From: Robert Klemme <shortcutter@...>
Date: 2009-09-11T04:08:17+09:00
Subject: Re: regex: multiline

2009/9/10 Ahmet Kilic <ahmedkilic@gmail.com>:
> Robert Klemme wrote:
>> 2009/9/10 7stud -- <bbxx789_05ss@yahoo.com>:
>>>> function TfrmExecMain.SetPrivilege(PrivilegeName: String;
>>>> for example;
>>>
>>> ENDOFSTRING
>>>
>>> md = str.match(/function.*?\):\s+(.*?;)/m)
>>> if md
>>>  puts md[1]
>>> end
>>>
>>> --output:--
>>> Boolean;
>>
>> Note that you do not need the /m flag for this to match.  This works
>> the same without it.  The only difference that /m makes is whether
>> newlines are matched by .:
>>
>> irb(main):001:0> s = "a\nb\n"
>> => "a\nb\n"
>> irb(main):002:0> s[/.*/]
>> => "a"
>> irb(main):003:0> s[/.*/m]
>> => "a\nb\n"
>>
>> It does not make any difference for \s:
>>
>> irb(main):004:0> s[/\s+/]
>> => "\n"
>> irb(main):005:0> s[/\s+/m]
>> => "\n"
>>
>> I'll attach an extended example.
>>
>> Kind regards
>>
>> robert
>
> Thank you Robert,
> Wooow! your code is seen perfect.
> I want to ask one more question; my file contains many many functions,
> vars and begins like this,
> for example;
> ...
> function TfrmExecMain.GetApiErrorMessage(ApiErrCode :DWORD) :String;
> var
>  Buf: array[0..511] of Char;
>  MsgCnt :DWORD;
> begin
> ...
> function TfrmExecMain.SetPrivilege(PrivilegeName: String;
>  Enable: Boolean): Boolean;
> var
>  tpPrev,
>  tp         : TTokenPrivileges;
>  token      : THandle;
>  dwRetLen   : DWord;
> begin
> ...
> how can I do it with your code?
> should I write new regexp for each function?

No, you simply need to replace the outer match with a call to String#scan.

Cheers

robert


-- 
remember.guy do |as, often| as.you_can - without end
http://blog.rubybestpractices.com/