sites that slip through and whose search engines are added

[gregsadetsky]

• kijiji.ca (after submitting the search form)

[cxw42]

@gregsadetsky Is this issue going to hold the master list referenced in #7? If so, may I suggest a wiki page instead? In an issue, commenters can only edit their own comments, and it will be harder to pull the data for automated testing.

All that said, we've seen issues on:

• bleepingcomputer.com (has two search boxes; both should be spoiled)
• gcmap.com
• pouet.net (because action="search.php" in the HTML, so it doesn't match the selector)
• en.cppreference.com
• perldoc.perl.org

[cxw42]

... Although, now that I think about it, if folks post live links, and all we care about in testing is that at least one form on the linked page is spoiled, we could scrape this just as well as we could scrape links off a wiki page. We just wouldn't be able to easily delete the links later if the sites changed how they searched.

In a wiki, we could have a table with URL and expected number of results (e.g., pouet 1; bleepingcomputer 2). We could scrape the table and check for that number. Anyway, let me know what you think.

[gregsadetsky]

great thoughts on wiki. basing off the automated testing on the wiki links is a really great one, but I'd be concerned that opening up the wiki to anyone & using those urls for testing could spell trouble. but we're not there yet. definitely let's do wiki! I'll create the page, move the links here (thank you!) there and then close this issue.

[gregsadetsky]

https://github.com/gregsadetsky/chrome-dont-add-custom-search-engines/wiki/Sites-known-to-be-added-as-custom-search-engines