rangify readLink and symlink#3967
Conversation
aG0aep6G
force-pushed
the
rangify-readLink,symlink
branch
from
5e80d30 to
d6f6a01
Compare
std/file.d
Outdated
| auto oz = original.tempCString(); | ||
| auto lz = link.tempCString(); | ||
| alias posixSymlink = core.sys.posix.unistd.symlink; | ||
| immutable int result = () @trusted {return posixSymlink(oz, lz);} (); |
There was a problem hiding this comment.
Nitpick, please put space both before and after { and }
|
The differences were just because I had done the two separately and squashed the commits later. Now using |
aG0aep6G
force-pushed
the
rangify-readLink,symlink
branch
from
5e879bb to
bf85715
Compare
|
Last activity was a week ago. Can I get some more review on this? |
aG0aep6G
force-pushed
the
rangify-readLink,symlink
branch
from
bf85715 to
f241204
Compare
|
Please rebase |
aG0aep6G
force-pushed
the
rangify-readLink,symlink
branch
from
f241204 to
e0062be
Compare
done |
And rebase again ... |
aG0aep6G
force-pushed
the
rangify-readLink,symlink
branch
from
e0062be to
0eff236
Compare
Done again. |
|
LGTM |
Ehm.. rebase again |
| auto dynamicBuffer = new char[](bufferLen * 3 / 2); | ||
|
|
||
| auto dynamicBuffer = new char[](bufferLen * 3 / 2); | ||
| foreach(i; 0 .. 10) |
There was a problem hiding this comment.
Why 10 ? And then what? At least a comment could help a bit
There was a problem hiding this comment.
Ditto. I just indented this.
Removing @safe from the signatures so that attribute inference takes care of things. There are @safe unittests already, ensuring that the functions themselves don't do anything unsafe. Replacing @trusted declarations with @trusted function literals. This way it's more clear that the specific call is @trusted, not the called function.
ElementType -> ElementEncodingType template constraints -> static if
aG0aep6G
force-pushed
the
rangify-readLink,symlink
branch
from
0eff236 to
863b251
Compare
Done again. The conflicts have only been in |
|
Auto-merge toggled on |
-> hopefully soon fixed by #4228 |
4 participants
Removing @safe from the signatures so that attribute inference takes
care of things. There are @safe unittests already, ensuring that the
functions themselves don't do anything unsafe.
Replacing @trusted declarations with @trusted function literals. This
way it's more clear that the specific call is @trusted, not the called
function.