docs: sync code to deepset workspace

.github/utils/deepset_sync.py

@@ -0,0 +1,194 @@
+# /// script
+# dependencies = [
+#   "requests",
+# ]
+# ///
+"""
+Upload Markdown files to Deepset API with transformed filenames.
+"""
+
+import os
+import sys
+import json
+import argparse
+import requests
+from pathlib import Path
+
+
+def transform_filename(filepath: Path) -> str:
+    """
+    Transform a file path to the required format:
+    - Replace path separators with underscores
+    """
+    # Convert to string and replace path separators with underscores
+    transformed = str(filepath).replace("/", "_").replace("\\", "_")
+
+    return transformed
+
+
+def upload_file_to_deepset(filepath: Path, api_key: str, workspace: str) -> bool:
+    """
+    Upload a single file to Deepset API.
+    """
+    # Read file content
+    try:
+        content = filepath.read_text(encoding="utf-8")
+    except Exception as e:
+        print(f"Error reading file {filepath}: {e}")
+        return False
+
+    # Transform filename
+    transformed_name = transform_filename(filepath)
+
+    # Prepare metadata
+    metadata: dict[str, str] = {"original_file_path": str(filepath)}
+
+    # Prepare API request
+    url = f"https://api.cloud.deepset.ai/api/v1/workspaces/{workspace}/files"
+    params: dict[str, str] = {"file_name": transformed_name, "write_mode": "OVERWRITE"}
+
+    headers: dict[str, str] = {
+        "accept": "application/json",
+        "authorization": f"Bearer {api_key}",
+    }
+
+    # Prepare multipart form data
+    files: dict[str, tuple[None, str, str]] = {
+        "meta": (None, json.dumps(metadata), "application/json"),
+        "text": (None, content, "text/plain"),
+    }
+
+    try:
+        response = requests.post(url, params=params, headers=headers, files=files)
+        response.raise_for_status()
+        print(f"Successfully uploaded: {filepath} as {transformed_name}")
+        return True
+    except requests.exceptions.HTTPError:
+        print(f"Failed to upload {filepath}: HTTP {response.status_code}")
+        print(f"   Response: {response.text}")
+        return False
+    except Exception as e:
+        print(f"Failed to upload {filepath}: {e}")
+        return False
+
+
+def delete_files_from_deepset(
+    filepaths: list[Path], api_key: str, workspace: str
+) -> bool:
+    """
+    Delete multiple files from Deepset API.
+    """
+    if not filepaths:
+        return True
+
+    # Transform filenames
+    transformed_names: list[str] = [transform_filename(fp) for fp in filepaths]
+
+    # Prepare API request
+    url = f"https://api.cloud.deepset.ai/api/v1/workspaces/{workspace}/files"
+
+    headers: dict[str, str] = {
+        "accept": "application/json",
+        "authorization": f"Bearer {api_key}",
+        "content-type": "application/json",
+    }
+
+    data: dict[str, list[str]] = {"names": transformed_names}
+
+    try:
+        response = requests.delete(url, headers=headers, json=data)
+        response.raise_for_status()
+        print(f"Successfully deleted {len(transformed_names)} file(s):")
+        for original, transformed in zip(filepaths, transformed_names):
+            print(f"   - {original} (as {transformed})")
+        return True
+    except requests.exceptions.HTTPError:
+        print(f"Failed to delete files: HTTP {response.status_code}")
+        print(f"   Response: {response.text}")
+        return False
+    except Exception as e:
+        print(f"Failed to delete files: {e}")
+        return False
+
+
+def main() -> None:
+    """
+    Main function to process and upload/delete files.
+    """
+    # Parse command line arguments
+    parser = argparse.ArgumentParser(
+        description="Upload/delete Markdown files to/from Deepset"
+    )
+    parser.add_argument(
+        "--changed", nargs="*", default=[], help="Changed or added files"
+    )
+    parser.add_argument("--deleted", nargs="*", default=[], help="Deleted files")
+    args = parser.parse_args()
+
+    # Get environment variables
+    api_key: str | None = os.environ.get("DEEPSET_API_KEY")
+    workspace: str = os.environ.get("DEEPSET_WORKSPACE")
+
+    if not api_key:
+        print("Error: DEEPSET_API_KEY environment variable not set")
+        sys.exit(1)
+
+    # Process arguments and convert to Path objects
+    changed_files: list[Path] = [Path(f.strip()) for f in args.changed if f.strip()]
+    deleted_files: list[Path] = [Path(f.strip()) for f in args.deleted if f.strip()]
+
+    if not changed_files and not deleted_files:
+        print("No Markdown files to process")
+        sys.exit(0)
+
+    print(f"Processing files in Deepset workspace: {workspace}")
+    print("-" * 50)
+
+    # Track results
+    upload_success: int = 0
+    upload_failed: list[Path] = []
+    delete_success: bool = False
+
+    # Handle deletions first
+    if deleted_files:
+        print(f"\nDeleting {len(deleted_files)} file(s)...")
+        delete_success = delete_files_from_deepset(deleted_files, api_key, workspace)
+
+    # Upload changed/new files
+    if changed_files:
+        print(f"\nUploading {len(changed_files)} file(s)...")
+        for filepath in changed_files:
+            if filepath.exists():
+                if upload_file_to_deepset(filepath, api_key, workspace):
+                    upload_success += 1
+                else:
+                    upload_failed.append(filepath)
+            else:
+                print(f"Skipping non-existent file: {filepath}")
+
+    # Summary
+    print("-" * 50)
+    print("Processing Summary:")
+    if changed_files:
+        print(
+            f"   Uploads - Successful: {upload_success}, Failed: {len(upload_failed)}"
+        )
+    if deleted_files:
+        print(
+            f"   Deletions - {'Successful' if delete_success else 'Failed'}: {len(deleted_files)} file(s)"
+        )
+
+    if upload_failed:
+        print("\nFailed uploads:")
+        for f in upload_failed:
+            print(f"   - {f}")
+
+    # Exit with error if any operation failed
+    if upload_failed or (deleted_files and not delete_success):
+        sys.exit(1)
+
+    print("\nAll operations completed successfully!")
+
+
+if __name__ == "__main__":
+    main()

.github/workflows/sync_code_to_deepset.yml

@@ -0,0 +1,55 @@
+name: Upload Documentation to Deepset
+on:
+  push:
+    branches:
+      - main
+jobs:
+  upload-files:
+    runs-on: ubuntu-latest
+
+    steps:
+    - name: Checkout repository
+      uses: actions/checkout@v4
+      with:
+        fetch-depth: 0  # Fetch all history for proper diff
+
+    - name: Set up Python
+      uses: actions/setup-python@v4
+      with:
+        python-version: '3.12'
+
+    - name: Install uv
+      run: |
+        pip install uv
+
+    - name: Get changed files
+      id: changed-files
+      uses: tj-actions/changed-files@v46
+      with:
+        files: |
+          haystack/**/*.py
+        separator: ' '
+
+    - name: Upload files to Deepset
+      if: steps.changed-files.outputs.any_changed == 'true' || steps.changed-files.outputs.any_deleted == 'true'
+      env:
+        DEEPSET_API_KEY: ${{ secrets.DEEPSET_API_KEY }}
+        DEEPSET_WORKSPACE: haystack-code
+      run: |
+        # Combine added and modified files for upload
+        CHANGED_FILES=""
+        if [ -n "${{ steps.changed-files.outputs.added_files }}" ]; then
+          CHANGED_FILES="${{ steps.changed-files.outputs.added_files }}"
+        fi
+        if [ -n "${{ steps.changed-files.outputs.modified_files }}" ]; then
+          if [ -n "$CHANGED_FILES" ]; then
+            CHANGED_FILES="$CHANGED_FILES ${{ steps.changed-files.outputs.modified_files }}"
+          else
+            CHANGED_FILES="${{ steps.changed-files.outputs.modified_files }}"
+          fi
+        fi
+        
+        # Run the script with changed and deleted files
+        uv run --no-project --no-config --no-cache .github/utils/deepset_sync.py \
+          --changed $CHANGED_FILES \
+          --deleted ${{ steps.changed-files.outputs.deleted_files }}