Proposing licensecheck change

[James McCoy]

On Tue, Jul 21, 2015 at 07:31:07PM +0200, Dominique Dumont wrote:
> Currently, licensecheck -r uses find to scan a directory and accepts files 
> based on their suffix (i.e. accepts .c .h .cxx ...)
> 
> This list of suffixes is a big regexp that must be updated regurlarly. Still 
> some files are missed like config.guess.
> 
> Maintaining this regexp is not efficient.

True.  The current mechanism is tedious and either requires intervention
from the user are requesting updates to the regexp.

> I'd like to propose to use 'file' command to decide whether to scan a file or 
> not. All files of mime type 'text/*' and 'application/xml' would be scanned.
> (Note that file is already used to find the charset of each scanned file.)

If I understand correctly, you're proposing something like the attached,
untested patch.  Keep the -c switch so users can still filter by
extension if they like, but by default find all files and filter based
on file type.

> This way, I think that licensecheck would miss less source files.

Sounds good to me.

Cheers,
-- 
James
GPG Key: 4096R/331BA3DB 2011-12-05 James McCoy <jamessan at debian.org>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: licensecheck-text.diff
Type: text/x-diff
Size: 1609 bytes
Desc: not available
URL: <http://lists.alioth.debian.org/pipermail/devscripts-devel/attachments/20150727/eff7856c/attachment.diff>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 949 bytes
Desc: Digital signature
URL: <http://lists.alioth.debian.org/pipermail/devscripts-devel/attachments/20150727/eff7856c/attachment.sig>