Copied file names becomes garbled

Question

I have PHP installed on a CentOS server, and I wanna transfer files from this (with PHP FTP) to a Windows FTP server that is running IIS. Now my problem, when the file name have non-latin characters it changes file names to something like 'ØªØ³Øª.3gp' (its name in CentOS is تست.3gp) I 've searched many times but no solutions. How can I fix this?

I think Windows use CP-1252 (Windows-1252) as charset and CentOS use UTF-8 as charset. But converting this with iconv() and mb_convert_encoding() didn't work.

see this page i answered this question there http://stackoverflow.com/questions/15055192/why-does-windows-need-to-utf8-decode-filenames-for-file-get-contents-to-work/15055869#15055869 — mohammad mohsenipur, Mar 02 '13 at 16:43
I try to user urlencode() but not working then try to convert file name too utf-16 but not working. in urlencode it is the same to do noting to file name, for utf-16 it says access denied for ftp_put() function — IVIR3zaM, Mar 02 '13 at 18:24
try first rename files inside of CentOS by Unicode utf-16 or Windows-1252 then send it to window server. — mohammad mohsenipur, Mar 02 '13 at 18:58
maybe you dont have that language installed n winshit or there is someting wrong with ftp .... — ddjikic, Mar 03 '13 at 01:07
I have installed all Asian languages and can make this file by remote desktop — IVIR3zaM, Mar 03 '13 at 08:32

score 0 · Answer 1 · answered Mar 03 '13 at 11:18

I have found my answer after many researches. windows use different charset for different languages. and for coding to these charset we can use this class: http://www.phpclasses.org/package/1360-PHP-Conversion-between-many-character-set-encodings.html

IVIR3zaM · Answer 2 · 2013-03-03T12:51:56.213

also we can use these codings in mb_convert_encoding():

// Cyrillic script such as Russian, Bulgarian, Serbian Cyrillic and other languages Windows-1251 ~= ISO-8859-5 

// Latin alphabet Windows-1252 ~= ISO-8859-1 // sometimes called incorrectly "ANSI"

//modern Greek. It is not capable of supporting the older polytonic Greek. It is not fully compatible with ISO 8859-7 because the letters like Ά are located at different byte values Windows-1253 ~= ISO-8859-7

// Turkish Windows-1254 = ISO-8859-9

// Hebrew Windows-1255 = ISO-8859-8

// Arabic script like Persian and Urdu Windows-1256 ~= ISO-8859-6

// Estonian, Latvian and Lithuanian languages Windows-1257 != ISO-8859-4

// Vietnamese texts. It is very similar to Windows-1252 with the differences being that s-caron and z-caron Windows-1258 ~= Windows-1252

I use these codes for persian file names:

$filename=iconv('UTF-8','Windows-1256//IGNORE',$filename);

Copied file names becomes garbled

2 Answers2