Difference between revisions of "Installation"

From Apertium
Jump to navigation Jump to search
Line 37: Line 37:
#*[[Prerequisites for Arch Linux|Arch Linux]]
#*[[Prerequisites for Arch Linux|Arch Linux]]
#*[[Prerequisites for Gentoo|Gentoo]]
#*[[Prerequisites for Gentoo|Gentoo]]
#*[[Prerequisites for FreeBSD|FreeBSD]] (untested)
#*[[Prerequisites for FreeBSD|FreeBSD]]
#*[[Prerequisites for Slackware|Slackware]]
#*[[Prerequisites for Slackware|Slackware]]

Revision as of 16:05, 20 December 2016

En français


Gnome-home.png Home PageBugs.png BugsInternet.png WikiGaim.png Chat

If you just want a simple way to translate offline

Most of the applications below are simple two-box "choose a translation pair and start typing" graphic user interfaces,

  • Unix users, e.g. GNU/Linux or BSD users, do not have a currently-maintained UI. However, Apertium tolk, a Python/DBus application, is in repositories and will install and run ok.

See Tools for more graphical user interfaces, services, plugins, etc.

If you want a simple way to see how Apertium works

  • Apertium-viewer is a Java-based multi-platform UI to view and edit the output of the various stages of an Apertium translation.

Installing extra languages in tools

Several tools can download language pairs from the web. Some can also install their downloaded language pairs.

Some tools, including those that do not download language pairs, seek language pairs in known places. For example, a Debian-packaged version of 'apertium-tolk' seeks pairs in '/usr/share/apertium'. This is where Debian installs pre-packaged pairs, and 'apertium-tolk' can find them. If language pairs are placed in this folder (not necessarily by the packaging system, though you will be told not to do this), 'apertium-tolk' will load them.

The language pairs offered by the tools are those cleared as stable or developed. These pairs have been moved to subversion 'trunk'. There are many more language pairs available, but in various stages of development. However intriguing they may seem, a person using a GUI will have little interest in these (some pairs will not run without custom build environments).

If you want to add language data / do more advanced stuff

Unix users (GNU/Linux, Mac, BSD) who wish to hack on Apertium should follow these three stages to installing the core packages and language data:

  1. Install prerequisites specific to your operating system:
  2. Install apertium and related packages from SVN.
  3. Fix any problems :)

For Windows users who wish to hack on Apertium, the best method is to download the Apertium VirtualBox, which lets you run a Unix in your Windows, and comes with Apertium from SVN (and one language pair) pre-installed. Once you've got the VirtualBox, you can install more language pairs as shown in Minimal installation from SVN. Alternatively, you can compile it for Windows using Cygwin; documentation for how to compile on Windows manually is at Apertium on Windows; there is also a script at Apertium guide for Windows users, but it is currently out-of-date and in need of updating.

Some language pairs require extra packages like Constraint Grammar (vislcg3) or HFST in addition to apertium/lttoolbox/apertium-lex-tools. For most systems, the "Prerequisites" pages will show how to install these through your package manager, but see the section on installing vislcg3 and HFST, respectively if you're on a not-yet-supported system. You can tell if a package requires CG if it has an .rlx file, and HFST if it has a .lexc file.

There are released tarball packages on the official download page. These are meant for packagers, and advanced users who are not planning to develop on apertium. These are installed like the SVN packages, except you use ./configure instead of ./autogen.sh. Many systems now also have pre-packaged apertium language pairs, but the official Debian/MacPorts/… repositories are often quite outdated compared to the SVN versions – use TinoDidriksen's nightly packages as instructed in the "Prerequisites" pages above.

Installation Videos

Most of these videos have been produced by Google Code-In students.