New run_sanity_checks mechanism, for SpatiLite

Moved VirtualSpatialIndex check into a new mechanism that should allow
us to add further sanity checks in the future.

To test this I've had to commit a binary sample SpatiaLite database to
the repository. I included a build script for creating that database.

Closes #466

.gitignore

@@ -9,9 +9,8 @@ scratchpad
 Pipfile
 Pipfile.lock
 
-# SQLite databases
-*.db
-*.sqlite
+fixtures.db
+*test.db
 
 # Byte-compiled / optimized / DLL files
 __pycache__/

datasette/app.py

@@ -344,6 +344,28 @@ class Datasette:
                 except ValueError:
                     # Plugin already registered
                     pass
+        # Run the sanity checks
+        asyncio.get_event_loop().run_until_complete(self.run_sanity_checks())
+
+    async def run_sanity_checks(self):
+        # Only one check right now, for Spatialite
+        for database_name, database in self.databases.items():
+            # Run pragma_info on every table
+            for table in await database.table_names():
+                try:
+                    await self.execute(
+                        database_name,
+                        "PRAGMA table_info({});".format(escape_sqlite(table)),
+                    )
+                except sqlite3.OperationalError as e:
+                    if e.args[0] == "no such module: VirtualSpatialIndex":
+                        raise click.UsageError(
+                            "It looks like you're trying to load a SpatiaLite"
+                            " database without first loading the SpatiaLite module."
+                            "\n\nRead more: https://datasette.readthedocs.io/en/latest/spatialite.html"
+                        )
+                    else:
+                        raise
 
     def config(self, key):
         return self._config.get(key, None)
@@ -530,29 +552,17 @@ class Datasette:
                 name = path.stem
                 if name in self._inspect:
                     raise Exception("Multiple files with same stem %s" % name)
-                try:
-                    with sqlite3.connect(
-                        "file:{}?mode=ro".format(path), uri=True
-                    ) as conn:
-                        self.prepare_connection(conn)
-                        self._inspect[name] = {
-                            "hash": inspect_hash(path),
-                            "file": str(path),
-                            "size": path.stat().st_size,
-                            "views": inspect_views(conn),
-                            "tables": inspect_tables(
-                                conn, (self.metadata("databases") or {}).get(name, {})
-                            ),
-                        }
-                except sqlite3.OperationalError as e:
-                    if e.args[0] == "no such module: VirtualSpatialIndex":
-                        raise click.UsageError(
-                            "It looks like you're trying to load a SpatiaLite"
-                            " database without first loading the SpatiaLite module."
-                            "\n\nRead more: https://datasette.readthedocs.io/en/latest/spatialite.html"
-                        )
-                    else:
-                        raise
+                with sqlite3.connect("file:{}?mode=ro".format(path), uri=True) as conn:
+                    self.prepare_connection(conn)
+                    self._inspect[name] = {
+                        "hash": inspect_hash(path),
+                        "file": str(path),
+                        "size": path.stat().st_size,
+                        "views": inspect_views(conn),
+                        "tables": inspect_tables(
+                            conn, (self.metadata("databases") or {}).get(name, {})
+                        ),
+                    }
         return self._inspect
 
     def register_custom_units(self):

tests/build_small_spatialite_db.py

@@ -0,0 +1,23 @@
+import sqlite3
+
+# This script generates the spatialite.db file in our tests directory.
+
+
+def generate_it(filename):
+    conn = sqlite3.connect(filename)
+    # Lead the spatialite extension:
+    conn.enable_load_extension(True)
+    conn.load_extension("/usr/local/lib/mod_spatialite.dylib")
+    conn.execute("select InitSpatialMetadata(1)")
+    conn.executescript("create table museums (name text)")
+    conn.execute("SELECT AddGeometryColumn('museums', 'point_geom', 4326, 'POINT', 2);")
+    # At this point it is around 5MB - we can shrink it dramatically by doing thisO
+    conn.execute("delete from spatial_ref_sys")
+    conn.execute("delete from spatial_ref_sys_aux")
+    conn.commit()
+    conn.execute("vacuum")
+    conn.close()
+
+
+if __name__ == "__main__":
+    generate_it("spatialite.db")

tests/test_cli.py

@@ -1,6 +1,7 @@
 from .fixtures import app_client
 from datasette.cli import cli
 from click.testing import CliRunner
+import pathlib
 import json
 
 
@@ -28,3 +29,12 @@ def test_inspect_cli_writes_to_file(app_client):
     assert 0 == result.exit_code, result.output
     data = json.load(open("foo.json"))
     assert ["fixtures"] == list(data.keys())
+
+
+def test_spatialite_error_if_attempt_to_open_spatialite():
+    runner = CliRunner()
+    result = runner.invoke(
+        cli, ["serve", str(pathlib.Path(__file__).parent / "spatialite.db")]
+    )
+    assert result.exit_code != 0
+    assert "trying to load a SpatiaLite database" in result.output