Testing Pull Request #87

### Context
I'm currently evaluating Pull Request #87 and performing some ad-hoc testing of the new program. This issue provides details of aspects of that testing in the hope it'll be useful to you and others if/when you engage in similar work.

I configured the PYTHONPATH as described in: https://github.com/commercetest/nlnet/wiki/Using-PYTHONPATH

### Testing input handling
The program accepts an optional command-line parameter for the input file name. The tests incrementally exercised the program starting with no input file, then by adding an empty file, then adding a single unrelated word to represent a column name that _wouldn't_ match any of the ones the program looks for, and finally by replacing this word with the name of the only column the program appears to need in order to run.

```
python src/DS/metrics_extraction.py
touch ~/NLnet-Projects/commercetest-nlnet/data/guessed_languages_rows.csv
python src/DS/metrics_extraction.py
echo column > ~/NLnet-Projects/commercetest-nlnet/data/guessed_languages_rows.csv
python src/DS/metrics_extraction.py
echo guessed_language > ~/NLnet-Projects/commercetest-nlnet/data/guessed_languages_rows.csv
python src/DS/metrics_extraction.py
```
By this point the program ran without reporting any errors.
```
2024-11-25 14:00:08.966 | INFO | __main__::306 - Working directory: /home/julian/NLnet-Projects/commercetest-nlnet
2024-11-25 14:00:08.971 | INFO | __main__::323 - Found 0 already processed files.
2024-11-25 14:00:08.971 | INFO | __main__::407 - Final dataframe saved to /home/julian/NLnet-Projects/commercetest-nlnet/data/test_metrics_df_guesslang.csv
```

### Good next tests?
What might be good next tests to perform as well as these tests? We can add these test ideas as comments to this issue.

Start with src/DS/metrics_extraction.py and the command sequence in the issue, using the linked PYTHONPATH instructions for setup. Done means recording concrete additional input-handling test ideas and their observed outcomes in this issue; no test file or implementation change is identified.

Written by the indexing model from the issue text.