Eclipse Community Forums: TMF (Xtext) » How to write terminals/rules with same subset of characters

Help

Home

Home » Modeling » TMF (Xtext) » How to write terminals/rules with same subset of characters

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

How to write terminals/rules with same subset of characters [message #1494778]

Mon, 01 December 2014 19:04

Grégoire Avot

Messages: 7
Registered: December 2014

Junior Member

Hello Forum,

I am new user on Xtext, I want to understand how to be able to parse the following input:

129
o 37

Each line is either
- a decimal formed of one or more DECIMAL_DIGIT: '0'..'9'
- an octal value: 'o' <optional space> OCTAL_DIGIT: '0'..'7'

I am not able to find an elegant solution, based either on terminal or parser rules.

Does anybody have the solution,

Thank you very much in advance,

Regards,

Grégoire.

Report message to a moderator

Re: How to write terminals/rules with same subset of characters [message #1495321 is a reply to message #1494778]

Tue, 02 December 2014 05:39

Christian Dietrich

Messages: 14665
Registered: July 2009

Senior Member

What about keeping the int rule and introducing a data type rule for the octal one?

Octal: "o" INT;

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

Report message to a moderator

Re: How to write terminals/rules with same subset of characters [message #1495350 is a reply to message #1495321]

Tue, 02 December 2014 06:14

Grégoire Avot

Messages: 7
Registered: December 2014

Junior Member

Hello Christian

Thank you for your reply.

I do not understand your reply: your proposed parser rule will accept "o 78", which is not correct.

Today the only solution I found is to implement a rule in the way you proposed, and use a Validator to check legal value. But in my mind I think there could be a purely grammatical solution.

Regards,

Grégoire.

Report message to a moderator

Re: How to write terminals/rules with same subset of characters [message #1495359 is a reply to message #1495350]

Tue, 02 December 2014 06:24

Christian Dietrich

Messages: 14665
Registered: July 2009

Senior Member

Hi pure grammatical solutions are pain. Especially if you want to support binary and hexadecimal as well. But the same principal should be possible if you introduce terminals for 0+1 and 2 3 4 5 6 7 and 8 9 and datatype rules for INT and OCTAL (and maybe ID as well) but as said before this wil be pita

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

Report message to a moderator

Re: How to write terminals/rules with same subset of characters [message #1495519 is a reply to message #1495359]

Tue, 02 December 2014 09:23

Grégoire Avot

Messages: 7
Registered: December 2014

Junior Member

Thank you Christian for your feedback. You are right, next step is Hex + Binary. Furthermore, real example I have to handle is much more complex.
So I will continue to use Validator for these local check to avoid a too complex Xtext rules. But I keep in mind the purpose of Validator is to handle semantic, no grammar.
Regards,
Grégoire.

Report message to a moderator

Previous Topic:	simple xbase project causes antlr timeout
Next Topic:	Internal "Literals" class not generated after grammar change

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

]

Current Time: Thu Apr 25 17:24:28 GMT 2024

.:: Contact :: Home ::.

Breadcrumbs

Sign up to our Newsletter