Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.php > #18908

Re: question about

Show all headers | View raw

On Mon, 04 Apr 2022 13:07:59 +0000, Lew Pitcher wrote:

> On Mon, 04 Apr 2022 14:26:01 +0200, J.O. Aho wrote:
> 
>> On 23/02/2022 15.54, alex wrote:
>>> $ echo center | sed -E 's#(.*)#start-\1-end#'
>>> start-center-end
>>> 
>>> I tried to do the same thing with php, but the result is different
>>> 
>>> $ php -r "echo preg_replace('#(.*)#', 'start-\0-end', 'center') . PHP_EOL;"
>>> start-center-endstart--end
>>>                  ^^^^^^^^^^
>>> 
>>> Why?
>> 
>> Why I'm not sure,
> 
> So, let's look at it and figure the "why" out, shall we?
> 
> The first three arguments to preg_replace() are
>   the regex pattern to test with
>   the substitution pattern to replace regex matches with, and
>   the string to be tested with the regex
> 
> In the OP's case, the regex pattern is
>   #(.*)#
> where the # characters simply delimit the edges of the regex.
> (Note that the convention is to use forward slashes, but PHP
> preg_replace() accepts other characters as well. So, you might
> normally see this pattern as /(.*)/ in other regex parsers.)
> 
> As the octothorpe signs are string delimiters rather than part
> of the regex match string, the OP's regex devolves to
>   (.*)
> which will match to ZERO OR MORE occurrences of ANY character.
> 
> The OP's substitution pattern is
>   start-\0-end
> which instructs preg_replace() to replace the matched input
> with ITSELF (\0), immediately preceded by "start-", and
> immediately followed by "-end".
> 
> Finally, we get the OP's input string
>   center
> 
> Combine these two, and we get
> - the regex #(.*)# matches the multi-character input string
>   'center'
> - preg_replace() replaces 'center' with the interpreted
>   replacement string, resulting in an output of
>   begin-center-end
> - the regex #(.*) matches EXACTLY ZERO characters in the
>   remaining input string
> - preg_replace() replaces EXACTLY ZERO characters with
>   the interpreted replacement string, resulting in an
>   additional output of
>     begin--end
>   (note the EXACTLY ZERO characters copied from the input
>   to the gap between the 'begin-' portion and the '-end'
>   portion)
> - the regex has run out of string, and preg_replace() stops.
> 
>> but figured out two solutions:
>> 
>> You set a number a limit to the replacement
> 
> Thus telling preg_replace() to stop before it matches the
> remaining nothing to (.*)
>  
>> $ php -r "echo preg_replace('#(.*)#', 'start-\1-end', 'center', 1) . 
>> PHP_EOL;"
>> 
>> or you can tell that it's from the beginning of the row
>
> Thus telling preg_replace() to /only/ match everything up to and
> including the remaining nothing at the end of the input.

Oops... strike that. My response was for a different (and bad) solution


>> $ php -r "echo preg_replace('#^(.*)#', 'start-\1-end', 'center') . PHP_EOL;

Thus telling preg_replace() to /only/ match everything from the BEGINNING
of the input (the remaining nothing isn't at the beginning, so it doesn't
match).

>> I would guess this has to do with the end of char array \0 (internally) 
>> and it's counted as a white space, so it's not included in the first 
>> match, but then as the second match, but that is just a wild guess.
> 
> 
> 
> HTH




-- 
Lew Pitcher
"In Skills, We Trust"

Back to comp.lang.php | Previous | Next — Previous in thread | Find similar | Unroll thread

Thread

question about alex <1j9448a02@lnx159sneakemail.com.invalid> - 2022-02-23 15:54 +0100
  Re: question about Kristjan Robam <he6655442211@hotmail.com> - 2022-04-04 04:06 -0700
  Re: question about "J.O. Aho" <user@example.net> - 2022-04-04 14:26 +0200
    Re: question about Lew Pitcher <lew.pitcher@digitalfreehold.ca> - 2022-04-04 13:07 +0000
      Re: question about Lew Pitcher <lew.pitcher@digitalfreehold.ca> - 2022-04-04 13:29 +0000

csiph-web