Add backup functionality

Dockerfile

@@ -10,7 +10,7 @@ COPY src ./src
 RUN python3 -m pip install .
 
 COPY "demo.py" .
-EXPOSE 7860
+EXPOSE 7860 8080
 COPY start.sh .
 RUN chmod +x start.sh
 

README.md

@@ -70,6 +70,10 @@ docker run -p 7860:7860 -e GRADIO_SERVER_NAME=0.0.0.0 -v $HOME/.cache/huggingfac
 
 Note that the Hugging Face cache folder is mounted as a docker volume to make use of potentially available local model cache instead of downloading the transformers models each time the container is started. To prevent that, remove the `-v ...` parameter. The database used for storing the results is ephemeral and will be deleted when the docker container is stopped.
 
+### Backup
+
+To download a backup of the backend DB, visit `localhost:8080/backup`. If you wish to start the app based on a backup, set the `DB_FILE_NAME` environment variable to the name of the backup.
+
 ## Checks
 
 ### Running tests

src/gistillery/db.py

@@ -86,7 +86,7 @@ def namedtuple_factory(cursor, row):  # type: ignore
     return cls._make(row)
 
 
-def _get_db_connection() -> sqlite3.Connection:
+def get_db_connection() -> sqlite3.Connection:
     global TABLES_CREATED
 
     # sqlite cannot deal with concurrent access, so we set a big timeout
@@ -116,7 +116,7 @@ def _get_db_connection() -> sqlite3.Connection:
 
 @contextmanager
 def get_db_cursor() -> Generator[sqlite3.Cursor, None, None]:
-    conn = _get_db_connection()
+    conn = get_db_connection()
     cursor = conn.cursor()
     try:
         yield cursor

src/gistillery/webservice.py

@@ -1,11 +1,14 @@
+import datetime as dt
 import logging
+import shutil
+import sqlite3
 import uuid
 
 from fastapi import FastAPI
+from fastapi.responses import FileResponse
 
 from gistillery.base import EntriesResult, JobStatus, JobStatusResult, RequestInput
-from gistillery.db import TABLES, get_db_cursor
-
+from gistillery.db import TABLES, get_db_connection, get_db_cursor
 
 logger = logging.getLogger(__name__)
 logger.setLevel(logging.DEBUG)
@@ -140,3 +143,26 @@ def clear() -> str:
         for table_name in TABLES:
             cursor.execute(f"DELETE FROM {table_name}")
     return "OK"
+
+
+@app.get("/backup/")
+def backup() -> FileResponse:
+    # create a backup and return it
+    def progress(status, remaining, total):
+        logger.debug(f"DB: Copied {total-remaining} of {total} pages...")
+
+    now = dt.datetime.now(dt.timezone.utc)
+    fname = f"sqlite-data_backup_{now.strftime('%Y-%m-%d_%H-%M-%S')}.db"
+
+    try:
+        conn = get_db_connection()
+        backup_db = sqlite3.connect(fname)
+        with backup_db:
+            conn.backup(backup_db, pages=1, progress=progress)
+    except Exception as e:
+        logger.error(f"Error creating backup: {e}")
+        conn.close()
+        backup_db.close()
+        raise e
+
+    return FileResponse(fname, media_type="application/octet-stream", filename=fname)

tests/test_app.py

@@ -1,5 +1,6 @@
 import datetime as dt
 import os
+import sqlite3
 from types import SimpleNamespace
 
 import pytest
@@ -311,3 +312,22 @@ class TestWebservice:
             ]
         )
         assert rows[0].input == expected
+
+    def test_backup(self, client, tmp_path):
+        # submit an entry, create a backup, check that the backup contains the entry
+        from gistillery.db import namedtuple_factory
+
+        client.post("/submit", json={"author": "Pie Test", "content": "this is a pie"})
+        resp = client.get("/backup")
+        assert resp.status_code == 200
+
+        with open(tmp_path / "backup.db", "wb") as f:
+            f.write(resp.content)
+            conn = sqlite3.connect(tmp_path / "backup.db")
+
+        conn.row_factory = namedtuple_factory
+        cursor = conn.cursor()
+        res = cursor.execute("select * from entries").fetchall()
+        assert len(res) == 1
+        assert is_roughly_now(res[0].created_at)
+        assert res[0].author == "Pie Test"