[Crum] Design a Pipeline for Semantic Sisters

[pishoyg]

Let's say you just read the word ⲙⲟⲩϣⲧ, which means examine or search. You want to learn the true sense of the word, so you look at examples. The examples give you an idea of what the word means.

But let's say you want to go further and learn how to talk about the concept of searching in Coptic in general. In such a case, it's helpful for you to look at words such as ⲕⲱϯ or ϧⲟⲧϧⲉⲧ. These are referred to as sisters or semantic sisters (to distinguish them from lexical or etymologic sisters or cognates).

Implementing search in Crum (#108) would make this task easier, because lots of sisters will be obtained with a simple for the translation. You can also use the search feature in Anki.

• The display might consist of a footnote, or a section at the bottom of the page called See also.
• This would be an appendix, thus it will go to our appendices sheet.
• The sisterhood relation should be symmetric, meaning that if ϧⲟⲧϧⲉⲧ shows as a sister of ⲙⲟⲩϣⲧ, then ⲙⲟⲩϣⲧ should show as a sister of ϧⲟⲧϧⲉⲧ. This symmetry should be enforced by a pre-commit.
• The sisterhood relation does NOT need to be transitive, although it often is.

[pishoyg]

NOTE: Both Crum and KELLIA (example) have cf. annotations. KELLIA often has a See also: section (example).

[pishoyg]

Here is what the pipeline will look like:

1. Add a sisters column to your appendices sheet.
2. The column should bear the keys to the sisters, in a symmetric manner. If a is sister to b, then b is sister to a. Symmetry can be enforced in the crum_appendices pipeline. It would be desirable to obtain it in the source sheet as well. You could do this through a GSHEET macro, or manually in the meantime.
3. Your flashcards pipeline should then introduce an optional See also section (and that's a better name for our use case than cf.); right after the Editor's Notes section, which has the list of sisters. I am inclined to use hover-link formatting, with the displayed text being the title (word-title column).

[pishoyg]

In our expansion of Crum's original data (currently planned in #204 and #121, and to some extent #9), we might want to include plans for the cf. entries that Crum mentions.

[pishoyg]

One task remaining! Add an argument to your appendices helper that makes it easy to populate the relations for a symmetric family.

For example, typing the following:

./dictionary/marcion.sourceforge.net/appendices_helper.py --sisters ${KEY_1} ${KEY_2} ${KEY_3}

should prompt the script to add ${KEY_2} and ${KEY_3} as sisters of ${KEY_1}; ${KEY_1} and ${KEY_3} as sisters of ${KEY_2}; and ${KEY_1} and ${KEY_2} as sisters of ${KEY_3}. The change should be reflected in the sheet, which you can then download to pick up the changes that you made through your shell.

[pishoyg]

Almost done! Assigning so I will get it done!

Semantic sisters are cool. :)

[pishoyg]

Sometimes, not all the dialects that a word belongs to have infinitive forms! In those cases, some dialects might end up being unrepresented! This is undesirable.

[pishoyg]

Filed #270 to track the fix. Reclosing this as done!

[pishoyg]

Make it possible to include anchors in sister links. Sometimes, you want to link a specific derivation or a sense, because the "default" meaning of the word doesn't make it a sister.

For example, both ⲕⲧⲏⲣ and ⲙⲁⲥⲓ mean calf. The latter comes under ⲙⲓⲥⲓ. But marking ⲙⲓⲥⲓ itself as a sister of ⲕⲧⲏⲣ is confusing, You should mark the derivation - ⲙⲁⲥⲓ.

[pishoyg]

Status:

DONE:

• Support fragments. (Commits yet to be pushed!)

TODO:

• Add validation for fragments.

[pishoyg]

TODO: Implement sister groups.

[pishoyg]

It was marked as closed by mistake! Reopening!