search Google and extract results directly. skip all the click-through links and other sketchiness
|size (curr.)||13 kB|
search Google and extract result urls directly. skip all the click-through links and other sketchiness
|command line usage||script usage|
optional arguments: -o, --offset page # to start from (default 0) -p, --pages number of pages to grab (default 1) -t, --time-window how far back in time to look (default all-time) -j, --exclude-junk skip youtube, FB, quora, pinterest links (default true)
time window follows a syntax used by google’s
tbs parameter with the
qdr option (read someone explain how it works here)
.edusites within the past
python3 degoogle.py "site:edu filetype:txt" -t m3
- with one
queryin a loop to perform multiple searches:
from degoogle import dg degoogler = dg() queries = ["site:edu", "site:gov", "filetype:txt"] for query in queries: print(query) degoogler.query = query results = degoogler.run() for result in results: print(result) print()
- one liner: make
dginstance (query set in constructor), search across 5months, format + print results, end:
[print(result['desc'],": ",result['url'],"\n") for result in (dg("intext:'begin rsa' site:*.edu.*",time_window="m5").run())]
this is an experiment meant to have benefits for both user privacy and broadened osint capabilities
idea: when you search google normally, your results will appear to be direct links to the target site, but what you’re really getting is more like this:
by using a link like this to access
example.com, you are providing more information about yourself than necessary. they don’t need to know what is done with the results after they are served; their job is done, the click-through is merely a suggestion.
example.com can identify that google was your referer; the page you’re clicking through from includes your search query as part of the URL, so they probably know what you searched for too. whether or not they’re looking at it is another question
if you navigate to
example.com directly, without a referer like google,
example.com knows that you have visited, but not where you came from, and google knows even less - even if you originally found out about
example.com by spotting it in a google search at some point.
google will obviously always know that the search took place, and which results were served, but I prefer to access results directly and not give anyone that click-through confirmation.
there are also utility benefits here for dorking - you might find a super juicy link on page 10 with a bunch of strange parameters cached, but when you follow google’s click-through you’re redirected to an index page or 404 without even having a chance to copy the link from the result to research more.
there are times where visually inspecting a URL is just as valuable as accessing the link. this tool ganks the good stuff, URL decodes and normalizes, and returns just the scraped URL + description