NinjaCz / crawley

The unix-way web crawler

crawley

Crawls web pages and prints any link it can find.

features

• fast html SAX-parser (powered by golang.org/x/net/html)
• small (below 1500 SLOC), idiomatic, 100% test covered codebase
• grabs most of useful resources urls (pics, videos, audios, forms, etc...)
• found urls are streamed to stdout and guranteed to be unique (with fragments omitted)
• scan depth (limited by starting host and path, by default - 0) can be configured
• can crawl rules and sitemaps from robots.txt
• brute mode - scan html comments for urls (this can lead to bogus results)
• make use of HTTP_PROXY / HTTPS_PROXY environment values + handles proxy auth
• directory-only scan mode (aka fast-scan)
• user-defined cookies, in curl-compatible format (i.e. -cookie "ONE=1; TWO=2" -cookie "ITS=ME" -cookie @cookie-file)
• user-defined headers, same as curl: -header "ONE: 1" -header "TWO: 2" -header @headers-file
• tag filter - allow to specify tags to crawl for (single: -tag a -tag form, multiple: -tag a,form, or mixed)
• url ignore - allow to ignore urls with matched substrings from crawling (i.e.: '-ignore logout')
• js parser - extract api endpoints from js files, this done by regexp, so results can be messy

examples

# print all links from first page:
crawley http://some-test.site

# print all js files and api endpoints:
crawley -depth -1 -tag script -js http://some-test.site

# print all endpoints from js:
crawley -js http://some-test.site/app.js

# download all png images from site:
crawley -depth -1 -tag img http://some-test.site | grep '\.png$' | wget -i -

# fast directory traversal:
crawley -headless -delay 0 -depth -1 -dirs only http://some-test.site

installation

• binaries / deb / rpm for Linux, FreeBSD, macOS and Windows.
• archlinux you can use your favourite AUR helper to install it, e. g. paru -S crawley-bin.

usage

crawley [flags] url

possible flags with default values:

-brute
    scan html comments
-cookie value
    extra cookies for request, can be used multiple times, accept files with '@'-prefix
-delay duration
    per-request delay (0 - disable) (default 150ms)
-depth int
    scan depth (-1 - unlimited)
-dirs string
    policy for non-resource urls: show / hide / only (default "show")
-header value
    extra headers for request, can be used multiple times, accept files with '@'-prefix
-headless
    disable pre-flight HEAD requests
-help
    this flags (and their defaults) description
-ignore value
    patterns (in urls) to be ignored in crawl process
-js
    scan js files for endpoints
-proxy-auth string
    credentials for proxy: user:password
-robots string
    policy for robots.txt: ignore / crawl / respect (default "ignore")
-silent
    suppress info and error messages in stderr
-skip-ssl
    skip ssl verification
-tag value
    tags filter, single or comma-separated tag names allowed
-user-agent string
    user-agent string
-version
    show version
-workers int
    number of workers (default - number of CPU cores)

flags autocompletion

Crawley can handle flags autocompletion in bash and zsh via complete:

complete -C "/full-path-to/bin/crawley" crawley