Parser for Regexp

[Andrea_44]

I am attempting to write a program to interpret PCRE (perl compatible regular expression) into literal text.

For example:

$./MyProgram /^MDTM(?!\n)\s[^\n]{100}/smi

Output:

MDTM 1234567890!@£$%^&*()ABCDEFGHIJaaaaaaaaaaaaaa...

Are there any existing program that can do this already?
or
Could you please give me some ideas on how I should approach the issues?
i.e. Can I use "Lex"? Is it a good idea to use C or Java? What libraries should I use?

Any help will be very nice.

Thanks,
Andrea

[nereid]

Why don't you use Perl? BTW there are some web regexp parsers around. You can give them a regexp and some text and they will highlight the matched pattern.

Edit:
A program which parses the regexp themselves is a little bit tricky to do, which requires some serious knowledge of them. If you've got the knowledge you don't need the program anymore ;)

[Andrea_44]

BTW there are some web regexp parsers around. You can give them a regexp and some text and they will highlight the matched pattern.


Thanks for the reply, but that is not quite what I meant.

What I meant was that you give my program a regexp and it will generate some text that will matched the pattern.

For example: MyProgram can parse in the Regexp below....

$./MyProgram /^MDTM(?!\n)\s[^\n]{100}/smi

And then generate the text below that will match the Regexp MyProgram has just parsed...

MDTM 1234567890!@£$%^&*()ABCDEFGHIJaaaaaaaaaaaaaa...

Is it even possible to do that?

Cheers~
Andrea

FYI, I am attempting to write a program that parse in Snort rules(containing PCRE) and generate packets (containing payload that matches the PCRE) which will trigger the Snort rules.

[nereid]

It should be possible but requires some serious thought. Which I ain't have the time for at the moment.

You should parse the regexp. Take a look at the modifiers behind them and watch out for modifiers in the regexp. Second look out for the ^ and the $. Then you'll have to model the lookahead and lookbehing modifier. And so on...This requires some heavy knowledge and in the end you'll build a compiler from regexp to plain text, which isn't easy.