yujunjun/firecrawl
1
0
Fork 0
mirror of https://github.com/mendableai/firecrawl.git synced 2025-11-03 19:43:01 +00:00
Code Issues Packages Projects Releases Wiki Activity

fix(crawl-redis): normalize URL before locking

Browse Source
This commit is contained in:
Gergő Móricz 2024-10-01 19:00:33 +02:00
parent 96245e387d
commit fe721fffbe
1 changed files with 24 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

24
apps/api/src/lib/crawl-redis.ts
View File

@ -1,5 +1,6 @@
import { WebCrawler } from "../scraper/WebScraper/crawler"; import { WebCrawler } from "../scraper/WebScraper/crawler";
import { redisConnection } from "../services/queue-service"; import { redisConnection } from "../services/queue-service";
import { Logger } from "./logger";
export type StoredCrawl = { export type StoredCrawl = {
originUrl: string; originUrl: string;
@ -88,6 +89,16 @@ export async function lockURL(id: string, sc: StoredCrawl, url: string): Promise
return false; return false;
} }
} }
try {
const urlO = new URL(url);
urlO.search = "";
urlO.hash = "";
url = urlO.href;
} catch (error) {
Logger.warn("Failed to normalize URL " + JSON.stringify(url) + ": " + error);
}
const res = (await redisConnection.sadd("crawl:" + id + ":visited", url)) !== 0 const res = (await redisConnection.sadd("crawl:" + id + ":visited", url)) !== 0
await redisConnection.expire("crawl:" + id + ":visited", 24 * 60 * 60, "NX"); await redisConnection.expire("crawl:" + id + ":visited", 24 * 60 * 60, "NX");
return res; return res;
@ -95,6 +106,19 @@ export async function lockURL(id: string, sc: StoredCrawl, url: string): Promise
/// NOTE: does not check limit. only use if limit is checked beforehand e.g. with sitemap /// NOTE: does not check limit. only use if limit is checked beforehand e.g. with sitemap
export async function lockURLs(id: string, urls: string[]): Promise<boolean> { export async function lockURLs(id: string, urls: string[]): Promise<boolean> {
urls = urls.map(url => {
try {
const urlO = new URL(url);
urlO.search = "";
urlO.hash = "";
return urlO.href;
} catch (error) {
Logger.warn("Failed to normalize URL " + JSON.stringify(url) + ": " + error);
}
return url;
});
const res = (await redisConnection.sadd("crawl:" + id + ":visited", ...urls)) !== 0 const res = (await redisConnection.sadd("crawl:" + id + ":visited", ...urls)) !== 0
await redisConnection.expire("crawl:" + id + ":visited", 24 * 60 * 60, "NX"); await redisConnection.expire("crawl:" + id + ":visited", 24 * 60 * 60, "NX");
return res; return res;