priorities for July 2020 - NomVallex production TODO
Please tick any item before you START working on it so that we do not have multiple people working on the same task.
These are the tasks we need to finish in order to use http://quest.ms.mff.cuni.cz/pyvallex/ as the webpage of NomVallex 1.0. Most of them should be fairly straighforward, I mark the easiest as JUNIOR:
Changes that go into master:
-
[frontend] proper alphabetical sorting (right now, lexemes ÚTĚCHA and ČETBA appear towards the end of the list) -
[LaTeX2html, assigned to @Ansa211] convert the NomVallex book intro into html form; on NomVallex production, replace the Vallex book intro, while on master, find a way of showing both
- tasks related to individual attributes:
-
[txt data] pdt-vallex: change in the source data: leave there only the actual linked id or the value "no" (possibly with aspect index, i.e. impf: no pf: no
is also ok) -
[backend and frontend] pdt-vallex: copy the file with pdt-vallex in Vallex format (svn: NomVallex/Data/pdt-vallex_3.0_in_vallex_format) to production and add it to the list of parsed files -
[frontend, maybe also backend; JUNIOR] derivedV (and possibly pdt-vallex): make sure the value vallex_no is shown in the frontend as Vallex-no
-
[frontend - Vue templates; JUNIOR] show the valdiff attribute as part of the derivedV line (i.e., after the list of verbal units from which the noun is derived)
-
Changes that go into nomvallex-production (a branch that we manually deploy to http://quest.ms.mff.cuni.cz/pyvallex/):
-
[frontend] EDIT: see #96 (comment 391712992) ; first implement that and THEN see what else needs to be done in pyvallex everywhere in the interface, ignore all but the following attributes:attributes coming from the txt data: lemma, frame, synon, derivedV, pdt-vallex, semcategory, examplerich, classdynamic attributes: valdiff, pos
- tasks related to individual attributes:
-
[frontend - Vue templates; JUNIOR] synon and valdiff: do not show the info about obligatoriness (the "obl" and "opt" upper index) -
[backend, frontend] EDIT: see #96 (comment 391712992) frame: completely ignore typical complementations (ideally remove them already when parsing the data so that the user never sees them; this should automatically remove typical complementations from valdiff as well) -
[backend and frontend] pdt-vallex: (once the pdt-vallex file is added to production and corresponding changes to the master branch) treat the file with the pdt-vallex data it in the same way as we treat the file v-vallex (the data from v-vallex is read but cannot be searched directly, but because it is read, the lexical units from v-vallex can be shown in the attribute derivedV; run git grep v-vallex
on the pyvallex repo and make edits around the occurrences of v-vallex in the files deploy/production/server.ini, vallex/data_structures/collections.py, vallex/example-pyvallex.ini, vallex/grep.py, vallex/server/frontend/src/Api.js, vallex/server/frontend/src/Store.js, vallex/server/frontend/src/components/SearchForm.vue) -
[frontend OR txt data] EDIT: see #96 (comment 389060961) derived: pretend that "derivedV" is called "derived" OR rename derivedV and derivedN to derived everywhere in the textual data (this needs to be discussed with Markéta and Vendulka and thought through)
-
Tasks that require some analysis and it's not obvious what to do:
-
[frontend?] with some attributes, the drop-down list gives a variant with a dot at the end, e.g. semcategory.
,pdt-vallex.
,status.
- why? we do not want this -
u blu-n-uvědomování-si-uvědomění-si-1 a blu-n-zakazování-zakázání-1 nevidim v examplerich priklady k impf, ale jen k pf (ovsem bez oznaceni pf:)
Edited by Anna Vernerová