Checking for ASCII 7-bit Cleaness

Checking for ASCII 7-bit Cleaness

[Nordlöw]

How do I check that a buffer is 7-bit ASCII-clean, that is it only
contains characters between 0 and 127 inclusively.

I guess it should all start with find-file-literally().

Performance is of high priority in this case.

Thanks in advance,
Nordlöw

Re: Checking for ASCII 7-bit Cleaness

[Nordlöw]

In C I would have stepped forward 4 bytes (except for the resting 1-3
bytes of the file of course) and for each such block x then tested
that (x & 0x80808080) equals zero. If so the file is 7-bit clean.
Could we do something similar with elisp regular expression
operations?

/Nordlöw

Re: Checking for ASCII 7-bit Cleaness

[Hallvard B Furuseth]

Nordlöw writes:
> In C I would have stepped forward 4 bytes (except for the resting 1-3
> bytes of the file of course) and for each such block x then tested
> that (x & 0x80808080) equals zero. If so the file is 7-bit clean.
> Could we do something similar with elisp regular expression
> operations?

Don't think so.  That's a C internals trick, abusing how integers and
bytes are represented.

(defun buffer-ascii-clean-p ()
  (save-excursion
    (goto-char (point-min))
    (skip-chars-forward "[\0-\177]")
    (eobp)))

-- 
Hallvard