Swac plugin for Anki

This plugin allows you to easily memorize word sounds from Swac collections.
There are currently more than 85000 words in 15 languages: Arabian, Bielorussian, Chinese (Mandarin), Czech, German, Dutch, English, French, Portuguese, Russian, Serbian, Spanish, Swedish, Ukrainian and Wolof.

The files audio sounds won't automatically be saved in ~/.anki/*.media/ as usually and other programs will also be able to use these files (for example: Swac-explore, Stardict, QStardict...)

This is the 0.8.4 version of the plugin (September 2009). It has been tested on Anki and Graphical user interface and this documentation have also been translated into French language.

Changes since the 0.8.0 release:


You must use a GNU/Linux distribution (because the Swac tools don't exist on Windows and MacOSX yet) and to install the programs:
python (>= 2.5), python-qt4 (>= 4.4.0), swac-get (>= 0.2), swac-play (>= 0.1), anki (>=

Install also: sox (=>14.1), mplayer, lame and python-pyaudio. The 2 last programs are usefull to record your voice witha microphone.


  1. Download and install the swac-get and swac-play tools
  2. Download and install at least one Swac audio collection (or don't install a collection if you want to use a online collection)
  3. Build your Swac database
  4. Install the Swac plugin:

    - for Anki >=
    File menu -> Download -> Shared Plugin -> search the pattern Swac

    - for Anki = or =, open a terminal, then:
    $ wget ftp://download.tuxfamily.org/polyglotte/archives/anki_swac0.8.4_0.9.9.6.zip
    $ mv anki_swac0.8.4_0.9.9.6.zip ~/.anki/plugins/ && cd ~/.anki/plugins/
    $ unzip anki_swac0.8.4_0.9.9.6.zip
    $ rm anki_swac0.8.4_0.9.9.6.zip



A new Swac Settings menu has appeared in Settings menu -> Plugins (a deck must be open to active this menu from Anki >=

Select the Swac Settings menu. The following window appears:

Swac model name, Swac field name and Swac database file path are the most important parameters:

Swac model creation

To create a Swac model: Settings menu -> Deck Properties -> Models & Priorities tabs -> Models section -> Add button -> Add Model window:

You has noticed a new item: Add: Swac. Select it and press OK button. The language selection window displays:

In this example, we have already installed three Swac collections (swe-balm-voc, eng-balm-verbs and eng-wcp-us) related to two languages: English (eng) and Swedish (swe). We select English language.

On Deck Properties dialog windows, click on Edit button to see the characteristics of your Swac model:

You can recognize four parameters already seen before in the Swac Settings window:

There is also another parameter on (2): eng. It's the ISO-639-3 symbol language of your Swac model (related to English of course).

By scrolling the Answer area, we would have seen the expression: %(Phonetic)s

You can also recognize two other parameters already seen before in the Swac Settings window:

You can notice that the Option combo boxes are different on the two images. It's logical because it's useful to type the Answer only with Recognition questions/answers (look at the Reviewing section).

Card creations

Click on:
Settings menu -> Plugins -> Swac Settings -> Add/Suppress cards
... or press the F12 key by defaults.

For each new expression selected from the "Swac expressions remaining" list then moved ('>' button) to the "Anki expressions" one, one or two new cards will be created, depending on the state of the button on the top of the window.

For each expression selected from the "Anki expressions" list then moved ('<' button) to the "Swac expressions remaining" one, one or two cards will be deleted (the state of button on the top doesn't matter here).

Try "Ctrl+A" on a list to select all their expressions.


Classical questions/answers

We start by a Reading question/answer card. Take your microphone.

Click on the "Record your voice" button (3) (or on the F10 key by default).

Press the "Enter" key to finish the recording, then one more time to listen the answer -> the Swac audio is playing.

Your voice is immediately played before the native speaker one in order to be able to compare them. Choose the answer "Again", "Hard", "Good" or "Easy" to finish the card.

It is possible to replay the Swac sound by clicking on the "Repeat audio" button (1) (or by pressing the "F5" key by default) or your voice by clicking on the "Repeat your voice" button (3) (or by pressing the "F11" key by default).

It is possible for you to use Anki without a microphone and recording your voice, but in this case your judgment will be less accurate.

The "Record your voice" and "Repeat your voice" icons don't appear in the next screenshots because I have chosen to hide them (look at the "Settings" dialog box to do the same thing). But the key shortcuts remain enables.

Then we continue with a Recognition question/answer card. You must recognize the expression played:

You can also replay the audio sound if you press the F5 key or if you click on the audio button.

Questions/Answers with homographs

Some words can be pronounced of several ways. For example "bass":

A clue is written in brackets. It's the clue.

Advanced usage

Exportation/Importation from other models/decks

Instead of manually select the expressions from the Expressions Selection dialog window, you can automatically add the expressions found to other models or decks. It's magic. ;-)

Open a deck (or keep your current deck if it contains other models) and click on Settings menu -> Plugins -> Swac Settings -> Export. Exemple:

By defaults, every word found both in your fields and in the Swac collection (related to the lang selected) will be export. You will afterward be able to import the related file in a Swac Expression field (in this deck or another) with:
File menu -> Import:

You must change the Field 2 content like on the image above.

"Suggest export" (it's the second item of the combo box on the bottom of the window) makes it possible for you to send a list of words to Shtooka project. If you are lucky, the related sounds will be quickly add to the Swac collection.


It's sometimes difficult to correctly answer to a Recognition question because several words can be homophons.

For exemple, the following sound:

... can be spelled: write, right or rite!

You can press on the Edit Current button to add the other homophons in the Homophon field:

Even if you have only got one homophon, you can answer Hard, Good or Easy.

How do you know that words are homophons?
you can:


If it is difficult for you to recognize a word or an expression or pronounce it, you can edit the fact and add it the phonetic.

You will be able to find a phonetic editor here. But you can add the phonetic by copying and pasting it from a software dictionary like StarDict or QStardict.

Good reviewing!

Useful links:
A French website about language learning: http://polyglotte.tuxfamily.org/
You can offer to record audio files here

Author: Alteo Gange
If you want to translate the graphical user interface in your language, send me an email.

License document: GNU GPL, version 3 or later
License of Swac extension for Anki: GNU GPL, version 3 or later
License of Swac tools: GNU GPL, version 3
License of Swac collections: usually free (but not always)