Help: divide html code in units

[GoRide_]

Hi,
i have file test.html and i need to load it's code and divide it in "units".

i need to get output like this:

if I open test.html in notepad and add line "bla bla bla", it should be recognized as new unit TEXT [part of text]....

it's hard for me to explain, but if you used kmx wap constructor, you know what i am talking about.

Please help,
thank you !

[dagon]

sounds like a job for the DOM ext

[GoRide_]

hmm, i can't because i must have option to divide wml, not only html. DOM doesn't support wml....

some idea?

[GoRide_]

i forgot to say, i do not need to define ALL html tags. i will define several tags which will be recognized as UNIT, other tags will be recognizes as UNKNOWN unit....

if it helps to you

sorry for bad eng

[dagon]

wml is based on xml so there is DOM XML and XML Parser; while they may get you so far you wil end up rolling your own at least in part

[GoRide_]

can i load wml file using dom->loadXML(); ?

[GoRide_]

is it possible with preg_replace and how i can find/replace tags?

[dagon]

What are you actually trying to achieve? What's the purpose of this?

[GoRide_]

i am making site where users can register and create html/wml pages, and have options to edit them. one of edit options must be UNITS edit, so i need to divide html code of that file in units.

that's why i asked for help. it is PROBABLY possible with preg_replace, so please, if somebody can help....

theese are tags which need to be recognized and divided into units:

<a href=""> </a> - UNIT LINK
<img src="" alt=""> - UNIT IMAGE
<p> </p> - UNIT PARAGRAPH
<br/> - UNIT BREAK

and when there is some text(without tags), it should be recognized as UNIT TEXT.
if tag from html file is not listed above, it should be shown as UNKNOWN UNIT.

thats it....

[dagon]

how are are defining UNIT it does not exist in the specs in the way you are using the word

[GoRide_]

ok, my english is bad, and i am confusing. i will try to explain it more simple.

i need php script which will open html file located on same server ...fopen("test.html") ..

then i need to read line by line and if php script find in html file, for example

<a href="http://google.com">GOOGLE</a>, i want to print that in next format:

LINK: google(http://google.com)

so, script need to recognize it is link and print in format above.

i need same for <img> tag, <p>, <br>.... if script find BR in text it will show BREAK LINE

i did that for A HREF

   if(preg_match('/<a href=\"(.*?)\">.*?<\/a>/i', $buff2))
   	   							{
      								$buff2 = preg_replace('/<a href=\"(.*?)\">(.*?)<\/a>/i', "\n".'LINK: $1 ($2)', $buff2);
								}

but, i tried to do for <img> and it doesn't work .

   elseif(preg_match('/<img src=\"(.*?)\" alt=\"(.*?)\"\/>/i', $buff2))
   	   							{
      								$buff2 = preg_replace('/<img src=\"(.*?)\" alt=\"(.*?)\"\/>/i', "\n".'IMAGE: $1', $buff2);
								}

also, when there is only text, which is not located between some tags, it need to be recognized as TEXT: part of text(10 characters for example).

      $buff2 = preg_replace("/(.+)\n/i", 'TEXT: $1' . "\n", $buff2);

problem is that, when there is empty line in html file, php script show it as TEXT: and there is nothing....

maybe now it's easier to understand.

[Weedpacket]

GoRide! wrote:

can i load wml file using dom->loadXML(); ?

Since WML is an application of XML, the answer is yes.

Surprisingly, the DOM extension can also read HTML, even though HTML is not XML.

[GoRide_]

hm okay...but, can any body tell me what's the problem with this code

 elseif(preg_match('/<img src=\"(.*?)\" alt=\"(.*?)\"\/>/i', $buff2))
                                      {
                                      $buff2 = preg_replace('/<img src=\"(.*?)\" alt=\"(.*?)\"\/>/i', "\n".'IMAGE: $1', $buff2);
                                } 

[Weedpacket]

Apart from the fact that you're not using DOM? 🙂

The most obvious thing to note is that the preg_match() test gains you nothing except having to do everything twice if there is a match.

It also requires

1. both src= and alt= attributes

2. in that order

3. no other attributes

4. a single space between the "img" and the start of the src= attribute

5. double-quoted attribute values

6. no spaces around the '=' separating either attribute name from its value

7. a single space between the ending quote on the src= attribute and the start of the alt= attribute

8. no space after the end of the alt= attribute

9. nothing between the end of the alt= attribute and the / at the end of the tag

10. and a / at the end of the tag.

If any of these conditions are not met then the pattern will not match.

Apart from that, there's nothing wrong with the pattern (the double quotes don't need to be escaped as they're not considered significant, but since they aren't significant the extra escapes get ignore anyway). Obviously it won't work if $buff2 is the wrong variable to begin with, but like I said, that's obvious.

[GoRide_]

Ok, so, i need to work with php DOM?
is there any example how to do this with dom?

[Weedpacket]

$doc = new DOMDocument();
$doc->loadHTMLFile('source.html');

$images = $doc->GetElementsByTagName('img');
while($images->length)
{
	$image = $images->item(0);
	$label = "IMAGE: ".$image->getAttribute('src');
	$image->parentNode->replaceChild($doc->CreateTextNode($label), $image);
}

$doc->saveHTMLFile('target.html');

Of course, that's only an example. The real thing would no doubt require a bit of thought.

[GoRide_]

hm, but how can I get atributes for all tags?

i am selecting all tags from html file with

$tags = $doc->GetElementsByTagName('*');

problem is that i cannot select only images or some tag, it must be sorted as in file.

if first line in file is <p>bla</p>, it must be first in "unit" list.

any help?

[Weedpacket]

The list of elements returned by GetElementsByTagName is in document order.

Now that you've got a list of all the elements in the document, the obvious thing to do would be to go through each element of the list and, depending on what sort of element it is, do something appropriate.

[GoRide_]

sorry, i bother

but, if you have time please write example for working with some tag (no matter which), when all tags are loaded?

[Weedpacket]

$doc = new DOMDocument();
$doc->loadHTMLFile('test.html');

$elements = $doc->GetElementsByTagName('*');
$length = $elements->length;
for($i=0; $i<$length; ++$i)
{
	$element = $elements->item($i);
	switch($element->tagName)
	{
	case 'img':
		echo "An image, src = ".$element->getAttribute('src')."\n";
		break;
	case 'p':
		echo "A paragraph\n";
		break;
	case 'br':
		echo "Break\n";
		break;
	}
}