Tough Regular Expressions string replace

replace string patterns

[Garoun]

I'll use two examples to display my problem. Basicaly I'm writing and application text type editor that sends its output in plain text to the database upon submitting. What I need is a regular expression that removes all of the special HTML tag formatting that MSWord is so kind as to paste in with.

(Side Note: I've thought about using the inner references which are another option 'currently', so suggestions to that extent are welcome also)

String 1 Example: (assume Str1 = )

Str1 = "<p class=MsoNormal><b style='mso-bidi-font-height:normal'>Glossary of assumed
terms:<o:p></o:p></b></p>"

String 2 Example: (assume Str2 = )

Str2 = "<li class=MsoNormal style='mso-list:l0 level1 lfo1;tab-stops:list .5in'>Desktop</li>"

Currently I don't care about lost formatting, I simply want all of the tags gone. and the plain text output. Hopefully the regular expressions are pretty straight forward and any ideas would be greatly appreciated ASAP.

Thanks

[claus]

>> I simply want all of the tags gone. and the plain text output

$someString = <li class=MsoNormal style='mso-list:...
$someString =~ s/<.*>//igm

Perl Regexp syntax, line two is what you need, line 1+2 is how to do it. (igm = ignore case, global, multiline)

/claus

---

Welcome to WebmasterWorld Garoun :)

[Garoun]

My fault on the terminology. You definately gave me a good starting point for deciphering what I need :) thanks a lot.

I shouldn't need the multiline flag but I'll keep that in mind if my data isn't kicking out quite right which will most likely be from the lovely carriage returns.

Now I just need to convert your perl RegExp over to javascript for use in html/javascript which won't be very hard... I"ll be sure to come up with a harder question later. :D

Thank you for the prompt reply, I was crossing my fingers.

I will get this regexp stuff figured out eventually, it's all too powerful of an option not to learn.

[timster]

$someString =~ s/<.*>//igm ;

That could remove more than the doctor intended (i.e., everything between the first "<" and the last ">" on a line).

These might be better:

$someString =~ s/<.*?>//gis ;
$someString =~ s/<[^>]*>//gis ;

[Garoun]

Yeah, the original peice Claus gave was doing a bit much once I tested it, so my 'current' implementation uses:

temp2_field_val = temp_field_val.replace(/<.*?>/ig," ");

Where temp_field_val is my text string I want stripped

[Storyteller]

Garoun, for this particular task you're much better off using HTML::TreeBuilder or even W3C's HTML Tidy utility (if all you want is to scratch MS Word artifacts).

With just the regular expressions you'll have a hard time covering numerous special cases that will totally screw such a straightforward "parser".

[Garoun]

Thanks for the info Storyteller, I'll be working more with my parse and such later this month. I will definately check out those items you mentioned to see how they could help with the issues I encounter as the application advances.

[timster]

Very true Storyteller. Still, I've never seen MS Word create any HTML that the one-liner can't handle.

[Garoun]

Well its already been asked of me to attempt to 'save' the formatting and convert it to a format I can use
aka:
<UL class=asfawref.sdfsd sdfasd sadfasd>Text</UL> -> <UL>Text</UL>

It won't be easy at all due to man variations word adds, but if it was easy there'd be more people doing what we do :D.

Garoun