[SVN] Re: Align Source

[Allan Odgaard]

I changed the /,/ pattern to /(,)(?!$)/ to avoid inserting trailing whitespace, e.g.:

  $headers = array(
    'Date'                      => gmdate(DATE_RFC2822),
    'MIME-Version'              => '1.0',
    'Content-Type'              => 'text/plain; charset=UTF-8; format=flowed',
    'Content-Transfer-Encoding' => 'quoted-printable'
  );

Though I realize that comma might get us too many false positives due to it also separating function arguments, e.g:

  a = foo(x, y)
  bb = g(42, "str")

We need to build a test suite for this :)


On 24 May 2011, at 22:59, Allan Odgaard wrote:

> On 24 May 2011, at 22:43, Mads Hartmann Jensen wrote:
> 
>> Just improved the code so it will sort the regular expressions such that the
>> regular expression whose captures has the lowest offset (across all lines)
>> is used to align first. Provides a better end result in many cases imo :)
>> 
>> This will however means that the developers have no change of choosing the
>> precedence of regular expressions (previously the regexps had to highest
>> precedence from left to right in the array).
> 
> Sounds acceptable — since these are regexps, one can actually influence it partially by matching a larger part of the line, for example use negative look-ahead if we don’t want a certain pattern to be active say, when a line contains a trailing line comment, etc.
> 
> It would be nice if the algorithm strips whitespace before doing the width calculation, for example:
> 
> 1. foo         = "bar"
> 2. abracadabra = "magic"
> 3. answer      = 42
> 
> Now if I remove the second line and run the aligner, it won’t decrease the excessive amount of whitespace.
> 
> _______________________________________________
> textmate-dev mailing list
> textmate-dev at lists.macromates.com
> http://lists.macromates.com/listinfo/textmate-dev