rasa-pro 3.13.1a18__py3-none-any.whl → 3.13.1a20__py3-none-any.whl

rasa/builder/project_generator.py

@@ -3,15 +3,21 @@
 import json
 import os
 import shutil
+import tarfile
+import tempfile
 from pathlib import Path
 from textwrap import dedent
 from typing import Any, Dict, List, Optional
 
+import aiofiles
+import aiohttp
 import structlog
 
+import rasa.version
 from rasa.builder import config
 from rasa.builder.exceptions import ProjectGenerationError, ValidationError
 from rasa.builder.llm_service import get_skill_generation_messages, llm_service
+from rasa.builder.logging_utils import capture_exception_with_context
 from rasa.builder.models import BotFiles
 from rasa.builder.validation_service import validate_project
 from rasa.cli.scaffold import ProjectTemplateName, create_initial_project
@@ -35,10 +41,11 @@ class ProjectGenerator:
         self.project_folder = Path(project_folder)
         self.project_folder.mkdir(parents=True, exist_ok=True)
 
-    def init_from_template(self, template: ProjectTemplateName) -> None:
+    async def init_from_template(self, template: ProjectTemplateName) -> None:
         """Create the initial project files."""
         self.cleanup()
         create_initial_project(self.project_folder.as_posix(), template)
+        await download_cache_for_template(template, self.project_folder.as_posix())
 
     async def generate_project_with_retries(
         self,
@@ -63,7 +70,7 @@ class ProjectGenerator:
         if max_retries is None:
             max_retries = config.MAX_RETRIES
 
-        self.init_from_template(template)
+        await self.init_from_template(template)
 
         project_data = self._get_bot_data_for_llm()
 
@@ -168,11 +175,9 @@ class ProjectGenerator:
                 domain_path = self.project_folder / "domain"
 
             return TrainingDataImporter.load_from_config(
-                config_path=self.project_folder / "config.yml",
-                domain_path=domain_path,
-                training_data_paths=[
-                    self.project_folder / "data",
-                ],
+                config_path=str(self.project_folder / "config.yml"),
+                domain_path=str(domain_path),
+                training_data_paths=[str(self.project_folder / "data")],
                 args={},
             )
 
@@ -280,3 +285,82 @@ class ProjectGenerator:
                     error=str(e),
                     file_path=file_path,
                 )
+
+
+CACHE_BUCKET_URL = "https://trained-templates.s3.us-east-1.amazonaws.com"
+
+
+async def download_cache_for_template(
+    template: ProjectTemplateName, project_folder: str
+) -> None:
+    # get a temp path for the cache file download
+    temporary_cache_file = tempfile.NamedTemporaryFile(suffix=".tar.gz", delete=False)
+
+    try:
+        url = f"{CACHE_BUCKET_URL}/{rasa.version.__version__}-{template.value}.tar.gz"
+        async with aiohttp.ClientSession() as session:
+            async with session.get(url) as response:
+                response.raise_for_status()
+                async with aiofiles.open(temporary_cache_file.name, "wb") as f:
+                    async for chunk in response.content.iter_chunked(1024 * 1024):
+                        await f.write(chunk)
+
+        # extract the cache to the project folder
+        with tarfile.open(temporary_cache_file.name, "r:gz") as tar:
+            tar.extractall(path=project_folder)
+
+        structlogger.info(
+            "project_generator.download_cache_for_template.success",
+            template=template,
+            event_info=(
+                f"Downloaded cache for template, extracted to {project_folder}."
+            ),
+        )
+    except aiohttp.ClientResponseError as e:
+        if e.status == 403:
+            structlogger.debug(
+                "project_generator.download_cache_for_template.no_cache_found",
+                template=template,
+                event_info=("No cache found for template, continuing without it."),
+            )
+        else:
+            structlogger.debug(
+                "project_generator.download_cache_for_template.response_error",
+                error=str(e),
+                status=e.status,
+                template=template,
+                event_info=(
+                    "Failed to download cache for template, continuing without it."
+                ),
+            )
+            capture_exception_with_context(
+                e,
+                "project_generator.download_cache_for_template.response_error",
+                tags={"template": template.value, "status": str(e.status)},
+            )
+    except Exception as exc:
+        structlogger.debug(
+            "project_generator.download_cache_for_template.unexpected_error",
+            error=str(exc),
+            template=template,
+            event_info=(
+                "Unexpected error when downloading cache for template, "
+                "continuing without it."
+            ),
+        )
+        capture_exception_with_context(
+            exc,
+            "project_generator.download_cache_for_template.unexpected_error",
+            tags={"template": template.value},
+        )
+    finally:
+        # Clean up the temporary file
+        try:
+            Path(temporary_cache_file.name).unlink(missing_ok=True)
+        except Exception as exc:
+            structlogger.debug(
+                "project_generator.download_cache_for_template.cleanup_error",
+                error=str(exc),
+                template=template,
+                event_info=("Failed to cleanup cache for template, ignoring."),
+            )

rasa/builder/scrape_rasa_docs.py

@@ -77,7 +77,7 @@ while to_visit and pages_scraped < MAX_PAGES:
             print("⏭️ Skipped (too short)")
             continue
 
-        file_name = save_as_markdown(text)
+        file_name = save_as_markdown(text, url)
         markdown_to_url[file_name] = url
         pages_scraped += 1