Closed bibstha closed 4 years ago
I'm sorry that ripper-tags is slow for you!
- Can ripper-tags be run on jruby - parallel and merge the output?
You could definitely manually run ripper-tags in parallel processes which would utilize multiple CPU cores. You would have to split the git ls-files
list in N pieces and then start N ripper-tags processes that output to separate files. Finally, you would have to merge the separate tags files into one and sort it. No jruby is needed for this.
2. Is there a way to run only on updates files from the last time it was run?
No, but I suspect it could be done like so with a script:
tags
file,git ls-files
that are newer than the mtime,tags.new
,tags
filetags
and tags.new
into a single file.ripper-tags right now doesn't have any tools for regenerating and merging tags
file, but it could be useful to have the regeneration functionality built-in. Until then, you'll have to resort to your own scripts.
P.S. you don't need --exclude="*.js" --exclude="*.sql"
because ripper-tags only ever processes *.rb
files.
2. Is there a way to run only on updates files from the last time it was run?
Now there is a way: call ripper-tags --append <files...>
(where <files>
is the list of updated files) and it will append the existing tags
file with new tags from updated files. https://github.com/tmm1/ripper-tags/pull/96
Hi,
We have a large codebase and it takes ripper tags around 25 minutes to run through the whole codebase. Is there a way to make it faster?
Significant portion of the code doesn't change so when I re-run it, it's mostly duplicating the task.
Some thoughts:
I run it as