Jiri Moskovcak <[email protected]> writes:
> On 02/13/2012 12:03 PM, Nikola Pajkovsky wrote:
>> test 'Attaching better backtrace' is broken. problem dir contains
>> backtrace_rating 1 and servers response is backtrace_rating 3 ->> never
>> upload.
>>
>> Signed-off-by: Nikola Pajkovsky<[email protected]>
>> ---
>> src/plugins/reporter-bugzilla.c | 16 +++++++++-------
>> src/plugins/rhbz.c | 20 ++++++++++----------
>> 2 files changed, 19 insertions(+), 17 deletions(-)
>>
>> diff --git a/src/plugins/reporter-bugzilla.c
>> b/src/plugins/reporter-bugzilla.c
>> index cd31f5b..c49ffe2 100644
>> --- a/src/plugins/reporter-bugzilla.c
>> +++ b/src/plugins/reporter-bugzilla.c
>> @@ -402,13 +402,14 @@ int main(int argc, char **argv)
>> strbuf_append_strf(full_desc, "%s: %s\n", FILENAME_RATING,
>> rating_str);
>> strbuf_append_strf(full_desc, "Package: %s\n", rhbz.b_package);
>> /* attach the architecture only if it's different from the
>> initial report */
>> - if ((strcmp(bz->bi_platform, "All") != 0)&&
>> - (strcmp(bz->bi_platform, "Unspecified") != 0)&&
>> - (strcmp(bz->bi_platform, rhbz.b_arch) !=0))
>> + if ((strcmp(bz->bi_platform, "All") != 0)
>> +&& (strcmp(bz->bi_platform, "Unspecified") != 0)
>> +&& (strcmp(bz->bi_platform, rhbz.b_arch) !=0))
>> strbuf_append_strf(full_desc, "Architecture: %s\n",
>> rhbz.b_arch);
>> else
>> {
>> - VERB3 log("not adding the arch: %s because rep_plat is %s",
>> rhbz.b_arch, bz->bi_platform);
>> + VERB3 log("not adding the arch: %s because rep_plat is %s",
>> + rhbz.b_arch, bz->bi_platform);
>> }
>> strbuf_append_strf(full_desc, "OS Release: %s\n",
>> rhbz.b_release);
>>
>
> - just an indent fixes, please don't do that, it just ruins git
> history, these kind of problems should be noted at the review time,
> once it's pushed it's too late
hmm, ok
>> @@ -418,8 +419,8 @@ int main(int argc, char **argv)
>>
>> "is_private");
>> */
>>
>> - int allow_comment = is_comment_dup(bz->bi_comments,
>> full_desc->buf);
>> - if (!allow_comment)
>> + int dup_comment = is_comment_dup(bz->bi_comments,
>> full_desc->buf);
>> + if (!dup_comment)
>> {
>> log(_("Adding new comment to bug %d"), bz->bi_id);
>> rhbz_add_comment(client, bz->bi_id, full_desc->buf, 0);
>> @@ -434,7 +435,8 @@ int main(int argc, char **argv)
>> /* python doesn't have rating file */
>> if (rating_str)
>> rating = xatou(rating_str);
>> - if (!allow_comment&& (bz->bi_best_bt_rating< rating))
>> +
>> + if (!dup_comment&& (rating> bz->bi_best_bt_rating))
>> {
>
> - here it's just a variable rename, so the same as the above applies,
> but since the previous name was really wrong (wasn't that mine??) I
> will look away and let you push it :)
it's mine stupid variable name ;)
>> char bug_id_str[sizeof(int)*3 + 2];
>> sprintf(bug_id_str, "%i", bz->bi_id);
>> diff --git a/src/plugins/rhbz.c b/src/plugins/rhbz.c
>> index fbf5bba..ba88110 100644
>> --- a/src/plugins/rhbz.c
>> +++ b/src/plugins/rhbz.c
>> @@ -86,22 +86,22 @@ static char *trim_all_whitespace(const char *str)
>> return trim;
>> }
>>
>> -int is_comment_dup(GList *comments, const char *comment)
>> +int is_comment_dup(GList *srv_comments, const char *usr_comment)
>> {
>> - for (GList *l = comments; l; l = l->next)
>> + char *trim_usr_comment = trim_all_whitespace(usr_comment);
>> + for (GList *l = srv_comments; l; l = l->next)
>> {
>> - char *comment_body = (char *) l->data;
>> - char *trim_comment_body = trim_all_whitespace(comment_body);
>> - char *trim_comment = trim_all_whitespace(comment);
>> - if (!strcmp(trim_comment_body, trim_comment))
>> + char *srv_comment = (char *) l->data;
>> + char *trim_srv_comment = trim_all_whitespace(srv_comment);
>> + if (!strcmp(trim_srv_comment, trim_usr_comment))
>> {
>> - free(trim_comment_body);
>> - free(trim_comment);
>> + free(trim_usr_comment);
>> + free(trim_srv_comment);
>> return 1;
>> }
>> }
>>
>
> - and here it's really not that useful, so please don't do it...
it's way better to read, but I can push another dummy commit with
variable renaming
>> - return 0;;
>> + return 0;
>> }
>>
>> static unsigned find_best_bt_rating_in_comments(GList *comments)
>> @@ -122,7 +122,7 @@ static unsigned find_best_bt_rating_in_comments(GList
>> *comments)
>> continue;
>> }
>>
>> - start_rating_line += strlen("rating: ");
>> + start_rating_line += strlen(FILENAME_RATING": ");
>
> - this is the actual fix? hm... looks too *easy* ;)
yes :)
>> char *end_rating_line = strchr(start_rating_line, '\n');
>> if (!end_rating_line)
>> VERB3 error_msg("broken comment body");
--
Nikola
