search

Signbank / Global-signbank

An online sign dictionary and sign database management system for research purposes. Developed originally by Steve Cassidy/ This repo is a fork for the Dutch version, previously called 'NGT-Signbank'.
http://signbank.cls.ru.nl
BSD 3-Clause "New" or "Revised" License
19 stars 12 forks source link

Improve variable names related to morphology #351

Open Woseseltops opened 7 years ago

Woseseltops commented 7 years ago

The gloss detail view clearly organizes the morphology of glosses: you have sequential morphology, which is about glueing full glosses together, and simultaneous morphology, which is about smaller more abstract morphemes that do not have meaning on their own.

The underlying code does not mirror this clear structure, mostly because it emerged so gradually. Now we can see the full picture, I propose the following name changes:

Type Old name New name Reason
Model MorphologyDefinition SequentialMorphologyDefinition There also exists a SimultaneousMorphologyDefinition
Attribute MorphologyDefinition.morpheme .child This is not a morpheme, this is a full gloss; 'morpheme' refers to the Morpheme model within Signbank
Related name MorphologyDefinition.parent_gloss : parent_glosses sequential_morphology Related names are for searching the relation backwards, so instead of starting with a (sequential) morph def to get the parent gloss, you can search from a particular gloss to see whether it consists of any parts . It seems more elegant to do this with gloss.sequential_morphology than with gloss.parent_glosses.
Related name MorphologyDefinition.morpheme : morphemes compound_parents Related names are for searching the relation backwards, so instead of starting with a (sequential) morph def to get the child gloss, you can search from a particular gloss to whether it's part of a compound. It seems more elegant to do this with gloss.compound_parent than with gloss.morphemes.

Also, Gloss.morphemePart will be removed in #350 .

Hard part about this issue: the names of various things in the database need to change too.

ocrasborn commented 7 years ago

Now you're at it, let's take into consideration a third type of morphology: a combination of glosses, but simultaneously rather than sequentially. See #280 for the issue.

susanodd commented 1 year ago

This looks like it's still relevant!

susanodd commented 9 months ago

I'm encountering this issue in #1071. I've added Gloss display methods for the relations the gloss is in (the gloss is a field of another class/model). All of the morphology models fall into this.

There are plenty examples of Sequential Morphology.

I tried to find examples in the database of Simultaneous Morphology (NGT). But no glosses have this! @ocrasborn several years ago, the morphemes were all deleted. Then they were added back again. Did some gloss data go missing?

For Blend Morphology (NGT), there are two glosses: WEEKEND-A and WEGDOEN.

susanodd commented 8 months ago

See #1101