[wp-trac] [WordPress Trac] #15955: move_uploaded_file mangles non-ascii characters on Windows platforms

[WordPress Trac]

#15955: move_uploaded_file mangles non-ascii characters on Windows platforms
-------------------------------------+------------------------------
 Reporter:  t31os_                   |       Owner:
     Type:  defect (bug)             |      Status:  new
 Priority:  normal                   |   Milestone:  Awaiting Review
Component:  Upload                   |     Version:  3.0.3
 Severity:  normal                   |  Resolution:
 Keywords:  has-patch needs-testing  |
-------------------------------------+------------------------------

Comment (by SergeyBiryukov):

 Took more time to investigate. Actually it's not just some characters and
 not just `move_uploaded_file()`. The characters on the screenshots are
 perfectly valid on Windows, since NTFS uses UTF-16.

 The problem is that on Windows, there's currently no way to tell PHP that
 we use UTF-8 in file names. It defaults to single-byte ANSI code page
 (Windows-1251, etc.) for console applications. Therefore `abcdef²³.jpg`
 becomes `abcdefВІВі.jpg`. Files with Cyrillic names get mangled too:
 `Картинка.jpg` becomes `Картинка.jpg`.

 The names need to be converted to that ANSI charset not just in
 `wp_handle_upload()`, but also in other places like
 `wp_generate_attachment_metadata()`, `wp_delete_attachment()`, etc.

 As a workaround, I wrote an experimental plugin that leverages current
 API.
 Tested file uploading, deleting and image editing (the most tricky part)
 on the examples from the screenshots.

-- 
Ticket URL: <http://core.trac.wordpress.org/ticket/15955#comment:11>
WordPress Trac <http://core.trac.wordpress.org/>
WordPress blogging software