extern/se-scraper
1
0
Fork 1
mirror of https://github.com/NikolaiT/se-scraper.git synced 2024-11-21 23:23:07 +01:00
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #60 from Monibrand/fix/unusable-proxy-file-option

Browse Source 
fix(scrape-manager): proxy_file options can't be used with proxies default value
This commit is contained in:
Nikolai Tschacher 2020-02-13 20:02:55 +01:00 committed by GitHub
commit 5633b10e50
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 3 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
src/node_scraper.js
View File

@ -139,7 +139,7 @@ class ScrapeManager {
custom_func: null,
throw_on_detection: false,
// List of proxies to use ['socks5://78.94.172.42:1080', 'http://localhost:1080']
proxies: [],
proxies: null,
// a file with one proxy per line. Example:
// socks5://78.94.172.42:1080
// http://118.174.233.10:48400
@ -180,7 +180,7 @@ class ScrapeManager {
throw new Error('Either use a proxy_file or specify a proxy for all connections. Do not use both options.');
}
if (fs.existsSync(this.config.proxy_file)) {
if (this.config.proxy_file) {
this.config.proxies = read_keywords_from_file(this.config.proxy_file);
this.logger.info(`${this.config.proxies.length} proxies read from file.`);
}
@ -253,7 +253,7 @@ class ScrapeManager {
// if we have at least one proxy, always use CONCURRENCY_BROWSER
// and set maxConcurrency to this.config.proxies.length + 1
// else use whatever this.configuration was passed
if (this.config.proxies.length > 0) {
if (this.config.proxies && this.config.proxies.length > 0) {
this.config.puppeteer_cluster_config.concurrency = Cluster.CONCURRENCY_BROWSER;
// because we use real browsers, we ran out of memory on normal laptops