How to change text inside an HTML tag?

I've got a (large, hairy, evil) database full of raw dumps from MySQL
that look like this:

<code>
+----+----------+
| k | t |
+----+----------+
| 1 | apple |
| 2 | orange |
| 3 | pear |
+----+----------+
3 rows in set (0.00 sec)
</code>

When displayed in HTML, the extra spaces don't show, leaving this:

+----+----------+
| k | t |
+----+----------+
| 1 | apple |
| 2 | orange |
| 3 | pear |
+----+----------+
3 rows in set (0.00 sec)

Using <pre> tags won't work, due to other complications with the way
carriage returns are being automatically converted to <br> tags at the
end of each line.

Can anyone recommend a PHP preg_replace statement that would change all
spaces within <code></code> tags to &nbsp; nonbreaking spaces? (This
seems like it ought to be out there somewhere in dozens of examples; if
it is, please feel free to rub my nose in my failure to RTFM.)

I really want to do this with a regular expression and a bunch of other
crawling around; each data record is already filtered through
preg_replace to pop external links up in new browser windows and so
forth. It would be outstanding if I could just add another find pattern
and another replace pattern to the existing module and be done.

Thanks very much,

--Kent

"Kent Brewster" <kent@mindsack.com> wrote in message
news:5K5vc.3381$7g5.1115@newssvr27.news.prodigy.co m...
> I've got a (large, hairy, evil) database full of raw dumps from MySQL
> that look like this:
>
> <code>
> +----+----------+
> | k | t |
> +----+----------+
> | 1 | apple |
> | 2 | orange |
> | 3 | pear |
> +----+----------+
> 3 rows in set (0.00 sec)
> </code>
>
> When displayed in HTML, the extra spaces don't show, leaving this:
>
> +----+----------+
> | k | t |
> +----+----------+
> | 1 | apple |
> | 2 | orange |
> | 3 | pear |
> +----+----------+
> 3 rows in set (0.00 sec)
>
> Using <pre> tags won't work, due to other complications with the way
> carriage returns are being automatically converted to <br> tags at the
> end of each line.
>
> Can anyone recommend a PHP preg_replace statement that would change all
> spaces within <code></code> tags to &nbsp; nonbreaking spaces? (This
> seems like it ought to be out there somewhere in dozens of examples; if
> it is, please feel free to rub my nose in my failure to RTFM.)
>
> I really want to do this with a regular expression and a bunch of other
> crawling around; each data record is already filtered through
> preg_replace to pop external links up in new browser windows and so
> forth. It would be outstanding if I could just add another find pattern
> and another replace pattern to the existing module and be done.
>
> Thanks very much,
>
> --Kent

Kent,
Presumably there would be fewer replacements - thus fewer CPU cycles -
involved in either preventing the conversion of newline to <br> for those
fragments, or post-processing the fragment to revert <br> to newline?
Just my $0.02.
Doug

--
Remove the blots from my address to reply

On Tue, 01 Jun 2004 20:33:05 GMT, Kent Brewster <kent@mindsack.com>
stated wide-eyed, with arms akimbo:

>I've got a (large, hairy, evil) database full of raw dumps from MySQL
>that look like this:
>
><code>
>+----+----------+
>| k | t |
>+----+----------+
>| 1 | apple |
>| 2 | orange |
>| 3 | pear |
>+----+----------+
>3 rows in set (0.00 sec)
></code>
>
>When displayed in HTML, the extra spaces don't show, leaving this:
>
>+----+----------+
>| k | t |
>+----+----------+
>| 1 | apple |
>| 2 | orange |
>| 3 | pear |
>+----+----------+
>3 rows in set (0.00 sec)
>
>Using <pre> tags won't work, due to other complications with the way
>carriage returns are being automatically converted to <br> tags at the
>end of each line.

Does your database contain discrete data for each column, or are
these text dumps within a database field you're extracting?

If text dumps, g'luck.

If it's discrete (k contains values 1, 2, 3; t contains values apple,
orange, pear, etc.), why not use a table and CSS text classes to make
set-pixel-width boxes, then drop the data into them with a "while"
structure?

Can you post a sample data set?


--
STOP LIVING LIKE VEAL
-----------------------
http://diversify.com Veal-free Websites

On Tue, 01 Jun 2004 20:33:05 GMT, Kent Brewster <kent@mindsack.com> wrote:

>I've got a (large, hairy, evil) database full of raw dumps from MySQL
>that look like this:
>
><code>
>+----+----------+
>| k | t |
>+----+----------+
>| 1 | apple |
>| 2 | orange |
>| 3 | pear |
>+----+----------+
>3 rows in set (0.00 sec)
></code>
>
>When displayed in HTML, the extra spaces don't show, leaving this:
>
>+----+----------+
>| k | t |
>+----+----------+
>| 1 | apple |
>| 2 | orange |
>| 3 | pear |
>+----+----------+
>3 rows in set (0.00 sec)

OK, that's expected.

>Using <pre> tags won't work, due to other complications with the way
>carriage returns are being automatically converted to <br> tags at the
>end of each line.
>
>Can anyone recommend a PHP preg_replace statement that would change all
>spaces within <code></code> tags to &nbsp; nonbreaking spaces? (This
>seems like it ought to be out there somewhere in dozens of examples; if
>it is, please feel free to rub my nose in my failure to RTFM.)
>
>I really want to do this with a regular expression and a bunch of other
>crawling around; each data record is already filtered through
>preg_replace to pop external links up in new browser windows and so
>forth. It would be outstanding if I could just add another find pattern
>and another replace pattern to the existing module and be done.

If you've already got newlines sussed, then basically all you want to do is
change any sequence of n spaces (n>1) into a space and (n-1) &nbsp; entities
(if you want to allow longish lines to wrap. If not, then ditch the leading
space and just output n &nbsp;s). Something like (untested):

preg_replace_callback('/( {2,})/',
create_function('$m',
'return " ".str_repeat("&nbsp;",strlen($matches[1])-1);'));

If you've got tabs in there, you'll probably want a prior step turning them
into 4 or 8 spaces depending on your preference. Then present the lot in a
monospace font.

Since you only want it between <code> tags, you probably want that function
called as a callback from another preg_replace_callback('/<code>(.*?)<\/code>',
....). At which point you're probably better of ditching the create_function's
and just making them ordinary functions (I tend to overuse create_function as I
think in Perl for this sort of thing, where anonymous subs have much cleaner
syntax).

So something vaguely like:

function munch_spaces($matches)
{
return ' '.str_repeat('&nbsp;', strlen($matches[1]) - 1);
}

function crunch_code($matches)
{
return preg_replace_callback(
'/( {2,})/',
'munch_spaces',
$matches[1]
);
}

preg_replace_callback(
'/<code>(.*?)<\/code>/',
'crunch_code',
$large_hairy_evil
);

--
Andy Hassall <andy@andyh.co.uk> / Space: disk usage analysis tool
http://www.andyh.co.uk / http://www.andyhsoftware.co.uk/space