path: root/docs/Samba3-HOWTO/TOSHARG-Unicode.xml

<?xml version="1.0" encoding="iso-8859-1"?>
<!DOCTYPE chapter PUBLIC "-//Samba-Team//DTD DocBook V4.2-Based Variant V1.0//EN" "http://www.samba.org/samba/DTD/samba-doc">
<chapter id="unicode">
<chapterinfo>
	&author.jelmer;
	&author.jht;
	<author>
		<firstname>TAKAHASHI</firstname><surname>Motonobu</surname>
		<affiliation>
		<address><email>monyo@home.monyo.com</email></address>
		</affiliation>
		<contrib>Japanese character support</contrib>
	</author>
	<pubdate>25 March 2003</pubdate>
</chapterinfo>

<title>Unicode/Charsets</title>

<sect1>
<title>Features and Benefits</title>

<para>
Every industry eventually matures. One of the great areas of maturation is in
the focus that has been given over the past decade to make it possible for anyone
anywhere to use a computer. It has not always been that way. In fact, not so long
ago, it was common for software to be written for exclusive use in the country of
origin.
</para>

<para>
Of all the effort that has been brought to bear on providing native
language support for all computer users, the efforts of the
<ulink url="http://www.openi18n.org/">Openi18n organization</ulink>
is deserving of special mention.
</para>

<para>
Samba-2.x supported a single locale through a mechanism called 
<emphasis>codepages</emphasis>. Samba-3 is destined to become a truly transglobal
file- and printer-sharing platform.
</para>

</sect1>

<sect1>
<title>What Are Charsets and Unicode?</title>

<para>
Computers communicate in numbers. In texts, each number is 
translated to a corresponding letter. The meaning that will be assigned 
to a certain number depends on the <emphasis>character set (charset)
</emphasis> that is used. 
</para>

<para>
A charset can be seen as a table that is used to translate numbers to 
letters. Not all computers use the same charset (there are charsets 
with German umlauts, Japanese characters, and so on). The American Standard Code
for Information Interchange (ASCII) encoding system has been the normative character
encoding scheme used by computers to date. This employs a charset that contains 
256 characters. Using this mode of encoding, each character takes exactly one byte.
</para>

<para>
There are also charsets that support extended characters, but those need at least
twice as much storage space as does ASCII encoding. Such charsets can contain
<command>256 * 256 = 65536</command> characters, which is more than all possible
characters one could think of. They are called multibyte charsets because they use
more then one byte to store one character. 
</para>

<para>
One standardized multibyte charset encoding scheme is known as
<ulink url="http://www.unicode.org/">unicode</ulink>.  A big advantage of using a
multibyte charset is that you only need one. There is no need to make sure two
computers use the same charset when they are communicating.
</para>

<para>Old Windows clients use single-byte charsets, named 
<parameter>codepages</parameter>, by Microsoft. However, there is no support for 
negotiating the charset to be used in the SMB/CIFS protocol. Thus, you 
have to make sure you are using the same charset when talking to an older client.
Newer clients (Windows NT, 200x, XP) talk Unicode over the wire.
</para>
</sect1>

<sect1>
<title>Samba and Charsets</title>

<para>
As of Samba-3, Samba can (and will) talk Unicode over the wire. Internally, 
Samba knows of three kinds of character sets: 
</para>

<variablelist>
	<varlistentry>
		<term><smbconfoption name="unix charset"/></term>
		<listitem><para>
		This is the charset used internally by your operating system. 
		The default is <constant>UTF-8</constant>, which is fine for most 
		systems and covers all characters in all languages. The default
		in previous Samba releases was to save filenames in the encoding of the 
		clients &smbmdash; for example, cp850 for Western European countries.
		</para></listitem>
	</varlistentry>

	<varlistentry>
		<term><smbconfoption name="display charset"/></term>
		<listitem><para>This is the charset Samba uses to print messages
		on your screen. It should generally be the same as the <parameter>unix charset</parameter>.
		</para></listitem>
	</varlistentry>

	<varlistentry>
		<term><smbconfoption name="dos charset"/></term>
		<listitem><para>This is the charset Samba uses when communicating with 
		DOS and Windows 9x/Me clients. It will talk Unicode to all newer clients.
		The default depends on the charsets you have installed on your system.
		Run <command>testparm -v | grep &quot;dos charset&quot;</command> to see 
		what the default is on your system. 
		</para></listitem>
	</varlistentry>
</variablelist>

</sect1>

<sect1>
<title>Conversion from Old Names</title>

<para>Because previous Samba versions did not do any charset conversion, 
characters in filenames are usually not correct in the UNIX charset but only 
for the local charset used by the DOS/Windows clients.</para>

<para>Bjoern Jacke has written a utility named <ulink url="http://j3e.de/linux/convmv/">convmv</ulink>
that can convert whole directory structures to different charsets with one single command. 
</para>

</sect1>

<sect1>
<title>Japanese Charsets</title>

<para>
Setting up Japanese charsets is quite difficult. This is mainly because:
</para>

<itemizedlist>
	<listitem><para>The Windows character set is extended from the original legacy Japanese
		standard (JIS X 0208) and is not standardized. This means that the strictly
		standardized implementation cannot support the full Windows character set.
	</para></listitem>

	<listitem><para> Mainly for historical reasons, there are several encoding methods in
		Japanese, which are not fully compatible with each other. There are
		two major encoding methods. One is the Shift_JIS series used in Windows
		and some UNIXes. The other is the EUC-JP series used in most UNIXes
		and Linux. Moreover, Samba previously also offered several unique encoding
		methods, named CAP and HEX, to keep interoperability with CAP/NetAtalk and
		UNIXes that can't use Japanese filenames.  Some implementations of the
		EUC-JP series can't support the full Windows character set.
	</para></listitem>

	<listitem><para>There are some code conversion tables between Unicode and legacy
		Japanese character sets. One is compatible with Windows, another one
		is based on the reference of the Unicode consortium, and others are 
		a mixed implementation. The Unicode consortium does not officially
		define any conversion tables between Unicode and legacy character
		sets, so there cannot be standard one.
	</para></listitem>

	<listitem><para>The character set and conversion tables available in iconv() depend
		on the iconv library that is available. Next to that, the Japanese locale 
		names may be different on different systems.  This means that the value of 
		the charset parameters depends on the implementation of iconv() you are using.
		</para>

		<para>Though 2-byte fixed UCS-2 encoding is used in Windows internally,
		Shift_JIS series encoding is usually used in Japanese environments
		as ASCII encoding is in English environments.
	</para></listitem>
</itemizedlist>

<sect2><title>Basic Parameter Setting</title>

	<para>
	The <smbconfoption name="dos charset"/> and 
	<smbconfoption name="display charset"/>
	should be set to the locale compatible with the character set 
	and encoding method used on Windows. This is usually CP932
	but sometimes has a different name.
	</para>

	<para>
	The <smbconfoption name="unix charset"/> can be either Shift_JIS series,
	EUC-JP series, or UTF-8. UTF-8 is always available, but the availability of other locales
	and the name itself depends on the system.
	</para>

	<para>
	Additionally, you can consider using the Shift_JIS series as the
	value of the <smbconfoption name="unix charset"/>
	parameter by using the vfs_cap module, which does the same thing as
	setting <quote>coding system = CAP</quote> in the Samba 2.2 series.
	</para>

	<para>
	Where to set <smbconfoption name="unix charset"/>
	to is a difficult question. Here is a list of details, advantages, and
	disadvantages of using a certain value.
	</para>

	<variablelist>
		<varlistentry><term>Shift_JIS series</term>
			<listitem><para>
			Shift_JIS series means a locale that is equivalent to <constant>Shift_JIS</constant>,
			used as a standard on Japanese Windows. In the case of <constant>Shift_JIS</constant>,
			for example, if a Japanese filename consists of 0x8ba4 and 0x974c
			(a 4-bytes Japanese character string meaning <quote>share</quote>) and <quote>.txt</quote>
			is written from Windows on Samba, the filename on UNIX becomes
			0x8ba4, 0x974c, <quote>.txt</quote> (an 8-byte BINARY string), same as Windows.
			</para>

			<para>Since Shift_JIS series is usually used on some commercial-based
			UNIXes; hp-ux and AIX as the Japanese locale (however, it is also possible
			to use the EUC-JP locale series). To use Shift_JIS series on these platforms,
			Japanese filenames created from Windows can be referred to also on
			UNIX.</para>

			<para>
			If your UNIX is already working with Shift_JIS and there is a user 
			who needs to use Japanese filenames written from Windows, the
			Shift_JIS series is the best choice.  However, broken filenames
			may be displayed, and some commands that cannot handle non-ASCII
			filenames may be aborted during parsing filenames. Especially, there
			may be <quote>\ (0x5c)</quote> in filenames, which need to be handled carefully.
			It is best to not touch filenames written from Windows on UNIX.
			</para>

			<para>
			Note that most Japanized free software actually works with EUC-JP
			only. It is good practice to verify that the Japanized free software can work
			with Shift_JIS.
			</para>
			</listitem>
		</varlistentry>

		<varlistentry><term>EUC-JP series</term>
			<listitem><para>
			EUC-JP series means a locale that is equivalent to the industry
			standard called EUC-JP, widely used in Japanese UNIX (although EUC
			contains specifications for languages other than Japanese, such as
			EUC-KR). In the case of EUC-JP series, for example, if a Japanese
			filename consists of 0x8ba4 and 0x974c and <quote>.txt</quote> is written from
			Windows on Samba, the filename on UNIX becomes 0xb6a6, 0xcdad,
			<quote>.txt</quote> (an 8-byte BINARY string). 
			</para>

			<para>
			Since EUC-JP is usually used on open source UNIX, Linux, and FreeBSD, and on commercial-based UNIX, Solaris,
			IRIX, and Tru64 UNIX as Japanese locale (however, it is also possible on Solaris to use Shift_JIS and UTF-8,
			and on Tru64 UNIX it is possible to use Shift_JIS). To use EUC-JP series, most Japanese filenames created from
			Windows can be referred to also on UNIX. Also, most Japanized free software work mainly with EUC-JP only.
			</para>

			<para>
			It is recommended to choose EUC-JP series when using Japanese filenames on UNIX.
			</para>

			<para>
			Although there is no character that needs to be carefully treated
			like <quote>\ (0x5c)</quote>, broken filenames may be displayed and some
			commands that cannot handle non-ASCII filenames may be aborted
			during parsing filenames.
			</para>

			<para>
			Moreover, if you built Samba using differently installed libiconv,
			the eucJP-ms locale included in libiconv and EUC-JP series locale
			included in the operating system may not be compatible. In this case, you may need to
			avoid using incompatible characters for filenames.
			</para>
			</listitem>
		</varlistentry>

		<varlistentry><term>UTF-8</term>
			<listitem><para>
			UTF-8 means a locale equivalent to UTF-8, the international standard defined by the Unicode consortium. In
			UTF-8, a <parameter>character</parameter> is expressed using 1 to 3 bytes. In case of the Japanese language,
			most characters are expressed using 3 bytes. Since on Windows Shift_JIS, where a character is expressed with 1
			or 2 bytes is used to express Japanese, basically a byte length of a UTF-8 string the length of the UTF-8
			string is 1.5 times that of the original Shift_JIS string. In the case of UTF-8, for example, if a Japanese
			filename consists of 0x8ba4 and 0x974c, and <quote>.txt</quote> is written from Windows on Samba, the filename
			on UNIX becomes 0xe585, 0xb1e6, 0x9c89, <quote>.txt</quote> (a 10-byte BINARY string).
			</para>

			<para>
			For systems where iconv() is not available or where iconv()'s locales
			are not compatible with Windows, UTF-8 is the only locale available.
			</para>

			<para> 
			There are no systems that use UTF-8 as the default locale for Japanese.
			</para>

			<para>
			Some broken filenames may be displayed, and some commands that
			cannot handle non-ASCII filenames may be aborted during parsing
			filenames. Especially, there may be <quote>\ (0x5c)</quote> in filenames, which
			must be handled carefully, so you had better not touch filenames
			written from Windows on UNIX.
			</para>

			<para>
			In addition, although it is not directly concerned with Samba, since
			there is a delicate difference between the iconv() function, which is
			generally used on UNIX, and the functions used on other platforms,
			such as Windows and Java, so far is concerens the conversion between
			Shift_JIS and Unicode UTF-8 must be done with care and recognition
			of the limitations involved in the process.
			</para>

			<para>
			Although Mac OS X uses UTF-8 as its encoding method for filenames,
			it uses an extended UTF-8 specification that Samba cannot handle, so
			UTF-8 locale is not available for Mac OS X.
			</para>
			</listitem>
		</varlistentry>

		<varlistentry><term>Shift_JIS series + vfs_cap (CAP encoding)</term>
			<listitem><para>
			CAP encoding means a specification used in CAP and NetAtalk, file
			server software for Macintosh. In the case of CAP encoding, for
			example, if a Japanese filename consists of 0x8ba4 and 0x974c, and
			<quote>.txt</quote> is written from Windows on Samba, the filename on UNIX
			becomes <quote>:8b:a4:97L.txt</quote> (a 14 bytes ASCII string). 
			</para>

			<para>
			For CAP encoding, a byte that cannot be expressed as an ASCII
			character (0x80 or above) is encoded in an <quote>:xx</quote> form. You need to take
			care of containing a <quote>\(0x5c)</quote> in a filename, but filenames are not
			broken in a system that cannot handle non-ASCII filenames.
			</para>

			<para>
			The greatest merit of CAP encoding is the compatibility of encoding
			filenames with CAP or NetAtalk. These are respectively the Columbia Appletalk
			Protocol, and the NetAtalk Open Source software project.
			Since these software applications write a file name on UNIX with CAP encoding, if a
			directory is shared with both Samba and NetAtalk, you need to use
			CAP encoding to avoid non-ASCII filenames from being broken.
			</para>

			<para>
			However, recently, NetAtalk has been
			patched on some systems to write filenames with EUC-JP (e.g., Japanese original Vine Linux).
			In this case, you need to choose EUC-JP series instead of CAP encoding.
			</para>

			<para>
			vfs_cap itself is available for non-Shift_JIS series locales for
			systems that cannot handle non-ASCII characters or systems that
			share files with NetAtalk.
			</para>

			<para>
			To use CAP encoding on Samba-3, you should use the unix charset parameter and VFS 
			as in Example 29.5.1:
			</para>

<example><title>VFS CAP</title>
	<smbconfblock>
<smbconfsection name="[global]"/>
<smbconfcomment>the locale name "CP932" may be different</smbconfcomment>
<smbconfoption name="dos charset">CP932</smbconfoption>
<smbconfoption name="unix charset">CP932</smbconfoption>

<smbconfsection name="[cap-share]"/>
<smbconfoption name="vfs option">cap</smbconfoption>
</smbconfblock>
</example>

			<para>
			You should set CP932 if using GNU libiconv for unix charset. With this setting,
			filenames in the <quote>cap-share</quote> share are written with CAP encoding.
			</para>
			</listitem>
		</varlistentry>
	</variablelist>

</sect2>

<sect2><title>Individual Implementations</title>

<para>
Here is some additional information regarding individual implementations:
</para>

	<variablelist>
		<varlistentry><term>GNU libiconv</term>
			<listitem><para>
			To handle Japanese correctly, you should apply the patch
			<ulink url="http://www2d.biglobe.ne.jp/~msyk/software/libiconv-patch.html">libiconv-1.8-cp932-patch.diff.gz</ulink>
			to libiconv-1.8.
			</para>
			
			<para>
			Using the patched libiconv-1.8, these settings are available:
			</para>


<!-- FIXME: Convert to diagram ? -->
<programlisting>
dos charset = CP932
unix charset = CP932 / eucJP-ms / UTF-8
		|       |
		|       +-- EUC-JP series
		+-- Shift_JIS series
display charset = CP932
</programlisting>

			<para>
			Other Japanese locales (for example, Shift_JIS and EUC-JP) should not
			be used because of the lack of the compatibility with Windows.
			</para>
			</listitem>
		</varlistentry>

		<varlistentry><term>GNU glibc</term>
			<listitem><para>
			To handle Japanese correctly, you should apply a <ulink url="http://www2d.biglobe.ne.jp/~msyk/software/glibc/">patch</ulink>
			to glibc-2.2.5/2.3.1/2.3.2 or should use the patch-merged versions, glibc-2.3.3 or later.
			</para>

			<para>
			Using the above glibc, these setting are available:
			</para>

<smbconfblock>
<smbconfoption name="dos charset">CP932</smbconfoption>
<smbconfoption name="unix charset">CP932 / eucJP-ms / UTF-8</smbconfoption>
<smbconfoption name="display charset">CP932</smbconfoption>
</smbconfblock>

			<para>
			Other Japanese locales (for example, Shift_JIS and EUC-JP) should not
			be used because of the lack of the compatibility with Windows.
			</para>
			</listitem>
		</varlistentry>
	</variablelist>

</sect2>

<sect2>
	<title>Migration from Samba-2.2 Series</title>

<para> 
Prior to Samba-2.2 series, the <quote>coding system</quote> parameter was used. The default codepage in Samba
2.x was code page 850. In the Samba-3 series this has been replaced with the <smbconfoption name="unix
charset"/> parameter.  <link linkend="japancharsets">Japanese Character Sets in Samba-2.2 and Samba-3</link>
shows the mapping table when migrating from the Samba-2.2 series to Samba-3.
</para>

	<table frame="all" id="japancharsets">
		<title>Japanese Character Sets in Samba-2.2 and Samba-3</title>

		<tgroup cols="2" align="center">
			<colspec align="center"/>
			<colspec align="center"/>
			<thead>
				<row><entry>Samba-2.2 Coding System</entry><entry>Samba-3 unix charset</entry></row>
			</thead>
			<tbody>
				<row><entry>SJIS</entry><entry>Shift_JIS series</entry></row>
				<row><entry>EUC</entry><entry>EUC-JP series</entry></row>
				<row><entry>EUC3<footnote><para>Only exists in Japanese Samba version</para></footnote></entry><entry>EUC-JP series</entry></row>
				<row><entry>CAP</entry><entry>Shift_JIS series + VFS</entry></row>
				<row><entry>HEX</entry><entry>currently none</entry></row>
				<row><entry>UTF8</entry><entry>UTF-8</entry></row>
				<row><entry>UTF8-Mac<footnote><para>Only exists in Japanese Samba version</para></footnote></entry><entry>currently none</entry></row>
				<row><entry>others</entry><entry>none</entry></row>
			</tbody>
		</tgroup>
	</table>

</sect2>

</sect1>

<sect1>
	<title>Common Errors</title>

	<sect2>
		<title>CP850.so Can't Be Found</title>

		<para><quote>Samba is complaining about a missing <filename>CP850.so</filename> file.</quote></para>

		<para>
		CP850 is the default <smbconfoption name="dos charset"/>.
		The <smbconfoption name="dos charset"/> is used to convert data to the codepage used by your DOS clients.
		If you do not have any DOS clients, you can safely ignore this message. </para>

		<para>
		CP850 should be supported by your local iconv implementation. Make sure you have all the required packages installed.
		If you compiled Samba from source, make sure that the configure process found iconv. This can be
		confirmed by checking the <filename>config.log</filename> file that is generated when
		<command>configure</command> is executed.</para> 
	</sect2>
</sect1>

</chapter>