Merge pull request #77 from linkml/enrich

Schema enricher and Docker config

Author: cmungall

Dockerfile

@@ -0,0 +1,29 @@
+# set base image (host OS)
+FROM python:3.9
+
+# https://stackoverflow.com/questions/53835198/integrating-python-poetry-with-docker
+ENV YOUR_ENV=${YOUR_ENV} \
+  PYTHONFAULTHANDLER=1 \
+  PYTHONUNBUFFERED=1 \
+  PYTHONHASHSEED=random \
+  PIP_NO_CACHE_DIR=off \
+  PIP_DISABLE_PIP_VERSION_CHECK=on \
+  PIP_DEFAULT_TIMEOUT=100 \
+  POETRY_VERSION=1.1.13
+
+# System deps:
+RUN pip install "poetry==$POETRY_VERSION"
+
+# set the working directory in the container
+WORKDIR /work
+
+RUN pip install schema-automator
+
+#COPY poetry.lock pyproject.toml /code/
+
+# Project initialization:
+#RUN poetry install
+
+
+# command to run on container start
+CMD [ "bash" ]

Makefile

@@ -1,7 +1,10 @@
+VERSION = $(shell git tag | tail -1)
+
 .PHONY: all clean test
 
 all: clean test target/soil_meanings.yaml
 
+
 clean:
 	rm -rf target/soil_meanings.yaml
 	rm -rf target/soil_meanings_generated.yaml
@@ -63,3 +66,41 @@ target/availabilities_g_s_strain_202112151116_org_meanings_curated.yaml: target/
 # this can be used outside the poetry environment
 bin/schemauto:
 	echo `poetry run which schemauto` '"$$@"' > $@ && chmod +x $@
+
+
+################################################
+#### Commands for building the Docker image ####
+################################################
+
+IM=linkml/schema-automator
+
+docker-build-no-cache:
+	@docker build --no-cache -t $(IM):$(VERSION) . \
+	&& docker tag $(IM):$(VERSION) $(IM):latest
+
+docker-build:
+	@docker build -t $(IM):$(VERSION) . \
+	&& docker tag $(IM):$(VERSION) $(IM):latest
+
+docker-build-use-cache-dev:
+	@docker build -t $(DEV):$(VERSION) . \
+	&& docker tag $(DEV):$(VERSION) $(DEV):latest
+
+docker-clean:
+	docker kill $(IM) || echo not running ;
+	docker rm $(IM) || echo not made 
+
+docker-publish-no-build:
+	@docker push $(IM):$(VERSION) \
+	&& docker push $(IM):latest
+
+docker-publish-dev-no-build:
+	@docker push $(DEV):$(VERSION) \
+	&& docker push $(DEV):latest
+
+docker-publish: docker-build
+	@docker push $(IM):$(VERSION) \
+	&& docker push $(IM):latest
+
+docker-run:
+	@docker run  -v $(PWD):/work -w /work -ti $(IM):$(VERSION) 

docs/conf.py

@@ -20,8 +20,8 @@
 # -- Project information -----------------------------------------------------
 
 project = 'Schema Automator'
-copyright = '2022, Chris Mungall'
-author = 'Chris Mungall, Harshad Hegde'
+copyright = '2022, LinkML Developers'
+author = 'Chris Mungall, Harshad Hegde, Mark Miller'
 
 # The full version, including alpha/beta/rc tags
 # release = '0.1.4'

docs/index.rst

@@ -1,12 +1,29 @@
 LinkML Schema Automator
 ============================================
 
-Schema Automator is a toolkit for bootstrapping and automatically enhancing LinkML schemas from a variety of sources
+Schema Automator is a toolkit for bootstrapping and automatically enhancing schemas from a variety of sources.
+
+Use cases include:
+
+1. Inferring an initial schema or data dictionary from a dataset that is a collection of TSVs
+2. Automatically annotating schema elements and enumerations using the BioPortal annotator
+3. Importing from a language like RDFS/OWL
+
+The primary output of Schema Automator is a `LinkML Schema <https://linkml.io/linkml>`_. This can be converted to other
+schema frameworks, including:
+
+* JSON-Schema
+* SQL DDL
+* SHACL
+* ShEx
+* RDFS/OWL
+* Python dataclasses or Pydantic
 
 .. toctree::
    :maxdepth: 3
    :caption: Contents:
 
+   index
    introduction
    install
    cli

docs/install.rst

@@ -1,4 +1,7 @@
 Installation
+======
+
+Direct Installation
 ------------
 
 ``schema-automator`` and its components require Python 3.9 or greater.
@@ -13,3 +16,19 @@ To check this works:
 
    schemauto --help
 
+Running via Docker
+------------
+
+You can use the `Schema Automator Docker Container <https://hub.docker.com/r/linkml/schema-automator>`_
+
+To start a shell
+
+.. code:: bash
+
+    docker run  -v $PWD:/work -w /work -ti linkml/schema-automator
+
+Within the shell you should see all your files, and you should have access:
+
+.. code:: bash
+
+   schemauto --help

docs/introduction.rst

@@ -1,4 +1,6 @@
-LinkML Schema Automator
+.. _introduction:
+
+Introduction
 =======================
 
 This is a toolkit that assists with generating and enhancing schemas and data models from a variety
@@ -17,17 +19,17 @@ See :ref:`generalizers`
 
 Generalizers allow you to *bootstrap* a schema by generalizing from existing data files
 
--  TSVs and spreadsheets
--  SQLite databases
--  RDF instance graphs
+* TSVs and spreadsheets
+* SQLite databases
+* RDF instance graphs
 
-Importing from alternative modeling framework
+Importing from alternative modeling frameworks
 ---------------------------------
 
 See :ref:`importers`
 
-- OWL (but this only works for schema-style OWL)
-- JSON-Schema
+* OWL (but this only works for schema-style OWL)
+* JSON-Schema
 
 In future other frameworks will be supported
 

schema_automator/annotators/schema_annotator.py

@@ -90,6 +90,25 @@ def annotate_schema(self, schema: Union[SchemaDefinition, str], curie_only=True)
 
         return sv.schema
 
+    def enrich(self, schema: Union[SchemaDefinition, str]) -> SchemaDefinition:
+        sv = SchemaView(schema)
+        oi = self.ontology_implementation
+        for elt_name, elt in sv.all_elements().items():
+            curies = [sv.get_uri(elt)]
+            for rel, ms in sv.get_mappings().items():
+                curies += ms
+            for x in curies:
+                print(f"X={x}")
+                if elt.description:
+                    break
+                try:
+                    defn = oi.get_definition_by_curie(x)
+                    if defn:
+                        elt.description = defn
+                except Exception:
+                    pass
+        return sv.schema
+
 
 @click.command()
 @click.argument('schema')

schema_automator/cli.py

@@ -245,6 +245,23 @@ def annotate_schema(schema: str, input: str, output: str, curie_only: bool, **ar
     write_schema(schema, output)
 
 
+@main.command()
+@click.argument('schema')
+@click.option('--input', '-i', help="OAK input ontology selector")
+@output_option
+def enrich_schema(schema: str, input: str, output: str, **args):
+    """
+    Annotate all elements of a schema
+
+    Requires Bioportal API key
+    """
+    impl = get_implementation_from_shorthand(input)
+    logging.basicConfig(level=logging.INFO)
+    annr = SchemaAnnotator(impl)
+    schema = annr.enrich(schema)
+    write_schema(schema, output)
+
+
 @main.command()
 @click.argument('schema')
 @output_option

tests/resources/so-mini.obo

@@ -0,0 +1,4 @@
+[Term]
+id: SO:0000704
+name: gene
+def: "A region (or regions) that includes all of the sequence elements necessary to encode a functional transcript. A gene may include regulatory regions, transcribed regions and/or other functional sequence regions." []

tests/test_annotators/test_schema_enricher.py

@@ -0,0 +1,32 @@
+# -*- coding: utf-8 -*-
+
+import logging
+import os
+import unittest
+from linkml.utils.schema_builder import SchemaBuilder
+from linkml_runtime.dumpers import yaml_dumper
+from linkml_runtime.linkml_model import SchemaDefinition, EnumDefinition, PermissibleValue
+from oaklib.implementations import BioportalImplementation
+from oaklib.selector import get_implementation_from_shorthand
+
+from schema_automator.annotators.schema_annotator import SchemaAnnotator
+from linkml.generators.yamlgen import YAMLGenerator
+from tests import INPUT_DIR, OUTPUT_DIR
+
+
+class SchemaEnricherTestCase(unittest.TestCase):
+
+    def setUp(self) -> None:
+        impl = get_implementation_from_shorthand(os.path.join(INPUT_DIR, "so-mini.obo"))
+        self.annotator = SchemaAnnotator(impl)
+
+    def test_enrich(self):
+        s = SchemaDefinition(id='test', name='test')
+        sb = SchemaBuilder(s)
+        sb.add_class('Gene', class_uri="SO:0000704").add_slot('part_of')
+        s = self.annotator.enrich(sb.schema)
+        #print(yaml_dumper.dumps(s))
+        assert s.classes['Gene'].description.startswith("A region")
+
+
+