One word a day keeps brain rot away.
Larousse's mini débutants dictionary (1990).
dailyword is a (very) simple bash computer program that unveils a new random word along with its definition(s) helping you review french words progressively, once a day, one at a time, straight from your terminal. Furthermore, the displayed words are not common ones. They have been carefully extracted by cross-referencing two datasets. You'll find more informations about the dataset building in the processing part and in the acknowledgments.
Pardon my french ;)
The idea occurred to me when I came accross a Guillemette Faure article. Why not having something like this? But not in the shape of another heavyweight app. There are already too much of these, and most just keep increasing our mental workload until we stop using them after a short time. Nah, I wanted to take over an unexplored space, in a simple and minimalist fashion that could spark my curiosity without being overly intrusive nor distracting. What's better space for that purpose than a terminal? So, after a weekend of tinkering, here it is! The first prototype. There is a broad scope of possibilities in there: I really invite you to take over the concept and make it your own daily greeting space somehow. That could be a mental challenge, an excerpt from your favourite book or just a song to listen to. Whatever, as long as it reflects a genuine part of your self and makes your day start with a mesmerizing smile.
Contents - click to expand
- The dailyword.deb file and dpkg (built-in Debian/Ubuntu based systems).
- Only Bash (or a compatible shell), that's it, really.
- Standard Unix utilities: date, awk (preferably GNU awk for FPAT), sed, printf, mv, grep and alias (built-in shell).
Note
In short: you should not need to install anything more than the dailyword program. In standard Unix systems, required utilities are built-in by default.
Only if you want to experiment with the data processing and build the definitions dataset from scratch. You'll need:
- Python (3.7 or newer)
- pandas (1.3.5 or newer)
- requests (2.31.0 or newer)
Open the terminal and in the processing folder, run:
pip install -r requirements.txt You can then run the processing script from you favourite IDE or manually with:
python processing.py Note
If you installed the .deb package, the computed dataset is already located in the opt/dailyword/data folder of your system. You can also find all the archived datasets in the src/data/datasets_backup folder of this repository.
Run the installation script:
curl -sL https://raw.githubusercontent.com/brooks-code/fuzzy-carnival/main/install_dailyword.sh | bashThis command automates the download and installation process of the dailyword.deb package.
in case you already have the .deb package on your disk - click to expand
You can then proceed like this:
sudo dpkg -i dailyword.debfrom the package location folder.
You can remove the software with this command:
sudo dpkg -r dailywordThis will delete all the installed files and restore the .bashrc file to its initial state.
Click to expand
If for any reason, you don't want to use the .deb package. On GNU/POSIX compatible systems, you can still set things up manually:
-
Download (and unpack) or clone this repository:
git clone https://github.com/brooks-code/fuzzy-carnival.git cd fuzzy-carnival-main # if you're not in the created directory already.
-
From your local folder, ensure
dailyword.shis executable, and copy the relevant contents of thesrcfolder to your system'soptfolder:chmod +x src/dailyword.sh && sudo mkdir -p /opt/dailyword && sudo cp src/dailyword.sh /opt/dailyword/ && sudo cp -r src/data /opt/dailyword/
Change permissions of the dictionary file's data folder:
sudo chown -R $(whoami):$(whoami) /opt/dailyword/data/
-
Optional. You can already test dailyword with this command:
/opt/dailyword/dailyword.sh
-
Copy the
daily_bashrc.shfile to your distro'setc/profile.dfolder:sudo cp src/daily_bashrc.sh /etc/profile.d/
NB: If you copied the source files in some other folder than the default
/opt/dailyword/mentioned above, you will have to update the content of thedaily_bash.rcfile accordingly. -
Update your distros
.bashrcfile equivalent by sourcingdaily_bashrc.sh:echo "source /etc/profile.d/daily_bashrc.sh" >> ~/.bashrc
From now on, your terminal will display a new word once a day at startup.
-
Remove unneeded files.
cd .. && rm -rf fuzzy-carnival-main
To uninstall dailyword manually, you can undo the steps from 5 to 1 mentioned above with this command.
sed -i '/source \/etc\/profile.d\/daily_bashrc.sh/d' ~/.bashrc && sudo rm -rf /etc/profile.d/daily_bashrc.sh && sudo rm -rf /opt/dailyword && rm -f ~/.dailyword && source ~/.bashrcThis command will:
- Remove the line sourcing
daily_bashrc.shfrom.bashrc. - Delete the
daily_bashrc.shfile from/etc/profile.d. - Remove the application files from
/opt/dailywordand delete the folder. - Delete the
.dailywordmarker file from the home folder if it exists. - Reload the
.bashrcfile to apply the changes.
The script displays a new word, once, every new day. You have nothing more to do! If you forgot your daily surprise, you can invoke it back during the current day with:
oogfWhy oogf? Well that's a long story, but for the mnemonic association, just remember that the acronym stands for: Oh oui, grand.e fou.e! (inclusive form for: oh yeah, crazy girl! or oh yeah, crazy boy!). That might help you remember the magic command.
The CSV file is updated after every display change, so you can still find a word you'd like to recall by it's display date (range).
dailyword/
├── DEBIAN
│ ├── control
│ ├── postinst
│ └── prerm
├── etc
│ └── profile.d
│ └── daily_bashrc.sh # source file for .bashrc
└── opt
└── dailyword
├── dailyword.sh # main bash script
└── data
└── daily_words.csv # datasetThe script utilizes the following configuration variables:
CSV_FILE: # Path to the CSV file containing the words and definitions.
# (default: (app_dir)/data/daily_words.csv)
ASCII_ART: # Decorative text art displayed at the top.
ASCII_COLOR: # Color used for displaying ASCII art header (default: copper).
# Gradient configuration variables:
GRADIENT_START: # The starting RGB values for the gradient (default: teal - [0, 128, 128])
GRADIENT_END: # The ending RGB values for the gradient (default: dark orange - [255, 140, 0])
SCALE: # Used for calculating the gradient steps.The colors have been selected to fit both dark and light terminal modes. You can adjust these variables directly in the script if you need some changes to better reflect your visual preferences and environment.
The main objective is to provide new words to be discovered with the challenge of avoiding too common words. To ensure that, the processing leverages two datasets: Lexique and Dico. The processing script keeps only pertinent columns in Lexique and computes a custom frequency index (a weighted average) that synthetizes frequencies based on different measures. After reading this (french), the choice has been made to emphasize the weights of the book-related frequency indexes since, vocabulary found in literature tends to be more nuanced and sophisticated than the one used in movies.
The process is followed by associating the data with matching definitions entries from the Dico dataset. The cleaning process removes duplicates and words considered as too common (i.e. too frequent) based on the custom frequency threshold. Also, definitions with keywords who potentially indicate a definition with less informational value will be filtered out. The script also corrects a bias occuring for some feminine definitions to ensure equivalent levels of description for both genders.
Finally, the sorted results, with their handled accented characters, are exported to a CSV file, thus providing an informative list ready to be processed by the dailyword program.
| Column Name | Description | Type |
|---|---|---|
| mot | The word or term being defined. | String |
| genre | Gender: (m)asculine, (f)eminine or missing (""). | String (1 char) |
| lemme | The base form or lemma of the word, often used in linguistic analysis. | String |
| phonétique | The phonetic representation of the word, indicating how it is pronounced. More information here (french). | String |
| custom_freq_index | A numerical score that represents a weighted average of several indexes related to frequencies. More weights have been added on the books related frequencies, read why here (french). Frequencies are occurrences per million words (pmw). | Float |
| définitions | The definitions or meanings of the word. Multiple definitions are separated by a pipe symbol. | String |
| display_date | The date on which the word and its information have been displayed. | Date (YYYY-MM-DD) |
Example:
"mot","genre","lemme","phonétique","custom_freq_index","définitions","display_date"
"abaca","m","abaca","abaka","0.002","Bananier des Philippines de la famille des musacées, dont les feuilles fournissent le chanvre de Manille, une matière textile. | Fibre textile tirée du bananier du même nom, appelée aussi chanvre de Manille ou tagal.",""
As you see can see above, in case of polysemy, definitions are separated by a | (pipe) symbol.
- A slight lag occurs while the dataset is being processed during the awk pass.
- The script depends on GNU awk, which supports the FPAT variable. Compatibility may vary on different systems.
Some improvements that might happen during the next releases:
- Better low content definitions cleaning.
- Keep optimizing the main script for better performance.
- Use more advanced NLP techniques to ensure a better semantic discoverability.
- If you get the
Error: CSV file not foundmessage, verify that the CSV file path in thedailyword.shscript points to the correct location and that the file is accessible. - If no definition is displayed and the output returns
No more definitions to display, ensure that there are still definitions left in the CSV with an empty value in thedisplay_datecolumn. - If the definitions are not appearing, each day, the first time you launch your terminal and the
oogfcommand is throwing back an error, please check if your.bashrcfile is correctly sourced. - Check that your terminal supports 24-bit (true color) for the gradient display to be displayed correctly.
- For compatibility issues with awk, ensure your version supports FPAT or consider installing GNU awk.
- Review Bash scripting fundamentals for text processing.
- ANSI escape codes cheatsheet (used for colorful terminal outputs).
- Explore how CSV files can be managed and updated via Unix command-line utilities. Link 1, Link 2.
Feel free to open issues or discuss any improvements. Contributions are welcome! Each contribution and feedback helps improve this project - it's always an honour :)
- Fork the repository.
- Create a branch for your feature or bug fix.
- Submit a pull request with your changes.
The source code is provided under the Unlicense license. See the LICENSE file for details.
This project is not affiliated with or endorsed by any third party. It is provided as is for educational and practical uses. Use it at your own risk (and always back up your data).
This software makes use of various open source libraries and Unix utilities. Special thanks to the developers and the community for providing these valuable tools.
Datasets:
- Lexique 3 (version 3.83) provided valuable insights about french words frequencies. More information here.
@article{New2001,
author = {New, B. and Pallier, C. and Ferrand, L. and Matos, R.},
title = {Une base de données lexicales du français contemporain sur internet: LEXIQUE},
journal = {L'Année Psychologique},
volume = {101},
pages = {447-462},
year = {2001},
url = {http://www.lexique.org}
}
@article{New2004,
author = {New, B. and Pallier, C. and Brysbaert, M. and Ferrand, L.},
title = {Lexique 2 : A New French Lexical Database},
journal = {Behavior Research Methods, Instruments, & Computers},
volume = {36},
number = {3},
pages = {516-524},
year = {2004}
}- Kartmaan's french wiktionary (wiktionnaire) CSV dump, for providing definitions. Also available on Kaggle.
Misc: