@apmantza/greedysearch-pi 1.8.0 → 1.8.2

package/src/search/chrome.mjs

@@ -1,223 +1,223 @@
-// src/search/chrome.mjs — Chrome launch, probe, port file management, and CDP wrapper
-//
-// Extracted from search.mjs to reduce file complexity.
-// Also used by coding-task.mjs (via import).
-//
-// cdp() is re-exported from extractors/common.mjs to avoid duplication.
-
-import { spawn } from "node:child_process";
-import {
-	existsSync,
-	readFileSync,
-	renameSync,
-	unlinkSync,
-	writeFileSync,
-} from "node:fs";
-import http from "node:http";
-import { join } from "node:path";
-
-import { GREEDY_PORT, ACTIVE_PORT_FILE, PAGES_CACHE } from "./constants.mjs";
-import { cdp as _cdp } from "../../extractors/common.mjs";
-
-const __dir = import.meta.dirname || new URL(".", import.meta.url).pathname.replace(/^\/([A-Z]:)/, "$1");
-
-/** Re-export cdp() from the canonical location in extractors/common.mjs */
-export const cdp = _cdp;
-
-export async function getAnyTab() {
-	const list = await cdp(["list"]);
-	const first = list.split("\n")[0];
-	if (!first) throw new Error("No Chrome tabs found");
-	return first.slice(0, 8);
-}
-
-export async function openNewTab() {
-	const anchor = await getAnyTab();
-	const raw = await cdp([
-		"evalraw",
-		anchor,
-		"Target.createTarget",
-		'{"url":"about:blank"}',
-	]);
-	const { targetId } = JSON.parse(raw);
-	return targetId;
-}
-
-export async function activateTab(targetId) {
-	try {
-		const anchor = await getAnyTab();
-		await cdp([
-			"evalraw",
-			anchor,
-			"Target.activateTarget",
-			JSON.stringify({ targetId }),
-		]);
-	} catch {
-		// best-effort
-	}
-}
-
-export async function closeTab(targetId) {
-	try {
-		const anchor = await getAnyTab();
-		await cdp([
-			"evalraw",
-			anchor,
-			"Target.closeTarget",
-			JSON.stringify({ targetId }),
-		]);
-	} catch {
-		/* best-effort */
-	}
-}
-
-export async function closeTabs(targetIds = []) {
-	for (const tid of targetIds) {
-		if (!tid) continue;
-		await closeTab(tid);
-	}
-	if (targetIds.length > 0) {
-		await new Promise((r) => setTimeout(r, 300));
-		await cdp(["list"]).catch(() => null);
-	}
-}
-
-export function getFullTabFromCache(engine, engineDomains) {
-	try {
-		if (!existsSync(PAGES_CACHE)) return null;
-		const pages = JSON.parse(readFileSync(PAGES_CACHE, "utf8"));
-		const found = pages.find((p) => p.url.includes(engineDomains[engine]));
-		return found ? found.targetId : null;
-	} catch {
-		return null;
-	}
-}
-
-export function probeGreedyChrome(timeoutMs = 3000) {
-	return new Promise((resolve) => {
-		const req = http.get(
-			`http://localhost:${GREEDY_PORT}/json/version`,
-			(res) => {
-				res.resume();
-				resolve(res.statusCode === 200);
-			},
-		);
-		req.on("error", () => resolve(false));
-		req.setTimeout(timeoutMs, () => {
-			req.destroy();
-			resolve(false);
-		});
-	});
-}
-
-export async function refreshPortFile() {
-	const LOCK_FILE = `${ACTIVE_PORT_FILE}.lock`;
-	const TEMP_FILE = `${ACTIVE_PORT_FILE}.tmp`;
-	const LOCK_STALE_MS = 5000;
-	const LOCK_WAIT_MS = 1000;
-
-	// File-based lock with exclusive create + stale lock recovery
-	const lockAcquired = await new Promise((resolve) => {
-		const start = Date.now();
-		const tryLock = () => {
-			try {
-				const payload = JSON.stringify({ pid: process.pid, ts: Date.now() });
-				writeFileSync(LOCK_FILE, payload, { encoding: "utf8", flag: "wx" });
-				resolve(true);
-			} catch (e) {
-				if (e?.code !== "EEXIST") {
-					if (Date.now() - start < LOCK_WAIT_MS) {
-						setTimeout(tryLock, 50);
-					} else {
-						resolve(false);
-					}
-					return;
-				}
-
-				try {
-					const lockRaw = readFileSync(LOCK_FILE, "utf8").trim();
-					const parsed = lockRaw.startsWith("{")
-						? JSON.parse(lockRaw)
-						: { ts: Number(lockRaw) };
-					const lockTime = Number(parsed?.ts) || 0;
-
-					if (lockTime > 0 && Date.now() - lockTime > LOCK_STALE_MS) {
-						try {
-							unlinkSync(LOCK_FILE);
-						} catch {}
-					}
-
-					if (Date.now() - start < LOCK_WAIT_MS) {
-						setTimeout(tryLock, 50);
-					} else {
-						resolve(false);
-					}
-				} catch {
-					if (Date.now() - start < LOCK_WAIT_MS) {
-						setTimeout(tryLock, 50);
-					} else {
-						resolve(false);
-					}
-				}
-			}
-		};
-		tryLock();
-	});
-
-	try {
-		const body = await new Promise((res, rej) => {
-			const req = http.get(
-				`http://localhost:${GREEDY_PORT}/json/version`,
-				(r) => {
-					let b = "";
-					r.on("data", (d) => (b += d));
-					r.on("end", () => res(b));
-				},
-			);
-			req.on("error", rej);
-			req.setTimeout(3000, () => {
-				req.destroy();
-				rej(new Error("timeout"));
-			});
-		});
-		const { webSocketDebuggerUrl } = JSON.parse(body);
-		const wsPath = new URL(webSocketDebuggerUrl).pathname;
-
-		// Atomic write: write to temp file, then rename
-		if (lockAcquired) {
-			writeFileSync(TEMP_FILE, `${GREEDY_PORT}\n${wsPath}`, "utf8");
-			try {
-				unlinkSync(ACTIVE_PORT_FILE);
-			} catch {}
-			renameSync(TEMP_FILE, ACTIVE_PORT_FILE);
-		}
-	} catch {
-		/* best-effort — launch.mjs already wrote the file on first start */
-	} finally {
-		if (lockAcquired) {
-			try {
-				unlinkSync(LOCK_FILE);
-			} catch {}
-		}
-	}
-}
-
-export async function ensureChrome() {
-	const ready = await probeGreedyChrome();
-	if (!ready) {
-		process.stderr.write(
-			`GreedySearch Chrome not running on port ${GREEDY_PORT} — auto-launching...\n`,
-		);
-		await new Promise((resolve, reject) => {
-			const proc = spawn("node", [join(__dir, "..", "..", "bin", "launch.mjs")], {
-				stdio: ["ignore", process.stderr, process.stderr],
-			});
-			proc.on("close", (code) =>
-				code === 0 ? resolve() : reject(new Error("launch.mjs failed")),
-			);
-		});
-	} else {
-		// Chrome already running — refresh the port file
-		await refreshPortFile();
-	}
+// src/search/chrome.mjs — Chrome launch, probe, port file management, and CDP wrapper
+//
+// Extracted from search.mjs to reduce file complexity.
+// Also used by coding-task.mjs (via import).
+//
+// cdp() is re-exported from extractors/common.mjs to avoid duplication.
+
+import { spawn } from "node:child_process";
+import {
+	existsSync,
+	readFileSync,
+	renameSync,
+	unlinkSync,
+	writeFileSync,
+} from "node:fs";
+import http from "node:http";
+import { join } from "node:path";
+
+import { GREEDY_PORT, ACTIVE_PORT_FILE, PAGES_CACHE } from "./constants.mjs";
+import { cdp as _cdp } from "../../extractors/common.mjs";
+
+const __dir = import.meta.dirname || new URL(".", import.meta.url).pathname.replace(/^\/([A-Z]:)/, "$1");
+
+/** Re-export cdp() from the canonical location in extractors/common.mjs */
+export const cdp = _cdp;
+
+export async function getAnyTab() {
+	const list = await cdp(["list"]);
+	const first = list.split("\n")[0];
+	if (!first) throw new Error("No Chrome tabs found");
+	return first.slice(0, 8);
+}
+
+export async function openNewTab() {
+	const anchor = await getAnyTab();
+	const raw = await cdp([
+		"evalraw",
+		anchor,
+		"Target.createTarget",
+		'{"url":"about:blank"}',
+	]);
+	const { targetId } = JSON.parse(raw);
+	return targetId;
+}
+
+export async function activateTab(targetId) {
+	try {
+		const anchor = await getAnyTab();
+		await cdp([
+			"evalraw",
+			anchor,
+			"Target.activateTarget",
+			JSON.stringify({ targetId }),
+		]);
+	} catch {
+		// best-effort
+	}
+}
+
+export async function closeTab(targetId) {
+	try {
+		const anchor = await getAnyTab();
+		await cdp([
+			"evalraw",
+			anchor,
+			"Target.closeTarget",
+			JSON.stringify({ targetId }),
+		]);
+	} catch {
+		/* best-effort */
+	}
+}
+
+export async function closeTabs(targetIds = []) {
+	for (const tid of targetIds) {
+		if (!tid) continue;
+		await closeTab(tid);
+	}
+	if (targetIds.length > 0) {
+		await new Promise((r) => setTimeout(r, 300));
+		await cdp(["list"]).catch(() => null);
+	}
+}
+
+export function getFullTabFromCache(engine, engineDomains) {
+	try {
+		if (!existsSync(PAGES_CACHE)) return null;
+		const pages = JSON.parse(readFileSync(PAGES_CACHE, "utf8"));
+		const found = pages.find((p) => p.url.includes(engineDomains[engine]));
+		return found ? found.targetId : null;
+	} catch {
+		return null;
+	}
+}
+
+export function probeGreedyChrome(timeoutMs = 3000) {
+	return new Promise((resolve) => {
+		const req = http.get(
+			`http://localhost:${GREEDY_PORT}/json/version`,
+			(res) => {
+				res.resume();
+				resolve(res.statusCode === 200);
+			},
+		);
+		req.on("error", () => resolve(false));
+		req.setTimeout(timeoutMs, () => {
+			req.destroy();
+			resolve(false);
+		});
+	});
+}
+
+export async function refreshPortFile() {
+	const LOCK_FILE = `${ACTIVE_PORT_FILE}.lock`;
+	const TEMP_FILE = `${ACTIVE_PORT_FILE}.tmp`;
+	const LOCK_STALE_MS = 5000;
+	const LOCK_WAIT_MS = 1000;
+
+	// File-based lock with exclusive create + stale lock recovery
+	const lockAcquired = await new Promise((resolve) => {
+		const start = Date.now();
+		const tryLock = () => {
+			try {
+				const payload = JSON.stringify({ pid: process.pid, ts: Date.now() });
+				writeFileSync(LOCK_FILE, payload, { encoding: "utf8", flag: "wx" });
+				resolve(true);
+			} catch (e) {
+				if (e?.code !== "EEXIST") {
+					if (Date.now() - start < LOCK_WAIT_MS) {
+						setTimeout(tryLock, 50);
+					} else {
+						resolve(false);
+					}
+					return;
+				}
+
+				try {
+					const lockRaw = readFileSync(LOCK_FILE, "utf8").trim();
+					const parsed = lockRaw.startsWith("{")
+						? JSON.parse(lockRaw)
+						: { ts: Number(lockRaw) };
+					const lockTime = Number(parsed?.ts) || 0;
+
+					if (lockTime > 0 && Date.now() - lockTime > LOCK_STALE_MS) {
+						try {
+							unlinkSync(LOCK_FILE);
+						} catch {}
+					}
+
+					if (Date.now() - start < LOCK_WAIT_MS) {
+						setTimeout(tryLock, 50);
+					} else {
+						resolve(false);
+					}
+				} catch {
+					if (Date.now() - start < LOCK_WAIT_MS) {
+						setTimeout(tryLock, 50);
+					} else {
+						resolve(false);
+					}
+				}
+			}
+		};
+		tryLock();
+	});
+
+	try {
+		const body = await new Promise((res, rej) => {
+			const req = http.get(
+				`http://localhost:${GREEDY_PORT}/json/version`,
+				(r) => {
+					let b = "";
+					r.on("data", (d) => (b += d));
+					r.on("end", () => res(b));
+				},
+			);
+			req.on("error", rej);
+			req.setTimeout(3000, () => {
+				req.destroy();
+				rej(new Error("timeout"));
+			});
+		});
+		const { webSocketDebuggerUrl } = JSON.parse(body);
+		const wsPath = new URL(webSocketDebuggerUrl).pathname;
+
+		// Atomic write: write to temp file, then rename
+		if (lockAcquired) {
+			writeFileSync(TEMP_FILE, `${GREEDY_PORT}\n${wsPath}`, "utf8");
+			try {
+				unlinkSync(ACTIVE_PORT_FILE);
+			} catch {}
+			renameSync(TEMP_FILE, ACTIVE_PORT_FILE);
+		}
+	} catch {
+		/* best-effort — launch.mjs already wrote the file on first start */
+	} finally {
+		if (lockAcquired) {
+			try {
+				unlinkSync(LOCK_FILE);
+			} catch {}
+		}
+	}
+}
+
+export async function ensureChrome() {
+	const ready = await probeGreedyChrome();
+	if (!ready) {
+		process.stderr.write(
+			`GreedySearch Chrome not running on port ${GREEDY_PORT} — auto-launching...\n`,
+		);
+		await new Promise((resolve, reject) => {
+			const proc = spawn("node", [join(__dir, "..", "..", "bin", "launch.mjs")], {
+				stdio: ["ignore", process.stderr, process.stderr],
+			});
+			proc.on("close", (code) =>
+				code === 0 ? resolve() : reject(new Error("launch.mjs failed")),
+			);
+		});
+	} else {
+		// Chrome already running — refresh the port file
+		await refreshPortFile();
+	}
 }

package/src/search/constants.mjs

@@ -1,38 +1,38 @@
-// src/search/constants.mjs — Shared constants for GreedySearch search pipeline
-
-import { tmpdir } from "node:os";
-
-export const GREEDY_PORT = 9222;
-export const GREEDY_PROFILE_DIR = `${tmpdir().replace(/\\/g, "/")}/greedysearch-chrome-profile`;
-export const ACTIVE_PORT_FILE = `${GREEDY_PROFILE_DIR}/DevToolsActivePort`;
-export const PAGES_CACHE = `${tmpdir().replace(/\\/g, "/")}/cdp-pages.json`;
-
-export const ALL_ENGINES = ["perplexity", "bing", "google"];
-
-export const ENGINE_DOMAINS = {
-	perplexity: "perplexity.ai",
-	bing: "copilot.microsoft.com",
-	google: "google.com",
-	gemini: "gemini.google.com",
-};
-
-export const ENGINES = {
-	perplexity: "perplexity.mjs",
-	pplx: "perplexity.mjs",
-	p: "perplexity.mjs",
-	bing: "bing-copilot.mjs",
-	copilot: "bing-copilot.mjs",
-	b: "bing-copilot.mjs",
-	google: "google-ai.mjs",
-	g: "google-ai.mjs",
-	gemini: "gemini.mjs",
-	gem: "gemini.mjs",
-};
-
-export const SOURCE_FETCH_CONCURRENCY = Math.max(
-	1,
-	parseInt(process.env.GREEDY_FETCH_CONCURRENCY || "2", 10) || 2,
-);
-
-// Tell cdp.mjs to prefer the GreedySearch Chrome profile's DevToolsActivePort
+// src/search/constants.mjs — Shared constants for GreedySearch search pipeline
+
+import { tmpdir } from "node:os";
+
+export const GREEDY_PORT = 9222;
+export const GREEDY_PROFILE_DIR = `${tmpdir().replace(/\\/g, "/")}/greedysearch-chrome-profile`;
+export const ACTIVE_PORT_FILE = `${GREEDY_PROFILE_DIR}/DevToolsActivePort`;
+export const PAGES_CACHE = `${tmpdir().replace(/\\/g, "/")}/cdp-pages.json`;
+
+export const ALL_ENGINES = ["perplexity", "bing", "google"];
+
+export const ENGINE_DOMAINS = {
+	perplexity: "perplexity.ai",
+	bing: "copilot.microsoft.com",
+	google: "google.com",
+	gemini: "gemini.google.com",
+};
+
+export const ENGINES = {
+	perplexity: "perplexity.mjs",
+	pplx: "perplexity.mjs",
+	p: "perplexity.mjs",
+	bing: "bing-copilot.mjs",
+	copilot: "bing-copilot.mjs",
+	b: "bing-copilot.mjs",
+	google: "google-ai.mjs",
+	g: "google-ai.mjs",
+	gemini: "gemini.mjs",
+	gem: "gemini.mjs",
+};
+
+export const SOURCE_FETCH_CONCURRENCY = Math.max(
+	1,
+	parseInt(process.env.GREEDY_FETCH_CONCURRENCY || "2", 10) || 2,
+);
+
+// Tell cdp.mjs to prefer the GreedySearch Chrome profile's DevToolsActivePort
 process.env.CDP_PROFILE_DIR = GREEDY_PROFILE_DIR;

package/src/search/defaults.mjs

@@ -1,15 +1,15 @@
-// src/search/defaults.mjs — Shared default values and timeouts
-//
-// Centralizes magic numbers used across the codebase.
-// Import from here instead of hardcoding values.
-
-export const DEFAULTS = {
-	CDP_TIMEOUT: 30000,             // Default CDP command timeout (ms)
-	CDP_TIMEOUT_SHORT: 15000,       // Short CDP timeout for search operations (ms)
-	NAV_TIMEOUT: 35000,             // Navigation timeout (ms)
-	STREAM_TIMEOUT: 30000,          // Stream completion timeout (ms)
-	COPY_TIMEOUT: 60000,            // Copy button appearance timeout (ms)
-	CODING_TASK_TIMEOUT: 180000,    // Coding task max duration (ms)
-	MAX_SOURCE_FETCH: 10,           // Max concurrent source fetches
-	DESCRIPTION_MAX_LENGTH: 300,    // Max answer length in truncated mode
+// src/search/defaults.mjs — Shared default values and timeouts
+//
+// Centralizes magic numbers used across the codebase.
+// Import from here instead of hardcoding values.
+
+export const DEFAULTS = {
+	CDP_TIMEOUT: 30000,             // Default CDP command timeout (ms)
+	CDP_TIMEOUT_SHORT: 15000,       // Short CDP timeout for search operations (ms)
+	NAV_TIMEOUT: 35000,             // Navigation timeout (ms)
+	STREAM_TIMEOUT: 30000,          // Stream completion timeout (ms)
+	COPY_TIMEOUT: 60000,            // Copy button appearance timeout (ms)
+	CODING_TASK_TIMEOUT: 180000,    // Coding task max duration (ms)
+	MAX_SOURCE_FETCH: 10,           // Max concurrent source fetches
+	DESCRIPTION_MAX_LENGTH: 300,    // Max answer length in truncated mode
 };

package/src/search/engines.mjs

@@ -9,7 +9,9 @@ import { ENGINES, GREEDY_PROFILE_DIR } from "./constants.mjs";
 
 export { ENGINES };
 
-const __dir = import.meta.dirname || new URL(".", import.meta.url).pathname.replace(/^\/([A-Z]:)/, "$1");
+const __dir =
+	import.meta.dirname ||
+	new URL(".", import.meta.url).pathname.replace(/^\/([A-Z]:)/, "$1");
 
 export function runExtractor(
 	script,
@@ -17,6 +19,7 @@ export function runExtractor(
 	tabPrefix = null,
 	short = false,
 	timeoutMs = null,
+	locale = null,
 ) {
 	// Gemini is slower - use longer timeout
 	if (timeoutMs === null) {
@@ -25,6 +28,7 @@ export function runExtractor(
 	const extraArgs = [
 		...(tabPrefix ? ["--tab", tabPrefix] : []),
 		...(short ? ["--short"] : []),
+		...(locale ? ["--locale", locale] : []),
 	];
 	return new Promise((resolve, reject) => {
 		const proc = spawn(
@@ -55,4 +59,4 @@ export function runExtractor(
 			}
 		});
 	});
-}
+}