From 8d594658ab1b394a24f62528e761b54bacde482c Mon Sep 17 00:00:00 2001 From: Vincent Ambo Date: Tue, 24 Jan 2023 01:54:29 +0300 Subject: [PATCH] feat(corp/data-import): map OC word grammemes to OR form types This table maps the grammemes for individual word forms (*not* for lemmata in either corpus!) to the corresponding grammemes from the other dataset. These have drastically different shapes, so the mapping is not perfect, but will help in determining which forms are intended to be the same on both sides. Change-Id: Ib0717e2f7a79d96bcb5e955a20f551e391fcd759 Reviewed-on: https://cl.tvl.fyi/c/depot/+/7918 Reviewed-by: tazjin Tested-by: BuildkiteCI Autosubmit: tazjin --- corp/russian/data-import/src/main.rs | 1 + corp/russian/data-import/src/mappings.rs | 84 ++++++++++++++++++++++++ 2 files changed, 85 insertions(+) create mode 100644 corp/russian/data-import/src/mappings.rs diff --git a/corp/russian/data-import/src/main.rs b/corp/russian/data-import/src/main.rs index 95a38e6e9..21da48e8d 100644 --- a/corp/russian/data-import/src/main.rs +++ b/corp/russian/data-import/src/main.rs @@ -111,6 +111,7 @@ use std::fs::File; use std::io::BufReader; mod db_setup; +mod mappings; mod oc_parser; mod or_parser; diff --git a/corp/russian/data-import/src/mappings.rs b/corp/russian/data-import/src/mappings.rs new file mode 100644 index 000000000..5d8d4f335 --- /dev/null +++ b/corp/russian/data-import/src/mappings.rs @@ -0,0 +1,84 @@ +//! Manual mapping of some data structures in OC/OR corpora. + +/// This maps the *names* of OpenRussian grammemes (the set of +/// `form_type` fields in the `word_forms` table) to the *names* +/// of OpenCorpora grammemes. +/// +/// The names of the OR grammemes are much easier to understand in +/// general, as the OC ones seem to have strange acronyms in them, +/// however the OC ones are much more structured. +/// +/// As these forms map to the word_forms table they lack the forms +/// attached to the lemmata. +pub const FORM_TYPES_GRAMMEMES: &'static [(&'static str, &'static [&'static str])] = &[ + ("ru_adj_comparative", &["Cmp2"]), + ("ru_adj_f_acc", &["femn", "sing", "accs"]), + ("ru_adj_f_dat", &["femn", "sing", "datv"]), + ("ru_adj_f_gen", &["femn", "sing", "gent"]), + ("ru_adj_f_inst", &["femn", "sing", "ablt"]), + ("ru_adj_f_nom", &["femn", "sing", "nomn"]), + ("ru_adj_f_prep", &["femn", "sing", "loct"]), + ("ru_adj_m_acc", &["masc", "sing", "accs"]), + ("ru_adj_m_dat", &["masc", "sing", "datv"]), + ("ru_adj_m_gen", &["masc", "sing", "gent"]), + ("ru_adj_m_inst", &["masc", "sing", "ablt"]), + ("ru_adj_m_nom", &["masc", "sing", "nomn"]), + ("ru_adj_m_prep", &["masc", "sing", "loct"]), + ("ru_adj_n_acc", &["neut", "sing", "accs"]), + ("ru_adj_n_dat", &["neut", "sing", "datv"]), + ("ru_adj_n_gen", &["neut", "sing", "gent"]), + ("ru_adj_n_inst", &["neut", "sing", "ablt"]), + ("ru_adj_n_nom", &["neut", "sing", "nomn"]), + ("ru_adj_n_prep", &["neut", "sing", "loct"]), + ("ru_adj_pl_acc", &["plur", "accs"]), + ("ru_adj_pl_dat", &["plur", "datv"]), + ("ru_adj_pl_gen", &["plur", "gent"]), + ("ru_adj_pl_inst", &["plur", "ablt"]), + ("ru_adj_pl_nom", &["plur", "nomn"]), + ("ru_adj_pl_prep", &["plur", "loct"]), + ("ru_adj_short_f", &["femn", "sing"]), + ("ru_adj_short_m", &["masc", "sing"]), + ("ru_adj_short_n", &["neut", "sing"]), + ("ru_adj_short_pl", &["plur"]), + ("ru_noun_pl_acc", &["plur", "accs"]), + ("ru_noun_pl_dat", &["plur", "datv"]), + ("ru_noun_pl_gen", &["plur", "gent"]), + ("ru_noun_pl_inst", &["plur", "ablt"]), + ("ru_noun_pl_nom", &["plur", "nomn"]), + ("ru_noun_pl_prep", &["plur", "loct"]), + ("ru_noun_sg_acc", &["sing", "accs"]), + ("ru_noun_sg_dat", &["sing", "datv"]), + ("ru_noun_sg_gen", &["sing", "gent"]), + ("ru_noun_sg_inst", &["sing", "ablt"]), + ("ru_noun_sg_nom", &["sing", "nomn"]), + ("ru_noun_sg_prep", &["sing", "loct"]), + ("ru_verb_gerund_past", &["past", "V-sh"]), + ("ru_verb_imperative_pl", &["plur", "impr"]), + ("ru_verb_imperative_sg", &["sing", "impr"]), + ("ru_verb_past_f", &["femn", "sing", "past"]), + ("ru_verb_past_m", &["masc", "sing", "past"]), + ("ru_verb_past_n", &["neut", "sing", "past"]), + ("ru_verb_past_pl", &["plur", "past"]), + ("ru_verb_presfut_pl1", &["plur", "1per", "pres"]), + ("ru_verb_presfut_pl2", &["plur", "2per", "pres"]), + ("ru_verb_presfut_pl3", &["plur", "3per", "pres"]), + ("ru_verb_presfut_sg1", &["sing", "1per", "pres"]), + ("ru_verb_presfut_sg2", &["sing", "2per", "pres"]), + ("ru_verb_presfut_sg3", &["sing", "3per", "pres"]), + // Unclear items, probably only useful tags on lemmata + ( + "ru_verb_gerund_present", + &["pres" /* prob. something missing? */], + ), + ( + "ru_adj_superlative", + &[/* TODO: unclear, random list of grammemes?! */], + ), + ("ru_base", &[/* TODO: unclear */]), + // These have no useful tags in the forms table, only gender & + // case tagging. + ("ru_verb_participle_active_past", &[]), + ("ru_verb_participle_active_present", &[]), + ("ru_verb_participle_passive_past", &[]), + ("ru_verb_participle_passive_present", &[]), +];