Intelligent replace in HTML source

[Ima2003]

Hello, gurus!

As I can see this question is very important for lot of developers and of course for me. So let's join our efforts to resolve this.

So. We have HTML source code like this.

.... <tag1> <tag2>...</tag2>
some text to replace or Replace or replaceme
<tagN>... </tagN> some text here ...

and so on. This is a arbitrary source.
Also we have dictionary with pair of words, for example
$pattern = ("replace", "text", ....)
$replace = ("rpl", "txt", ...)

it means that we should find all whole words "replace" and replace them to "rpl", find all whole words "text" and replace them to "txt" and so on.. OUT OF HTML TAGS!

This task looks like a little web translator word-by-word.

So. The question. How can I solve this task? This translator should work VERY VERY fast. Much faster than preg_replace with two arrays on whole html source.

Thanks!

[dalecosp]

[man]str_replace[/man] is faster than any regular expression call.

Now, can you clarify your original question?

By way of example, I'm thinking you want to replace every reference to the word "size" in an HTML page, but not every reference to <font size>?

Is that correct ... if so, Hmm....

[Ima2003]

By way of example, I'm thinking you want to replace every >reference to the word "size" in an HTML page, but not every >reference to <font size>?

Yes, that is correct.

But str_replace is case sensitive and also replace partial word like "oversize". And this function replace everywere (inside html tags)

So is this clear for you? And what is proposal to solve this?

Thanks!

[Ima2003]

To simplify a problem let's translate not html but simply text.
Ok?

[Bijan]

I really don't think that you have any other way but using reg exp, me thinks!

[Ima2003]

I'we already use regexp but it is too slow on big pages...

Any comments?