Simplify upload data for task (#5498)

It's possible to specify only the manifest file and filename pattern for
creating task with cloud storage data.
The special characters supported now for the pattern are `*`, `?`,
`[seq]`, `[!seq]`.
Please see
[here](8898a8b264/tests/python/rest_api/test_tasks.py (L686))
for some examples of how to use this functionality.

Co-authored-by: Maxim Zhiltsov <zhiltsov.max35@gmail.com>

CHANGELOG.md

@@ -7,7 +7,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## \[2.4.0] - Unreleased
 ### Added
-- TDB
+- Filename pattern to simplify uploading cloud storage data for a task (<https://github.com/opencv/cvat/pull/5498>)
 
 ### Changed
 - TDB

cvat/apps/engine/serializers.py

@@ -371,12 +371,13 @@ class DataSerializer(WriteOnceMixin, serializers.ModelSerializer):
     use_cache = serializers.BooleanField(default=False)
     copy_data = serializers.BooleanField(default=False)
     cloud_storage_id = serializers.IntegerField(write_only=True, allow_null=True, required=False)
+    filename_pattern = serializers.CharField(allow_null=True, required=False)
 
     class Meta:
         model = models.Data
         fields = ('chunk_size', 'size', 'image_quality', 'start_frame', 'stop_frame', 'frame_filter',
             'compressed_chunk_type', 'original_chunk_type', 'client_files', 'server_files', 'remote_files', 'use_zip_chunks',
-            'cloud_storage_id', 'use_cache', 'copy_data', 'storage_method', 'storage', 'sorting_method')
+            'cloud_storage_id', 'use_cache', 'copy_data', 'storage_method', 'storage', 'sorting_method', 'filename_pattern')
 
     # pylint: disable=no-self-use
     def validate_frame_filter(self, value):
@@ -396,6 +397,7 @@ class DataSerializer(WriteOnceMixin, serializers.ModelSerializer):
         if 'start_frame' in attrs and 'stop_frame' in attrs \
             and attrs['start_frame'] > attrs['stop_frame']:
             raise serializers.ValidationError('Stop frame must be more or equal start frame')
+
         return attrs
 
     def create(self, validated_data):

cvat/apps/engine/task.py

@@ -5,6 +5,7 @@
 # SPDX-License-Identifier: MIT
 
 import itertools
+import fnmatch
 import os
 import sys
 from rest_framework.serializers import ValidationError
@@ -127,7 +128,7 @@ def _save_task_to_db(db_task, extractor):
     db_task.data.save()
     db_task.save()
 
-def _count_files(data, manifest_files=None):
+def _count_files(data):
     share_root = settings.SHARE_ROOT
     server_files = []
 
@@ -158,7 +159,7 @@ def _count_files(data, manifest_files=None):
             if mime in counter:
                 counter[mime].append(rel_path)
             elif rel_path.endswith('.jsonl'):
-                manifest_files.append(rel_path)
+                continue
             else:
                 slogger.glob.warn("Skip '{}' file (its mime type doesn't "
                     "correspond to supported MIME file type)".format(full_path))
@@ -177,6 +178,12 @@ def _count_files(data, manifest_files=None):
 
     return counter
 
+def _find_manifest_files(data):
+    manifest_files = []
+    for files in ['client_files', 'server_files', 'remote_files']:
+        manifest_files.extend(list(filter(lambda x: x.endswith('.jsonl'), data[files])))
+    return manifest_files
+
 def _validate_data(counter, manifest_files=None):
     unique_entries = 0
     multiple_entries = 0
@@ -207,10 +214,10 @@ def _validate_data(counter, manifest_files=None):
 
     return counter, task_modes[0]
 
-def _validate_manifest(manifests, root_dir, is_in_cloud, db_cloud_storage):
+def _validate_manifest(manifests, root_dir, is_in_cloud, db_cloud_storage, data_storage_method):
     if manifests:
         if len(manifests) != 1:
-            raise Exception('Only one manifest file can be attached with data')
+            raise ValidationError('Only one manifest file can be attached to data')
         manifest_file = manifests[0]
         full_manifest_path = os.path.join(root_dir, manifests[0])
         if is_in_cloud:
@@ -221,8 +228,10 @@ def _validate_manifest(manifests, root_dir, is_in_cloud, db_cloud_storage):
                     < cloud_storage_instance.get_file_last_modified(manifest_file):
                 cloud_storage_instance.download_file(manifest_file, full_manifest_path)
         if is_manifest(full_manifest_path):
+            if not (settings.USE_CACHE or data_storage_method != models.StorageMethodChoice.CACHE):
+                raise ValidationError("Manifest file can be uploaded only if 'Use cache' option is also selected")
             return manifest_file
-        raise Exception('Invalid manifest was uploaded')
+        raise ValidationError('Invalid manifest was uploaded')
     return None
 
 def _validate_url(url):
@@ -291,6 +300,26 @@ def _download_data(urls, upload_dir):
 def _get_manifest_frame_indexer(start_frame=0, frame_step=1):
     return lambda frame_id: start_frame + frame_id * frame_step
 
+def _create_task_manifest_based_on_cloud_storage_manifest(
+    sorted_media,
+    cloud_storage_manifest_prefix,
+    cloud_storage_manifest,
+    manifest
+):
+    if cloud_storage_manifest_prefix:
+        sorted_media_without_manifest_prefix = [
+            os.path.relpath(i, cloud_storage_manifest_prefix) for i in sorted_media
+        ]
+        sequence, raw_content = cloud_storage_manifest.get_subset(sorted_media_without_manifest_prefix)
+        def _add_prefix(properties):
+            file_name = properties['name']
+            properties['name'] = os.path.join(cloud_storage_manifest_prefix, file_name)
+            return properties
+        content = list(map(_add_prefix, raw_content))
+    else:
+        sequence, content = cloud_storage_manifest.get_subset(sorted_media)
+    sorted_content = (i[1] for i in sorted(zip(sequence, content)))
+    manifest.create(sorted_content)
 
 @transaction.atomic
 def _create_thread(db_task, data, isBackupRestore=False, isDatasetImport=False):
@@ -300,69 +329,80 @@ def _create_thread(db_task, data, isBackupRestore=False, isDatasetImport=False):
     slogger.glob.info("create task #{}".format(db_task.id))
 
     db_data = db_task.data
-    upload_dir = db_data.get_upload_dirname()
+    upload_dir = db_data.get_upload_dirname() if db_data.storage != models.StorageChoice.SHARE else settings.SHARE_ROOT
     is_data_in_cloud = db_data.storage == models.StorageChoice.CLOUD_STORAGE
 
     if data['remote_files'] and not isDatasetImport:
         data['remote_files'] = _download_data(data['remote_files'], upload_dir)
 
-    manifest_files = []
-    media = _count_files(data, manifest_files)
-    media, task_mode = _validate_data(media, manifest_files)
-
-    if data['server_files']:
-        if db_data.storage == models.StorageChoice.LOCAL:
-            _copy_data_from_source(data['server_files'], upload_dir, data.get('server_files_path'))
-        elif db_data.storage == models.StorageChoice.SHARE:
-            upload_dir = settings.SHARE_ROOT
-
+    # find and validate manifest file
+    manifest_files = _find_manifest_files(data)
     manifest_root = None
-    if db_data.storage in {models.StorageChoice.LOCAL, models.StorageChoice.SHARE}:
+
+    # we should also handle this case because files from the share source have not been downloaded yet
+    if data['copy_data']:
+        manifest_root = settings.SHARE_ROOT
+    elif db_data.storage in {models.StorageChoice.LOCAL, models.StorageChoice.SHARE}:
         manifest_root = upload_dir
     elif is_data_in_cloud:
         manifest_root = db_data.cloud_storage.get_storage_dirname()
 
     manifest_file = _validate_manifest(
         manifest_files, manifest_root,
-        is_data_in_cloud, db_data.cloud_storage if is_data_in_cloud else None
+        is_data_in_cloud, db_data.cloud_storage if is_data_in_cloud else None,
+        db_data.storage_method,
     )
-    if manifest_file and (not settings.USE_CACHE or db_data.storage_method != models.StorageMethodChoice.CACHE):
-        raise Exception("File with meta information can be uploaded if 'Use cache' option is also selected")
 
-    if data['server_files'] and is_data_in_cloud:
+    if is_data_in_cloud:
         cloud_storage_instance = db_storage_to_storage_instance(db_data.cloud_storage)
-        sorted_media = sort(media['image'], data['sorting_method'])
-
-        data_size = len(sorted_media)
-        segment_step, *_ = _get_task_segment_data(db_task, data_size)
-        for start_frame in range(0, data_size, segment_step):
-            first_sorted_media_image = sorted_media[start_frame]
-            cloud_storage_instance.download_file(first_sorted_media_image, os.path.join(upload_dir, first_sorted_media_image))
 
-        # prepare task manifest file from cloud storage manifest file
-        # NOTE we should create manifest before defining chunk_size
-        # FIXME in the future when will be implemented archive support
         manifest = ImageManifestManager(db_data.get_manifest_path())
         cloud_storage_manifest = ImageManifestManager(
             os.path.join(db_data.cloud_storage.get_storage_dirname(), manifest_file),
             db_data.cloud_storage.get_storage_dirname()
         )
-        cloud_storage_manifest_prefix = os.path.dirname(manifest_file)
         cloud_storage_manifest.set_index()
-        if cloud_storage_manifest_prefix:
-            sorted_media_without_manifest_prefix = [
-                os.path.relpath(i, cloud_storage_manifest_prefix) for i in sorted_media
-            ]
-            sequence, raw_content = cloud_storage_manifest.get_subset(sorted_media_without_manifest_prefix)
-            def _add_prefix(properties):
-                file_name = properties['name']
-                properties['name'] = os.path.join(cloud_storage_manifest_prefix, file_name)
-                return properties
-            content = list(map(_add_prefix, raw_content))
+        cloud_storage_manifest_prefix = os.path.dirname(manifest_file)
+
+    # update list with server files if task creation approach with pattern and manifest file is used
+    if is_data_in_cloud and data['filename_pattern']:
+        if 1 != len(data['server_files']):
+            l = len(data['server_files']) - 1
+            raise ValidationError(
+                'Using a filename_pattern is only supported with a manifest file, '
+                f'but others {l} file{"s" if l > 1 else ""} {"were" if l > 1 else "was"} found'
+                'Please remove extra files and keep only manifest file in server_files field.'
+            )
+
+        cloud_storage_manifest_data = list(cloud_storage_manifest.data) if not cloud_storage_manifest_prefix \
+            else [os.path.join(cloud_storage_manifest_prefix, f) for f in cloud_storage_manifest.data]
+        if data['filename_pattern'] == '*':
+            server_files = cloud_storage_manifest_data
         else:
-            sequence, content = cloud_storage_manifest.get_subset(sorted_media)
-        sorted_content = (i[1] for i in sorted(zip(sequence, content)))
-        manifest.create(sorted_content)
+            server_files = fnmatch.filter(cloud_storage_manifest_data, data['filename_pattern'])
+        data['server_files'].extend(server_files)
+
+    # count and validate uploaded files
+    media = _count_files(data)
+    media, task_mode = _validate_data(media, manifest_files)
+
+    if data['server_files']:
+        if db_data.storage == models.StorageChoice.LOCAL:
+            _copy_data_from_source(data['server_files'], upload_dir, data.get('server_files_path'))
+        elif is_data_in_cloud:
+            sorted_media = sort(media['image'], data['sorting_method'])
+
+            # download previews from cloud storage
+            data_size = len(sorted_media)
+            segment_step, *_ = _get_task_segment_data(db_task, data_size)
+            for preview_frame in range(0, data_size, segment_step):
+                preview = sorted_media[preview_frame]
+                cloud_storage_instance.download_file(preview, os.path.join(upload_dir, preview))
+
+            # Define task manifest content based on cloud storage manifest content and uploaded files
+            _create_task_manifest_based_on_cloud_storage_manifest(
+                sorted_media, cloud_storage_manifest_prefix,
+                cloud_storage_manifest, manifest)
 
     av_scan_paths(upload_dir)
 

tests/python/rest_api/test_tasks.py

@@ -4,8 +4,12 @@
 # SPDX-License-Identifier: MIT
 
 import json
+import os.path as osp
+import subprocess
 from copy import deepcopy
+from functools import partial
 from http import HTTPStatus
+from tempfile import TemporaryDirectory
 from time import sleep
 
 import pytest
@@ -13,6 +17,7 @@ from cvat_sdk.api_client import apis, models
 from cvat_sdk.core.helpers import get_paginated_collection
 from deepdiff import DeepDiff
 
+import shared.utils.s3 as s3
 from shared.utils.config import get_method, make_api_client, patch_method
 from shared.utils.helpers import generate_image_files
 
@@ -675,6 +680,120 @@ class TestPostTaskData:
             self._USERNAME, task_spec, data_spec, content_type="application/json", org=org
         )
 
+    @pytest.mark.with_external_services
+    @pytest.mark.parametrize("cloud_storage_id", [1])
+    @pytest.mark.parametrize(
+        "manifest, filename_pattern, sub_dir, task_size",
+        [
+            ("manifest.jsonl", "*", True, 3),  # public bucket
+            ("manifest.jsonl", "test/*", True, 3),
+            ("manifest.jsonl", "test/sub*1.jpeg", True, 1),
+            ("manifest.jsonl", "*image*.jpeg", True, 3),
+            ("manifest.jsonl", "wrong_pattern", True, 0),
+            ("abc_manifest.jsonl", "[a-c]*.jpeg", False, 2),
+            ("abc_manifest.jsonl", "[d]*.jpeg", False, 1),
+            ("abc_manifest.jsonl", "[e-z]*.jpeg", False, 0),
+        ],
+    )
+    @pytest.mark.parametrize("org", [""])
+    def test_create_task_with_file_pattern(
+        self,
+        cloud_storage_id,
+        manifest,
+        filename_pattern,
+        sub_dir,
+        task_size,
+        org,
+        cloud_storages,
+        request,
+    ):
+        # prepare dataset on the bucket
+        prefixes = ("test_image_",) * 3 if sub_dir else ("a_", "b_", "d_")
+        images = generate_image_files(3, prefixes=prefixes)
+        s3_client = s3.make_client()
+
+        cloud_storage = cloud_storages[cloud_storage_id]
+
+        for image in images:
+            s3_client.create_file(
+                data=image,
+                bucket=cloud_storage["resource"],
+                filename=f"{'test/sub/' if sub_dir else ''}{image.name}",
+            )
+            request.addfinalizer(
+                partial(
+                    s3_client.remove_file,
+                    bucket=cloud_storage["resource"],
+                    filename=f"{'test/sub/' if sub_dir else ''}{image.name}",
+                )
+            )
+
+        with TemporaryDirectory() as tmp_dir:
+            for image in images:
+                with open(osp.join(tmp_dir, image.name), "wb") as f:
+                    f.write(image.getvalue())
+
+            command = [
+                "docker",
+                "run",
+                "--rm",
+                "-u",
+                "root:root",
+                "-v",
+                f"{tmp_dir}:/local",
+                "--entrypoint",
+                "python3",
+                "cvat/server",
+                "utils/dataset_manifest/create.py",
+                "--output-dir",
+                "/local",
+                "/local",
+            ]
+            subprocess.run(command, check=True)
+            with open(osp.join(tmp_dir, "manifest.jsonl"), mode="rb") as m_file:
+                s3_client.create_file(
+                    data=m_file.read(),
+                    bucket=cloud_storage["resource"],
+                    filename=f"test/sub/{manifest}" if sub_dir else manifest,
+                )
+                request.addfinalizer(
+                    partial(
+                        s3_client.remove_file,
+                        bucket=cloud_storage["resource"],
+                        filename=f"test/sub/{manifest}" if sub_dir else manifest,
+                    )
+                )
+
+        task_spec = {
+            "name": f"Task with files from cloud storage {cloud_storage_id}",
+            "labels": [
+                {
+                    "name": "car",
+                }
+            ],
+        }
+
+        data_spec = {
+            "image_quality": 75,
+            "use_cache": True,
+            "cloud_storage_id": cloud_storage_id,
+            "server_files": [f"test/sub/{manifest}" if sub_dir else manifest],
+            "filename_pattern": filename_pattern,
+        }
+
+        if task_size:
+            task_id = self._test_create_task(
+                self._USERNAME, task_spec, data_spec, content_type="application/json", org=org
+            )
+
+            with make_api_client(self._USERNAME) as api_client:
+                (task, response) = api_client.tasks_api.retrieve(task_id, org=org)
+                assert response.status == HTTPStatus.OK
+                assert task.size == task_size
+        else:
+            status = self._test_cannot_create_task(self._USERNAME, task_spec, data_spec)
+            assert "No media data found" in status.message
+
     @pytest.mark.with_external_services
     @pytest.mark.parametrize(
         "cloud_storage_id, manifest, org",

tests/python/shared/utils/helpers.py

@@ -18,10 +18,11 @@ def generate_image_file(filename="image.png", size=(50, 50), color=(0, 0, 0)):
     return f
 
 
-def generate_image_files(count) -> List[BytesIO]:
+def generate_image_files(count, prefixes=None) -> List[BytesIO]:
     images = []
     for i in range(count):
-        image = generate_image_file(f"{i}.jpeg", color=(i, i, i))
+        prefix = prefixes[i] if prefixes else ""
+        image = generate_image_file(f"{prefix}{i}.jpeg", color=(i, i, i))
         images.append(image)
 
     return images