GoldenDict tools

A set of helpful programs to enhance goldendict for immersion learning.

Table of Contents

• Installation
• gd-marisa
• gd-mecab
• gd-images
• gd-strokeorder
• gd-handwritten
• gd-massif
• gd-ankisearch

Installation

First, install goldendict-ng.

Pacman-based distros

Install gd-tools-git from the AUR.

Debian-based distros

Ubuntu and Debian users can install the package attached to the latest GitHub release.

wget https://github.com/Ajatt-Tools/gd-tools/releases/latest/download/gd-tools.deb
sudo apt install ./gd-tools.deb

Compile from source if apt reports an incompatible or unavailable dependency. The DEB package does not include the optional Mandarin dictionary resources.

Other distros

If you want to package gd-tools for your distro and know how to do it, please create a pull request.

Manual installation

This method is not recommended because it's easy to make a mistake during the process and end up with a broken install. If a package for your system isn't available, you should create it first.

1. Install Xmake 2.9.3 or newer and GCC 14.1 or newer.
2. Install mecab (mecab-git from the AUR).
3. Ubuntu users also need GCC 14 and libcurl headers: sudo apt install g++-14 libcurl4-openssl-dev.
4. Run ./scripts/quickinstall.sh.

Note: You can run ./scripts/quickinstall.sh --local to install the program locally (to ~/.local/ ).

To uninstall a system-wide manual installation, run ./scripts/quickuninstall.sh. If you installed with --local or --user, run ./scripts/quickuninstall.sh --local instead. This removes files installed by the scripts and preserves other files in their directories.

Note: Ubuntu's standard repository may provide an outdated version of Xmake. Install Xmake from the official Xmake PPA instead:

sudo add-apt-repository ppa:xmake-io/xmake
sudo apt update
sudo apt install xmake

Setup

Open GoldenDict, press "Edit" > "Dictionaries" > "Programs" and add the installed executables. Set type to html. Command Line: <name of script> --word %GDWORD% --sentence %GDSEARCH%. Optionally add arguments, such as: gd-mecab --word %GDWORD% --sentence %GDSEARCH% --user-dict <path> --font-size 20px. These programs are treated as dictionaries and you can add them under "Dictionaries" or "Groups".

gd-marisa

This script outputs the sentence with clickable characters and searches for the longest available dictionary entry (from a predefined list) beginning at that character. For deinflection it currently relies on rdricpp. It also provides links of available entries of smaller substrings.

Usage

gd-marisa --word %GDWORD% --sentence %GDSEARCH% --path-to-dic [PATH_TO_DIC_FILE]

The path to the .dic is an optional argument and defaults to /usr/share/gd-tools/marisa_words.dic

Dependencies

marisa-trie. The official Arch Linux package is called marisa, but it's already a dependency of goldendict.

Building an own index from a set of words

If you would like to make changes to found words, you can also create an own index from a newline-separated list of words (here called keyset.txt):

marisa-build < keyset.txt > keyset.dic

More information at https://www.s-yata.jp/marisa-trie/docs/readme.en.html

gd-mecab

This script passes a sentence through mecab in order to make every part of the sentence clickable.

screencast-area-2023-.3.-18_20-56-12.mp4

Dependencies

This script requires MeCab and the IPA dictionary to be installed. If you are on an Arch Linux system you can simply install the AUR package mecab-ipadic to obtain both.

Command format

Add this script to GoldenDict under "Dictionaries" > "Programs" (format HTML), like this:

gd-mecab --word %GDWORD% --sentence %GDSEARCH%

Optional arguments

• --font-size SIZE the font size to be used, e.g. 30px.
• --user-dict FILE full path to the user_dic.dic file. This is done automatically if you install via make.

gd-images

This script shows the top 5 pictures from Bing images for the given search string.

gd-images --word %GDWORD%

gd-strokeorder

This script shows the search string in the KanjiStrokeOrders font.

gd-strokeorder --word %GDWORD%

Font source: https://www.nihilist.org.uk/

Arguments:

• --max-len 5 maximum size of the input string.
• --font-size 10rem font size. It has to be large in order to see the stroke numbers.

How to call:

gd-strokeorder --word %GDWORD%

gd-handwritten

This script displays the handwritten form of each character

gd-handwritten --word %GDWORD% --font-size 4rem

Bundled font: ArmedLemon.ttf. The original usage terms and an English summary are included in res/ArmedLemon-LICENSE.txt.

How to call:

gd-handwritten --word %GDWORD%

gd-massif

This script shows example sentences from https://massif.la/

gd-massif --word %GDWORD%

gd-ankisearch

This script searches Anki cards in your collection that contain %GDWORD%.

Arguments:

• --field-name NAME optional field to limit search to.
• --deck-name NAME optional deck to limit search to.
• --show-fields VocabKanji,SentKanji optional comma-separated list of fields to show.

Example invocation:

gd-ankisearch --field-name VocabKanji --show-fields VocabKanji,SentKanji,Image,SentAudio --word %GDWORD%

gd-translate

Usage

gd-translate --sentence %GDSEARCH%
gd-translate --spoiler yes --to fr --sentence %GDSEARCH%

Dependencies

This script requires Argos Translate and the JA -> target language package to be installed.

pipx install argostranslate
argospm install translate-ja_en

argos-translate -f ja --t en "Hello World!"

If you can't install argostranslate due to an error, try an older version of python.

pipx install --python /usr/bin/python3.9 argostranslate

gd-mandarin

This script passes a sentence through mecab in order to make every part of the sentence clickable. It also automatically converts the sentence to traditional characters.

To use gd-mandarin, you need to install gd-tools by running ./scripts/quickinstall.sh --mandarin.