"You can't parse HTML with regex" WHY?

Question

I know My question seems Like duplicated(duplicate)... But exceptionally I'm not trying to parse a "True/Valid" HTML, its a part of javascript variables / or of HTML tags, see my example:

                <td width="10"></td>
            <td ><div id="combo_rub" style="CURSOR:pointer"></div><input type="hidden" id="rech_cod_cat"  name="rech_cod_cat" value="1"/></td>
            <td ><div id="combo_typ" style="CURSOR:pointer"></div></td>
            <td ><div id="combo_sou_typ" style="CURSOR:pointer"></div></td>

            <script type="text/javascript">
            d_rech_cod_rub       = "101";
            d_rech_cod_typ       = "1";
            d_rech_cod_sou_typ   = "1020121";

            //ComboBox sous_type
            var z_sou_typ=new dhtmlXCombo("combo_sou_typ","rech_cod_sou_typ",160);

So what i want in my code PHP is to retrieve the value of(d_rech_cod_rub, d_rech_cod_typ, d_rech_cod_sou_typ) from Javascript part, not from HTML as you can see...

the second scenario is:

<td><a href="AnnoncesImmobilier.asp?rech_cod_rub=101&amp;rech_cod_typ=1">
<font color="#2154eb">Location (6920)</font></a> </td></tr></table></td></tr>
<tr><td><table width="100%" cellpadding="0" cellspacing="0"><tr>   
<td><a href="AnnoncesImmobilier.asp?rech_cod_rub=101&amp;rech_cod_typ=5"></td>

Same here i need to retrieve the '101' value...

Finally, The full story: I'm using very frequently the Curl extension of PHP then I come across with a content like the above examples... the problem is when i use a HTML-Parser like (http://simplehtmldom.sourceforge.net/) i can not done my work with this, because it work only with a valid HTML..:( my question: there is any solution to retrieve the values in a stable manner ? what are you doing if you come across a case like me ?

Thanks guys.

EDIT : So according to most answers, using RegX do the trick...

possible duplicate: http://stackoverflow.com/questions/3577641/how-to-parse-and-process-html-with-php — Jocelyn, Aug 24 '12 at 13:33
Use a more forgiving html parser. What do you think the browser is doing when it meets that html? — Esailija, Aug 24 '12 at 13:33
@Jocelyn: As i already said in my question is not a duplicated question, i'm not trying to parse a HTML tags, just a little part of text — Marwen Trabelsi, Aug 24 '12 at 13:39
Since your text is within a file containing HTML tags, I don't think using a regex is a good idea. This is what is explained on the page I linked to. — Jocelyn, Aug 24 '12 at 13:40
The problem with JS is almost exactly the same as the problem with (X)HTML. JS is not a regular language (it's context-free, AFAIK), and REs tend to suck at parsing non-regular languages. Past a certain point, even with our modern souped-up "RE"s, it's not possible...and even before that, you'll be tearing your hair out in frustration. If you have a *known sample*, and are ok with not being able to tolerate deviations, then you can easily grab a specific chunk of text...but the more general you want it, the less useful REs get -- and the easier it is to use a real parser. — cHao, Aug 24 '12 at 13:47

score 2 · Answer 1 · answered Aug 24 '12 at 13:37

2

If all you want is that value it is quite easy to use a regexp just for that:

d_rech_cod_rub\s*=\s*"([0-9]+)";

Where the first capture group is your value.

answered Aug 24 '12 at 13:37

newts

85
4

If you want it to work on all the d_rech variables you can add another capture group: `(d_rech_cod_[a-z_]+)\s*=\s*"([0-9]+)";` – newts Aug 24 '12 at 13:40

score 2 · Accepted Answer · answered Aug 24 '12 at 13:38

2

The following regexes do the trick

d_rech_cod_[a-z_ ]*= "([0-9]*)";
rech_cod_rub=([0-9]*)

Not sure about PHP but you might be able to evaluate it using the following code:

preg_match($regex, $content, $matches);
print($matches[1]);

answered Aug 24 '12 at 13:38

tbl

771
3
11

score 2 · Answer 3 · edited Sep 24 '12 at 02:02

2

In a more specific answer, regular expressions cannot deal with nested information.

For example ...

<span>Hello <span>World</span></span>

A regular expression that looks for <span>.*<\/span> will return:

<span>Hello <span>World</span>

We won't even mention trying to process invalid HTML with missing tags.

Something like this (loadHTML) might be what you want instead.

edited Sep 24 '12 at 02:02

uınbɐɥs

7,236
5
26
42

answered Aug 24 '12 at 13:40

Jeremy J Starcher

23,369
6
54
74

score 1 · Answer 4 · answered Aug 24 '12 at 13:36

I'm not sure whether it's possible to do what you want with one regular expression, but certainly possible with one regular expression per variable. Also it's Possible to use strpos function to parse string. Here the second approach will probably be more efficient way.

"You can't parse HTML with regex" WHY?

4 Answers4