Re: [PATCH] Document the grep workaround.

[Eric Blake]

On 05/20/2010 02:34 PM, Paolo Bonzini wrote:
>>>> +but works with @code{grep -c '^'}.  Another alternative for counting
>>>> +lines is to use @code{wc -l}.
>>>
>>> I think the most portable way to count lines is "sed -n '$='".
>>
>> Is there a decided advantage over 'wc -l' such as allowed maximum or so?
> 
> No, but wc is not including in the GNU coding standards' list of
> portable utilities for configure/make.

Good point.  I'll push this patch shortly:

diff --git i/ChangeLog w/ChangeLog
index f7f50e9..fc0c3cc 100644
--- i/ChangeLog
+++ w/ChangeLog
@@ -1,3 +1,10 @@
+2010-05-20  Eric Blake  <address@hidden>
+
+       Mention another line-counting alternative.
+       * doc/autoconf.texi (Limitations of Usual Tools) <grep>: Mention
+       how to use sed to count lines.
+       Suggested by Paolo Bonzini.
+
 2010-02-25  Eric Blake  <address@hidden>

        Properly quote AC_PREREQ during autoupdate.
diff --git i/doc/autoconf.texi w/doc/autoconf.texi
index 51fe0f4..1c85191 100644
--- i/doc/autoconf.texi
+++ w/doc/autoconf.texi
@@ -17869,8 +17869,8 @@ Limitations of Usual Tools
 status of @code{grep} to determine whether it found a match.

 The QNX4 implementation fails to count lines with @code{grep -c '$'},
-but works with @code{grep -c '^'}.  Another alternative for counting
-lines is to use @code{wc -l}.
+but works with @code{grep -c '^'}.  Other alternatives for counting
+lines are to use @code{sed -n '$='} or @code{wc -l}.

 Some traditional @command{grep} implementations do not work on long
 input lines.  On AIX the default @code{grep} silently truncates long


-- 
Eric Blake   address@hidden    +1-801-349-2682
Libvirt virtualization library http://libvirt.org

signature.asc
Description: OpenPGP digital signature