From: Michael Weller <michael@...>
Date: 2004-10-08T22:38:51+09:00
Subject: Re: [ANN] svn_auto_add

Jonathan Paisley wrote:
> On Fri, 08 Oct 2004 21:31:21 +0900, Michael Weller wrote:
> 
> 
>>I wrote a little script for svn users. It will detect any file added to 
>>a working copy which svn doesn't know about. In other words: it calls 
>>'svn add' for any newly added file.
>>I thought I share it,just in case it might be helpful to somebody... (I 
>>actually don't think so because I'm a Nuby and everybody could write 
>>this script, but you never know :-)
> 
> 
>   [snip code]
> 
> 
>>If somebody reads the script and thinks "Why is he doing this that way?"
>>feel free to comment, I'm always open for improvements...
> 
> 
> I wonder whether it might be simpler to just ask the command line tool
> what files it doesn't know about (which will take into account the
> relevant ignore settings for *.o, *.class etc). For example:
> 
> $ svn status
> ?      somefile
> ?      otherfile
> M      modifiedfile
> 
> By looking for lines that begin with a '?', we get precisely the list that
> needs to be added. These lines could be turned into the appropriate 'svn
> add' commands.
> 
> One way (not particularly pleasant to look at) using ruby would be:
> 
> $ svn status | ruby -ne 'puts $_[7..-2] if $_[0] == ??'
> 
> to actually run the 'svn add' command you could use:
> 
> $ svn status | ruby -e 'system("svn","add",$_[7..-2]) if $_[0] == ??'
> 
> I think I would normally do it with a shell pipeline, since I find it
> easier to compose the simpler steps than come up with the ruby script for
> the whole thing:
> 
> $ svn add `svn status | grep '^?' | cut -c8-`
>  (assuming no spaces in the names of the files to be added)
> 
> Hope this makes sense!
> 

Actually it makes a lot of sense... Much more than my approach :-((
But like I said I'm a nuby and happy that I got my script working my 
(much more complicated) way...
Yes, shell pipes would be great :-((

Thanks for your input!

Michael