bison-patches
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [PATCH]: Suggested documentation about working with Bison versions.


From: Paul Eggert
Subject: Re: [PATCH]: Suggested documentation about working with Bison versions.
Date: Wed, 14 Oct 2020 11:38:14 -0700
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101 Thunderbird/68.10.0

On 10/13/20 5:11 PM, Kaz Kylheku wrote:

     doc: document best deployment practices.

Thanks, this sort of thing should be helpful. I see you haven't signed copyright papers for Bison, though. I assume you know how to jump through the hoops? If not, I can send you email as to how to get the ball rolling.

+Bison provides a Yacc compatibility mode in which it strives to conform with
+the POSIX standard. Grammar files which are written to the POSIX standard, and
+do not take advantage of any of the special capabilities of Bison, are very
+likely to work with many version of Bison without modification.

very likely to work with many version of->
should work with

Also, the Texinfo input should have two spaces after ".", if that's the style used elsewhere in the Bison manual.

+Some features of Bison have been, or are being adopted into other Yacc-like
+programs. Therefore it might seem that is a good idea to write grammar code
+which targets multiple implementations, similarly to the way C programs are
+often written to target multiple compilers and language versions. This practice
+is not highly recommended, however.

Remove "This practice is not highly recommended, however." It conveys little info and isn't needed in the context of the paragraph.

+Developers who strive to make their Bison code simultaneously
+compatible with other parser generators are encouraged to nevertheless use
+specific versions of all generators, and still follow the recommended practice
+of shipping generated output.

Not sure what is being recommended here. A developer is supposed to list all versions of all generates the grammar is capable with, and test them all, but generate a tarball with one of them? Or generate multiple tarballs?

Perhaps supply a small example of the issue. That would clarify the intent, I expect.



reply via email to

[Prev in Thread] Current Thread [Next in Thread]