Create lexmatch using oaklib

.gitignore

@@ -64,4 +64,7 @@ src/scripts/pattern-matches/matches_old/
 src/scripts/pattern-matches/generic_matches/
 src/scripts/pattern-matches/upheno_matches/
 src/scripts/pattern-matches/upheno_patterns/
-src/scripts/pattern-matches/ontologies/
+src/scripts/pattern-matches/ontologies/
+src/ontology/.template.db
+src/ontology/tmp/upheno_all_with_relations.*
+src/ontology/tmp/upheno.db.lexical.yaml

src/ontology/config/upheno-match-rules.yaml

@@ -0,0 +1,75 @@
+rules:
+  - description: default
+    postconditions:
+      predicate_id: skos:closeMatch
+      weight: 0.0
+
+  - description: exact to exact
+    preconditions:
+      subject_match_field_one_of:
+        - oio:hasExactSynonym
+        - rdfs:label
+        - skos:prefLabel
+      object_match_field_one_of:
+        - oio:hasExactSynonym
+        - rdfs:label
+        - skos:prefLabel
+    postconditions:
+      predicate_id: skos:exactMatch
+      weight: 2.0
+
+  - description: >-
+     label to label; note this is additive with the exact to exact rule,
+      so the score just represents an additional small boost
+    preconditions:
+      subject_match_field_one_of:
+        - rdfs:label
+      object_match_field_one_of:
+        - rdfs:label
+    postconditions:
+      predicate_id: skos:exactMatch
+      weight: 0.5
+
+  - description: xref match
+    preconditions:
+      subject_match_field_one_of:
+        - oio:hasDbXref
+        - skos:exactMatch
+      object_match_field_one_of:
+        - oio:hasDbXref
+        - skos:exactMatch
+    postconditions:
+      predicate_id: skos:exactMatch
+      weight: 4.0
+
+  - preconditions:
+      subject_match_field_one_of:
+        - oio:hasExactSynonym
+        - rdfs:label
+      object_match_field_one_of:
+        - oio:hasBroadSynonym
+    postconditions:
+      predicate_id: skos:broadMatch
+      weight: 2.0
+
+  - preconditions:
+      subject_match_field_one_of:
+        - oio:hasExactSynonym
+        - rdfs:label
+      object_match_field_one_of:
+        - oio:hasNarrowSynonym
+    postconditions:
+      predicate_id: skos:narrowMatch
+      weight: 2.0
+
+  - synonymizer:
+      the_rule: Remove parentheses bound info from the label.
+      match: r'\([^)]*\)'
+      match_scope: "*"
+      replacement: ""
+
+  - synonymizer:
+      the_rule: Remove box brackets bound info from the label.
+      match: r'\[[^)]*\]'
+      match_scope: "*"
+      replacement: ""

src/ontology/metadata/upheno.sssom.config.yml

@@ -0,0 +1,40 @@
+curie_map:
+  MP: http://purl.obolibrary.org/obo/MP_
+  HP: http://purl.obolibrary.org/obo/HP_
+  WBPhenotype: http://purl.obolibrary.org/obo/WBPhenotype_
+  XPO: http://purl.obolibrary.org/obo/XPO_
+  PLANP: http://purl.obolibrary.org/obo/PLANP_
+  ZP: http://purl.obolibrary.org/obo/ZP_
+  DPO: http://purl.obolibrary.org/obo/FBcv_
+  FYPO: http://purl.obolibrary.org/obo/FYPO_
+  DDPHENO: http://purl.obolibrary.org/obo/DDPHENO_
+  phipo: http://purl.obolibrary.org/obo/PHIPO_
+  MGPO: http://purl.obolibrary.org/obo/MGPO_
+  APO: http://purl.obolibrary.org/obo/APO_
+subject_prefixes:
+  - MP
+  - HP
+  - WBPhenotype
+  - XPO
+  - PLANP
+  - ZP
+  - DPO
+  - FYPO
+  - DDPHENO
+  - phipo
+  - MGPO
+  - APO
+
+relations:
+  - oboInOwl:hasDbXref
+  - skos:exactMatch
+  - skos:broadMatch
+  - skos:narrowMatch
+  - skos:closeMatch
+  - skos:relatedMatch
+
+global_metadata:
+  subject_source: http://purl.obolibrary.org/obo/upheno_all_with_relations.owl
+  license: "https://creativecommons.org/publicdomain/zero/1.0/"
+  mapping_date: "2017-09-20"
+  mapping_set_id: "GLOBAL" # Apparently this is the one that matters.

src/ontology/upheno.Makefile

@@ -224,3 +224,18 @@ reports/phenotype_trait.sssom.tsv: $(UPHENO_RELEASE_FILE_ANALYSIS)
 	sed -i 's/[?]//g' $@
 	sed -i 's/<http:[/][/]purl[.]obolibrary[.]org[/]obo[/]/obo:/g' $@
 	sed -i 's/>//g' $@
+
+#############################
+#### Lexical matching #######
+#############################
+SCRIPTSDIR=../scripts
+tmp/upheno_all_with_relations.db: tmp/upheno_all_with_relations.owl
+	semsql make $@
+
+../../mappings/upheno-lexical.sssom.tsv: $(SCRIPTSDIR)/upheno-lexmatch.py
+	python $^ run tmp/upheno_all_with_relations.db -c metadata/upheno.sssom.config.yml -r config/upheno-match-rules.yaml --no-recreate -o $@
+
+## ../../mappings/upheno-lexical.sssom.tsv: tmp/upheno_all_with_relations.db
+##	runoak -i sqlite:$< lexmatch -R config/upheno-match-rules.yaml --no-recreate -o $@
+
+lexical_matches: ../../mappings/upheno-lexical.sssom.tsv

src/scripts/upheno-lexmatch.py

@@ -0,0 +1,104 @@
+import logging
+from pathlib import Path
+from oaklib.resource import OntologyResource
+from oaklib.implementations.sqldb.sql_implementation import SqlImplementation
+from oaklib.utilities.lexical.lexical_indexer import (
+    create_lexical_index,
+    lexical_index_to_sssom,
+    load_mapping_rules,
+    save_lexical_index,
+    load_lexical_index
+)
+import sys
+import click
+import yaml
+
+from sssom.constants import SUBJECT_ID, OBJECT_ID
+from sssom.util import filter_prefixes
+from sssom.parsers import parse_sssom_table
+from sssom.writers import write_table
+from sssom.io import get_metadata_and_prefix_map, filter_file
+
+SRC = Path(__file__).resolve().parents[1]
+ONTOLOGY_DIR = SRC / "ontology"
+OUT_INDEX_DB = ONTOLOGY_DIR / "tmp/upheno.db.lexical.yaml"
+TEMP_DIR = ONTOLOGY_DIR / "tmp"
+
+input_argument = click.argument("input", required=True, type=click.Path())
+output_option = click.option(
+    "-o",
+    "--output",
+    help="Path for output file.",
+    default=sys.stdout,
+)
+
+
+@click.group()
+@click.option("-v", "--verbose", count=True)
+@click.option("-q", "--quiet")
+def main(verbose: int, quiet: bool):
+    """Run the SSSOM CLI."""
+    logger = logging.getLogger()
+    if verbose >= 2:
+        logger.setLevel(level=logging.DEBUG)
+    elif verbose == 1:
+        logger.setLevel(level=logging.INFO)
+    else:
+        logger.setLevel(level=logging.WARNING)
+    if quiet:
+        logger.setLevel(level=logging.ERROR)
+
+
+@main.command()
+@input_argument
+@click.option(
+    "-c",
+    "--config",
+    help="YAML file containing metadata.",
+)
+@click.option(
+    "-r",
+    "--rules",
+    help="Ruleset for mapping.",
+)
+@click.option(
+    "--recreate/--no-recreate",
+    default=True,
+    show_default=True,
+    help="if true and lexical index is specified, always recreate, otherwise load from index",
+)
+@output_option
+def run(input: str, config: str, rules: str, recreate:bool, output: str):
+    # Implemented `meta` param in `lexical_index_to_sssom`
+
+    meta = get_metadata_and_prefix_map(config)
+    with open(config, "r") as f:
+        yml = yaml.safe_load(f)
+
+    prefix_of_interest = yml["subject_prefixes"]
+
+    resource = OntologyResource(slug=f"sqlite:///{Path(input).absolute()}")
+    oi = SqlImplementation(resource=resource)
+    ruleset = load_mapping_rules(rules)
+    syn_rules = [x.synonymizer for x in ruleset.rules if x.synonymizer]
+    if not recreate and Path(OUT_INDEX_DB).exists():
+        lexical_index = load_lexical_index(OUT_INDEX_DB)
+    else:
+        lexical_index = create_lexical_index(oi=oi, synonym_rules=syn_rules)
+        save_lexical_index(lexical_index, OUT_INDEX_DB)
+
+    if rules:
+        msdf = lexical_index_to_sssom(
+            oi, lexical_index, ruleset=load_mapping_rules(rules), meta=meta
+        )
+    else:
+        msdf = lexical_index_to_sssom(oi, lexical_index, meta=meta)
+
+    msdf.df = filter_prefixes(
+        df=msdf.df, filter_prefixes=prefix_of_interest, features=[SUBJECT_ID, OBJECT_ID]
+    )
+    with open(str(Path(output)), "w", encoding="utf8") as f:
+        write_table(msdf, f)
+
+if __name__ == "__main__":
+    main()