dcrawl is a simple, but smart, multi-threaded web crawler for randomly gathering huge lists of unique domain names.
How it works?
dcrawl takes one site URL as input and detects all
<a href=...>links in the site's body. Each found link is put into the queue. Successively, each queued link is crawled in the same way, branching out to more URLs found in links on each site's body.
How smart crawling works:
- Branching out only to predefined number of links found per one hostname.
- Maximum number of allowed different hostnames per one domain (avoids subdomain crawling hell e.g. blogspot.com).
- Can be restarted with same list of domains - last saved domains are added to the URL queue.
- Crawls only sites that return text/html Content-Type in HEAD response.
- Retrieves site body of maximum 1MB size.
- Does not save inaccessible domains.
How to run?
go build dcrawl.go
./dcrawl -url http://wired.com -out ~/domain_lists/domains1.txt -t 8
__| _/________________ __ _ _| |
/ __ |/ ___\_ __ \__ \\ \/ \/ / |
/ /_/ \ \___| | \// __ \\ /| |__
\____ |\___ >__| (____ /\/\_/ |____/
\/ \/ \/ v.1.0
usage: dcrawl -url URL -out OUTPUT_FILE -t THREADS
maximum different subdomains for one domain (def. 10) (default 10)
maximum number of links to spider per hostname (def. 5) (default 5)
output file to save hostnames to
number of concurrent threads (def. 8) (default 8)
URL to start scraping from
verbose (default false)