parse2.com

[OrangeHopper]

Well, after many an hour snatched here and there, out of the hands of my demanding children, I have finally got my website up and running.

parse2.com

I would appreciate it if people could give it a once over and look for typos, missing links, bits that don't format properly etc. If you do experience any formatting problems could you let me know which browser you are using.

Many thanks in advance.

[Cowboy Bob]

Website looks fine and the project looks really interesting. Before really diving in, I've got a couple of questions:-

I've used ANTLR before - http://www.antlr.org/

What are the main differences between your approach and theirs?
Why should I choose your software over ANTLR?

[OrangeHopper]

Hello Cowboy,

When I started out I looked at ANTLR and JavaCC but couldn't get either of them to deal with the types of protocol/messages I was trying to parse. This may have been due to ignorance of the products but I then started looking at parsers in general and didn't try them again.

The main difference that I was interested in was that Augmented BNF has "variable repetition".

<a>*<b>element

This is not present in Extended BNF which most other parsers use.

I have been working on a messaging system for a few years where the messages do not comform to a grammar where every field is delimited by a punctuation character. Identifying one field from the next is often based on length and format which is much easier to define with Augmented BNF.

The other comment is that I found the above compiler compilers relatively difficult to master. I wanted a system where the grammar file was all you needed. Probably find I will learn the lessons they learned many years ago and mine will get more complicated.

Sorry, not a particularly well thought out answer but it is the power of the "variable repetition" construct that I needed.

[TheRightStuff]

I wish i was that clever. you is like einstein in it.

[OrangeHopper]

I wish i was that clever. you is like einstein in it.

Don't you mean, "You is like Threaded".

[MrsGoof]

Don't you mean, "You is like Threaded".

If that were to be true then more pies would be required

[Cowboy Bob]

The main difference that I was interested in was that Augmented BNF has "variable repetition".

<a>*<b>element

This is not present in Extended BNF which most other parsers use.

Sounds interesting. I'll play around with it a bit, and you never know, it might end up on my "possible solutions" list.

[OrangeHopper]

Ta Cowboy.

Obviously let me know via the parse2 email if you have any questions or find any bugs.

[Cowboy Bob]

Any chance of the code? Or are you unable to open source it? I prefer my libraries to come with source code.

[OrangeHopper]

Sorry but for the time being I'm not going to make the code available.