Merge pull request #6 from INCATools/add-bioregistry

Convert IRI -> CURIE using `bioregistry` while `parse`-ing a KGCL command.

.gitignore

@@ -5,3 +5,4 @@ __pycache__
 .pytest_cache
 .idea
 tmp/
+tests/outputs/*

pyproject.toml

@@ -14,6 +14,7 @@ readme = "README.md"
 python = "^3.8"
 linkml-runtime = "^1.1.24"
 lark = "^1.1.2"
+bioregistry = "^0.5.49"
 
 [tool.poetry.dev-dependencies]
 linkml = "^1.2.15"

src/kgcl_schema/grammar/parser.py

@@ -1,20 +1,24 @@
 """KGCL parser."""
 import logging
+import re
 import sys
 from pathlib import Path
 from typing import List
 
 import click
-from kgcl_schema.utils import to_json, to_rdf, to_yaml
-from lark import Lark, Token
-
-from kgcl_schema.datamodel.kgcl import (ClassCreation, EdgeCreation, EdgeDeletion,
-                                        NewSynonym, NodeAnnotationChange, NodeCreation,
+from bioregistry import parse_iri, get_preferred_prefix, curie_to_str
+from kgcl_schema.datamodel.kgcl import (Change, ClassCreation, EdgeCreation,
+                                        EdgeDeletion, NewSynonym,
+                                        NodeAnnotationChange, NodeCreation,
                                         NodeDeepening, NodeDeletion, NodeMove,
-                                        NodeObsoletion, NodeRename, NodeShallowing,
-                                        NodeUnobsoletion, PlaceUnder, PredicateChange,
-                                        RemovedNodeFromSubset, RemoveUnder, Change, Session)
+                                        NodeObsoletion, NodeRename,
+                                        NodeShallowing, NodeUnobsoletion,
+                                        PlaceUnder, PredicateChange,
+                                        RemovedNodeFromSubset, RemoveUnder,
+                                        Session)
 from kgcl_schema.datamodel.ontology_model import Edge
+from kgcl_schema.utils import to_json, to_rdf, to_yaml
+from lark import Lark, Token
 
 
 def id_generator():
@@ -54,6 +58,16 @@ def parse_statement(input: str) -> Change:
 
     Return an instantiated dataclass object from model.kgcl_schema.
     """
+    regex = r'http[s]?://(?:[a-zA-Z]|[0-9]|[$-_@.&+]|[!*\(\),]|(?:%[0-9a-fA-F][0-9a-fA-F]))+'
+    uri_list = re.findall(regex, input)
+    if uri_list:
+        # curie = curie_from_iri(uri[0].replace("<", "").replace(">",""))
+        for _, uri in enumerate(uri_list):
+            pref, i = parse_iri(uri)
+            pref = get_preferred_prefix(pref)
+            curie = curie_to_str(pref, i)
+            input = input.replace(uri, curie)
+
     tree = kgcl_parser.parse(input)
     id = "kgcl_change_id_" + str(next(id_gen))
 

src/kgcl_schema/grammar/render_operations.py

@@ -20,13 +20,13 @@ def render_entity(entity, rdf_type):
     if rdf_type is None:
         return entity
     elif rdf_type == "uri":
-        return "<" + entity + ">"
+        return entity
     elif rdf_type == "label":
         if "'" in entity:
             # TODO: replacing quotes with backticks
             # is only a temporary workaround
             entity = entity.replace("'", "`")
-        return "'" + entity + "'"
+        return entity
     elif rdf_type == "literal":
         # TODO: test this
         if '"' not in entity:

tests/cases.py

@@ -6,21 +6,21 @@
 
 TODO_TOKEN = "TODO"
 PART_OF = "BFO:0000050"
-PART_OF_URI = "<http://purl.obolibrary.org/obo/BFO_0000050>"
+PART_OF_URI = "http://purl.obolibrary.org/obo/BFO_0000050"
 IS_A = "rdfs:subClassOf"
-IS_A_URI = "<http://www.w3.org/2000/01/rdf-schema#subClassOf>"
+IS_A_URI = "http://www.w3.org/2000/01/rdf-schema#subClassOf"
 NUCLEUS = "GO:0005634"
-NUCLEUS_URI = "<http://purl.obolibrary.org/obo/GO_0005634>"
+NUCLEUS_URI = "http://purl.obolibrary.org/obo/GO_0005634"
 MITOCHONDRION = "GO:0005739"
-MITOCHONDRION_URI = "<http://purl.obolibrary.org/obo/GO_0005739>"
+MITOCHONDRION_URI = "http://purl.obolibrary.org/obo/GO_0005739"
 IMBO = "GO:0043231"
-IMBO_URI = "<http://purl.obolibrary.org/obo/GO_0043231>"
+IMBO_URI = "http://purl.obolibrary.org/obo/GO_0043231"
 NUCLEAR_ENVELOPE = "GO:0005635"
-NUCLEAR_ENVELOPE_URI = "<http://purl.obolibrary.org/obo/GO_0005635>"
+NUCLEAR_ENVELOPE_URI = "http://purl.obolibrary.org/obo/GO_0005635"
 NEW_TERM = "GO:9999999"
-NEW_TERM_URI = "<http://purl.obolibrary.org/obo/GO_9999999>"
+NEW_TERM_URI = "http://purl.obolibrary.org/obo/GO_9999999"
 RESPONSE_TO_UV = "GO:0009411"
-RESPONSE_TO_UV_URI = "<http://purl.obolibrary.org/obo/GO_0009411>"
+RESPONSE_TO_UV_URI = "http://purl.obolibrary.org/obo/GO_0009411"
 
 UID = "CHANGE:001"
 TERM = "GO:123"
@@ -124,10 +124,10 @@
         #f"create node {NEW_TERM_URI} 'foo'",
         TODO_TOKEN,
         NodeCreation(id=UID,
-                     node_id=NEW_TERM_URI,   ## TODO: remove this
-                     about_node=NEW_TERM_URI,
+                     node_id=NEW_TERM,   ## TODO: remove this
+                     about_node=NEW_TERM,
                      name="'foo'",
-                     about_node_representation='uri'),
+                     about_node_representation='curie'),
         None
     ),
     (