binarycloud-dev

[binarycloud-dev] Regex for bc:tags - problem with nesting

[<bin...@li...>]

Hi,

i'm currently on the BccHrefCompiler and the following problem occured. The
Regex to grab the tags does not handle nested tags properly.

preg_match_all("!<(bc:.*?)>.*?</\\1.*?>!si", $source, $match);

i.e.:
...
<bc:href>
  <id>123</id>
  <content>
    <bc:img>
      <id>321</id>
    </bc:img>
  </content>
</bc:href>
...

The regex now extracts the whole bc:href stuff including the bc:img tag.
Then the href compiler is called and its output replaces the whole thing.
The desired behavior is, that it first replaces the inner tag(s) calling the
ImgCompiler then calling the Href Compiler.

A possible solution is to call the ImgCompiler from within the Href
Compiler, what I dislike. Another might be using some kind of recrusion or
the php native xml_parser functions using some kind of event based parsing.

Any ideas?

andi

Re: [binarycloud-dev] Regex for bc:tags - problem with nesting

[<bin...@li...>]

> Hi,
> 
> i'm currently on the BccHrefCompiler and the following problem occured. The
> Regex to grab the tags does not handle nested tags properly.
> 
> preg_match_all("!<(bc:.*?)>.*?</\\1.*?>!si", $source, $match);
> 
> i.e.:
> ...
> <bc:href>
> <id>123</id>
> <content>
>   <bc:img>
>     <id>321</id>
>   </bc:img>
> </content>
> </bc:href>
> ...
> 
> The regex now extracts the whole bc:href stuff including the bc:img tag.
> Then the href compiler is called and its output replaces the whole thing.
> The desired behavior is, that it first replaces the inner tag(s) calling the
> ImgCompiler then calling the Href Compiler.

I believe the href tag is the only tag we have right now which can accept
other bc tags as params... so I think the internal 'sub-parsing' is 100%
applicable, and valid + clean.

_alex





> 
> A possible solution is to call the ImgCompiler from within the Href
> Compiler, what I dislike. Another might be using some kind of recrusion or
> the php native xml_parser functions using some kind of event based parsing.
> 
> Any ideas?
> 
> andi
> 
> 
> _______________________________________________
> binarycloud-dev mailing list
> bin...@li...
> https://lists.sourceforge.net/lists/listinfo/binarycloud-dev
> http://www.binarycloud.com
> http://www.sf.net/projects/binarycloud/
> 

RE: [binarycloud-dev] Regex for bc:tags - problem with nesting

[<bin...@li...>]

Hi Alex,

>I believe the href tag is the only tag we have right now which
>can accept other bc tags as params... so I think the
>internal 'sub-parsing' is 100% applicable, and valid + clean.

Ok, I'll do it that way. But there is another application where to accept bc
tags as params: Everywhere where lang dependend strings are needed.

Andi

Re: [binarycloud-dev] Regex for bc:tags - problem with nesting

[<bin...@li...>]

> Ok, I'll do it that way. But there is another application where to accept bc
> tags as params: Everywhere where lang dependend strings are needed.

because we can _always_ assume that <bc:lang strings need to be process
first, i think it's fine to process that first, and use a separate regex
that will only match <bc:lang

_a

RE: [binarycloud-dev] Regex for bc:tags - problem with nesting

[<bin...@li...>]

Hi Alex,

>> Ok, I'll do it that way. But there is another application where
>> to accept bc tags as params: Everywhere where lang dependend
>> strings are needed.

> because we can _always_ assume that <bc:lang strings
> need to be process first, i think it's fine to
> process that first, and use a separate regex
> that will only match <bc:lang

Oh yes. Cool. We can introduce precedence to bc tags and unsing a "multi
pass" technique :-)

No recrusion, no much extra code overhead, no extra parser. Simple yet
powerful.


Andi