Edition interface for TEI dictionaries

The XML format allows you to build interoperable lexicons. For instance, you can have a flexional lexicon whose forms are linked to those of a bilingual lexicon. This will allow things like displaying the translation of a verb when the user has searched for a conjugated form, even if there is only its infinitive form in the bilingual dictionary.

But there are always risks of confusion when two lemmas have the same spelling or two flexional forms from different lemmas are written the same. For instance, in Occitan, "codar" can be either conjugated as "còli", "còlas", "còla"... and mean "to glue" or be conjugated as "coli", "colas", "cola"... and mean "to flow". We wouldn't want to attribute the translation "to glue" with the conjugation "coli", "colas", "cola".

We can manage these cases thanks to the "homograph ID" field of the forms. Each form in all the lexicons is identified with the group spelling + language (and dialect) + detailed category + general category + homograph ID. These fields constitute the form unique identifier which allows linking forms in many lexicons.

In my example, we will choose which homograph ID to attribute to "codar" - "to glue" (i.e., 1) and which one to attribute to "codar" - "to flow" (i.e., 2). These homograph IDs will be attributed to the right form in both bilingual lexicon and flexional lexicon. Then, you will use the forms' unique identifier to link the right translation to the right conjugation.

If you don't need your lexicons to be interoperables, you can let "1" in the homograph field of all the forms.