If I search for something with multiple words in (so that it doesn't just do a tag search and instead uses /search?q=a+b ) then the results have raw html in where there should be links.

e.g. https://stackoverflow.com/search?q=perl+gzip ..second link ("Hidden features of Perl") shows a load of aref rel="nofollow" junk in the output which I wouldn't expect.

This doesn't happen on the tag search, so just typing "perl" into the search box doesn't trigger this problem.


I suggest adding single HTML tags to the search ignore list. For example searching for "assembly strong name" returns a lot of results that matched "<strong>". This has no semantic value and just clutters the results.

| improve this answer | |

Well, they're valid hits in the full text engine. We can hide them, but it's arguably better to show you, so you know why a certain post has the word "perl" a zillion times and thus ranks very highly in searches for that word.

(also, perl doesn't map to [perl] because it's not a top 40 tag in the system as of yet. auto-mapping of multi-word search terms to tags requires that the tag be in the top 40)

| improve this answer | |
  • Fair enough, cheers :) – Legooolas Jul 31 '09 at 10:30

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .