Grzegorz

- improve `wordlist`: accept ranges, e.g. `3000:5000`, which returns only the
3000th up to the 5000th most common words
- add `--numproc` option to `fetchipa`, to explicitly set the number of
concurrent processes that should handle the wordlist
- fix Anki flashcard template: highlight the correct word in Card 1, as it is
highlighted in Card 2
- optimise `fetchipa` by removing time-expensive code
- change: replace JSON with a custom place-text format
- change: make `fetchipa` not lose progress whenever it is cancelled, by writing
(appending) progressively to the output file

- add diacritics parsing
- add `analyse` command, for phonologically analysing a given IPA transcription
- add `check` command, for determining whether two IPA transcriptions form a
minimal pair
- add `list-languages` command, for printing out all languages (and their codes)
for which a wordlist can be fetched
- add `--keep-failed` option to `fetchipa`
- add `--filter-file` (`-f`) option to `generate`; this way, you can specify
which phonemes may form a minimal pair instead of always relying on a
default
- improve minimal pair generation performance
- fix bugs related to `fullmake` trying to generate even if the given language
is invalid
- fix bug where long sounds at the end of a syllable would make the entire
syllable not be registered
- fix chroneme and stress contrast checkers, which weren't very reliable for any
words larger than one syllable
- improve phoneme contrast checker by allowing phonemes to have different
lengths; ergo, more minimal pairs found
- change: specify the output file of the `makedeck` command instead of always
using the same name (`grzegorz-anki-deck.apkg`)
- change: keep chroneme and stress contrasts by default during minimal pair
generation; replace `--keep-chronemes` with `--no-chronemes` and
`--keep-stress` with `--no-stress`
- refactor to remove redundancies + add unit tests
- completely change the documentation structure

- fix ModuleNotFound error where `grzegorz` wouldn't run at all because import
weren't prefixed with the module name

- generator:
- parse every word's IPA phonologically, delimiting it into syllables and
sounds
- increase performance when generating minimal pairs based on syllable
stress and chronemes (due to the above)
- add a few unit tests
- fix: update relative imports, removing starting dots

- Anki flashcards:
- fix error that autoplayed audio on both the back and the front of the
cards
- change appearance of boxes containing words; use black (on light theme)
and white (on dark theme)

- fix `AttributeError` when running `fetchipa` for e.g. Greek