NAME

uniconv - convert Netatalk volume encoding

SYNOPSIS

uniconv [-ndv] -c cnidbackend -f fromcode -t tocode [-m maccode] volumepath

DESCRIPTION

uniconv converts the volume encoding of volumepath from the fromcode to the tocode encoding.

OPTIONS

-c

CNID backend used on this volume, usually cdb or dbd. Should match the backend selected with afpd for this volume. If not specified, the default CNID backend "dbd" is used

-d

don't HEX encode leading dots (:2e), equivalent to use dots = yes in afp.conf(5)

-f

encoding to convert from, use ASCII for HEX encoded volumes

-h

display help

-m

Macintosh client codepage, required for HEX encoded volumes. Defaults to "MAC_ROMAN"

-n

"dry run", don't do any real changes

-t

volume encoding to convert to, e.g. UTF8

-v

verbose output, use twice for maximum logging.

-V

print version and exit

WARNING

Setting the wrong options might render your data unusable!!! Make sure you know what you are doing. Always backup your data first.

It is *strongly* recommended to do a "dry run" first and to check the output for conversion errors.

afpd(8) should not be running while you change the volume encoding. Remember to change unix charset or vol charset in afp.conf(5) to the new codepage, before restarting afpd.

In case of MacChineseTraditional, MacJapanese or MacKorean, uniconv cannot be used.

USE AT YOUR OWN RISK!!!

SELECTABLE CHARSETS

Netatalk provides internal support for UTF-8 (pre- and decomposed) and HEX. If you want to use other charsets, they must be provided by iconv(1)

uniconv also knows iso-8859.adapted, an old style 1.x NLS widely used. This is only intended for upgrading old volumes, afpd(8) cannot handle iso-8859.adapted anymore.

CNID BACKGROUND

The CNID backends maintains name to ID mappings. If you change a filename outside afpd(8) (shell, samba), the CNID db, i.e. the DIDNAME index, gets inconsistent. Netatalk tries to recover from such inconsistencies as gracefully as possible. The mechanisms to resolve such inconsistencies may fail sometimes, though, as this is not an easy task to accomplish. I.e. if several names in the path to the file or directory have changed, things may go wrong.

If you change a lot of filenames at once, chances are higher that the afpds fallback mechanisms fail, i.e. files will be assigned new IDs, even though the file hasn't changed. uniconv therefore updates the CNID entry for each file/directory directly after it changes the name to avoid inconsistencies. The two supported backends for volumes, dbd and cdb, use the same CNID db format. Therefore, you could use uniconv with cdb and afpd with dbd later.

Warning: There must not be two processes opening the CNID database using different backends at once! If a volume is still opened with dbd (cnid_metad/cnid_dbd) and you start uniconv with cdb, the result will be a corrupted CNID database, as the two backends use different locking schemes. You might run into additional problems, e.g. if dbd is compiled with transactions, cdb will not update the transaction logs.

In general, it is recommended to use the same backend for uniconv you are using with afpd(8).

EXAMPLES

convert 1.x CAP encoded volume to UTF-8, clients used MacRoman codepage, cnidscheme is dbd:

example% uniconv -c dbd -f ASCII -t UTF8 -m MAC_ROMAN /path/to/share

convert iso8859-1 volume to UTF-8, cnidscheme is cdb:

example% uniconv -c cdb -f ISO-8859-1 -t UTF8 -m MAC_ROMAN /path/to/share

convert 1.x volume using iso8859-1 adapted NLS to HEX encoding:

example% uniconv -f ISO-8859-ADAPTED -t ASCII -m MAC_ROMAN/path/to/share

convert UTF-8 volume to HEX, for MacCyrillic clients:

example% uniconv -f UTF8 -t ASCII -m MAC_CYRILLIC /path/to/share