[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: CVS Bison (2002-10-20) creates empty structures and initializers
From: |
Akim Demaille |
Subject: |
Re: CVS Bison (2002-10-20) creates empty structures and initializers |
Date: |
23 Oct 2002 11:07:38 +0200 |
User-agent: |
Gnus/5.0808 (Gnus v5.8.8) XEmacs/21.4 (Honest Recruiter) |
>>>>> "Paul" == Paul Eggert <address@hidden> writes:
>> From: Akim Demaille <address@hidden> Date: 23 Oct 2002 09:14:27
>> +0200
>> | it's easier for non-M4 programmers to follow what's going on.
>>
>> But people trying to copy-cat a CPP skeleton for a CPP-less
>> language will fail.
Paul> That's OK, and indeed expected; people using other programming
Paul> languages will expect to have to write skeletons in those
Paul> languages, and not in C.
Paul> Anyway, I was more worried about Bison users (i.e., ordinary
Paul> programmers trying to follow what a parser is doing by reading
Paul> skeletons and/or debugging the generated code) than about the
Paul> relatively few experts who are writing skeletons.
Honestly, I find that the parsers that we produce are much more
readable than they used to be. Novice can read readable things, we
have much less cascades of #if to cope with the combinatoric explosion
of options. Using M4, I do believe we produce better code (understand
``more legible''), and in most cases, more readable skeletons.
But undoubtedly, had we had a better M4, we could do even better :(
It seems that stealing some of M5 ideas to put into M4 would be a
great win. Not only for clarity, but I tend to think that for speed
too.