HTML generation issue in sqlite.org's search engine
(1) By anonymous on 2021-09-28 00:45:24
The following search query is generating oddly corrupted HTML for me as of the time of writing: https://www.sqlite.org/search?s=d&q=regexp It seems to be going off the rails around line 198 (as I write this) in the generated HTML, when what looks like it's meant to be part of a CSS color specification (`rgb(0, 0, 0);`) gets plunked into the page text: ``` <b> ... </b>0,0,0);" /> ``` Things go downhill from there; firefox ends up interpreting a fair chunk of HTML as attributes belonging to the next <path> element, and the same general pattern repeats until the end of the results list. I apologize if this is not the right place to report bugs with the webpages, and would be happy to follow pointers to that right place.
(2) By Richard Hipp (drh) on 2021-09-29 13:14:52 in reply to 1 [link]
Should be fixed now.
(3) By anonymous on 2021-09-29 14:27:52 in reply to 2 [link]
The search results page numbering also needs fixing, on the second page(or any other page after the first) of the search results page one is highlighted but not clickable.
(4) By Richard Hipp (drh) on 2021-09-29 14:52:46 in reply to 3 [link]
I was not able to reproduce the anomaly with a quick check. Do you have an example of a search that generates the erroneous behavior?
(5) By anonymous on 2021-09-29 16:09:02 in reply to 4 [link]
I'm not able to reproduce it exactly after your fix went live, but the link is <https://www.sqlite.org/search?q=sqlite&i=20>.
(6) By anonymous on 2021-09-29 18:19:35 in reply to 3 [link]
[For clarity, I am the original poster and the above anonymous is not] My recollection was that my broken search acted like this, but those numbers are all after the point at which the markup went off the rails. (I don't see multiple pages at all anymore for my original query, which appears to be as it should have been in the first place.)