Searching "topicinfo" will find (almost?) all topics.
If you do this on twiki.org it will have a big impact on the server. This is a normal search, not a META search, so the main problem is that search does look into
text as well.
I am not sure if we should do anything with this. If we want to exclude some meta data we will pay with slower perfomance. Personally I suggest to let this be as a "not-to-be-fixed" item.
Just to pinpoint the problem: where would this hit performance?
The current websearch is based on the simple command line
, which doesn't care for the different components of a TWiki topic.
If you want to avoid this, you'd have to replace
by some "intelligent" function which knows about the text and
parts, and if you implement this function in Perl, it is likely to be significantly slower than a system
be passed a regex pattern...
Yes, it can. But that would be a brute-force regex which would be applied to every search. I'd bet that if you design a regex which safeguards a match against being in a
line (not an easy task in itself, or is it?) then a search based on this regex would be at least a factor of two slower than a plain
Besides speed people use SEARCH to search meta data as well. And we cannot break those applications.