Switch gender annotation over to reference separate lexemes in a loop

Terms

• I have searched open and closed feature requests
• I agree to follow Scribe-iOS' Code of Conduct

Description

Scribe will be switching over its data process to be more directly based on one lexemes per data entry. At this time we combine lexemes together based on the individual strings, so in German the word Schild means sign and shield, but is one entry for us. In order to simplify the data formatting process, we'll need to remove this, which further means that the way we store genders will be different.

The current way is that if a string has multiple genders, then we'll store each of them separated by slashed, so F/M/N/C/PL and all the variants. We'll soon have a situation where we'll have one entry for every lexeme and their plural. What this means is that rather then checking to see if the string has a dash in it and then separating it, we'll need to get the gender and check to see if the string/lexeme occurs more time and then append those genders.

• Note that this is blocked by the new formatting processes in https://github.com/scribe-org/Scribe-Data/issues/142

Contribution

Happy to discuss the work for this and help with implementation or work on it myself at some point!

Start by reviewing the new formatting process in Scribe-Data issue 142, which blocks this work, then locate the current gender annotation handling in Scribe-iOS. Compare the existing slash-separated approach with the separate-lexeme format described here; done means genders are derived across repeated lexemes without the current string-splitting behavior.