replace only first & last ocurrence of string

rw22

I have a list that looks like this..

*bread
*milk
*eggs
*butter

and i want to convert it to this

<ul>
<li>bread</li>
<li>milk</li>
<li>eggs</li>
<li>butter</li>
</ul>

I'm currently parsing it like so :

$text = preg_replace ("/^\*(.*)/m", "<li>\\1</li>", $text);

And that gives me :

<li>milk</li>
<li>eggs</li>
<li>butter</li>

But how do i get it so <ul> & </ul> are at the start and end of this part of the string, because this is part of a much larger multi-line string.

volleytotal_ch

why don't you just add the ul-stuff to the string:

$text = "<ul>".$text."</ul>";

rw22

because the string looks like this :

h2. This is a title\r\n\r\nh3. This is a subhead\r\n\r\nThis is some 
text of dubious character. Isn't the use of \"quotes\" just lazy writing -- and theft of 'intellectual property' besides? I think the 
time has come to see a block quote.\r\n\r\nbq. This is a block quote. I'll admit it's not the most exciting block quote ever 
devised.\r\n\r\nWell, that went well. How about we insert an <a href=\"/\" title=\"watch out\">old-fashioned hypertext link</a>? 
Will the quote marks in the tags get messed up? No!\r\n\r\n\"This is a link (optional title)\":[url]http://www.textism.com\r\n\r\nAn
[/url] image:\r\n\r\n!/common/textist.gif(optional alt text)!\r\n\r\n# Librarians rule\r\n# Yes they do\r\n# But you knew that\r\n\r\nSome more text of dubious character. Here is a
 noisome string of CAPITAL letters. Here is something we want to _emphasize_. \r\nThat was a linebreak. And something to indicate *strength*. Of course I could use <em>my own HTML tags</em> if I <strong>felt</strong> like it.\r\n\r\nh3. Coding\r\n\r\nThis <code>is some code, \"isn't it\"</code>. 
Watch those quote marks! Now for some preformatted text:\r\n\r\n<pre>\r\n<code>\r\n\t\$text = str_replace(\"<p>%::%</p>\",\"\",\$text);\r\n\t\$text = str_replace(\"%::%</p>\",\"\",\$text);\r\n\t\$text = str_replace(\"%::%\",\"\",\$text);\r\n\r\n</code>\r\n</pre>\r\n\r\nThis isn't 
code.\r\n\r\n\r\nSo you see, my friends:\r\n\r\n* The time is now\r\n* The time is not later\r\n* The time is not yesterday\r\n* We must act");

rw22

any ideas

diego25

Well, '*' is a special character for regexps, so you should escape it:

$text = preg_replace ("/^*(.*)/m", "<li>\1</li>", $text);

Also, are you sure it's \1? I thought it was \1....but if it works, ok...

Diego

rw22

No your right it is \1 it just didn't seem to copy correctly.

Can i put this question across differently??

Right then, i have a template that looks like this :

h1. Heading
h2. subheading
bq. blockquote
*sex
*drugs
*rock
*roll

a basic paragraph

And the current code i'm using to parse it looks like this.

function textParser ($text) {
	// converts hn. text to <hn>text</hn>
	$text = preg_replace ("/^h+([0-6])\.(.*)/m", "<h\\1>\\2</h\\1>", $text);

// converts bq. text to <blockquote>text</blockquote>
$text = preg_replace ("/^bq+\.(.*)/m", "<blockquote>\\1</blockquote>", $text);

// converts *text to <li></li>
$text = preg_replace ("/^\*(.*)/m", "<li>\\1</li>", $text);

// converts anything that doesn't start with either a bracket or space to <p>text</p>
$text = preg_replace ("/^([^<|[:space:]].*)/m", "<p>\\1</p>", $text);

echo $text;
};

But my problem lies where i need to preg_replace the * character for a <li> tag. How do i get it so it put's an <ul> tag just before the first <li> and </ul> at the end of the last </li>.

So instead of this :

<h1>Heading</h1>
<h2>SubHeading</h2>
<blockquote>blockquote</blockquote>
<li>sex</li>
<li>drugs</li>
<li>rock</li>
<li>roll</li>

I would get this :

<h1>Heading</h1>
<h2>SubHeading</h2>
<blockquote>blockquote</blockquote>
<ul>
<li>sex</li>
<li>drugs</li>
<li>rock</li>
<li>roll</li>
</ul>

It seems so simple yet i can't find out how to do it.

xblue

might be easiest to do it in 2 steps:

first, find a sequence of lines starting with *, and place <ul> </ul> around it. then add the <li> </li> tags at each line as before.

e.g.

$text = preg_replace ("/(^\*.*\n)+/m", "<ul>$0</ul>", $text);

btw, the "code" formatting option does not seem to drop the backslashes.

rw22

that almost works but i get this :

<ul>* The time is now
<li> The time is not later
</li>
<li> The time is not yesterday
</li>
</ul>* We must act

i know its easy enough to provide a line-break so the first * is parsed but what about at the bottom

xblue

yes, linebreaks are a good idea, didn't think about that in my example.

simply change the replacement argument to
"<ul>\n$0\n</ul>"

Also, I think the last element is not caught because I defined that it must be followed by a line break. Marking this as optional might solve the problem (hopefully):
("/(^{*.*$\n?)+/m",}

rw22

thankyou i'll try this out now

rw22

yep that works like a dream, thankyou very much..

i was really smacking my head over that one.