mrgreekgeek / abbott-smith-greek-lexicon-online

A little JS app to make Abbott-Smith searchable and usable with a focus on good UI and UX
https://lexicon.mrgreekgeek.com
MIT No Attribution
2 stars 0 forks source link

Words with variant spellings aren't properly searched in lexicon #12

Open welcome-me opened 3 weeks ago

welcome-me commented 3 weeks ago

Problem

Some words in the New Testament have multiple spellings. Searching for the variant spelling does not find any results.

Example

Clicking on the last word in Luke 18:1 should pull up this word... signal-2024-09-20-21-01-45-504

But instead, nothing is displayed.

Screenshot_20240920-203803_Via

Expected Behavior

Clicking the last word of Luke 18:1 should take us to the word with the variant spelling.

Additional Info

This is a tricky problem for me to fix alone, @mrgreekgeek because I know nothing about Greek word forms.

I have no way of knowing that ἐνκακέω, ἐκκακέω, and ἐκκακεῖν are various forms of the same word. Insofar as I can tell, ἐκκακέω, -ῶ, v.s. ἐνκακέω says nothing about ἐκκακεῖν.

mrgreekgeek commented 3 weeks ago

@welcome-me thanks for bringing this to my attention. I think it's fairly standard practice for lexicons to list alternate spellings for a word and then redirect the user to the most common spelling to actually find the definition. But since we're doing this lexicon thing digitally, we can make things a lot easier for the user.

I just created a branch missing_entries where we can experiment on the best way to fix this. I was able to locate 209 entries that use "v.s." to redirect the user to a different Greek word. Not all of them are mere spelling variations, but it gives us some data to work with so we can figure out what the problem is, and what the best solution will be.

I also created a file of all entries that have no Strongs number (452 of them). Not sure why there are so many, but I'll have to go through that data and see what I can find out.

Processed 5896 entries.
Found 209 empty entries.
Found 452 with no Strongs number.
Results saved to ../parse-lexicon/empty_entries.json and ../parse-lexicon/entries_no_strongs.json

One quick fix that I thought of is to automatically create a link for the word with the canonical spelling so the user can click it to go to the canonical lemma rather than having to copy/paste it and search. You'll see that change in index.html on the branch, but I'm sure it can be improved.