructor(private config: Config) { }
+
+    async discover(cwd: string): Promise<GoldenRecord> {
+        const anchors: ProjectAnchor[] = [];
+        const files = await FileScanner.findFiles({
+            cwd,
+            patterns: [
+                '**/*.{ts,js,py,yaml,yml,json}',
+                '.env*',
+                '**/.env*',
+                '**/package.json',
+                '**/Dockerfile',
+                '**/*.tf'
+            ]
+        });
+
+        const limit = this.config.gates.context?.mining_depth || 100;
+        const samples = files.slice(0, limit);
+
+        const envVars = new Map<string, number>();
+        let scannedFiles = 0;
+
+        for (const file of samples) {
+            try {
+                const content = await fs.readFile(path.join(cwd, file), 'utf-8');
+                scannedFiles++;
+                this.mineEnvVars(content, file, envVars);
+            } catch (e) { }
+        }
+
+        console.log(`[ContextEngine] Discovered ${envVars.size} env anchors`);
+
+        // Convert envVars to anchors
+        for (const [name, count] of envVars.entries()) {
+            const confidence = count >= 2 ? 1 : 0.5;
+            anchors.push({
+                id: name,
+                type: 'env',
+                pattern: name,
+                occurrences: count,
+                confidence
+            });
+        }
+
+        return {
+            anchors,
+            metadata: {
+                scannedFiles,
+                detectedCasing: 'unknown', // TODO: Implement casing discovery
+            }
+        };
+    }
+
+    private mineEnvVars(content: string, file: string, registry: Map<string, number>) {
+        const isAnchorSource = file.includes('.env') || file.includes('yml') || file.includes('yaml');
+
+        if (isAnchorSource) {
+            const matches = content.matchAll(/^\s*([A-Z0-9_]+)\s*=/gm);
+            for (const match of matches) {
+                // Anchors from .env count for more initially
+                registry.set(match[1], (registry.get(match[1]) || 0) + 2);
+            }
+        }
+
+        // Source code matches (process.env.VAR or process.env['VAR'])
+        const tsJsMatches = content.matchAll(/process\.env(?:\.([A-Z0-9_]+)|\[['"]([A-Z0-9_]+)['"]\])/g);
+        for (const match of tsJsMatches) {
+            const name = match[1] || match[2];
+            this.incrementRegistry(registry, name);
+        }
+
+        // Python matches (os.environ.get('VAR') or os.environ['VAR'])
+        const pyMatches = content.matchAll(/os\.environ(?:\.get\(|\[)['"]([A-Z0-9_]+)['"]/g);
+        for (const match of pyMatches) {
+            this.incrementRegistry(registry, match[1]);
+        }
+    }
+
+    private incrementRegistry(registry: Map<string, number>, key: string) {
+        registry.set(key, (registry.get(key) || 0) + 1);
+    }
+}

package/src/templates/index.ts

@@ -190,9 +190,22 @@ export const UNIVERSAL_CONFIG: Config = {
             max_files_changed_per_cycle: 10,
             protected_paths: ['.github/**', 'docs/**', 'rigour.yml'],
         },
+        context: {
+            enabled: true,
+            sensitivity: 0.8,
+            mining_depth: 100,
+            ignored_patterns: [],
+        },
+        environment: {
+            enabled: true,
+            enforce_contracts: true,
+            tools: {},
+            required_env: [],
+        },
     },
     output: {
         report_path: 'rigour-report.json',
     },
     planned: [],
+    ignore: [],
 };

package/src/types/index.ts

@@ -35,6 +35,18 @@ export const GatesSchema = z.object({
         max_files_changed_per_cycle: z.number().optional().default(10),
         protected_paths: z.array(z.string()).optional().default(['.github/**', 'docs/**', 'rigour.yml']),
     }).optional().default({}),
+    context: z.object({
+        enabled: z.boolean().optional().default(true),
+        sensitivity: z.number().min(0).max(1).optional().default(0.8), // 0.8 correlation threshold
+        mining_depth: z.number().optional().default(100), // Number of files to sample
+        ignored_patterns: z.array(z.string()).optional().default([]),
+    }).optional().default({}),
+    environment: z.object({
+        enabled: z.boolean().optional().default(true),
+        enforce_contracts: z.boolean().optional().default(true), // Auto-discovery of versions from truth sources
+        tools: z.record(z.string()).optional().default({}), // Explicit overrides
+        required_env: z.array(z.string()).optional().default([]),
+    }).optional().default({}),
 });
 
 export const CommandsSchema = z.object({
@@ -54,12 +66,17 @@ export const ConfigSchema = z.object({
         report_path: z.string().default('rigour-report.json'),
     }).optional().default({}),
     planned: z.array(z.string()).optional().default([]),
+    ignore: z.array(z.string()).optional().default([]),
 });
 
 export type Gates = z.infer<typeof GatesSchema>;
 export type Commands = z.infer<typeof CommandsSchema>;
 export type Config = z.infer<typeof ConfigSchema>;
 
+export type RawGates = z.input<typeof GatesSchema>;
+export type RawCommands = z.input<typeof CommandsSchema>;
+export type RawConfig = z.input<typeof ConfigSchema>;
+
 export const StatusSchema = z.enum(['PASS', 'FAIL', 'SKIP', 'ERROR']);
 export type Status = z.infer<typeof StatusSchema>;
 

package/src/utils/scanner.test.ts

@@ -0,0 +1,37 @@
+import { describe, it, expect, vi } from 'vitest';
+import { FileScanner } from './scanner.js';
+import { globby } from 'globby';
+
+vi.mock('globby', () => ({
+    globby: vi.fn(),
+}));
+
+describe('FileScanner', () => {
+    it('should merge default ignores with user ignores', async () => {
+        const options = {
+            cwd: '/test',
+            ignore: ['custom-ignore']
+        };
+
+        await FileScanner.findFiles(options);
+
+        const call = vi.mocked(globby).mock.calls[0];
+        const ignore = (call[1] as any).ignore;
+
+        expect(ignore).toContain('**/node_modules/**');
+        expect(ignore).toContain('custom-ignore');
+    });
+
+    it('should normalize paths to forward slashes', async () => {
+        const options = {
+            cwd: 'C:\\test\\path',
+            patterns: ['**\\*.ts']
+        };
+
+        await FileScanner.findFiles(options);
+
+        const call = vi.mocked(globby).mock.calls[1];
+        expect(call[0][0]).toBe('**/*.ts');
+        expect(call[1]?.cwd).toBe('C:/test/path');
+    });
+});

package/src/utils/scanner.ts

@@ -20,16 +20,22 @@ export class FileScanner {
     ];
 
     static async findFiles(options: ScannerOptions): Promise<string[]> {
-        return globby(options.patterns || this.DEFAULT_PATTERNS, {
-            cwd: options.cwd,
-            ignore: options.ignore || this.DEFAULT_IGNORE,
+        const patterns = (options.patterns || this.DEFAULT_PATTERNS).map(p => p.replace(/\\/g, '/'));
+        const userIgnore = options.ignore || [];
+        const ignore = [...new Set([...this.DEFAULT_IGNORE, ...userIgnore])].map(p => p.replace(/\\/g, '/'));
+        const normalizedCwd = options.cwd.replace(/\\/g, '/');
+
+        return globby(patterns, {
+            cwd: normalizedCwd,
+            ignore: ignore,
         });
     }
 
     static async readFiles(cwd: string, files: string[]): Promise<Map<string, string>> {
         const contents = new Map<string, string>();
         for (const file of files) {
-            const filePath = path.join(cwd, file);
+            const normalizedFile = file.replace(/\//g, path.sep);
+            const filePath = path.isAbsolute(normalizedFile) ? normalizedFile : path.join(cwd, normalizedFile);
             contents.set(file, await fs.readFile(filePath, 'utf-8'));
         }
         return contents;