gcc/libstdc++-v3/docs/html/22_locale/messages.html
Benjamin Kosnik 183490ecc1 messages.html: Remove angle brackets.
2002-01-22  Benjamin Kosnik  <bkoz@redhat.com>

	* docs/html/22_locale/messages.html: Remove angle brackets.
	* docs/html/17_intro/TODO: Add.

From-SVN: r49119
2002-01-23 02:02:21 +00:00

405 lines
12 KiB
HTML

<html>
<head>
<h1>
Notes on the messages implementation.
</h1>
</head>
<I>
prepared by Benjamin Kosnik (bkoz@redhat.com) on August 8, 2001
</I>
<p>
<h2>
1. Abstract
</h2>
<p>
The std::messages facet implements message retrieval functionality
equivalent to Java's java.text.MessageFormat .using either GNU gettext
or IEEE 1003.1-200 functions.
</p>
<p>
<h2>
2. What the standard says
</h2>
The std::messages facet is probably the most vaguely defined facet in
the standard library. It's assumed that this facility was built into
the standard library in order to convert string literals from one
locale to the other. For instance, converting the "C" locale's
<code>const char* c = "please"</code> to a German-localized <code>"bitte"</code>
during program execution.
<BLOCKQUOTE>
22.2.7.1 - Template class messages [lib.locale.messages]
</BLOCKQUOTE>
This class has three public member functions, which directly
correspond to three protected virtual member functions.
The public member functions are:
<p>
<code>catalog open(const string&, const locale&) const</code>
<p>
<code>string_type get(catalog, int, int, const string_type&) const</code>
<p>
<code>void close(catalog) const</code>
<p>
While the virtual functions are:
<p>
<code>catalog do_open(const string&, const locale&) const</code>
<BLOCKQUOTE>
<I>
-1- Returns: A value that may be passed to get() to retrieve a
message, from the message catalog identified by the string name
according to an implementation-defined mapping. The result can be used
until it is passed to close(). Returns a value less than 0 if no such
catalog can be opened.
</I>
</BLOCKQUOTE>
<p>
<code>string_type do_get(catalog, int, int, const string_type&) const</code>
<BLOCKQUOTE>
<I>
-3- Requires: A catalog cat obtained from open() and not yet closed.
-4- Returns: A message identified by arguments set, msgid, and dfault,
according to an implementation-defined mapping. If no such message can
be found, returns dfault.
</I>
</BLOCKQUOTE>
<p>
<code>void do_close(catalog) const</code>
<BLOCKQUOTE>
<I>
-5- Requires: A catalog cat obtained from open() and not yet closed.
-6- Effects: Releases unspecified resources associated with cat.
-7- Notes: The limit on such resources, if any, is implementation-defined.
</I>
</BLOCKQUOTE>
<p>
<h2>
3. Problems with &quot;C&quot; messages: thread safety,
over-specification, and assumptions.
</h2>
A couple of notes on the standard.
<p>
First, why is <code>messages_base::catalog</code> specified as a typedef
to int? This makes sense for implementations that use
<code>catopen</code>, but not for others. Fortunately, it's not heavily
used and so only a minor irritant.
<p>
Second, by making the member functions <code>const</code>, it is
impossible to save state in them. Thus, storing away information used
in the 'open' member function for use in 'get' is impossible. This is
unfortunate.
<p>
The 'open' member function in particular seems to be oddly
designed. The signature seems quite peculiar. Why specify a <code>const
string& </code> argument, for instance, instead of just <code>const
char*</code>? Or, why specify a <code>const locale&</code> argument that is
to be used in the 'get' member function? How, exactly, is this locale
argument useful? What was the intent? It might make sense if a locale
argument was associated with a given default message string in the
'open' member function, for instance. Quite murky and unclear, on
reflection.
<p>
Lastly, it seems odd that messages, which explicitly require code
conversion, don't use the codecvt facet. Because the messages facet
has only one template parameter, it is assumed that ctype, and not
codecvt, is to be used to convert between character sets.
<p>
It is implicitly assumed that the locale for the default message
string in 'get' is in the "C" locale. Thus, all source code is assumed
to be written in English, so translations are always from "en_US" to
other, explicitly named locales.
<p>
<h2>
4. Design and Implementation Details
</h2>
This is a relatively simple class, on the face of it. The standard
specifies very little in concrete terms, so generic implementations
that are conforming yet do very little are the norm. Adding
functionality that would be useful to programmers and comparable to
Java's java.text.MessageFormat takes a bit of work, and is highly
dependent on the capabilities of the underlying operating system.
<p>
Three different mechanisms have been provided, selectable via
configure flags:
<ul>
<li> generic
<p>
This model does very little, and is what is used by default.
</p>
<li> gnu
<p>
The gnu model is complete and fully tested. It's based on the
GNU gettext package, which is part of glibc. It uses the functions
<code>textdomain, bindtextdomain, gettext</code>
to implement full functionality. Creating message
catalogs is a relatively straight-forward process and is
lightly documented below, and fully documented in gettext's
distributed documentation.
</p>
<li> ieee_1003.1-200x
<p>
This is a complete, though untested, implementation based on
the IEEE standard. The functions
<code>catopen, catgets, catclose</code>
are used to retrieve locale-specific messages given the
appropriate message catalogs that have been constructed for
their use. Note, the script <code> po2msg.sed</code> that is part
of the gettext distribution can convert gettext catalogs into
catalogs that <code>catopen</code> can use.
</p>
</ul>
<p>
A new, standards-conformant non-virtual member function signature was
added for 'open' so that a directory could be specified with a given
message catalog. This simplifies calling conventions for the gnu
model.
<p>
The rest of this document discusses details of the GNU model.
<p>
The messages facet, because it is retrieving and converting between
characters sets, depends on the ctype and perhaps the codecvt facet in
a given locale. In addition, underlying "C" library locale support is
necessary for more than just the <code>LC_MESSAGES</code> mask:
<code>LC_CTYPE</code> is also necessary. To avoid any unpleasantness, all
bits of the "C" mask (ie <code>LC_ALL</code>) are set before retrieving
messages.
<p>
Making the message catalogs can be initially tricky, but become quite
simple with practice. For complete info, see the gettext
documentation. Here's an idea of what is required:
<ul>
<LI > Make a source file with the required string literals
that need to be translated. See
<code>intl/string_literals.cc</code> for an example.
<p>
<li> Make initial catalog (see "4 Making the PO Template File"
from the gettext docs).
<p>
<code> xgettext --c++ --debug string_literals.cc -o libstdc++.pot </code>
<p>
<li> Make language and country-specific locale catalogs.
<p>
<code>cp libstdc++.pot fr_FR.po</code>
<p>
<code>cp libstdc++.pot de_DE.po</code>
<p>
<li> Edit localized catalogs in emacs so that strings are
translated.
<p>
<code>emacs fr_FR.po</code>
<p>
<li> Make the binary mo files.
<p>
<code>msgfmt fr_FR.po -o fr_FR.mo</code>
<p>
<code>msgfmt de_DE.po -o de_DE.mo</code>
<p>
<li> Copy the binary files into the correct directory structure.
<p>
<code>cp fr_FR.mo (dir)/fr_FR/LC_MESSAGES/libstdc++-v3.mo</code>
<p>
<code>cp de_DE.mo (dir)/de_DE/LC_MESSAGES/libstdc++-v3.mo</code>
<p>
<li> Use the new message catalogs.
<p>
<code>locale loc_de("de_DE");</code>
<p>
<code>
use_facet<messages<char> >(loc_de).open("libstdc++", locale(), dir);
</code>
</ul>
<p>
<h2>
5. Examples
</h2>
<ul>
<li> message converting, simple example using the GNU model.
<pre>
#include <locale>
void test01()
{
using namespace std;
typedef std::messages<char>::catalog catalog;
// Set to the root directory of the libstdc++.mo catalogs.
const char* dir = LOCALEDIR;
locale loc_de("de_DE");
// Cache the messages facet.
const messages<char>& mssg_de = use_facet<messages<char> >(loc_de);
// Check German (de_DE) locale.
catalog cat_de = mssg_de.open("libstdc++", loc_c, dir);
string s01 = mssg_de.get(cat_de, 0, 0, "please");
string s02 = mssg_de.get(cat_de, 0, 0, "thank you");
// s01 == "bitte"
// s02 == "danke"
mssg_de.close(cat_de);
}
</pre>
</ul>
More information can be found in the following testcases:
<ul>
<li> testsuite/22_locale/messages.cc
<li> testsuite/22_locale/messages_byname.cc
<li> testsuite/22_locale/messages_char_members.cc
</ul>
<p>
<h2>
6. Unresolved Issues
</h2>
<ul>
<li> Things that are sketchy, or remain unimplemented:
<ul>
<li>_M_convert_from_char, _M_convert_to_char are in
flux, depending on how the library ends up doing
character set conversions. It might not be possible to
do a real character set based conversion, due to the
fact that the template parameter for messages is not
enough to instantiate the codecvt facet (1 supplied,
need at least 2 but would prefer 3).
<li> There are issues with gettext needing the global
locale set to extract a message. This dependence on
the global locale makes the current "gnu" model non
MT-safe. Future versions of glibc, ie glibc 2.3.x will
fix this, and the C++ library bits are already in
place.
</ul>
<p>
<li> Development versions of the GNU "C" library, glibc 2.3 will allow
a more efficient, MT implementation of std::messages, and will
allow the removal of the _M_name_messages data member. If this
is done, it will change the library ABI. The C++ parts to
support glibc 2.3 have already been coded, but are not in use:
once this version of the "C" library is released, the marked
parts of the messages implementation can be switched over to
the new "C" library functionality.
<p>
<li> At some point in the near future, std::numpunct will probably use
std::messages facilities to implement truename/falename
correctly. This is currently not done, but entries in
libstdc++.pot have already been made for "true" and "false"
string literals, so all that remains is the std::numpunct
coding and the configure/make hassles to make the installed
library search its own catalog. Currently the libstdc++.mo
catalog is only searched for the testsuite cases involving
messages members.
<p>
<li> The following member functions:
<p>
<code>
catalog
open(const basic_string<char>& __s, const locale& __loc) const
</code>
<p>
<code>
catalog
open(const basic_string<char>&, const locale&, const char*) const;
</code>
<p>
Don't actually return a "value less than 0 if no such catalog
can be opened" as required by the standard in the "gnu"
model. As of this writing, it is unknown how to query to see
if a specified message catalog exists using the gettext
package.
</ul>
<p>
<h2>
7. Acknowledgments
</h2>
Ulrich Drepper for the character set explanations, gettext details,
and patient answering of late-night questions, Tom Tromey for the java details.
<p>
<h2>
8. Bibliography / Referenced Documents
</h2>
Drepper, Ulrich, GNU libc (glibc) 2.2 manual. In particular, Chapters
&quot;7 Locales and Internationalization&quot;
<p>
Drepper, Ulrich, Thread-Aware Locale Model, A proposal. This is a
draft document describing the design of glibc 2.3 MT locale
functionality.
<p>
Drepper, Ulrich, Numerous, late-night email correspondence
<p>
ISO/IEC 9899:1999 Programming languages - C
<p>
ISO/IEC 14882:1998 Programming languages - C++
<p>
Java 2 Platform, Standard Edition, v 1.3.1 API Specification. In
particular, java.util.Properties, java.text.MessageFormat,
java.util.Locale, java.util.ResourceBundle.
http://java.sun.com/j2se/1.3/docs/api
<p>
System Interface Definitions, Issue 7 (IEEE Std. 1003.1-200x)
The Open Group/The Institute of Electrical and Electronics Engineers, Inc.
In particular see lines 5268-5427.
http://www.opennc.org/austin/docreg.html
<p> GNU gettext tools, version 0.10.38, Native Language Support
Library and Tools.
http://sources.redhat.com/gettext
<p>
Langer, Angelika and Klaus Kreft, Standard C++ IOStreams and Locales,
Advanced Programmer's Guide and Reference, Addison Wesley Longman,
Inc. 2000. See page 725, Internationalized Messages.
<p>
Stroustrup, Bjarne, Appendix D, The C++ Programming Language, Special Edition, Addison Wesley, Inc. 2000