[changeset] uninitialized value bug in strncmp

[John W. Eaton]

On  5-Jan-2009, Thorsten Meyer wrote:

| Jaroslav Hajek wrote:
| > On Thu, Jan 1, 2009 at 1:20 PM, Thorsten Meyer <thorsten.meyier at gmx.de> wrote:
| >   
| >> Hi,
| >>
| >> here is a patch (against mercurial) for the following bug in strncmp:
| >>
| >> comparing a string with a cell array, the return value for non-character cells is random:
| >>
| >> octave:1> strncmp("abc", {"abcd", 10}, 2)
| >> ans =
| >>
| >>     1   164
| >>
| >> octave:2>
| >> octave:2> strncmp("abc", {"abcd", 10}, 2)
| >> ans =
| >>
| >>     1   191
| >>
| >> octave:3> strncmp("abc", {"abcd", 10}, 2)
| >> ans =
| >>
| >>     1   184
| >>
| >> octave:4> strncmp("abc", {"abcd", 10}, 2)
| >> ans =
| >>
| >>     1   138
| >>
| >> The reason is that in this particular case, the output value is left uninitialized.
| >>
| >> regards
| >>
| >> Thorsten
| >>
| >> _______________________________________________
| >> Bug-octave mailing list
| >> Bug-octave at octave.org
| >> https://www-old.cae.wisc.edu/mailman/listinfo/bug-octave
| >>
| >>     
| >
| > Attachment missing?
| >   
| Yes, sorry, here it is.

It seems better to me to simply initialize all elements of the output
vector to false when the value is constructed so I checked in the
following change instead.

Thanks,

jwe


-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: diffs
Url: https://www-old.cae.wisc.edu/pipermail/bug-octave/attachments/20090112/66eb68aa/attachment.ksh