[TxMt] Find/Grep In Project - beta tester wanted for a new trial: Hypersearch

Hans-Jörg Bibiko bibiko at eva.mpg.de
Sat Sep 15 19:50:36 UTC 2007


Dear all,

in the past there're some mails regarding to Find In Project.

Henrik Nyh wrote the wonderful command 'Grep In Project'. This  
inspired me to go one step further.

Instead of using 'grep' in fixed string mode I used 'onigrep'. I  
wrote the command line tool 'onigrep' in C which includes Kosako
's Oniguruma regexp engine (5.8.0) and works ONLY on utf-8/ASCII text  
files.
You can find it within the bundle in folder ./bin/ppc or /bin/i386.  
More details on that 'onigrep --help'.
(By myself I copied onigrep to /usr/bin to use it system-wide ;) but  
please note: onigrep is work in progress too!)

The matches, also for regexp!, are highlighted as usual.

Furthermore I used a self-made dialog with history list etc. The  
search function is working not only in a project but also for a  
single document.

If you select 'Search in the entire project' it will exclude the  
paths '*/.svn' '*/vendor/rails' '*/build'.

If you select the root folder of your project in the drawer, and you  
select 'Search in Selection' it will search in all subfolders.

The option 'Ignore diacritics' is up to now only working if you saves  
the entire text according to the canonical Unicode decomposition.  
Then you can search for 'arger' or 'facade' or た and you will get  
words like 'Ärger', 'façade', で. A command line tool which does it  
will be included in the near future.

So, if someone has a bit time I would be appreciated if s/he can test  
it. My main question is whether it runs also on Intel Macs. I  
compiled 'onigrep' for both architectures, and I don't know whether  
my ruby code got rid of it.


Some words about the speed of 'onigrep'. Of course, to deal with  
utf-8 regexp, well it takes a bit more time. I compared 'onigrep'  
with 'LC_ALL=en_US.UTF-8 grep' and the search speed was almost  
identical, but 'onigrep' is 'ignore case' safe and it uses the  
Oniguruma syntax ;)

You can download the BETA VERSION! of the Hypersearch here:

http://email.eva.mpg.de/~bibiko/dt/Hypersearch.tmbundle.zip (256k)

Any other suggestions, bugs, etc. are highly welcomed for both the  
Hypersearch bundle and for onigrep.

Many thanks in advance

Hans





More information about the textmate mailing list