home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!sdrc!thor!scjones
- From: scjones@thor.sdrc.com (Larry Jones)
- Newsgroups: comp.lang.c
- Subject: Re: Questions about token merging and trigraphs
- Message-ID: <2404@sdrc.COM>
- Date: 16 Dec 92 22:47:40 GMT
- References: <42098@pprg.eece.unm.edu.pprg.unm.edu> <1TA04E3@cdis-1.compu.com>
- Sender: news@sdrc.COM
- Lines: 27
-
- In article <1TA04E3@cdis-1.compu.com>, tanner@cdis-1.compu.com (Dr. T. Andrews) writes:
- > Trigraphs are an unpleasant invention from the successor to the
- > horse committee. You will recall that the horse committee gave
- > us the camel. Well, X3J11 gave us trigraphs. I believe that the
- > intention was to prove that standards groups do better when they
- > stick with endorsing widely-used practice.
-
- If you have a better solution to not having some of the C source
- characters available in your local character set, I'd like to hear it.
- Remember that it must work for preprocessor commands (since '#' is one
- of the questionable characters) and in character and string literals as
- well as in more ordinary contexts. X3J11 recognizes that trigraphs are
- a terrible solution to the problem, but they're better than any other
- solution we could find.
-
- > The change means that certain old programs may have strings, most
- > likely error messages, compiled into something unexpected.
-
- The specification for trigraphs was carefully crafted to avoid as much
- breakage as possible. The most likely problem is a string containing
- the punctuation "??!" which will be interpreted as "|", but this is a
- non-standard punctuation ("?!?" and "!?!" are much more common) -- lots
- of people grep'ed lots of code and I think found exactly one occurrance.
- ----
- Larry Jones, SDRC, 2000 Eastman Dr., Milford, OH 45150-2789 513-576-2070
- larry.jones@sdrc.com or ...uunet!sdrc!larry.jones
- I think your train of thought is a runaway. -- Calvin's Mom
-