im allways in trouble about filenames with german umlauts or portuguese characters.
Now I have a special case wich cannot be resolved without understanding who this works…
Maybe someone can help me understand this from the scratch…
Imagine I have a file wich contains a german umlaut like äüö. Im running Suse opensuse 10.3. If I connect with my putty ssh i see the filename broken, with a ? or something else…
The filesystem may be mounted with a charset , right? Does the file itself contain a charset information, too ?
How can I see wich Charset is used for this filesystem?
If I connect with my putty, i can adjust some charset information before connecting, too. I think here should be choosen the same like the file system is mounted?
Now Im trying to copy the file with rsync to a Windows mashine (fat32), wich doesnt work at all. I can give rsync an option (–iconv) where I have to specify the remote and the local filesystem. I think I have to figure out, wich charset is acutally used ( see question 1) and than I need to know wich charset is used by windows, or the mounted harddisk there…
summary:
the charset deals only with the filesystem and not with the file ? how can I store files wich contain german umlauts next to files from Brazil ?
I know I asked a lot , but mybe just a few answers are enough to clearify my world ;))
I believe the easiest way to ENSURE trouble free file names is to NOT use
“special characters” or blanks. That is especially true if you want to access
Windows named files in Linux, or vice versa.
From your story it is not clear to me where you are talking about file names and where about file contents. These are completely different things.
The file name problem is already difficult enough and I think FishFryDay’s advice is true enough. But when you just get these filenames from a CD or so …??
But even when you have a filename in plain ASCII the contents, being a ‘plain text’ file or not, can contain characters in any coding and in any encoding. It is up to the application to solve that (helped by you by installing fonts, etc.).
In the meantime, i know ,that im talking about filenames. The problem is, usualy I take some files from someone else how gave the filenames. I even dont know how they are encoded …
Now I have the directory full of files from different languages with maybe different encodings.
I saw , there is a programm wich does convertings : convmv. but you have to give the original encoding , and the output encoding. fe UTF-8. But how can I know the original encoding? I tried some encodings but without success.
The next thing is : why can I choose an encoding, if I mount a filesystem ? If every file may have its encoding, and the shell can interpret it while giving the right enconding to it …
ok , lets say my files are UTF-8 encoded. If I see list such a file in my putty console , I allways se garbage…
if I change putty to UTF-8 , or anything else , it seems that it doesnt have anything to do with , b/c I allways see the same wrong chars…
I tried to convert the files with convmv trying several combinations of charsets without success… for me it seems that I still dont understand how this works…;(
the contents, being a ‘plain text’ file or not, can contain characters in any coding and in any encoding. It is up to the application to solve that (helped by you by installing fonts, etc.).