Test the migrations on database content

tests/test_migrations.py

@@ -11,7 +11,8 @@ import alembic.autogenerate
 import alembic.command
 import alembic.config
 from alembic.migration import MigrationContext
-from sqlalchemy import create_engine
+import sqlalchemy as sa
+from sqlalchemy import create_engine, sql
 
 from mutalyzer import db
 
@@ -36,13 +37,13 @@ def test_migrations():
         if database_uri != 'sqlite://':
             db.Base.metadata.drop_all(connection)
 
-        # TODO: We might also want to test this with data. An implementation
-        #   might look like this:
-        #
-        #   1. Create initial schema by running the first migration.
-        #   2. Add some database content.
-        #   3. Run the remaining migrations.
+        # Create initial schema by running the first migration.
+        alembic.command.upgrade(alembic_config, 'ea660b66f26')
 
+        # Add some database content to run the migrations on.
+        add_database_content(connection)
+
+        # Run the remaining migrations.
         alembic.command.upgrade(alembic_config, 'head')
 
         context = MigrationContext.configure(connection)
@@ -50,3 +51,134 @@ def test_migrations():
             context, db.Base.metadata)
 
     engine.dispose()
+
+
+def add_database_content(connection):
+    """
+    Add some content to the database.
+    """
+    # We only define tables and columns we actually need, so this is not a
+    # complete mapping of the schema.
+
+    assemblies = sql.table(
+        'assemblies',
+        sql.column('id', sa.Integer),
+        sql.column('name', sa.String(30)),
+        sql.column('alias', sa.String(10)),
+        sql.column('taxonomy_id', sa.Integer),
+        sql.column('taxonomy_common_name', sa.String(50)))
+
+    chromosomes = sql.table(
+        'chromosomes',
+        sql.column('id', sa.Integer),
+        sql.column('assembly_id', sa.Integer),
+        sql.column('name', sa.String(30)),
+        sql.column('accession', sa.String(30)),
+        sql.column('organelle', sa.Enum('nucleus', 'mitochondrion',
+                                        name='organelle')))
+
+    transcript_mappings = sql.table(
+        'transcript_mappings',
+        sql.column('chromosome_id', sa.Integer),
+        sql.column('reference_type', sa.Enum('refseq', 'lrg',
+                                             name='reference_type')),
+        sql.column('accession', sa.String(20)),
+        sql.column('gene', sa.String(30)),
+        sql.column('transcript', sa.Integer),
+        sql.column('orientation', sa.Enum('forward', 'reverse',
+                                          name='orentation')),
+        sql.column('start', sa.Integer),
+        sql.column('stop', sa.Integer),
+        sql.column('exon_starts', sa.Text),
+        sql.column('exon_stops', sa.Text),
+        sql.column('select_transcript', sa.Boolean),
+        sql.column('source', sa.Enum('ucsc', 'ncbi', 'reference',
+                                     name='source')))
+
+    transcript_protein_links = sql.table(
+        'transcript_protein_links',
+        sql.column('transcript_accession', sa.String(30)),
+        sql.column('protein_accession', sa.String(30)))
+
+    # Add some common data.
+    connection.execute(
+        assemblies.insert(),
+        name='GRCh37',
+        taxonomy_id=9606,
+        taxonomy_common_name='Homo sapiens',
+        alias='hg19')
+    hg19_id = connection.execute(
+        assemblies.select(assemblies.c.alias == 'hg19')
+        .with_only_columns([assemblies.c.id])
+    ).fetchone()[0]
+
+    connection.execute(
+        chromosomes.insert(),
+        assembly_id=hg19_id,
+        name='chr1',
+        accession='NC_000001.10',
+        organelle='nucleus')
+    chr1_id = connection.execute(
+        chromosomes.select(chromosomes.c.name == 'chr1')
+        .with_only_columns([chromosomes.c.id])
+    ).fetchone()[0]
+
+    # Data for migration 402ff01b0d5d:
+    # Fix GRCm38 chromosome accession number versions.
+    connection.execute(
+        chromosomes.insert(),
+        assembly_id=hg19_id,
+        name='chr11',
+        accession='NC_000077.60',
+        organelle='nucleus')
+
+    # Data for migration 2e062969eb54:
+    # Rename GRCh36 assembly to NCBI36.
+    connection.execute(
+        assemblies.insert(),
+        name='GRCh36',
+        taxonomy_id=9606,
+        taxonomy_common_name='Homo sapiens',
+        alias='hg18')
+
+    # Data for migration 4bafcc5086dd:
+    # Fix zero-exon transcript mappings.
+    connection.execute(
+        transcript_mappings.insert(),
+        chromosome_id=chr1_id,
+        reference_type='refseq',
+        accession='NC_001807',
+        gene='ATP6',
+        transcript=1,
+        orientation='forward',
+        start=8528,
+        stop=9208,
+        exon_starts='8528',
+        exon_stops='9208',
+        select_transcript=True,
+        source='ncbi')
+    connection.execute(
+        transcript_mappings.insert(),
+        chromosome_id=chr1_id,
+        reference_type='refseq',
+        accession='NC_001807',
+        gene='ATP8',
+        transcript=1,
+        orientation='forward',
+        start=8367,
+        stop=8573,
+        exon_starts='',
+        exon_stops='',
+        select_transcript=True,
+        source='ncbi')
+
+    # Data for migration 3492d2ee8884:
+    # Transcript protein links have nullable transcript and unique protein.
+    connection.execute(
+        transcript_protein_links.insert(),
+        transcript_accession='NM_052818',
+        protein_accession='NP_438169')
+    connection.execute(
+        transcript_protein_links.insert(),
+        transcript_accession='NM_001079691',
+        protein_accession=None)