Speed up searching for exclusions

[jmarceli]

Thanks for that tool, exactly something I was looking for.

One thing that may make it even better would be skipping some common folders while searching for development dependencies like:

• ~/Library
• ~/Applications
• ~/Music etc.

[rowanbeentje]

https://github.com/stevegrunwell/asimov/pull/18 excludes ~/Library as part of other speedup work - we can add others in the same way I suppose, but I wonder if there's a point where the exclusion checks make the overall process slower than just quickly scanning directories!

[jmarceli]

I was thinking rather about exclusion directly through find, something like (according to https://stackoverflow.com/questions/4210042/how-to-exclude-a-directory-in-find-command):

find ~ -type d -name "${parts[0]}" -not -path "/Users/*/Library/*"

I think that this shouldn't hit overall performance (I'm not a find expert so I may be wrong).

Also ~/.Trash exclusion would be nice, because even if I have trash excluded in TimeMachine I still have to wait for Asimov to scan through.

Regarding the speed. Maybe excluding nested node_modules with find rather than inside the dependency_file_exists() would save some time? What I mean is:

 find ~ -type d -name "${parts[0]}" -not -path "*/node_modules/*/node_modules*"

[rowanbeentje]

https://github.com/stevegrunwell/asimov/pull/17 (included in https://github.com/stevegrunwell/asimov/pull/18 as well!) excludes subdeps, and yeah, that -not -path is the approach used to exclude ~/Library - I just can't help feeling that will have a cost.

(Trashes folders are a great one though)

[jmarceli]

I'm still waiting for ~/.Trash to complete processing (about 20min) despite the fact that I've excluded trash in TM anyway...

The first run of Asimov takes about 2:30 hours, but the overall effect is great. Thanks for this script.

[FlorianWendelborn]

Also ~/.Trash exclusion would be nice, because even if I have trash excluded in TimeMachine I still have to wait for Asimov to scan through.

Wouldn't it be better to just automatically skip all folders that are excluded from time machine then?

[muuvmuuv]

Yep, why not exclude everything inside the TimeMaschine ignore list from the find process? I mean the ignore list that you can manually edit in the TimeMaschin pane in settings.