Eclipse Community Forums: TMF (Xtext) » Keyword as ID (context sensitive)

Help

Home

Home » Modeling » TMF (Xtext) » Keyword as ID (context sensitive)

Show: Today's Messages :: Show Polls :: Message Navigator

Keyword as ID (context sensitive) [message #954726]

Tue, 23 October 2012 07:44

Bryce Brinkley

Messages: 8
Registered: October 2011

Junior Member

Hello!

I am trying to implement a rather lengthy parameter file syntax in Xtext. I have defined a terminal for arbitrary IDs which looks like this:

terminal ID returns ecore::EString:
	('a'..'z' | 'A'..'Z' | '_') ('a'..'z' | 'A'..'Z' | '0'..'9' | '_' | '*' | '-' | ':' | '.')*;

This means an ID has to start with a letter or "_" and can then contain any number of letters, numbers, "_", "*" or "-".

Now I run into problems as soon as there comes an ID which is already defined as a keyword. The grammar contains around 270 keywords: For example words like "n", "Min", "Angle" or "Energy".
At some places I even _want_ an ID to be one of these keywords.

Is there an easy way to tell Xtext to ignore keywords at certain places and treat them as IDs? (enable context sensitive parsing) As far as I see it should be clear from the context at which places that is.

Thanks in advance!

B.

Report message to a moderator

Re: Keyword as ID (context sensitive) [message #954759 is a reply to message #954726]

Tue, 23 October 2012 08:15

Vlad Dumitrescu

Messages: 431
Registered: July 2009
Location: Gothenburg

Senior Member

Hi!

You can use

KW: 'n' | 'Min' | 'Angle' | ...
NAME: ID | KW;

and instead of ID, use NAME.

best regards,
Vlad

[Updated on: Tue, 23 October 2012 08:15]

Report message to a moderator

Re: Keyword as ID (context sensitive) [message #954960 is a reply to message #954759]

Tue, 23 October 2012 11:47

Bryce Brinkley

Messages: 8
Registered: October 2011

Junior Member

Vlad: Thank you for the quick response.
It's not the nicest solution, but it works like that.
Many thanks!

B.

[Updated on: Tue, 23 October 2012 11:47]

Report message to a moderator

Re: Keyword as ID (context sensitive) [message #955965 is a reply to message #954960]

Wed, 24 October 2012 06:11

Markus Wolf

Messages: 153
Registered: July 2009

Senior Member

Hi, I have a similar problem.
The rules KW and NAME you state here, are they dataTypeRules or terminal rules?

Report message to a moderator

Re: Keyword as ID (context sensitive) [message #955974 is a reply to message #955965]

Wed, 24 October 2012 06:17

Christian Dietrich

Messages: 14669
Registered: July 2009

Senior Member

Data type rules

--
Need training, onsite consulting or any other kind of help for Xtext?
Go visit http://xtext.itemis.com or send a mail to xtext at itemis dot de

Twitter : @chrdietrich
Blog : https://www.dietrich-it.de

Report message to a moderator

Re: Keyword as ID (context sensitive) [message #955975 is a reply to message #955974]

Wed, 24 October 2012 06:23

Markus Wolf

Messages: 153
Registered: July 2009

Senior Member

Thanks, it works.
Is there a cleaner solution? I mean I could stick with that, but I'm really new to Xtext and would like to learn bit more in detail.

Report message to a moderator

Re: Keyword as ID (context sensitive) [message #955984 is a reply to message #955975]

Wed, 24 October 2012 06:27

Christian Dietrich

Messages: 14669
Registered: July 2009

Senior Member

Hi a bit more concise would be: myid: "mykw" | ID;
Never the less it is THE solution

--
Need training, onsite consulting or any other kind of help for Xtext?
Go visit http://xtext.itemis.com or send a mail to xtext at itemis dot de

Twitter : @chrdietrich
Blog : https://www.dietrich-it.de

Report message to a moderator

Re: Keyword as ID (context sensitive) [message #957449 is a reply to message #955984]

Thu, 25 October 2012 08:21

Henrik Lindberg

Messages: 2509
Registered: July 2009

Senior Member

On 2012-24-10 8:27, Christian Dietrich wrote:
> Hi a bit more concise would be: myid: "mykw" | ID;
> Never the less it is THE solution
>
Actually, it is THE solution under normal circumstances.

The alternative is to treat the problem as a lexical issue and implement
an external lexer. This way they grammar does not see keyword tokens
where it is not supposed to see them.

The lexer however has no notion of the grammar, so it has to be driven
by lexical content alone.

- henrik

Report message to a moderator

Previous Topic:	Xtext as Plug-In Editor
Next Topic:	How to provide a different error message for 'Couldn't resolve reference to..' error

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

]

Current Time: Fri Apr 26 22:06:06 GMT 2024

.:: Contact :: Home ::.

Breadcrumbs

Sign up to our Newsletter