Prevent no branches #1074
Prevent no branches #1074
Conversation
|
This pull request fixes 1 alert when merging f2bb649 into 5d3c500 - view on LGTM.com fixed alerts:
|
|
Sorry I think I broke the CI with é in my last name 😅 |
|
Yeah. Perhaps my script could be setup to decode utf8. Try changing |
src/branches.cpp
Outdated
| @@ -269,6 +272,10 @@ namespace l | |||
| int | |||
| erase_begin(Branches::Impl *branches_) | |||
| { | |||
| if (branches_->size() <= 1) { | |||
There was a problem hiding this comment.
If you don't mind updating the code to use a more GNU style for braces and such. And I'm fine with something like
if(branches_->size() <= 1)
return -ENOTSUP;
There was a problem hiding this comment.
OK, updated the code to better match the style
rienafairefr
force-pushed
the
prevent-no-branches
branch
from
f2bb649 to
ece486d
Compare
|
This pull request fixes 1 alert when merging ad6c93f into 5d3c500 - view on LGTM.com fixed alerts:
|
|
This pull request fixes 1 alert when merging 5d61a5b into 5d3c500 - view on LGTM.com fixed alerts:
|
rienafairefr
force-pushed
the
prevent-no-branches
branch
from
5d61a5b to
80e0752
Compare
|
This pull request fixes 1 alert when merging 80e0752 into 5d3c500 - view on LGTM.com fixed alerts:
|
rienafairefr
force-pushed
the
prevent-no-branches
branch
2 times, most recently
from
f90ab5f to
694b819
Compare
|
This pull request fixes 1 alert when merging 694b819 into 5d3c500 - view on LGTM.com fixed alerts:
|
rienafairefr
force-pushed
the
prevent-no-branches
branch
from
694b819 to
f283cba
Compare
|
This pull request fixes 1 alert when merging f283cba into 5d3c500 - view on LGTM.com fixed alerts:
|
|
Ugh. Sorry. Python is a major pita regarding char encoding. I can try to fix it and then you can submit the PR for the original issue. Or if you want to see it through feel free. |
|
Can't you just set the ENV to UTF-8 encoding? Either through PYTHONIOENCODING="UTF-8" or the LANG/LANGUAGE/LC_ALL? |
as discussed in #1073, this seems to fix it.
returns "not supported" ENOTSUP when a change would result in no branches.
incidentally, the code for branches
from_stringwas already raising/erroring out in case of setting completely empty branches, that's fixed. I've added unit test cases to test all that.On my system, the
-fsanitize=undefinedoption in the compiler added some problems that seem to be fixed by also adding the-fsanitize=undefinedflag on linking, not sure why, that's why the change is in there,