Converts the conversion into a database migration

src/documents/migrations/1021_webp_thumbnail_conversion.py

@@ -0,0 +1,107 @@
+# Generated by Django 4.0.5 on 2022-06-11 15:40
+import logging
+import multiprocessing.pool
+import shutil
+import tempfile
+import time
+from pathlib import Path
+
+from django.conf import settings
+from django.db import migrations
+from documents.parsers import run_convert
+
+logger = logging.getLogger("paperless.migrations")
+
+
+def _do_convert(work_package):
+    existing_thumbnail, converted_thumbnail = work_package
+    try:
+
+        logger.info(f"Converting thumbnail: {existing_thumbnail}")
+
+        # Run actual conversion
+        run_convert(
+            density=300,
+            scale="500x5000>",
+            alpha="remove",
+            strip=True,
+            trim=False,
+            auto_orient=True,
+            input_file=f"{existing_thumbnail}[0]",
+            output_file=str(converted_thumbnail),
+        )
+
+        # Copy newly created thumbnail to thumbnail directory
+        shutil.copy(converted_thumbnail, existing_thumbnail.parent)
+
+        # Remove the PNG version
+        existing_thumbnail.unlink()
+
+        logger.info(
+            "Conversion to WebP completed, "
+            f"replaced {existing_thumbnail.name} with {converted_thumbnail.name}",
+        )
+
+    except Exception as e:
+        logger.error(f"Error converting thumbnail (existing file unchanged): {e}")
+
+
+def _convert_thumbnails_to_webp(apps, schema_editor):
+    start = time.time()
+
+    with tempfile.TemporaryDirectory() as tempdir:
+
+        work_packages = []
+
+        for file in Path(settings.THUMBNAIL_DIR).glob("*.png"):
+            existing_thumbnail = file.resolve()
+
+            # Change the existing filename suffix from png to webp
+            converted_thumbnail_name = existing_thumbnail.with_suffix(
+                ".webp",
+            ).name
+
+            # Create the expected output filename in the tempdir
+            converted_thumbnail = (
+                Path(tempdir) / Path(converted_thumbnail_name)
+            ).resolve()
+
+            # Package up the necessary info
+            work_packages.append(
+                (existing_thumbnail, converted_thumbnail),
+            )
+
+        if len(work_packages):
+
+            logger.info(
+                "\n\n"
+                "  This is a one-time only migration to convert thumbnails for all of your\n"
+                "  documents into WebP format.  If you have a lot of documents though, \n"
+                "  this may take a while, so a coffee break may be in order."
+                "\n",
+            )
+
+            with multiprocessing.pool.Pool(
+                processes=min(multiprocessing.cpu_count(), 4),
+                maxtasksperchild=4,
+            ) as pool:
+                pool.map(_do_convert, work_packages)
+
+        end = time.time()
+        duration = end - start
+
+    logger.info(f"Conversion completed in {duration:.3f}s")
+
+
+class Migration(migrations.Migration):
+
+    dependencies = [
+        ("documents", "1020_merge_20220518_1839"),
+    ]
+
+    operations = [
+        migrations.RunPython(
+            code=_convert_thumbnails_to_webp,
+            reverse_code=migrations.RunPython.noop,
+        ),
+    ]