Add Mkstemp (fixed for rust 1.2) #428
Conversation
… to be *mut u8 whereas x86_64 and i686 expect it to be *mut i8)
|
r? @kamalmarhubi @arcnmx Strings are involved, that's your expertise. |
|
ping to @kamalmarhubi @arcnmx: could you check if that implementation is safe memory wise? I guess the question is if |
|
Yeah no that string is dropped by the time you use |
| let fd = libc::mkstemp(p); | ||
| let pathname = OsStr::from_bytes(CStr::from_ptr(p).to_bytes()); | ||
| try!(Errno::result(fd)); | ||
| Ok((fd, PathBuf::from(pathname).to_owned())) |
There was a problem hiding this comment.
PathBuf::from already creates an owned PathBuf, so to_owned() is just making an additional copy. Also, it's unnecessary since path is already owned in the first place; you can just strip out the trailing nul byte and use OsString::from_vec
There was a problem hiding this comment.
thanks, removed the double copying and used OsString::from_vec to avoid the constructing of the string from pointer. Nice side effects: less includes and less code in the unsafe block. Is there a nicer way to not have to use .pop()?
There was a problem hiding this comment.
The nice way would've been to use CString::into_bytes (OsString::from_vec(pathname.into_bytes())) but that's new since Rust 1.7.
pop() seems fine to me, maybe with a debug_assert! that it == Some(b'\0')?
There was a problem hiding this comment.
thanks for the tip, I added the debug_assert
…ing::from_vec on existing path var rather than construct string from pointer
…espite its documentation\!), checking the failure of a directory now
|
Looks good to me. Thank you and @antifuchs. @homu r+ |
|
📌 Commit b47120d has been approved by |
|
⚡ Test exempted - status |
Add Mkstemp (fixed for rust 1.2) I fixed @antifuchs addition of `mkstmp` in #365 by making it compile in Rust 1.2 and adding documentation. A few remarks: - made it working with Rust 1.2 which needed `to_bytes_with_nul()`. I think the implementation is memory safe but would like to have a pair of eyes checking that - replaced Path by PathBuf so it's more in line with getcwd - I didn't move it into another module. If this still the consensus then I would like to do that but in a separate PR (probably moving all stdio methods out) - it's true that unistd doesn't need mkstmp since there is the crate tempdir and tempfile, but I'd love to see this here for completeness
I fixed @antifuchs addition of
mkstmpin #365 by making it compile in Rust 1.2 and adding documentation.A few remarks:
to_bytes_with_nul(). I think the implementation is memory safe but would like to have a pair of eyes checking that