[Bug]: Wrong syllable count

[stribor14]

Operating system

Linux

Download source

PyPi

Python distribution

Anaconda

Python version

Python 3.9.6

What happened?

When checking the results, some interesting stuff came to my attention. I was checking the return value of difficult_words for single-word inputs. To skip the part of reading through code, checking the easy_word_list in resources, etc. Here is what I found.

textstat.syllable_count("faeries") = 1
textstat.syllable_count("relived") = 3

both of those words have 2 syllables

Other information

Failing tests

def test_the_bug_i_found():
    assert (textstat.syllable_count("faeries") == 2) and (textstat.syllable_count("relived") == 2)

Code of Conduct

• I agree to follow this project's Code of Conduct

[LKirst]

That's probably an issue with Pyphen because textstat uses Pyphen to count syllables.

[alxwrd]

That's right, we use pyphen.positions(word) to calculate syllables.

https://github.com/textstat/textstat/blob/ad80df1/textstat/textstat.py#L326L363

Although this stackexchange answer suggests that Pyphen "[is] only 54% accurate", so there might be some work to do there: https://datascience.stackexchange.com/a/89312