Fluent Forever Vocab Builder

Video demo

This is a web app that searches and scrapes info for a vocab word and builds a card for your anki deck, based off the FF model deck.

• Web app because Anki add-ons are not well documented and this separates concerns a bit. Was tkinter but it's not great to work with long-term.

Basically a quick way to do the vocab step step in the FF method.

Decreases card creation time from 6+ minutes to 30 seconds.

Gathers the following info:

• IPA (Wiktionary)
• Pronunciation recording (Wiktionary)
• Gender/Word Usage (Wiktionary)
• Images (Google)
  • Resized automatically to reduce AnkiWeb sync time
• Personal notes (you)

Rationale

• Because the app is in disarray right now (v1.0.0, April 2019) and review sessions are a pain.
  • Words are repeated over and over, sometimes endlessly
  • Update, May 2019: Bugs are fixed, but I think I like this better anyway.
• I found I like having editable notes that I have full control over.
• App is maybe worth it ($10/mo) for its pre-made content, grammar difficulty levels, and hopes of tutors. But other than that, Anki is much more mature and flexible.
• We can also use Google Images here over Bing b/c of small scale.

Requirements

• Python 3
• ffmpeg installed and added to system PATH (for previewing pronunciation audio)
• Model deck from FF imported into Anki
  • This program only creates the "2. Picture Words" note type at the moment
    • Will generate the spelling (optional), production, and comprehension cards automatically
• AnkiConnect add-on (code 2055492159)
• Anki needs to be open before starting the app

Installation

preparation

Zlib and libjpeg are required by Pillow https://pillow.readthedocs.io/en/latest/installation.html

macOS

brew install zlib
# see output, set shell env-vars, for example:
export LDFLAGS="-L/usr/local/opt/zlib/lib"
export CPPFLAGS="-I/usr/local/opt/zlib/include"

follow

• git clone https://github.com/cofinley/FluentForeverVocabBuilder.git
• Go into the project's directory
• python3 -m venv venv
  • Python 3
• venv\Scripts\activate
  • on Linux/Mac source venv/bin/activate
• pip3 install --upgrade wheel pip
• pip3 install -r requirements.txt

Usage

1. Open Anki in the background
2. Go into project's directory
3. Click on (or run in command line) the start program for your OS
   • start.bat for Windows
   • start.sh for Linux/Mac
     • Might have to run chmod +x ./start.sh if it's not executable
4. Go to 127.0.0.1:5000 in your browser
5. Select language
   • This is saved for the next session
6. Select destination deck
   • Also saved
7. Enter word to search
8. Search (takes time to download images)
9. Tweak any values
10. Select pic(s)
    • You can paste images from your clipboard into the selected images area with Ctrl+V
    • You can also drag files in to the area
11. Enter any notes you want
12. Submit (takes time to store everything into Anki)
13. Repeat

Updating

• Go to the installation directory for the project
• git pull
• pip3 install -r requirements.txt

Development

pip3 install pip-tools

Specify major (not exact) package versions in file requirements.in then compile requirements.txt from it. pip-compile requirements.in

requirements.txt holds exact versions

Known Issues

• Phrases are not supported; currently used for single words that you could find in Wiktionary
  • Can still use program, just no IPA, audio, or definition choices if Wiktionary doesn't have it.

Todo

• Easier installation and startup
  • Batch/bash files created, a whole installer would be nice
• More recordings from Forvo, maybe
• Other note types in the model deck (i.e. grammar, minimal pairs)

Image was not working see hardikvasa/google-images-download#298 solutions:

• now, see requirements.in (may need to update from time to time)
• maybe try fix it by using https://github.com/RiddlerQ/simple_image_download