feat: add global workspace search with regex and file filtering

2026-04-20 19:30:42 -04:00
parent d7c5f18544
commit b36391b520
4 changed files with 440 additions and 0 deletions
@@ -1,8 +1,32 @@
 import fs from "fs/promises";
+import { createReadStream } from "fs";
 import path from "path";
+import readline from "readline";

 const WORKSPACE_PATH = process.env.WORKSPACE_PATH || "/workspace";

+export interface SearchMatch {
+  file: string;
+  line: number;
+  column: number;
+  text: string;
+  matchLength: number;
+}
+
+export interface SearchResult {
+  matches: SearchMatch[];
+  totalMatches: number;
+  filesSearched: number;
+}
+
+interface SearchOptions {
+  regex?: boolean;
+  caseSensitive?: boolean;
+  includePattern?: string;
+  excludePattern?: string;
+  maxResults?: number;
+}
+
 export interface FileNode {
  name: string;
  path: string;
@@ -74,6 +98,111 @@ export async function deleteFile(
  await fs.unlink(fullPath);
 }

+const BINARY_EXTENSIONS = new Set([
+  ".mod", ".hak", ".bic", ".tlk", ".png", ".jpg", ".tga",
+  ".wav", ".bmu", ".ncs", ".gic",
+]);
+
+function matchesExtPattern(filePath: string, pattern: string): boolean {
+  if (pattern.startsWith("*.")) {
+    return filePath.endsWith(pattern.slice(1));
+  }
+  return filePath.endsWith(pattern);
+}
+
+async function walkFiles(dir: string, relativeTo: string): Promise<string[]> {
+  const results: string[] = [];
+  const entries = await fs.readdir(dir, { withFileTypes: true });
+  for (const entry of entries) {
+    const fullPath = path.join(dir, entry.name);
+    if (entry.isDirectory()) {
+      if (IGNORED_DIRS.has(entry.name)) continue;
+      results.push(...(await walkFiles(fullPath, relativeTo)));
+    } else {
+      const ext = path.extname(entry.name).toLowerCase();
+      if (!BINARY_EXTENSIONS.has(ext)) {
+        results.push(path.relative(relativeTo, fullPath));
+      }
+    }
+  }
+  return results;
+}
+
+export async function searchFiles(
+  repo: string,
+  query: string,
+  options: SearchOptions = {},
+): Promise<SearchResult> {
+  const maxResults = options.maxResults ?? 1000;
+  const repoRoot = resolveRepoPath(repo, "");
+  const files = await walkFiles(repoRoot, repoRoot);
+
+  const matches: SearchMatch[] = [];
+  let filesSearched = 0;
+
+  let pattern: RegExp | null = null;
+  if (options.regex) {
+    pattern = new RegExp(query, options.caseSensitive ? "g" : "gi");
+  }
+
+  for (const file of files) {
+    if (matches.length >= maxResults) break;
+
+    if (options.includePattern && !matchesExtPattern(file, options.includePattern)) continue;
+    if (options.excludePattern && matchesExtPattern(file, options.excludePattern)) continue;
+
+    const fullPath = path.join(repoRoot, file);
+    filesSearched++;
+
+    const rl = readline.createInterface({
+      input: createReadStream(fullPath, { encoding: "utf-8" }),
+      crlfDelay: Infinity,
+    });
+
+    let lineNum = 0;
+    for await (const line of rl) {
+      lineNum++;
+      if (matches.length >= maxResults) {
+        rl.close();
+        break;
+      }
+
+      if (pattern) {
+        pattern.lastIndex = 0;
+        let m: RegExpExecArray | null;
+        while ((m = pattern.exec(line)) !== null) {
+          matches.push({
+            file,
+            line: lineNum,
+            column: m.index,
+            text: line,
+            matchLength: m[0].length,
+          });
+          if (matches.length >= maxResults) break;
+        }
+      } else {
+        const haystack = options.caseSensitive ? line : line.toLowerCase();
+        const needle = options.caseSensitive ? query : query.toLowerCase();
+        let startIdx = 0;
+        let idx: number;
+        while ((idx = haystack.indexOf(needle, startIdx)) !== -1) {
+          matches.push({
+            file,
+            line: lineNum,
+            column: idx,
+            text: line,
+            matchLength: needle.length,
+          });
+          startIdx = idx + 1;
+          if (matches.length >= maxResults) break;
+        }
+      }
+    }
+  }
+
+  return { matches, totalMatches: matches.length, filesSearched };
+}
+
 function resolveRepoPath(repo: string, filePath: string): string {
  const repoRoot = path.resolve(path.join(WORKSPACE_PATH, "repos", repo));
  const resolved = path.resolve(repoRoot, filePath);