Universal Character Set Detector C Library (libucd)

What is it?

This library provides a highly accurate set of heuristics that attempt to determine the character set used to encode some input text. This is extremely useful when your program has to handle an input file which is supplied without any encoding metadata.

The original code of libucd was written by Netscape Communications Corporation, is available at http://lxr.mozilla.org/seamonkey/source/extensions/universalchardet/, but unfortunately the Firefox project removed most of the encoding detecting functions in their latest version. While the multi-language detector is widely used in other open source projects. So it's urgent to maintain a standalone version of the library that supports most of the language detecting. And this project was setup, and now also extended more languages, utilities and packaging.

Pulls together:

A command line interface to the library, which also compare the results from alternative libraries, such as libicu.
The UCD library itself from the Mozilla seamonkey source tree
The extended languages detection from the project https://bitbucket.org/medoc/uchardet-enhanced/

Why do you need this library?

Integrated with patches and improvements of the users over the Internet;
Provided thread-safe APIs;
Support multiple packages compiling, includs RPM/DEB/PACMAN/ANDROID etc;
Including the test case data and tools, you may confident to release the new code after running test cases whenever you improved the code by yourself;
Added more languages and encoding support;
Provided document for APIs, man pages;

Supported Encodings

Unicode: UTF-8, UTF-16 (2 variants), UTF-32 (4 variants)
Traditional and Simplified Chinese: Big5, GB18030, EUC-TW, HZ-GB-2312, ISO-2022-CN
Japanese: EUC-JP, SHIFT_JIS, ISO-2022-JP
Korean: EUC-KR, ISO-2022-KR
Cyrillic: KOI8-R, MacCyrillic, IBM855, IBM866, ISO-8859-5, WINDOWS-1251
Hungarian: ISO-8859-2, WINDOWS-1250
Bulgarian: ISO-8859-5, WINDOWS-1251
English: WINDOWS-1252
Greek: ISO-8859-7, WINDOWS-1253
Visual and Logical Hebrew: ISO-8859-8, WINDOWS-1255
Thai: TIS-620
Czech: ISO-8859-2
Finnish: WINDOWS-1252
French: WINDOWS-1252
German: WINDOWS-1252
Polish: ISO-8859-2
Spanish: WINDOWS-1252
Swedish: WINDOWS-1252
Turkish: ISO-8859-9

Building

We have a build system based on autoconf/automake, simply do this incantation:

./configure
make

It also supports building for Linux distributions, such as RedHat/CentOS, Debian/Ubuntu, Arch Linux etc.

RedHat/CentOS

./autogen.sh make rpm
Debian/Ubuntu

./autogen.sh debuild -c -uc -us
Pacman

cd pacman makepkg -Asf
Android

Add a line in your Android.mk file in the folder jni, for example:

include jni/libucd/Android.mk

and then run ndk-build

API

See libucd.h or man pages, and utils/sample.c for the example.

Directory contents:

debian/, rpm/, pacman/

the configure files for various types of packages.

doc/

Contains the docuement described the general idea of auto detection.

man/

man pages for library and utils

include/

The C API header file

src/

The C API from the reference above, with the enhanced mozilla code.

utils/

A command line universal character set detector can process files both by file name or by reading from STDIN.

test/

Wikipedia index pages in target languages, sometimes in multiple encodings. The pages were manually stripped of english and boilerplate content, in hope that the remaining is significant and typical text.
Used to check how the detection works.

langstats/

Data and code used to produce the bigram frequencies for a language/encoding pair, used for the "Two char Distribution Method" from the reference article (neither the article nor the mozilla module publish the scripts used to generate the tables or the reference data).

License

The library is subject to the GNU General Public License Version 2. Alternatively, the library may be used under the terms of the GNU Lesser General Public License 2.1.

ycflash/cpp-libucd