ol-openedx-course-translations 0.1.0__py3-none-any.whl → 0.3.5__py3-none-any.whl

ol_openedx_course_translations/management/commands/translate_course.py

@@ -2,25 +2,64 @@
 Management command to translate course content to a specified language.
 """
 
-import json
 import logging
 import shutil
-import tarfile
+import time
 from pathlib import Path
-from typing import Any
 
-import deepl
-from defusedxml import ElementTree
+from celery import group
 from django.conf import settings
 from django.core.management.base import BaseCommand, CommandError
 
+from ol_openedx_course_translations.models import CourseTranslationLog
+from ol_openedx_course_translations.tasks import (
+    translate_file_task,
+    translate_grading_policy_task,
+    translate_policy_json_task,
+)
+from ol_openedx_course_translations.utils.constants import PROVIDER_DEEPL
+from ol_openedx_course_translations.utils.course_translations import (
+    create_translated_archive,
+    create_translated_copy,
+    extract_course_archive,
+    generate_course_key_from_xml,
+    get_translatable_file_paths,
+    update_course_language_attribute,
+    validate_course_inputs,
+)
+
 logger = logging.getLogger(__name__)
 
+# Task configuration
+TASK_TIMEOUT_SECONDS = 3600 * 2  # 2 hour total timeout for all tasks
+TASK_POLL_INTERVAL_SECONDS = 2  # Poll every 2 seconds for task completion
+
 
 class Command(BaseCommand):
     """Translate given course content to the specified language."""
 
-    help = "Translate course content to the specified language."
+    help = (
+        "Translate course content to the specified language.\n\n"
+        "Configuration:\n"
+        "All translation providers should be configured in TRANSLATIONS_PROVIDERS:\n"
+        "{\n"
+        '    "deepl": {"api_key": "<YOUR_DEEPL_API_KEY>"},\n'
+        '    "openai": {"api_key": "<KEY>", "default_model": "gpt-5.2"},\n'
+        '    "gemini": {"api_key": "<KEY>", "default_model": "gemini-3-pro-preview"},\n'
+        '    "mistral": {"api_key": "<KEY>", "default_model": "mistral-large-latest"}\n'
+        "}\n"
+    )
+
+    def __init__(self, *args, **kwargs):
+        """Initialize the command with empty task list."""
+        super().__init__(*args, **kwargs)
+        self.tasks = []
+        self.translated_course_dir = None
+        self.content_provider_name = None
+        self.content_model = None
+        self.srt_provider_name = None
+        self.srt_model = None
+        self.glossary_directory = None
 
     def add_arguments(self, parser) -> None:
         """Entry point for subclassed commands to add custom arguments."""
@@ -34,8 +73,8 @@ class Command(BaseCommand):
             ),
         )
         parser.add_argument(
-            "--translation-language",
-            dest="translation_language",
+            "--target-language",
+            dest="target_language",
             required=True,
             help=(
                 "Specify the language code in ISO format "
@@ -44,545 +83,503 @@ class Command(BaseCommand):
         )
         parser.add_argument(
             "--course-dir",
-            dest="course_directory",
+            dest="course_archive_path",
             required=True,
             help="Specify the course directory (tar archive).",
         )
+        parser.add_argument(
+            "--content-translation-provider",
+            dest="content_translation_provider",
+            required=True,
+            help=(
+                "Translation provider for content (XML/HTML and text). "
+                "Format: 'deepl', 'PROVIDER', or 'PROVIDER/MODEL' "
+                "(e.g., 'openai', 'openai/gpt-5.2', 'gemini', 'gemini/gemini-3-pro-preview'). "  # noqa: E501
+                "If model is not specified, uses the default model from settings."
+            ),
+        )
+        parser.add_argument(
+            "--srt-translation-provider",
+            dest="srt_translation_provider",
+            required=True,
+            help=(
+                "Translation provider for SRT subtitles. "
+                "Format: 'deepl', 'PROVIDER', or 'PROVIDER/MODEL' "
+                "(e.g., 'openai', 'openai/gpt-5.2', 'gemini', 'gemini/gemini-3-pro-preview'). "  # noqa: E501
+                "If model is not specified, uses the default model from settings."
+            ),
+        )
+        parser.add_argument(
+            "--glossary-dir",
+            dest="glossary_directory",
+            required=False,
+            help=(
+                "Path to glossary directory containing "
+                "language-specific glossary files."
+            ),
+        )
+
+    def _parse_and_validate_provider_spec(
+        self, provider_spec: str
+    ) -> tuple[str, str | None]:
+        """
+        Parse and validate provider specification into provider name and model.
+
+        Resolves model from settings if not provided in specification.
+
+        Args:
+            provider_spec: Provider specification
+
+        Returns:
+            Tuple of (provider_name, model_name). model_name is None for DeepL or
+            resolved from settings if not specified.
+
+        Raises:
+            CommandError: If provider specification format is invalid
+            or model and api_key cannot be resolved
+        """
+        # Parse the specification
+        if "/" in provider_spec:
+            parts = provider_spec.split("/", 1)
+            if len(parts) != 2 or not parts[0] or not parts[1]:  # noqa: PLR2004
+                error_msg = (
+                    f"Invalid provider specification: {provider_spec}. "
+                    "Use format 'PROVIDER' or 'PROVIDER/MODEL' "
+                    "(e.g., 'openai', 'openai/gpt-5.2')"
+                )
+                raise CommandError(error_msg)
+            provider_name = parts[0].lower()
+            model_name = parts[1]
+        else:
+            provider_name = provider_spec.lower()
+            model_name = None
+
+        # Try to get default model from settings
+        providers_config = getattr(settings, "TRANSLATIONS_PROVIDERS", {})
+        if provider_name not in providers_config:
+            error_msg = (
+                f"Provider '{provider_name}' not configured in TRANSLATIONS_PROVIDERS. "
+                f"Available providers: {', '.join(providers_config.keys())}"
+            )
+            raise CommandError(error_msg)
+
+        provider_config = providers_config[provider_name]
+        api_key = provider_config.get("api_key")
+        if not api_key:
+            error_msg = (
+                f"API key for provider '{provider_name}' is not configured in "
+                "TRANSLATIONS_PROVIDERS. Please set the 'api_key' in settings."
+            )
+            raise CommandError(error_msg)
+
+        # DeepL doesn't use models
+        if provider_name == PROVIDER_DEEPL:
+            return provider_name, None
+
+        # If model is explicitly provided, return it
+        if model_name:
+            return provider_name, model_name
+
+        default_model = provider_config.get("default_model")
+        if not default_model:
+            error_msg = (
+                f"No model specified for provider '{provider_name}' and no "
+                f"default_model found in TRANSLATIONS_PROVIDERS['{provider_name}']. "
+                f"Either specify a model (e.g., '{provider_name}/gpt-5.2') or "
+                f"configure a default_model in settings."
+            )
+            raise CommandError(error_msg)
+
+        return provider_name, default_model
 
     def handle(self, **options) -> None:
         """Handle the translate_course command."""
         try:
-            self._validate_inputs(options)
+            start_time = time.perf_counter()
+            course_archive_path = Path(options["course_archive_path"])
+            source_language = options["source_language"].upper()
+            target_language = options["target_language"].upper()
+            content_provider_spec = options["content_translation_provider"]
+            srt_provider_spec = options["srt_translation_provider"]
+            glossary_directory = options.get("glossary_directory")
+
+            # Parse and validate provider specifications (includes validation)
+            content_provider_name, content_model = (
+                self._parse_and_validate_provider_spec(content_provider_spec)
+            )
+            srt_provider_name, srt_model = self._parse_and_validate_provider_spec(
+                srt_provider_spec
+            )
 
-            course_dir = Path(options["course_directory"])
-            source_language = options["source_language"]
-            translation_language = options["translation_language"]
+            # Log the resolved configuration
+            if content_model:
+                self.stdout.write(
+                    f"Content provider: {content_provider_name}/{content_model}"
+                )
+            else:
+                self.stdout.write(f"Content provider: {content_provider_name}")
+
+            if srt_model:
+                self.stdout.write(f"SRT provider: {srt_provider_name}/{srt_model}")
+            else:
+                self.stdout.write(f"SRT provider: {srt_provider_name}")
+
+            # Validate inputs
+            validate_course_inputs(course_archive_path)
+
+            # Store provider names and models
+            self.content_provider_name = content_provider_name
+            self.content_model = content_model
+            self.srt_provider_name = srt_provider_name
+            self.srt_model = srt_model
+            self.glossary_directory = glossary_directory
 
             # Extract course archive
-            extracted_dir = self._extract_course_archive(course_dir)
+            extracted_course_dir = extract_course_archive(course_archive_path)
 
             # Create translated copy
-            translated_dir = self._create_translated_copy(
-                extracted_dir, translation_language
+            translated_course_dir = create_translated_copy(
+                extracted_course_dir, target_language
             )
 
+            # Store for cleanup on failure
+            self.translated_course_dir = translated_course_dir
+
             # Delete extracted directory after copying
-            if extracted_dir.exists():
-                shutil.rmtree(extracted_dir)
+            if extracted_course_dir.exists():
+                shutil.rmtree(extracted_course_dir)
 
-            # Translate content
-            billed_chars = self._translate_course_content(
-                translated_dir, source_language, translation_language
+            # Translate content asynchronously
+            self._translate_course_content_async(
+                translated_course_dir, source_language, target_language
             )
 
+            # Wait for all tasks and report status
+            command_stats = self._wait_and_report_tasks()
+            total_time_taken_msg = (
+                f"Command finished in: {time.perf_counter() - start_time:.2f} seconds."
+            )
+            self.stdout.write(self.style.SUCCESS(total_time_taken_msg))
+            command_stats.append(total_time_taken_msg)
+
+            # Add translation log entry
+            self._add_translation_log_entry(
+                source_language=source_language,
+                target_language=target_language,
+                command_stats=command_stats,
+            )
             # Create final archive
-            archive_path = self._create_translated_archive(
-                translated_dir, translation_language, course_dir.stem
+            translated_archive_path = create_translated_archive(
+                translated_course_dir, target_language, course_archive_path.stem
             )
-
-            self.stdout.write(
-                self.style.SUCCESS(
-                    f"Translation completed. Archive created: {archive_path}"
-                )
+            success_msg = (
+                f"Translation completed successfully. Translated archive created: "
+                f"{translated_archive_path}"
             )
-            logger.info("Total billed characters: %s", billed_chars)
+            self.stdout.write(self.style.SUCCESS(success_msg))
 
         except Exception as e:
             logger.exception("Translation failed")
+
+            # Cleanup translated course directory on failure
+            if self.translated_course_dir and self.translated_course_dir.exists():
+                self.stdout.write(
+                    self.style.WARNING(
+                        f"Cleaning up translated course directory: {self.translated_course_dir}"  # noqa: E501
+                    )
+                )
+                shutil.rmtree(self.translated_course_dir)
+
             error_msg = f"Translation failed: {e}"
             raise CommandError(error_msg) from e
 
-    def get_supported_archive_extension(self, filename: str) -> str | None:
-        """
-        Return the supported archive extension if filename ends with one, else None.
+    def _translate_course_content_async(
+        self, course_dir: Path, source_language: str, target_language: str
+    ) -> None:
         """
-        for ext in settings.OL_OPENEDX_COURSE_TRANSLATIONS_SUPPORTED_ARCHIVE_EXTENSIONS:
-            if filename.endswith(ext):
-                return ext
-        return None
+        Translate all course content using Celery tasks.
 
-    def _validate_inputs(self, options: dict[str, Any]) -> None:
-        """Validate command inputs."""
-        course_dir = Path(options["course_directory"])
+        Args:
+            course_dir: Path to the course directory
+            source_language: Source language code
+            target_language: Target language code
 
-        if not course_dir.exists():
-            error_msg = f"Course directory not found: {course_dir}"
-            raise CommandError(error_msg)
-
-        if self.get_supported_archive_extension(course_dir.name) is None:
-            supported_exts = ", ".join(
-                settings.OL_OPENEDX_COURSE_TRANSLATIONS_SUPPORTED_ARCHIVE_EXTENSIONS
-            )
-            error_msg = f"Course directory must be a tar file: {supported_exts}"
-            raise CommandError(error_msg)
-
-        if not hasattr(settings, "DEEPL_API_KEY") or not settings.DEEPL_API_KEY:
-            error_msg = "DEEPL_API_KEY setting is required"
-            raise CommandError(error_msg)
-
-    def _extract_course_archive(self, course_dir: Path) -> Path:
-        """Extract course archive to working directory."""
-        # Use the parent directory of the source file as the base extraction directory
-        extract_base_dir = course_dir.parent
-
-        # Get base name without extension
-        ext = self.get_supported_archive_extension(course_dir.name)
-        tarball_base = course_dir.name[: -len(ext)] if ext else course_dir.name
-
-        extracted_dir = extract_base_dir / tarball_base
-
-        if not extracted_dir.exists():
-            try:
-                with tarfile.open(course_dir, "r:*") as tar:
-                    # Validate tar file before extraction
-                    self._validate_tar_file(tar)
-                    tar.extractall(path=extracted_dir, filter="data")
-            except (tarfile.TarError, OSError) as e:
-                error_msg = f"Failed to extract archive: {e}"
-                raise CommandError(error_msg) from e
-
-        logger.info("Extracted course to: %s", extracted_dir)
-        return extracted_dir
-
-    def _validate_tar_file(self, tar: tarfile.TarFile) -> None:
-        """Validate tar file contents for security."""
-        for member in tar.getmembers():
-            # Check for directory traversal attacks
-            if member.name.startswith("/") or ".." in member.name:
-                error_msg = f"Unsafe tar member: {member.name}"
-                raise CommandError(error_msg)
-            # Check for excessively large files
-            if (
-                member.size > 512 * 1024 * 1024
-            ):  # 0.5GB limit because courses on Production are big
-                error_msg = f"File too large: {member.name}"
-                raise CommandError(error_msg)
-
-    def _create_translated_copy(
-        self, source_dir: Path, translation_language: str
-    ) -> Path:
-        """Create a copy of the course for translation."""
-        base_name = source_dir.name
-        new_dir_name = f"{translation_language}_{base_name}"
-        new_dir_path = source_dir.parent / new_dir_name
+        Raises:
+            CommandError: If course directory is not found
+        """
+        course_directory = course_dir / "course"
 
-        if new_dir_path.exists():
-            error_msg = f"Translation directory already exists: {new_dir_path}"
+        if not course_directory.exists() or not course_directory.is_dir():
+            error_msg = f"Course directory not found: {course_directory}"
             raise CommandError(error_msg)
 
-        shutil.copytree(source_dir, new_dir_path)
-        logger.info("Created translation copy: %s", new_dir_path)
-        return new_dir_path
-
-    def _translate_course_content(
-        self, course_dir: Path, source_language: str, translation_language: str
-    ) -> int:
-        """Translate all course content and return total billed characters."""
-        total_billed_chars = 0
-
-        # Translate files in main directories
-        for search_dir in [course_dir, course_dir.parent]:
-            total_billed_chars += self._translate_files_in_directory(
-                search_dir, source_language, translation_language, recursive=False
-            )
+        # Update language attributes in course XML, doing this
+        # because tasks can override the XML files
+        update_course_language_attribute(course_directory, target_language)
 
-            # Translate files in target subdirectories
-            for dir_name in settings.OL_OPENEDX_COURSE_TRANSLATIONS_TARGET_DIRECTORIES:
-                target_dir = search_dir / dir_name
-                if target_dir.exists() and target_dir.is_dir():
-                    total_billed_chars += self._translate_files_in_directory(
-                        target_dir,
-                        source_language,
-                        translation_language,
-                        recursive=True,
-                    )
+        # Collect all tasks
+        self.tasks = []
 
-        # Translate special JSON files
-        total_billed_chars += self._translate_grading_policy(
-            course_dir, source_language, translation_language
-        )
-        total_billed_chars += self._translate_policy_json(
-            course_dir, source_language, translation_language
+        # Add translation tasks for files in course directory
+        self._add_file_translation_tasks(
+            course_directory, source_language, target_language, recursive=False
         )
 
-        return total_billed_chars
+        # Add translation tasks for target subdirectories
+        for target_dir_name in settings.COURSE_TRANSLATIONS_TARGET_DIRECTORIES:
+            target_directory = course_directory / target_dir_name
+            if target_directory.exists() and target_directory.is_dir():
+                self._add_file_translation_tasks(
+                    target_directory, source_language, target_language, recursive=True
+                )
+
+        # Add tasks for special JSON files
+        self._add_grading_policy_tasks(course_dir, target_language)
+        self._add_policy_json_tasks(course_dir, target_language)
 
-    def _translate_files_in_directory(
+    def _add_file_translation_tasks(
         self,
-        directory: Path,
+        directory_path: Path,
         source_language: str,
-        translation_language: str,
+        target_language: str,
         *,
         recursive: bool = False,
-    ) -> int:
-        """Translate files in a directory."""
-        total_billed_chars = 0
-
-        if recursive:
-            file_paths: list[Path] = []
-            for ext in settings.OL_OPENEDX_COURSE_TRANSLATIONS_TRANSLATABLE_EXTENSIONS:
-                file_paths.extend(directory.rglob(f"*{ext}"))
-        else:
-            file_paths = [
-                f
-                for f in directory.iterdir()
-                if f.is_file()
-                and any(
-                    f.name.endswith(ext)
-                    for ext in settings.OL_OPENEDX_COURSE_TRANSLATIONS_TRANSLATABLE_EXTENSIONS  # noqa: E501
-                )
-            ]
-
-        for file_path in file_paths:
-            try:
-                total_billed_chars += self._translate_file(
-                    file_path, source_language, translation_language
-                )
-            except (OSError, UnicodeDecodeError) as e:
-                logger.warning("Failed to translate %s: %s", file_path, e)
-
-        return total_billed_chars
+    ) -> None:
+        """
+        Add Celery tasks for file translation to the task list.
 
-    def _translate_file(
-        self, file_path: Path, source_language: str, translation_language: str
-    ) -> int:
-        """Translate a single file and return billed characters."""
-        try:
-            content = file_path.read_text(encoding="utf-8")
-            logger.debug("Translating: %s", file_path)
+        Args:
+            directory_path: Path to directory containing files to translate
+            source_language: Source language code
+            target_language: Target language code
+            recursive: Whether to search for files recursively
+        """
+        translatable_file_paths = get_translatable_file_paths(
+            directory_path, recursive=recursive
+        )
 
-            translated_content, billed_chars = self._translate_text(
-                content, source_language, translation_language, file_path.name
+        for file_path in translatable_file_paths:
+            task = translate_file_task.s(
+                str(file_path),
+                source_language,
+                target_language,
+                self.content_provider_name,
+                self.content_model,
+                self.srt_provider_name,
+                self.srt_model,
+                self.glossary_directory,
             )
+            self.tasks.append(("file", str(file_path), task))
+            logger.info("Added translation task for: %s", file_path)
 
-            # Handle XML display_name translation
-            if file_path.suffix == ".xml":
-                translated_content = self._translate_display_name(
-                    translated_content, source_language, translation_language
-                )
-
-            file_path.write_text(translated_content, encoding="utf-8")
-        except (OSError, UnicodeDecodeError) as e:
-            logger.warning("Failed to translate %s: %s", file_path, e)
-            return 0
-        else:
-            return billed_chars
-
-    def _translate_grading_policy(
-        self, course_dir: Path, source_language: str, translation_language: str
-    ) -> int:
-        """Translate grading_policy.json files."""
-        total_billed_chars = 0
-        policies_dir = course_dir / "course" / "policies"
+    def _add_grading_policy_tasks(self, course_dir: Path, target_language: str) -> None:
+        """
+        Add Celery tasks for grading_policy.json translation to the task list.
 
-        if not policies_dir.exists():
-            return 0
+        Args:
+            course_dir: Path to the course directory
+            target_language: Target language code
+        """
+        course_policies_dir = course_dir / "course" / "policies"
 
-        for child_dir in policies_dir.iterdir():
-            if not child_dir.is_dir():
-                continue
+        if not course_policies_dir.exists():
+            return
 
-            grading_policy_path = child_dir / "grading_policy.json"
-            if not grading_policy_path.exists():
+        for policy_child_dir in course_policies_dir.iterdir():
+            if not policy_child_dir.is_dir():
                 continue
 
-            try:
-                grading_policy = json.loads(
-                    grading_policy_path.read_text(encoding="utf-8")
-                )
-                updated = False
-
-                for item in grading_policy.get("GRADER", []):
-                    if "short_label" in item:
-                        translated_label, billed_chars = self._translate_text(
-                            item["short_label"], source_language, translation_language
-                        )
-                        item["short_label"] = translated_label
-                        total_billed_chars += billed_chars
-                        updated = True
-
-                if updated:
-                    grading_policy_path.write_text(
-                        json.dumps(grading_policy, ensure_ascii=False, indent=4),
-                        encoding="utf-8",
-                    )
-            except (OSError, json.JSONDecodeError) as e:
-                logger.warning(
-                    "Failed to translate grading policy in %s: %s", child_dir, e
+            grading_policy_file = policy_child_dir / "grading_policy.json"
+            if grading_policy_file.exists():
+                task = translate_grading_policy_task.s(
+                    str(grading_policy_file),
+                    target_language,
+                    self.content_provider_name,
+                    self.content_model,
+                    self.glossary_directory,
                 )
+                self.tasks.append(("grading_policy", str(grading_policy_file), task))
+                logger.info("Added grading policy task for: %s", grading_policy_file)
 
-        return total_billed_chars
-
-    def _translate_policy_json(
-        self, course_dir: Path, source_language: str, translation_language: str
-    ) -> int:
-        """Translate policy.json files."""
-        total_billed_chars = 0
-        policies_dir = course_dir / "course" / "policies"
+    def _add_policy_json_tasks(self, course_dir: Path, target_language: str) -> None:
+        """
+        Add Celery tasks for policy.json translation to the task list.
 
-        if not policies_dir.exists():
-            return 0
+        Args:
+            course_dir: Path to the course directory
+            target_language: Target language code
+        """
+        course_policies_dir = course_dir / "course" / "policies"
 
-        for child_dir in policies_dir.iterdir():
-            if not child_dir.is_dir():
-                continue
+        if not course_policies_dir.exists():
+            return
 
-            policy_path = child_dir / "policy.json"
-            if not policy_path.exists():
+        for policy_child_dir in course_policies_dir.iterdir():
+            if not policy_child_dir.is_dir():
                 continue
 
-            try:
-                policy_data = json.loads(policy_path.read_text(encoding="utf-8"))
-                updated = False
-
-                for course_obj in policy_data.values():
-                    if not isinstance(course_obj, dict):
-                        continue
-
-                    # Translate various fields
-                    billed_chars, field_updated = self._translate_policy_fields(
-                        course_obj, source_language, translation_language
-                    )
-                    total_billed_chars += billed_chars
-                    updated = updated or field_updated
-
-                if updated:
-                    policy_path.write_text(
-                        json.dumps(policy_data, ensure_ascii=False, indent=4),
-                        encoding="utf-8",
-                    )
-            except (OSError, json.JSONDecodeError) as e:
-                logger.warning("Failed to translate policy in %s: %s", child_dir, e)
-
-        return total_billed_chars
-
-    def _translate_policy_fields(
-        self,
-        course_obj: dict[str, Any],
-        source_language: str,
-        translation_language: str,
-    ) -> tuple[int, bool]:
-        """Translate specific fields in policy object."""
-        total_billed_chars = 0
-        updated = False
-
-        # Translate simple string fields
-        billed_chars, field_updated = self._translate_string_fields(
-            course_obj, source_language, translation_language
-        )
-        total_billed_chars += billed_chars
-        updated = updated or field_updated
+            policy_file = policy_child_dir / "policy.json"
+            if policy_file.exists():
+                task = translate_policy_json_task.s(
+                    str(policy_file),
+                    target_language,
+                    self.content_provider_name,
+                    self.content_model,
+                    self.glossary_directory,
+                )
+                self.tasks.append(("policy", str(policy_file), task))
+                logger.info("Added policy.json task for: %s", policy_file)
 
-        # Translate discussion topics
-        billed_chars, field_updated = self._translate_discussion_topics(
-            course_obj, source_language, translation_language
-        )
-        total_billed_chars += billed_chars
-        updated = updated or field_updated
+    def _wait_and_report_tasks(self) -> list[str]:  # noqa: C901, PLR0915, PLR0912
+        """
+        Execute all tasks as a Celery group and wait for completion.
 
-        # Translate learning info and tabs
-        billed_chars, field_updated = self._translate_learning_info_and_tabs(
-            course_obj, source_language, translation_language
-        )
-        total_billed_chars += billed_chars
-        updated = updated or field_updated
+        Uses Celery's group primitive to execute tasks in parallel and
+        provides detailed progress reporting.
 
-        # Translate XML attributes
-        billed_chars, field_updated = self._translate_xml_attributes(
-            course_obj, source_language, translation_language
+        Raises:
+            CommandError: If any tasks fail
+        """
+        stats = []
+        if not self.tasks:
+            self.stdout.write("No tasks to execute.")
+            return []
+
+        total_tasks = len(self.tasks)
+        self.stdout.write(
+            f"\nExecuting {total_tasks} translation tasks in parallel...\n"
         )
-        total_billed_chars += billed_chars
-        updated = updated or field_updated
 
-        return total_billed_chars, updated
+        # Extract task signatures and create mappings
+        task_signatures = [task_sig for _, _, task_sig in self.tasks]
+        task_metadata = {
+            i: (task_type, file_path)
+            for i, (task_type, file_path, _) in enumerate(self.tasks)
+        }
 
-    def _translate_string_fields(
-        self,
-        course_obj: dict[str, Any],
-        source_language: str,
-        translation_language: str,
-    ) -> tuple[int, bool]:
-        """Translate simple string fields."""
-        total_billed_chars = 0
-        updated = False
-
-        string_fields = ["advertised_start", "display_name", "display_organization"]
-        for field in string_fields:
-            if field in course_obj:
-                translated, billed_chars = self._translate_text(
-                    course_obj[field], source_language, translation_language
-                )
-                course_obj[field] = translated
-                total_billed_chars += billed_chars
-                updated = True
+        # Create and execute group
+        job = group(task_signatures)
+        result = job.apply_async()
 
-        return total_billed_chars, updated
+        # Wait for all tasks to complete with progress reporting
+        completed_count = 0
+        self.stdout.flush()
 
-    def _translate_discussion_topics(
-        self,
-        course_obj: dict[str, Any],
-        source_language: str,
-        translation_language: str,
-    ) -> tuple[int, bool]:
-        """Translate discussion topics."""
-        total_billed_chars = 0
-        updated = False
-
-        if "discussion_topics" in course_obj:
-            topics = course_obj["discussion_topics"]
-            if isinstance(topics, dict):
-                new_topics = {}
-                for topic_key, value in topics.items():
-                    translated_key, billed_chars = self._translate_text(
-                        topic_key, source_language, translation_language
-                    )
-                    new_topics[translated_key] = value
-                    total_billed_chars += billed_chars
-                course_obj["discussion_topics"] = new_topics
-                updated = True
-
-        return total_billed_chars, updated
-
-    def _translate_learning_info_and_tabs(
-        self,
-        course_obj: dict[str, Any],
-        source_language: str,
-        translation_language: str,
-    ) -> tuple[int, bool]:
-        """Translate learning info and tabs."""
-        total_billed_chars = 0
-        updated = False
-
-        # Learning info
-        if "learning_info" in course_obj and isinstance(
-            course_obj["learning_info"], list
-        ):
-            translated_info = []
-            for item in course_obj["learning_info"]:
-                translated, billed_chars = self._translate_text(
-                    item, source_language, translation_language
-                )
-                translated_info.append(translated)
-                total_billed_chars += billed_chars
-            course_obj["learning_info"] = translated_info
-            updated = True
-
-        # Tabs
-        if "tabs" in course_obj and isinstance(course_obj["tabs"], list):
-            for tab in course_obj["tabs"]:
-                if isinstance(tab, dict) and "name" in tab:
-                    translated, billed_chars = self._translate_text(
-                        tab["name"], source_language, translation_language
-                    )
-                    tab["name"] = translated
-                    total_billed_chars += billed_chars
-                    updated = True
-
-        return total_billed_chars, updated
-
-    def _translate_xml_attributes(
-        self,
-        course_obj: dict[str, Any],
-        source_language: str,
-        translation_language: str,
-    ) -> tuple[int, bool]:
-        """Translate XML attributes."""
-        total_billed_chars = 0
-        updated = False
-
-        if "xml_attributes" in course_obj and isinstance(
-            course_obj["xml_attributes"], dict
-        ):
-            xml_attrs = course_obj["xml_attributes"]
-            xml_fields = [
-                "diplay_name",
-                "info_sidebar_name",
-            ]  # Note: keeping typo as in original
-            for field in xml_fields:
-                if field in xml_attrs:
-                    translated, billed_chars = self._translate_text(
-                        xml_attrs[field], source_language, translation_language
+        try:
+            # Poll for completion and show progress
+            while not result.ready():
+                # Count completed tasks
+                new_completed = sum(1 for r in result.results if r.ready())
+                if new_completed > completed_count:
+                    completed_count = new_completed
+                    self.stdout.write(
+                        f"\rProgress: {completed_count}/{total_tasks} tasks completed",
+                        ending="",
                     )
-                    xml_attrs[field] = translated
-                    total_billed_chars += billed_chars
-                    updated = True
-
-        return total_billed_chars, updated
-
-    def _create_translated_archive(
-        self, translated_dir: Path, translation_language: str, original_name: str
-    ) -> Path:
-        """Create tar.gz archive of translated course."""
-        # Remove all archive extensions from the original name
-        ext = self.get_supported_archive_extension(original_name)
-        clean_name = original_name[: -len(ext)] if ext else original_name
-
-        tar_gz_name = f"{translation_language}_{clean_name}.tar.gz"
-        tar_gz_path = translated_dir.parent / tar_gz_name
-
-        # Remove existing archive
-        if tar_gz_path.exists():
-            tar_gz_path.unlink()
+                    self.stdout.flush()
 
-        # Create tar.gz archive containing only the 'course' directory
-        course_dir_path = translated_dir / "course"
-        with tarfile.open(tar_gz_path, "w:gz") as tar:
-            tar.add(course_dir_path, arcname="course")
+                # Sleep before next poll (don't use join with timeout)
+                time.sleep(TASK_POLL_INTERVAL_SECONDS)
 
-        # Delete extracted directory after copying
-        if translated_dir.exists():
-            shutil.rmtree(translated_dir)
+            # Final update
+            self.stdout.write(
+                f"\rProgress: {total_tasks}/{total_tasks} tasks completed\n"
+            )
 
-        logger.info("Created tar.gz archive: %s", tar_gz_path)
-        return tar_gz_path
+            # Get all results (this will raise exceptions if propagate=True)
+            results = result.get(timeout=TASK_TIMEOUT_SECONDS, propagate=False)
 
-    def _translate_text(
-        self,
-        text: str,
-        source_language: str,
-        target_language: str,
-        filename: str | None = None,
-    ) -> tuple[str, int]:
-        """Translate text using DeepL API."""
-        if not text or not text.strip():
-            return text, 0
-
-        try:
-            deepl_client = deepl.Translator(settings.DEEPL_API_KEY)
-
-            tag_handling = None
-            if filename:
-                extension = Path(filename).suffix.lstrip(".")
-                if extension in ["html", "xml"]:
-                    tag_handling = extension
-
-            result = deepl_client.translate_text(
-                text,
-                source_lang=source_language,
-                target_lang=target_language,
-                tag_handling=tag_handling,
-            )
+        except Exception as e:
+            logger.exception("Task execution failed")
+            error_msg = f"Task execution timeout or error: {e}"
+            raise CommandError(error_msg) from e
 
-            return result.text, result.billed_characters  # noqa: TRY300
-        except (deepl.exceptions.DeepLException, OSError) as e:
-            logger.warning("Translation failed for text: %s... Error: %s", text[:50], e)
-            return text, 0
+        # Process results
+        completed_tasks = 0
+        failed_tasks = 0
+        skipped_tasks = 0
+
+        for i, task_result in enumerate(results):
+            task_type, file_path = task_metadata[i]
+
+            if isinstance(task_result, dict):
+                status = task_result.get("status", "unknown")
+                if status == "success":
+                    completed_tasks += 1
+                    msg = f"✓ {task_type}: {file_path}"
+                    stats.append(msg)
+                    self.stdout.write(self.style.SUCCESS(msg))
+                elif status == "skipped":
+                    skipped_tasks += 1
+                    reason = task_result.get("reason", "Skipped")
+                    msg = f"⊘ {task_type}: {file_path} - {reason}"
+                    stats.append(msg)
+                    self.stdout.write(self.style.WARNING(msg))
+                elif status == "error":
+                    failed_tasks += 1
+                    error = task_result.get("error", "Unknown error")
+                    msg = f"✗ {task_type}: {file_path} - {error}"
+                    stats.append(msg)
+                    self.stdout.write(self.style.ERROR(msg))
+                else:
+                    failed_tasks += 1
+                    msg = f"✗ {task_type}: {file_path} - Unknown status: {status}"
+                    stats.append(msg)
+                    self.stdout.write(self.style.ERROR(msg))
+            else:
+                # Task raised an exception
+                failed_tasks += 1
+                error_msg = str(task_result) if task_result else "Task failed"
+                msg = f"✗ {task_type}: {file_path} - {error_msg}"
+                stats.append(msg)
+                self.stdout.write(self.style.ERROR(msg))
+
+        # Print summary
+        self.stdout.write("\n" + "=" * 60)
+        successful_tasks_stats = (
+            f"Total tasks: {total_tasks}\nCompleted: {completed_tasks}"
+        )
+        stats.append(successful_tasks_stats)
+        self.stdout.write(self.style.SUCCESS(successful_tasks_stats))
+        if skipped_tasks > 0:
+            skipped_tasks_stats = f"Skipped: {skipped_tasks}"
+            stats.append(skipped_tasks_stats)
+            self.stdout.write(self.style.WARNING(skipped_tasks_stats))
+        if failed_tasks > 0:
+            failed_tasks_stats = f"Failed: {failed_tasks}"
+            stats.append(failed_tasks_stats)
+            self.stdout.write(self.style.ERROR(failed_tasks_stats))
+        self.stdout.write("=" * 60 + "\n")
+
+        if failed_tasks > 0:
+            error_msg = f"{failed_tasks} translation tasks failed"
+            raise CommandError(error_msg)
 
-    def _translate_display_name(
-        self, xml_content: str, source_language: str, target_language: str
-    ) -> str:
-        """Extract and translate the display_name attribute of the root element."""
-        try:
-            root = ElementTree.fromstring(xml_content)
-            display_name = root.attrib.get("display_name")
+        return stats
 
-            if display_name:
-                translated_name, _ = self._translate_text(
-                    display_name, source_language, target_language
-                )
-                root.set("display_name", translated_name)
-                return ElementTree.tostring(root, encoding="unicode")
-        except ElementTree.ParseError as e:
-            logger.warning("Could not translate display_name: %s", e)
+    def _add_translation_log_entry(
+        self, source_language, target_language, command_stats=None
+    ) -> None:
+        """
+        Add a log entry for the course translation operation.
 
-        return xml_content
+        Args:
+            source_language: Source language code
+            target_language: Target language code
+            command_stats: List of command statistics/logs
+        """
+        source_course_id = generate_course_key_from_xml(
+            course_dir_path=self.translated_course_dir
+        )
+        command_stats_str = "\n".join(command_stats) if command_stats else ""
+
+        CourseTranslationLog.objects.create(
+            source_course_id=source_course_id,
+            source_course_language=source_language,
+            target_course_language=target_language,
+            srt_provider_name=self.srt_provider_name,
+            srt_provider_model=self.srt_model or "",
+            content_provider_name=self.content_provider_name,
+            content_provider_model=self.content_model or "",
+            command_stats=command_stats_str,
+        )