String parsing question...

Hi,

I'm trying to do something with PHP and I'm not 100% familiar with PHP
as I am with VBScript. So if you could bear with me on what is likely a
stupid question, I'd appreciate it!

I have a chunk of text with a variety of tags inside the text. I want
to perform the following process to this chunk of text:

First, I want to grab data between two markers that I define (e.g.
<start> ... data here ... </start> and strip of the text before the
first marker (<start> in this example) and after the second marker
(</start>) in this example. That would leave me with the "... data here
...." chunk with my markers either included (worst case) or removed
(best case, saving me the third step below).

Is this one PHP function or two functions? I see that strstr will get
me everything to the right of <start> but I cannot figure out how to
remove everything to the right of </start> so that I only have the data
chunk I want (what's between these two markers).

Second, I want to substitute values for values found in the data chunk.
I know str_replace does that just fine.

Third, I then want to strip out the markers from my data chunk. The
<start> marker has elements to it (e.g. limit=) so I'd need something
that would grab everything from <start to the close of the > bracket
(e.g. remove <start limit=1>) to remove it from my data chunk. And
finally I would then want to remove the </start> marker from my data
chunk.

Is this do-able in PHP with a couple functions? Or does it require lots
of string manipulation and lots of functions? Or is it impossible?

Thanks in advance for any insight or pointers to PHP string functions
that'll help!

Tim

timslavin wrote:

> First, I want to grab data between two markers that I define (e.g.
> <start> ... data here ... </start> and strip of the text before the first
> marker (<start> in this example) and after the second marker (</start>) in
> this example. That would leave me with the "... data here ..." chunk with
> my markers either included (worst case) or removed (best case, saving me
> the third step below).

$pieces = preg_split('/\<(\/)?start\>/', $input);
$chunk = $pieces[1];

Assuming that $input is your input data, $chunk will contain your "data
here" segment. What this does is to split the data into an array; the
regular expression passed to preg_split() matches both the <start> tag and
the </start> tag, so the array has three elements. The 0th element contains
everything before <start>, the 1st contains everything between the tags, and
the 2nd contains everything afterwards. (Note that this is untested; my
regular expression might be wrong. Looking at [1] should help you with the
regex syntax expected by preg_split().)

[1] http://perldoc.perl.org/perlre.html

> Is this one PHP function or two functions? I see that strstr will get me
> everything to the right of <start> but I cannot figure out how to remove
> everything to the right of </start> so that I only have the data chunk I
> want (what's between these two markers).

You could probably call strstr() twice and then substr(), but IMO using
preg_split() is way easier.

> Second, I want to substitute values for values found in the data chunk. I
> know str_replace does that just fine.

Yep. If you need even more advanced replacing functionality look at
ereg_replace() and preg_replace().

> Third, I then want to strip out the markers from my data chunk.

This will be done as a side effect of preg_split().

HTH,
--
Benjamin D. Esham
bdesham@gmail.com | AIM: bdesham128 | Jabber: same as e-mail
"...more and more of our imports are coming from overseas."
— George W. Bush

"Benjamin Esham" <bdesham@gmail.com> wrote in message
news:pan.2006.07.27.03.58.44.371018@gmail.com...
> timslavin wrote:
>
>> First, I want to grab data between two markers that I define (e.g.
>> <start> ... data here ... </start> and strip of the text before the first
>> marker (<start> in this example) and after the second marker (</start>)
>> in
>> this example. That would leave me with the "... data here ..." chunk with
>> my markers either included (worst case) or removed (best case, saving me
>> the third step below).
>
> $pieces = preg_split('/\<(\/)?start\>/', $input);
> $chunk = $pieces[1];
>
> Assuming that $input is your input data, $chunk will contain your "data
> here" segment. What this does is to split the data into an array; the
> regular expression passed to preg_split() matches both the <start> tag and
> the </start> tag, so the array has three elements. The 0th element
> contains
> everything before <start>, the 1st contains everything between the tags,
> and
> the 2nd contains everything afterwards. (Note that this is untested; my
> regular expression might be wrong. Looking at [1] should help you with
> the
> regex syntax expected by preg_split().)
>
> [1] http://perldoc.perl.org/perlre.html
>
>> Is this one PHP function or two functions? I see that strstr will get me
>> everything to the right of <start> but I cannot figure out how to remove
>> everything to the right of </start> so that I only have the data chunk I
>> want (what's between these two markers).
>
> You could probably call strstr() twice and then substr(), but IMO using
> preg_split() is way easier.
>
>> Second, I want to substitute values for values found in the data chunk.
>> I
>> know str_replace does that just fine.
>
> Yep. If you need even more advanced replacing functionality look at
> ereg_replace() and preg_replace().
>
>> Third, I then want to strip out the markers from my data chunk.
>
> This will be done as a side effect of preg_split().
>
> HTH,
> --
> Benjamin D. Esham
> bdesham@gmail.com | AIM: bdesham128 | Jabber: same as e-mail
> "...more and more of our imports are coming from overseas."
> - George W. Bush
>

Assuming that you are only using <start> and </start> tags, and no other <>
</> tag pairs, in the line you're checking, you could use the strip_tags()
command, eg:

$variable_name=STRIP_TAGS($line_read_from_file[$optional_line_counter]);

to do it.
Your tags don't actually have to be called <start> and </start> for this to
work, ANY tag will be stripped.
Tags you want kept will have to be listed as allowable tags, eg:

$variable_name=STRIP_TAGS($line_read_from_file[$optional_line_counter],
"$allowable_tags", "$allowable_tags" );

xhtml tags wont be allowed unless you use the html tag name
eg <br /> should be <br>
You also need to be sure your tags are properly formatted with both < and >
characters or you could get some strange results.

I use strip_tags() in an xml parser and it reduced my code considerably.


Phil

Benjamin Esham wrote:
> timslavin wrote:
>
> > First, I want to grab data between two markers that I define (e.g.
> > <start> ... data here ... </start> and strip of the text before the first
> > marker (<start> in this example) and after the second marker (</start>) in
> > this example. That would leave me with the "... data here ..." chunk with
> > my markers either included (worst case) or removed (best case, saving me
> > the third step below).
>
> $pieces = preg_split('/\<(\/)?start\>/', $input);
> $chunk = $pieces[1];
>
> Assuming that $input is your input data, $chunk will contain your "data
> here" segment. What this does is to split the data into an array; the
> regular expression passed to preg_split() matches both the <start> tag and
> the </start> tag, so the array has three elements. The 0th element contains
> everything before <start>, the 1st contains everything between the tags, and
> the 2nd contains everything afterwards. (Note that this is untested; my
> regular expression might be wrong. Looking at [1] should help you with the
> regex syntax expected by preg_split().)
>
> [1] http://perldoc.perl.org/perlre.html
>
> > Is this one PHP function or two functions? I see that strstr will get me
> > everything to the right of <start> but I cannot figure out how to remove
> > everything to the right of </start> so that I only have the data chunk I
> > want (what's between these two markers).
>
> You could probably call strstr() twice and then substr(), but IMO using
> preg_split() is way easier.
>
> > Second, I want to substitute values for values found in the data chunk. I
> > know str_replace does that just fine.
>
> Yep. If you need even more advanced replacing functionality look at
> ereg_replace() and preg_replace().
>
> > Third, I then want to strip out the markers from my data chunk.
>
> This will be done as a side effect of preg_split().
>
> HTH,
> --
> Benjamin D. Esham
> bdesham@gmail.com | AIM: bdesham128 | Jabber: same as e-mail
> "...more and more of our imports are coming from overseas."
> - George W. Bush

Thanks, Benjamin, and for the Bush quote: very obvious and funny.

Probably it's the fact my mind goes blank when reading about regular
expressions but I'm not able to make the preg_split work. If you have
time/interest, I'd appreciate any additional thoughts.

Basically I'm pulling a template from a database field then performing
operations on that data. Within the template I have this data:

.... stuff here ...

<@content limit="" ... more elements ... >
<h2><$Title$></h2>
<$Content$>
</content@>

.... more stuff here ...

So I'm trying to grab everything between the end of <@content and
</content@> as a single data chunk that I can then perform operations
on (like replacing <$Title$> and <$Content$> with result set data from
another query).

What modifications to the preg_split do I need to make this work? Is
there a cleaner way to set up the <content> tags, like </content>
instead of </content@> that would make the regular expression more
efficient? I like using the @ as a flag to find the start marker, on
the premise that makes false results less likely, but maybe I'm
deluded.

I appreciate your help so far! Thank you.

Tim

Fred!head wrote:

> Benjamin Esham wrote:
>
> > timslavin wrote:
> >
> > > First, I want to grab data between two markers that I define (e.g.
> > > <start> ... data here ... </start> and strip of the text before the
> > > first marker (<start> in this example) and after the second marker
> > > (</start>) in this example. That would leave me with the "... data
> > > here ..." chunk with my markers either included (worst case) or
> > > removed (best case, saving me the third step below).
> >
> > $pieces = preg_split('/\<(\/)?start\>/', $input);
>
> Probably it's the fact my mind goes blank when reading about regular
> expressions but I'm not able to make the preg_split work. If you have
> time/interest, I'd appreciate any additional thoughts.

Whoops, I completely forgot that your opening tag has attributes! Sorry
about that. Try this:

$pieces = preg_split('/\<(@content[^>]*|\/content@)\>/', $input);

> What modifications to the preg_split do I need to make this work? Is there
> a cleaner way to set up the <content> tags, like </content> instead of
> </content@> that would make the regular expression more efficient?

Actually, if you used, for example, <@content> for both the start and the
end, you could simply do

$pieces = explode('<@content>', $input);

and bypass regular extensions altogether. The resulting array will be set
up the same as before. If you are able to modify the input to make both
tags the same, this would probably be the best solution.

HTH,
--
Benjamin D. Esham
bdesham@gmail.com | AIM: bdesham128 | Jabber: same as e-mail
....and that's why I'm not wearing any pants.

Thanks!

Tim


Benjamin Esham wrote:
> Fred!head wrote:
>
> > Benjamin Esham wrote:
> >
> > > timslavin wrote:
> > >
> > > > First, I want to grab data between two markers that I define (e.g.
> > > > <start> ... data here ... </start> and strip of the text before the
> > > > first marker (<start> in this example) and after the second marker
> > > > (</start>) in this example. That would leave me with the "... data
> > > > here ..." chunk with my markers either included (worst case) or
> > > > removed (best case, saving me the third step below).
> > >
> > > $pieces = preg_split('/\<(\/)?start\>/', $input);
> >
> > Probably it's the fact my mind goes blank when reading about regular
> > expressions but I'm not able to make the preg_split work. If you have
> > time/interest, I'd appreciate any additional thoughts.
>
> Whoops, I completely forgot that your opening tag has attributes! Sorry
> about that. Try this:
>
> $pieces = preg_split('/\<(@content[^>]*|\/content@)\>/', $input);
>
> > What modifications to the preg_split do I need to make this work? Is there
> > a cleaner way to set up the <content> tags, like </content> instead of
> > </content@> that would make the regular expression more efficient?
>
> Actually, if you used, for example, <@content> for both the start and the
> end, you could simply do
>
> $pieces = explode('<@content>', $input);
>
> and bypass regular extensions altogether. The resulting array will be set
> up the same as before. If you are able to modify the input to make both
> tags the same, this would probably be the best solution.
>
> HTH,
> --
> Benjamin D. Esham
> bdesham@gmail.com | AIM: bdesham128 | Jabber: same as e-mail
> ...and that's why I'm not wearing any pants.