[RESOLVED] preg_match_all : replace multiple words, with different words

[ungovernable]

$post[message] is the text of a post from my forum.

I wanted to take out all the text and keep only the text between [TAG1] tags

here is a part of the code i am using to do that:

<?php
$string = $post[message];
$tag = "TAG1";

preg_match_all("#\[".$tag."\](.*)\[/".$tag."\]#Ui", $string, $matches); 
$xmatches = $matches[1];   

$count = 1;


foreach ($xmatches as $value) {

echo "ITEM NUMBER $count : ";

echo $value;
echo "<br>";
}
$count = $count + 1;
}
?>

the code works perfectly

but i would like to also keep the text between another tag ( [TAG2] )

AND i want to make a link with the [TAG2] replacements

exemple:

with this text:

[TAG1]Hello[/TAG1] John [TAG2]how[/TAG2] are [TAG1]you[/TAG1] ?

the result would be:

ITEM NUMBER 1 : Hello
ITEM NUMBER 2 : <a href="how.php">how</a>
ITEM NUMBER 3 : you

could someone please help me ??

thanks a lot

[ungovernable]

bump

[nrg_alpha]

ungovernable;10917314 wrote:

exemple:

with this text:

[TAG1]Hello[/TAG1] John [TAG2]how[/TAG2] are [TAG1]you[/TAG1] ?

the result would be:

ITEM NUMBER 1 : Hello
ITEM NUMBER 2 : <a href="how.php">how</a> 
ITEM NUMBER 3 : you

You mean something like this?

$str = '[TAG1]Hello[/TAG1] John [TAG2]how[/TAG2] are [TAG1]you[/TAG1] ...some text... [TAG1]Goodbye[/TAG1] Mark [TAG2]Do[/TAG2] your [TAG1]worst[/TAG1]?';

$tag1 = 'TAG1';
$tag2 = 'TAG2';

preg_match_all('#\[' . $tag1 . '\]([^[]+)\[/' . $tag1 . '\].*?\[' . $tag2 . '\]([^[]+)\[/' . $tag2 . '\].*?\[' . $tag1 . '\]([^[]+)\[/' . $tag1 . '\]#si', $str, $matches);
$count = count($matches[0]);
for ($a = 0 ; $a < $count ; $a++) {
	echo 'ITEM NUMBER 1 : ' . $matches[1][$a]."<br />\n";
	echo 'ITEM NUMBER 2 : <a href="'. $matches[2][$a] .'.php">'. $matches[2][$a] . '</a>' . "<br />\n";
	echo 'ITEM NUMBER 3 : ' . $matches[3][$a] . "<br /><br />\n\n";
}

Output:

ITEM NUMBER 1 : Hello
ITEM NUMBER 2 : how
ITEM NUMBER 3 : you

ITEM NUMBER 1 : Goodbye
ITEM NUMBER 2 : Do
ITEM NUMBER 3 : worst

[ungovernable]

thanks for your answer

it works, but it will work only with your string.... if there is more tags, or if they are placed differently, it will not work

basically $str will be the content of a message from my forum, so the tags emplacements will vary...

i need something that can work anywhere the tags are placed....

also TAG1 is not always followed by TAG2, sometimes there can be TAG1 then a dozen of TAG2, then another TAG1, then just two TAG2, etc....

i hope you understand

[nrg_alpha]

ungovernable;10917462 wrote:

thanks for your answer

it works, but it will work only with your string.... if there is more tags, or if they are placed differently, it will not work

Actually, it's you string (in the sense that I am going by the format that you gave).Read this carefully: I can only work with what you explain / give as an example. Remember, ten times out of ten, peoples' crystal balls are broken and in the shop!

ungovernable wrote:

basically $str will be the content of a message from my forum, so the tags emplacements will vary...

i need something that can work anywhere the tags are placed....

also TAG1 is not always followed by TAG2, sometimes there can be TAG1 then a dozen of TAG2, then another TAG1, then just two TAG2, etc....

Well, this info is new, and related to what I was saying above. It's great that you gave an example (a before and after kind of thing), but if there are additional conditionals, it would be best to say that upfront in your initial post so that we don't have these back and forth rounds of miscommunication (which is in essence what this is).

So getting back to the issue at hand.. If I understand the format correctly, there would be a TAG1, then god knows how many TAG2's, then another TAG1, followed by 2 TAG2's.

The question becomes clear.... which TAG1 and TAG2 values get stored into ITEM NUMBER 1, 2 and 3 respectively? In your first example, it is clear which values get stored into what.. but now, with the introduction of multiple TAG1s and TAG2s, it becomes obfuscated rather quickly...

Perhaps giving an example (like you did before), but with all the multiple TAG1 and TAG2s and highlight in bold which ones you are intending on matching, and explain [i[why[/i] those ones are what you are looking for (in other words, help me establish the format within the multiple tags)..

[johanafm]

This will work with any number of tags in any order. Anything between TAG1 tags will end up in one array, while anything between TAG2 tags will end up in another one.
However, you will need to do additional processing on each array element in either array in case the tags can be nested:

[TAG1]Hello John [TAG2]how[/TAG2][TAG1]

Assuming there is no nesting of tags, this will work for you no matter in what order the tags show up, how many of each there are etc:

$str = "[TAG1]Hello[/TAG1] John [TAG2]how[/TAG2] are [TAG1]you[/TAG1]";
$re = "/\[TAG(\d)](.*)\[\/TAG\\1]/Uus";

preg_match_all($re, $str, $match);
print_r($match);
if (count($match[0]) > 0) {
	for ($i = 0; $i < count($match[1]); ++$i) {
		if ($match[1][$i] == "1") {
			$word[] = $match[2][$i];
		}
		else if ($match[1][$i] == "2") {
			$link[] = $match[2][$i];
		}
	}
}
print_r($word);
print_r($link);

output:

Array {
            0 => {
                        0 => [TAG1]Hello[/TAG1]
                        1 => [TAG2]how[/TAG2]
                        2 => [TAG1]you[/TAG1]
            }
            1 => {
                        0 => 1
                        1 => 2
                        2 => 1
            }
            2 => {
                        0 => Hello
                        1 => how
                        2 => you
            }
}

Array {
            0 => Hello
            1 => you
}

Array {
            0 => how
}

[ungovernable]

@johnanafm

thanks a lot, i got it to work to do exactly what i needed!

i only have one more question

i would like to use the same script on a different page, but my tags are differents

i use [NOUVEAU_CHAPITRE][/NOUVEAU_CHAPITRE] instead of TAG1
and [SOUS_CHAPITRE][/SOUS_CHAPITRE] instead of TAG2

i tryed to modify the script to use the new tags, but since i'm a newbie i always get errors 🙁

could you help me to modify the script to use the new tags?

thanks a lot!

So getting back to the issue at hand.. If I understand the format correctly, there would be a TAG1, then god knows how many TAG2's, then another TAG1, followed by 2 TAG2's.

Sorry, i should have given better explainations at first

What you said is correct, but there can also be a lot of TAG1, followed by a lot of TAG2, then a lot of TAG1, etc....

Here is my situation: The string i am modifying comes from the content of a message inside my forum. This message is the text of a eBook, so i want to build an index of chapters. TAG1 is used inside the message to mark the chapters (chapter name is inside the brackets). And TAG2 is to mark the sub-chapters.

Since i can't modify the forum itself, or the database structure, the only way i can build the "chapters index" is by taking the text of the ebook, and remove all the text except the chapters and the sub-chapters.

[johanafm]

Glad to hear it helped you out.

Very little change is required to make it work with the second set of tags. A little bit of trivia to make you understand how it works, and in the future you should be able to do this yourself.

The first pattern, "/[TAG(\d)](.)[\/TAG\1]/Uus", uses two subpatterns, which are the parts enclosed in parenthesis (). You can refer to subpatterns from within the pattern itself, using back references \1, \2 etc. Matches against these subpatterns are also put in arrays which are found in the third function argument of preg_match_all.
That is, $match[1] contains all matches of (\d), and $match[2] contains all matches of (.), which you could have referred to in the pattern with \2.

Do note that in a double quote enclosed strings in PHP, you need to escape the backslash \ to get the two characters \ and 1:
"\1" becomes \1, which is not the same as the character "\1".

So, (\d), is a subpattern matching one digit, which is back referenced in the closing tag above. All that's needed to match the new tags is to create a subpattern for several characters and put that at the beginning of the tag.

$str = "[NOUVEAU_CHAPITRE]Ma première femme[/NOUVEAU_CHAPITRE][SOUS_CHAPITRE]Rencontre[/SOUS_CHAPITRE]Il m'est à peu près impossible aujourd'hui de me souvenir...[NOUVEAU_CHAPITRE]L'infortuné comique[/NOUVEAU_CHAPITRE]";
$re = "/\[(.*)_CHAPITRE](.*)\[\/\\1_CHAPITRE]/Uus";

preg_match_all($re, $str, $match);
print_r($match);
if (count($match[0]) > 0) {
    for ($i = 0; $i < count($match[1]); ++$i) {
        if ($match[1][$i] == "NOUVEAU") {
            $word[] = $match[2][$i];
        }
        else if ($match[1][$i] == "SOUS") {
            $link[] = $match[2][$i];
        }
    }
}
print_r($word);
print_r($link);
exit;

Output:

Array {
            0 => {
                        0 => [NOUVEAU_CHAPITRE]Ma première femme[/NOUVEAU_CHAPITRE]
                        1 => [SOUS_CHAPITRE]Rencontre[/SOUS_CHAPITRE]
                        2 => [NOUVEAU_CHAPITRE]L'infortuné comique[/NOUVEAU_CHAPITRE]
            }
            1 => {
                        0 => NOUVEAU
                        1 => SOUS
                        2 => NOUVEAU
            }
            2 => {
                        0 => Ma première femme
                        1 => Rencontre
                        2 => L'infortuné comique
            }
}

Array {
            0 => Ma première femme
            1 => L'infortuné comique
}

Array {
            0 => Rencontre
}

[ungovernable]

Thanks for your help and for these infos,

I still have one problem (sorry, i'm not hyper skilled in php)

with the previous code working with TAG1 and TAG2 i added this code to display the array :

$str = "[TAG1]Hello[/TAG1] John [TAG2]how[/TAG2] are [TAG1]you[/TAG1]";
$re = "/[TAG(\d)](.*)[\/TAG\1]/Uus";

preg_match_all($re, $str, $match);

if (count($match[0]) > 0) {
for ($i = 0; $i < count($match[1]); ++$i) {
if ($match[1][$i] == "1") {
$word[] = $match[2][$i];
foreach ($match[1] as $zqwe) {
echo "CHAPITRE: $zqwe";
}
}
else if ($match[1][$i] == "2") {
$link[] = $match[2][$i];
foreach ($match[0] as $zqwe) {
echo "SOUS-CHAPITRE: <a href=\"#zqwe\">$zqwe</a><br>";
}
}
}
}

I tryed doing the same thing with the new code, however i get a blank page:

$str = "[NOUVEAU_CHAPITRE]Ma première femme[/NOUVEAU_CHAPITRE][SOUS_CHAPITRE]Rencontre[/SOUS_CHAPITRE]Il m'est à peu près impossible aujourd'hui de me souvenir...[NOUVEAU_CHAPITRE]L'infortuné comique[/NOUVEAU_CHAPITRE]";
$re = "/[(.)_CHAPITRE](.)[\/\1_CHAPITRE]/Uus";

preg_match_all($re, $str, $match);
if (count($match[0]) > 0) {
for ($i = 0; $i < count($match[1]); ++$i) {
if ($match[1][$i] == "NOUVEAU") {
$word[] = $match[2][$i];
foreach ($match[1] as $zqwe) {
echo "$zqwe<br>";
}

    }
    else if ($match[1][$i] == "SOUS") {
        $link[] = $match[2][$i];
    }
}

}
exit;

PS: do you speak french ? 😛

[johanafm]

Il fait douze out treize ans que je n'ai pas eu l'occasion de parler francais (et en plus il me manque le cedille sur mon clavier), mais je n'ai pas tout oublié 🙂

Anyway, let's stick to the forum's language. The opening and closing php tags were not included in your pasted code. Could it be that you simply didn't copy them, or did you actually forget to include "<?php" and "?>"?

Apart from that, the code works, but possibly not the way you intended it to. Since you loop over all the matches with the for ($i = 0...) loop, and inside of that loops over that same array once more with foreach ($match[1]...), whenever the current array element is "NOUVEAU" you will actually go over the whole $match[1] array twice with the given example string.

This is probably what you intended:

<?php
$str = "[NOUVEAU_CHAPITRE]Ma première femme[/NOUVEAU_CHAPITRE][SOUS_CHAPITRE]Rencontre[/SOUS_CHAPITRE]Il m'est à peu près impossible aujourd'hui de me souvenir...[NOUVEAU_CHAPITRE]L'infortuné comique[/NOUVEAU_CHAPITRE]";
$re = "/\[(.*)_CHAPITRE](.*)\[\/\\1_CHAPITRE]/Uus";

preg_match_all($re, $str, $match);
if (count($match[0]) > 0) {
	for ($i = 0; $i < count($match[1]); ++$i) {
		echo $match[1][$i].": ".$match[2][$i]."<br/>";
	}
}
exit;
?>

Also, please indent your code and use &#91;php&#93; ... &#91;/php&#93; tags to get colour highlighting. It makes it easier to read the code.

[ungovernable]

I tryed to use the code you gave to me (without touching it), but i always get a blank page when trying to use the code... i don't understand why

no, i didn't forget to include the <?php ?> tags 😉

[johanafm]

Oh, could you possibly be saving your file in some encoding other than UTF-8? It might screw it up.

[ungovernable]

I'm using normal encoding, i usually don't have problems about encoding, i use the same text editor as i always used for php scripts... I don't think it is the problem

if i save the code in a .php file, upload it on my site and try to open it with my browser, i get a blank page

if i try to put the code right inside my forum (using vbulletin template editor -- so there shouldn't be character encoding problems since i used the same interface to build a few php scripts) nothing is displayed

note: if i want to put the code right in the templates of my forum, i have to remove the "exit;" part or else it won't let me save the page. I think it is because it terminate vbulletin's scripts. I hope the script will work without this?

[johanafm]

Well, if I use my editors default encoding, I get a blank page as well, until I either save the script as UTF-8 or remove all accents aigus et graves, which is what I've done below.

And yes, you can of course remove the exit statement.

$str = "[NOUVEAU_CHAPITRE]Ma premire femme[/NOUVEAU_CHAPITRE][SOUS_CHAPITRE]Rencontre[/SOUS_CHAPITRE]Il m'est a peu pres impossible aujourd'hui de me souvenir...[NOUVEAU_CHAPITRE]L'infortune comique[/NOUVEAU_CHAPITRE]";
$re = "/[(.)_CHAPITRE](.)[\/\1_CHAPITRE]/Uus";

preg_match_all($re, $str, $match);
if (count($match[0]) > 0) {
for ($i = 0; $i < count($match[1]); ++$i) {
echo $match[1][$i].": ".$match[2][$i]."<br/>";
}
}

[ungovernable]

Guess you were right, it is working perfectly now!

Sorry for bothering you so much

you were of a great help!

[ungovernable]

i still have a problem with accents

i have put the code inside the template of my forums (the template are saved in database, and i use a web-based text editor to edit the templates... so encoding shouldn't be a problem)

the only part of the code i modified is this:

$str = $post[message];

so it gets the string to do replacements on from my database instead of the example string we were using

I was doing exactly the same thing with the first version of the code i posted above and it was working even with accents

But now as soon as there are accents (é, è, ê, à, etc..) inside the post message it will not display anything, even if the accents are not between my tags

[ungovernable]

ok i found the solution for the accents

$str = htmlentities($post[message]);

now it works with characters like "é" but it does the same thing with single quotes

is there a way to solve this problem without having results like "je mange l\'orange" ?

edit: i tryed to replace " ' " by " ", then " ’ " by " ", it worked, but now i have same problem with "œ"... looks like it will never end.... is there a better way to fix all special characters, something like "htmlentities" but for all weird characters?

edit2:

$pstring = htmlentities($string);
$str = utf8_encode($pstring);

this seems to fix most of the issues

but " ` " will show " Â’ "