Setters for ctime, mtime and atime. #50
Conversation
|
Please keep the Git history clean. git remote add upstream "https://github.com/Changaco/python-libarchive-c.git"
git fetch upstream
git rebase -i upstream/master
git push -f |
added atime, mtime and ctime properties
|
Sure thing! Is it OK now? |
| @@ -127,9 +127,12 @@ def ffi(name, argtypes, restype, errcheck=None): | |||
| ffi('entry_rdevmajor', [c_archive_entry_p], c_uint) | |||
| ffi('entry_rdevminor', [c_archive_entry_p], c_uint) | |||
|
|
|||
| ffi('entry_set_size', [c_archive_entry_p, c_longlong], c_int) | |||
| ffi('entry_set_size', [c_archive_entry_p, c_longlong], c_longlong) | |||
There was a problem hiding this comment.
Are you sure about this change?
There was a problem hiding this comment.
yes, this change is necessary to allow compressing big files (our files are huge).
BTW longlong is the type that actually matches the C signature.
There was a problem hiding this comment.
No, it doesn't match the signature. The archive_entry_set_* functions don't actually return anything (void), so the last argument to ffi should probably be None. I'm not sure why it's currently c_int.
There was a problem hiding this comment.
Sorry, I thought about this commit f301a92
yes, it's completely unnecessary.
There was a problem hiding this comment.
@Changaco should I change it back to c_int or to None?
There was a problem hiding this comment.
I've changed them to None. The tests passed.
There was a problem hiding this comment.
While I was at it I changed the argument type of entry_set_filetype from c_int to c_uint since the C signature says unsigned int.
| @@ -240,5 +243,7 @@ def ffi(name, argtypes, restype, errcheck=None): | |||
| c_int, check_int) | |||
| ffi('write_finish_entry', [c_archive_p], c_int, check_int) | |||
|
|
|||
| ffi('write_fail', [c_archive_p], c_int, check_int) | |||
|
|
|||
There was a problem hiding this comment.
This function doesn't seem to be used, why do you want to add it?
There was a problem hiding this comment.
It's not used directly by any of context managers in python-libarchive-c, but we use it in our project to allow cancelling creation of archive. Since this library is a convenient wrapper around libarchive we think that it's the best place for it (and make this wrapper more complete).
Turns out that if you want to discard the archive you were creating, you can not just stop creating it. Context managers will close the archive, and libarchive will try to fill the missing contents with 000000000000000000..., and that can cause memory overflows. You have to mark the archive as failed (by that function) to tell libarchive that it should give up any work on it.
There was a problem hiding this comment.
Then shouldn't the context managers call write_fail when an exception is raised?
There was a problem hiding this comment.
I've created a separate issue for that: #51.
Hey!
I have added setters for ctime, mtime and atime. Hope it might be useful!