Re: adding to the LSR

[Graham Percival]

On Wed, Apr 29, 2009 at 03:17:46PM +0200, Mats Bengtsson wrote:
>
> Graham Percival wrote:
>>
>> I'm not certain this is necessary.  OK, it might be good to use
>> word-matching for "\octave " rather than string matching
>> "\octave*", but that's no unique to this occurrance.
>> (no, I don't know the proper regex terms for these)
>>
>> In general, we cannot guarantee that tweaks or tweak names will
>> work after convert-ly.
>>   
> In this case and in many other cases, it's very easy to fix the  
> convert-ly rules and doing so will help several other people. The  
> upgrade of LSR is an excellent opportunity to find and fix such problems  
> in convert-ly.

Ok.

Is it worth defining our own function
  replaceOnly("\\octave", ...)
which does
  re.sub("\\octave[?a-z,A-Z]", ...)
or whatever the regex was?

I imagine that almost every convert-ly rule would benefit from
"match this command, and only this command" rules, so this could
save typing.  And, more to the point, confusion amongst
non-regex-savvy people.

Cheers,
- Graham