spidersouris / termic

A replacement for Microsoft Terminology Search

termic: an alternative to Microsoft Terminology Search

A deployed and ready-to-use version of termic is available on https://termic.me.

Data Collection

Data used by termic is available for download on Dropbox.

Translation Memory

The 2020+ translation memory was retrieved from Visual Studio Dev Essentials as .csv files. Those files were merged using a custom script, merge_csv.py, that is available in the termic-data GitHub project.

To download Microsoft's translation memory, follow these steps:

1. Go to Visual Studio Dev Essentials.
2. Search "Translation and UI Strings Glossaries September 2020" in the "Search downloads" search bar.
3. The link should appear as you start typing; click on it and search.
4. Choose your language on the right and click on the "Download" button.

In addition, this dataset was expanded with VSCode strings (which are not available in the TM provided by Microsoft). vscode_data.py was used for extraction.

Glossaries

Glossaries were retrieved from the Microsoft Terminology Collection. Those are .tbx files that were converted to .xlsx using Xbench.

Requirements

Website

• Python (>=3.10)
• Flask (>=2.3.1)
• psycopg2 (>=2.9.6)

Scripts

• Python (>=3.10)
• requests (>=2.29.0)
• pandas (>=2.0.1)

Usage

1. git clone https://github.com/spidersouris/termic.git
2. cd termic
3. pip install -r requirements.txt
4. python termic.py
5. Go to http://localhost:5000

Using a database

If you want to run termic locally with the data available for download on Dropbox or with your own terminology data, using a local database is recommended.

You can change the connection string in config/db_config.py to connect to your database.

Using local data files

Using local data files with termic is far from being ideal. However, if it is your only option, here are a few tips:

• You can use pandas (or any other alternative Python data processing librairies) to process the .csv (translation memory) and .xlsx (glossary) files.
• If you do use pandas, consider converting the .csv and .xlsx files to the binary feather format to reduce disk usage and improve search time. To do so, you can use the convert_to_feather.py script, that is available in the termic-data GitHub project.
• You can use termic_pandas.py as a basis for your local deployment.

Development

flask --app termic run --debug

Deployment

You can use Heroku for deployment.

1. heroku login
2. heroku create --app termic
3. git push heroku master

Contributors

• benediktkr
• MK

termic's Issues

"Highlight Results" doesn't stay active after new search

1. Perform a search.
2. Click the "Highlight Results" option.
   
3. Click "Search" again without changing the search terms.
4. The highlight is gone even though the search term ("button" in this case) is still present in the results.
   

Regarding the table scrolling on mobile devices

It's currently only possible to scroll the search result tables on mobile devices by swiping on the table header (area with green border) instead of the table body (area with blue border). I'm guessing this isn't the indented behavior.

OS: Android 11
Browser: Edge, Firefox and Chrome

The 2017 data period with Turkish support

Hi,

Thank you for this wonderful project. Is it possible to add Turkish language to 2017 data? I really need Windows string search. Their 2020+ data is very limited. I miss their old terminology search website.