From 92cf702ce49f2066110c589d97b57b0ed3ab8fa4 Mon Sep 17 00:00:00 2001 From: Prins Kumar Date: Mon, 17 Nov 2025 17:07:34 +0530 Subject: [PATCH 01/15] Add S3 Bedrock BDA ingestion support with user confirmation and pymupdf4llm integration --- common/requirements.txt | 3 +- common/utils/image_data_extractor.py | 163 +++--------- common/utils/markdown_parsing.py | 63 +++++ common/utils/text_extractors.py | 254 +++++++++--------- graphrag-ui/src/pages/Setup.tsx | 370 +++++++++++++-------------- 5 files changed, 403 insertions(+), 450 deletions(-) create mode 100644 common/utils/markdown_parsing.py diff --git a/common/requirements.txt b/common/requirements.txt index 562c2f6..f0022f3 100644 --- a/common/requirements.txt +++ b/common/requirements.txt @@ -110,7 +110,8 @@ packaging==24.2 pandas==2.2.3 #pathtools==0.1.2 pillow==11.2.1 -PyMuPDF==1.26.4 +#PyMuPDF==1.26.4 +pymupdf4llm==0.2.0 platformdirs==4.3.8 pluggy==1.6.0 prometheus_client==0.22.1 diff --git a/common/utils/image_data_extractor.py b/common/utils/image_data_extractor.py index bde9c97..74e8d2f 100644 --- a/common/utils/image_data_extractor.py +++ b/common/utils/image_data_extractor.py @@ -11,155 +11,54 @@ logger = logging.getLogger(__name__) - - -def describe_image_with_llm(image_input): +def describe_image_with_llm(file_path): """ - Send image (pixmap or PIL image) to LLM vision model and return description. - Uses multimodal_service from config if available, otherwise falls back to completion_service. - Currently supports: OpenAI, Azure OpenAI, Google GenAI, and Google VertexAI + Read image file and convert to base64 to send to LLM. """ try: + from PIL import Image as PILImage + client = get_multimodal_service() if not client: return "[Image: Failed to create multimodal LLM client]" - + + # Read image and convert to base64 + pil_image = PILImage.open(file_path) buffer = io.BytesIO() - # Convert to RGB if needed for better compatibility - if image_input.mode != 'RGB': - image_input = image_input.convert('RGB') - image_input.save(buffer, format="JPEG", quality=95) - b64_img = base64.b64encode(buffer.getvalue()).decode("utf-8") + if pil_image.mode != 'RGB': + pil_image = pil_image.convert('RGB') + pil_image.save(buffer, format="JPEG", quality=95) + image_base64 = base64.b64encode(buffer.getvalue()).decode('utf-8') - # Build messages (system + human) messages = [ - SystemMessage( - content="You are a helpful assistant that describes images concisely for document analysis." - ), - HumanMessage( - content=[ - { - "type": "text", - "text": ( - "Please describe what you see in this image and " - "if the image has scanned text then extract all the text. " - "if the image has any logo, icon, or branding element, try to describe it with text. " - "Focus on any text, diagrams, charts, or other visual elements." - "If the image is purely a logo, icon, or branding element, start your response with 'LOGO:' or 'ICON:'." - ), - }, - { - "type": "image_url", - "image_url": {"url": f"data:image/jpeg;base64,{b64_img}"}, - }, - ] - ), + SystemMessage( + content="You are a helpful assistant that describes images concisely for document analysis." + ), + HumanMessage( + content=[ + { + "type": "text", + "text": ( + "Please describe what you see in this image and " + "if the image has scanned text then extract all the text. " + "If the image has any graph, chart, table, or other diagram, describe it. " + ), + }, + { + "type": "image_url", + "image_url": {"url": f"data:image/jpeg;base64,{image_base64}"}, + }, + ], + ), ] - # Get response from LangChain LLM client - # Access the underlying LangChain client langchain_client = client.llm response = langchain_client.invoke(messages) - return response.content if hasattr(response, 'content') else str(response) + return response.content if hasattr(response, "content") else str(response) except Exception as e: logger.error(f"Failed to describe image with LLM: {str(e)}") return "[Image: Error processing image description]" -def save_image_and_get_markdown(image_input, context_info="", graphname=None): - """ - Save image locally to static/images/ folder and return markdown reference with description. - - LEGACY/OLD APPROACH: Used for backward compatibility with JSONL-based loading. - Images are saved as files and served via /ui/images/ endpoint with img:// protocol. - - For NEW direct loading approach, images are stored in Image vertex as base64 - and served via /ui/image_vertex/ endpoint with image:// protocol. - - Args: - image_input: PIL Image object - context_info: Optional context (e.g., "page 3 of invoice.pdf") - graphname: Graph name to organize images by graph (optional) - - Returns: - dict with: - - 'markdown': Markdown string with img:// reference - - 'image_id': Unique identifier for the saved image - - 'image_path': Path where image was saved to static/images/ - """ - try: - # FIRST: Get description from LLM to check if it's a logo - description = describe_image_with_llm(image_input) - - # Check if the image is a logo, icon, or decorative element BEFORE saving - # These should be filtered out as they're not content-relevant - description_lower = description.lower() - logo_indicators = ['logo', 'icon', 'branding', 'watermark', 'trademark', 'company logo', 'brand logo'] - - if any(indicator in description_lower for indicator in logo_indicators): - logger.info(f"Detected logo/icon in image, skipping: {description[:100]}") - return None - - # If not a logo, proceed with saving the image - # Generate unique image ID using hash of image content - buffer = io.BytesIO() - if image_input.mode != 'RGB': - image_input = image_input.convert('RGB') - image_input.save(buffer, format="JPEG", quality=95) - image_bytes = buffer.getvalue() - - # Create hash-based ID (deterministic for same image) - image_hash = hashlib.sha256(image_bytes).hexdigest()[:16] - image_id = f"{image_hash}.jpg" - - # Save image to local storage directory organized by graphname - project_root = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__)))) - - # If graphname is provided, organize images by graph - if graphname: - images_dir = os.path.join(project_root, "static", "images", graphname) - # Include graphname in the image reference for URL construction - image_reference = f"{graphname}/{image_id}" - else: - images_dir = os.path.join(project_root, "static", "images") - image_reference = image_id - - os.makedirs(images_dir, exist_ok=True) - - image_path = os.path.join(images_dir, image_id) - - # Save image file (skip if already exists with same hash) - if not os.path.exists(image_path): - with open(image_path, 'wb') as f: - f.write(image_bytes) - logger.info(f"Saved content image to: {image_path}") - else: - logger.debug(f"Image already exists: {image_path}") - - # Generate markdown with custom img:// protocol (will be replaced later) - # Format: ![description](img://graphname/image_id) or ![description](img://image_id) - markdown = f"![{description}](img://{image_reference})" - - logger.info(f"Created image reference: {image_reference} with description") - - return { - 'markdown': markdown, - 'image_id': image_reference, - 'image_path': image_path, - 'description': description - } - - except Exception as e: - logger.error(f"Failed to save image and generate markdown: {str(e)}") - # Fallback to text description only - fallback_desc = f"[Image: {context_info} - processing failed]" - return { - 'markdown': fallback_desc, - 'image_id': None, - 'image_path': None, - 'description': fallback_desc - } - - diff --git a/common/utils/markdown_parsing.py b/common/utils/markdown_parsing.py new file mode 100644 index 0000000..7c8c476 --- /dev/null +++ b/common/utils/markdown_parsing.py @@ -0,0 +1,63 @@ +import re +import os +import pymupdf4llm + +class MarkdownProcessor: + """ + A helper class to extract markdown image entries and + update descriptions based on image_id. + """ + + # regex for markdown images: ![alt](path) + _pattern = re.compile(r'!\[([^\]]*)\]\(([^)\s]+)\)') + + @classmethod + def extract_images(cls, md_text): + """ + Returns list of {"path": path, "image_id": image_id} + image_id = basename without extension + """ + images = [] + for m in cls._pattern.finditer(md_text): + path = m.group(2) + basename = os.path.basename(path) + image_id = os.path.splitext(basename)[0] + images.append({"path": path, "image_id": image_id}) + return images + + @classmethod + def insert_description_by_id(cls, md_text, image_id, description): + """ + Replace the description for an image whose basename == image_id. + """ + + def repl(m): + old_path = m.group(2) + candidate_id = os.path.splitext(os.path.basename(old_path))[0] + + if candidate_id == image_id: + # Insert new description + return f'![{description}]({old_path})' + + return m.group(0) + + return cls._pattern.sub(repl, md_text) + + @classmethod + def replace_path_with_tg_protocol(cls, md_text, image_id, tg_reference): + """ + Replace the file path for an image whose basename == image_id with tg:// protocol reference. + tg_reference should be like 'Graphs_image_1' + """ + def repl(m): + old_path = m.group(2) + candidate_id = os.path.splitext(os.path.basename(old_path))[0] + + if candidate_id == image_id: + # Replace path with tg:// protocol reference + alt_text = m.group(1) + return f'![{alt_text}](tg://{tg_reference})' + + return m.group(0) + + return cls._pattern.sub(repl, md_text) \ No newline at end of file diff --git a/common/utils/text_extractors.py b/common/utils/text_extractors.py index da3e22d..b900cae 100644 --- a/common/utils/text_extractors.py +++ b/common/utils/text_extractors.py @@ -183,137 +183,154 @@ def extract_text_from_file_with_images_as_docs(file_path, graphname=None): def _extract_pdf_with_images_as_docs(file_path, base_doc_id, graphname=None): """ - Extract PDF as ONE markdown document with inline image references. + Extract PDF as ONE markdown document with inline image references using pymupdf4llm. + Uses unique temporary folder per PDF to allow parallel processing. + After processing, delete the extracted image folder. """ + # Use unique folder per PDF to allow parallel processing without conflicts + unique_folder_id = uuid.uuid4().hex[:12] + image_output_folder = Path(f"tg_temp_{unique_folder_id}") + try: - import fitz # PyMuPDF + import pymupdf4llm from PIL import Image as PILImage + from common.utils.image_data_extractor import describe_image_with_llm + from common.utils.markdown_parsing import MarkdownProcessor + + # Ensure clean slate - remove folder if it exists from failed previous run + if image_output_folder.exists(): + shutil.rmtree(image_output_folder, ignore_errors=True) + + # Convert PDF to markdown with extracted image files + try: + markdown_content = pymupdf4llm.to_markdown( + file_path, + write_images=True, + image_path=str(image_output_folder), # unique folder per PDF + force_text=False, + margins=0, + image_size_limit=0.08, + ) + except Exception as e: + logger.error(f"pymupdf4llm failed for {file_path}: {e}") + # Cleanup folder if it was created + if image_output_folder.exists(): + shutil.rmtree(image_output_folder, ignore_errors=True) + return [{ + "doc_id": base_doc_id, + "doc_type": "markdown", + "content": f"[PDF extraction failed: {e}]", + "position": 0 + }] + + if not markdown_content or not markdown_content.strip(): + logger.warning(f"No content extracted from PDF: {file_path}") + + # Extract image references from markdown + image_refs = MarkdownProcessor.extract_images(markdown_content) + + if not image_refs: + # cleanup folder anyway + if image_output_folder.exists(): + shutil.rmtree(image_output_folder, ignore_errors=True) + + return [{ + "doc_id": base_doc_id, + "doc_type": "markdown", + "content": markdown_content, + "position": 0 + }] - doc = fitz.open(file_path) - markdown_parts = [] image_entries = [] image_counter = 0 - for page_num, page in enumerate(doc, start=1): - if page_num > 1: - markdown_parts.append("\n\n") - markdown_parts.append(f"--- Page {page_num} ---\n") #Avoid to be splitted as a single chunk - - blocks = page.get_text("blocks", sort=True) - text_blocks_with_pos = [] - - for block in blocks: - block_type = block[6] if len(block) > 6 else 0 - if block_type == 0: - text = block[4].strip() - if text: - y_pos = block[1] - text_blocks_with_pos.append({'type': 'text', 'content': text, 'y_pos': y_pos}) - - image_list = page.get_images(full=True) - images_with_pos = [] - - if image_list: - for img_index, img_info in enumerate(image_list): - try: - xref = img_info[0] - base_image = doc.extract_image(xref) - image_bytes = base_image["image"] - image_ext = base_image["ext"] - - img_rects = page.get_image_rects(xref) - y_pos = img_rects[0].y0 if img_rects else 999999 - - pil_image = PILImage.open(io.BytesIO(image_bytes)) - if pil_image.width < 100 or pil_image.height < 100: - continue - - from common.utils.image_data_extractor import describe_image_with_llm - description = describe_image_with_llm(pil_image) - description_lower = description.lower() - logo_indicators = [ - 'logo:', 'icon:', 'logo', 'icon', 'branding', - 'watermark', 'trademark', 'stylized letter', - 'stylized text', 'word "', "word '" - ] - if any(indicator in description_lower for indicator in logo_indicators): - continue - - buffer = io.BytesIO() - if pil_image.mode != 'RGB': - pil_image = pil_image.convert('RGB') - pil_image.save(buffer, format="JPEG", quality=95) - image_base64 = base64.b64encode(buffer.getvalue()).decode('utf-8') - - image_counter += 1 - image_doc_id = f"{base_doc_id}_image_{image_counter}" - - images_with_pos.append({ - 'type': 'image', - 'image_doc_id': image_doc_id, - 'description': description, - 'y_pos': y_pos, - 'image_data': image_base64, - 'image_format': image_ext, - 'width': pil_image.width, - 'height': pil_image.height - }) - except Exception as img_error: - logger.warning(f"Failed to extract image on page {page_num}: {img_error}") - - all_elements = text_blocks_with_pos + images_with_pos - all_elements.sort(key=lambda x: x['y_pos']) - - for element in all_elements: - if element['type'] == 'text': - markdown_parts.append(element['content']) - markdown_parts.append("\n\n") - else: - # Add image description as text, then markdown image reference - # Use short alt text in markdown, full description as regular text - markdown_parts.append(f"![{element['description']}](tg://{element['image_doc_id']})\n\n") - - image_entries.append({ - "doc_id": element['image_doc_id'], - "doc_type": "image", - "image_description": element['description'], - "image_data": element['image_data'], - "image_format": element['image_format'], - "parent_doc": base_doc_id, - "page_number": page_num, - "width": element['width'], - "height": element['height'], - "position": int(element['image_doc_id'].split('_')[-1]) - }) - - doc.close() - - markdown_content = "".join(markdown_parts) if markdown_parts else "" #No content extracted from PDF - if not markdown_content: - return [] + for img_ref in image_refs: + try: + img_path = Path(img_ref["path"]) # convert to Path + image_id = img_ref["image_id"] + + # Image description + description = describe_image_with_llm(str(img_path)) + + markdown_content = MarkdownProcessor.insert_description_by_id( + markdown_content, + image_id, + description + ) + + # Convert image to base64 + pil_image = PILImage.open(img_path) + buffer = io.BytesIO() + + if pil_image.mode != "RGB": + pil_image = pil_image.convert("RGB") + + pil_image.save(buffer, format="JPEG", quality=95) + image_base64 = base64.b64encode(buffer.getvalue()).decode("utf-8") + + image_counter += 1 + image_doc_id = f"{base_doc_id}_image_{image_counter}" + + # Replace file path with tg:// protocol reference in markdown + markdown_content = MarkdownProcessor.replace_path_with_tg_protocol( + markdown_content, + image_id, + image_doc_id + ) + + image_entries.append({ + "doc_id": image_doc_id, + "doc_type": "image", + "image_description": description, + "image_data": image_base64, + "image_format": "jpg", + "parent_doc": base_doc_id, + "page_number": 0, + "width": pil_image.width, + "height": pil_image.height, + "position": image_counter + }) + + except Exception as img_error: + logger.warning(f"Failed to process image {img_ref.get('path')}: {img_error}") + + # FINAL CLEANUP — delete folder after processing everything + if image_output_folder.exists() and image_output_folder.is_dir(): + try: + shutil.rmtree(image_output_folder) + logger.debug(f"Deleted image folder: {image_output_folder}") + except Exception as delete_err: + logger.warning(f"Failed to delete folder {image_output_folder}: {delete_err}") + # Build final result result = [{ "doc_id": base_doc_id, - "doc_type": "", + "doc_type": "markdown", "content": markdown_content, "position": 0 }] result.extend(image_entries) + return result - except ImportError: - logger.error("PyMuPDF not available") + except ImportError as import_err: + logger.error(f"Required library missing: {import_err}") + # Cleanup on import error + if image_output_folder.exists(): + shutil.rmtree(image_output_folder, ignore_errors=True) return [{ "doc_id": base_doc_id, - "doc_type": "", - "content": "[PDF extraction requires PyMuPDF]", + "doc_type": "markdown", + "content": "[PDF extraction requires pymupdf4llm and PyMuPDF]", "position": 0 }] except Exception as e: logger.error(f"Error extracting PDF: {e}") + # Cleanup on any other error + if image_output_folder.exists(): + shutil.rmtree(image_output_folder, ignore_errors=True) raise - def _extract_standalone_image_as_doc(file_path, base_doc_id, graphname=None): """ Extract standalone image file as ONE markdown document with inline image reference. @@ -324,25 +341,15 @@ def _extract_standalone_image_as_doc(file_path, base_doc_id, graphname=None): pil_image = PILImage.open(file_path) if pil_image.width < 100 or pil_image.height < 100: - return [{ - "doc_id": base_doc_id, - "doc_type": "", - "content": f"[Skipped small image: {file_path.name}]", - "position": 0 - }] + pass - description = describe_image_with_llm(pil_image) + description = describe_image_with_llm(str(Path(file_path).absolute())) description_lower = description.lower() logo_indicators = ['logo:', 'icon:', 'logo', 'icon', 'branding', 'watermark', 'trademark', 'stylized letter', 'stylized text', 'word "', "word '"] if any(indicator in description_lower for indicator in logo_indicators): - return [{ - "doc_id": base_doc_id, - "doc_type": "", - "content": f"[Skipped logo/icon: {file_path.name}]", - "position": 0 - }] + return [] buffer = io.BytesIO() if pil_image.mode != 'RGB': @@ -353,7 +360,6 @@ def _extract_standalone_image_as_doc(file_path, base_doc_id, graphname=None): image_id = f"{base_doc_id}_image_1" # Put description as text, then markdown image reference with short alt text content = f"![{description}](tg://{image_id})" - return [ { "doc_id": base_doc_id, @@ -379,7 +385,7 @@ def _extract_standalone_image_as_doc(file_path, base_doc_id, graphname=None): logger.error(f"Error extracting image: {e}") return [{ "doc_id": base_doc_id, - "doc_type": "", + "doc_type": "markdown", "content": f"[Image extraction failed: {str(e)}]", "position": 0 }] @@ -441,12 +447,10 @@ def get_doc_type_from_extension(extension): if extension in ['.html', '.htm']: return 'html' - elif extension in ['.md']: - return 'markdown' elif extension in ['.jpg', '.jpeg', '.png', '.gif', '.bmp', '.tiff', '.webp']: return 'image' else: - return '' + return 'markdown' def get_supported_extensions(): diff --git a/graphrag-ui/src/pages/Setup.tsx b/graphrag-ui/src/pages/Setup.tsx index b7d357d..2aaee99 100644 --- a/graphrag-ui/src/pages/Setup.tsx +++ b/graphrag-ui/src/pages/Setup.tsx @@ -2,7 +2,7 @@ import React, { useState, useEffect } from "react"; import { useNavigate } from "react-router-dom"; import { Button } from "@/components/ui/button"; import { Input } from "@/components/ui/input"; -import { Database, Upload, RefreshCw, Loader2, Trash2, FolderUp, Cloud, ArrowLeft, CloudDownload, CloudLightning } from "lucide-react"; +import { Database, Upload, RefreshCw, Loader2, Trash2, FolderUp, Cloud, ArrowLeft, CloudDownload, CloudCog } from "lucide-react"; import { Dialog, DialogContent, @@ -56,7 +56,6 @@ const Setup = () => { const [uploadMessage, setUploadMessage] = useState(""); const [isIngesting, setIsIngesting] = useState(false); const [ingestMessage, setIngestMessage] = useState(""); - const [activeTab, setActiveTab] = useState("upload"); // Refresh state const [refreshOpen, setRefreshOpen] = useState(false); @@ -67,12 +66,13 @@ const Setup = () => { const [isCheckingStatus, setIsCheckingStatus] = useState(false); // S3 state + const [fileFormat, setFileFormat] = useState<"json" | "multi">("json"); const [awsAccessKey, setAwsAccessKey] = useState(""); const [awsSecretKey, setAwsSecretKey] = useState(""); + const [dataPath, setDataPath] = useState(""); const [inputBucket, setInputBucket] = useState(""); const [outputBucket, setOutputBucket] = useState(""); const [regionName, setRegionName] = useState(""); - const [skipBDAProcessing, setSkipBDAProcessing] = useState(false); // Cloud Download state const [cloudProvider, setCloudProvider] = useState<"s3" | "gcs" | "azure">("s3"); @@ -458,7 +458,7 @@ const Setup = () => { } const createData = await createResponse.json(); - //console.log("Create ingest response:", createData); + console.log("Create ingest response:", createData); // Step 2: Run ingest setIngestMessage("Step 2/2: Running document ingest..."); @@ -484,7 +484,7 @@ const Setup = () => { } const ingestData = await ingestResponse.json(); - //console.log("Ingest response:", ingestData); + console.log("Ingest response:", ingestData); setIngestMessage(`✅ Data ingested successfully! Processed documents from ${folderPath}/`); } catch (error: any) { @@ -495,8 +495,8 @@ const Setup = () => { } }; - // Ingest files from S3 with Amazon BDA - const handleAmazonBDAIngest = async () => { + // Ingest files from S3 with Bedrock BDA + const handleS3BedrockIngest = async () => { if (!ingestGraphName) { setIngestMessage("Please select a graph"); return; @@ -508,112 +508,92 @@ const Setup = () => { return; } - if (skipBDAProcessing) { - // When skipping BDA, only output bucket and region are required - if (!outputBucket || !regionName) { - setIngestMessage("❌ Please provide Output Bucket and Region Name"); - return; - } - } else { - // When using BDA, all fields are required + if (fileFormat === "multi") { if (!inputBucket || !outputBucket || !regionName) { setIngestMessage("❌ Please provide Input Bucket, Output Bucket, and Region Name"); return; } - } - // Ask for confirmation - const confirmMessage = skipBDAProcessing - ? `You're skipping Amazon BDA processing and will ingest directly from the output bucket (${outputBucket}). Please confirm to proceed.` - : `You're using Amazon BDA for multimodal document processing. This will trigger Amazon BDA to process your documents from the input bucket (${inputBucket}) and store the results in the output bucket (${outputBucket}) and then ingest them into your knowledge graph. Please confirm to proceed.`; - - const shouldProceed = await confirm(confirmMessage); - if (!shouldProceed) { - setIngestMessage("Operation cancelled by user."); - return; + // Ask for confirmation if using Bedrock (multi format) + const shouldProceed = await confirm( + `Are you using AWS Bedrock for multimodal document processing? This will trigger AWS Bedrock BDA to process your documents from the input bucket (${inputBucket}) and store the results in the output bucket (${outputBucket}).` + ); + if (!shouldProceed) { + setIngestMessage("Operation cancelled by user."); + return; + } + } else if (fileFormat === "json") { + if (!dataPath) { + setIngestMessage("❌ Please provide Data Path (e.g., s3://bucket-name/path/to/data)"); + return; + } } setIsIngesting(true); + setIngestMessage("Step 1/2: Creating ingest job..."); try { const creds = localStorage.getItem("creds"); - let loadingInfo: any = {}; - if (skipBDAProcessing) { - // Skip BDA processing - create ingest job that reads directly from output bucket - const runIngestConfig: any = { - data_source: "bda", + // Step 1: Create ingest job + const createIngestConfig: any = { + data_source: "s3", + data_source_config: { aws_access_key: awsAccessKey, aws_secret_key: awsSecretKey, - output_bucket: outputBucket, - region_name: regionName, - bda_jobs:[], - loader_config: { - doc_id_field: "doc_id", - content_field: "content", - doc_type: "markdown", - }, - file_format: "multi" - }; - - setIngestMessage("Step 1/2: Creating ingest job from output bucket..."); - - // Run ingest directly - loadingInfo = { - load_job_id: "load_documents_content_json", - data_source_id: runIngestConfig, - file_path: outputBucket, - }; - setIngestMessage(`Step 2/2: Running document ingestion for all files in ${outputBucket}...`); - } else { - // Step 1: Create ingest job with BDA processing - const createIngestConfig: any = { - data_source: "bda", - data_source_config: { - aws_access_key: awsAccessKey, - aws_secret_key: awsSecretKey, - input_bucket: inputBucket, - output_bucket: outputBucket, - region_name: regionName, - }, - loader_config: { - doc_id_field: "doc_id", - content_field: "content", - doc_type: "markdown", - }, - file_format: "multi" - }; + }, + loader_config: { + doc_id_field: "doc_id", + content_field: "content", + doc_type: fileFormat === "multi" ? "markdown" : "", + }, + file_format: fileFormat + }; - setIngestMessage("Step 1/2: Triggering Amazon BDA processing and creating ingest job..."); + // Add format-specific configuration + if (fileFormat === "multi") { + createIngestConfig.data_source_config.input_bucket = inputBucket; + createIngestConfig.data_source_config.output_bucket = outputBucket; + createIngestConfig.data_source_config.region_name = regionName; + setIngestMessage("Step 1/2: Creating ingest job and triggering AWS Bedrock BDA processing..."); + } else if (fileFormat === "json") { + createIngestConfig.loader_config.doc_id_field = "url"; + } - const createResponse = await fetch(`/ui/${ingestGraphName}/create_ingest`, { - method: "POST", - headers: { - "Content-Type": "application/json", - Authorization: `Basic ${creds}`, - }, - body: JSON.stringify(createIngestConfig), - }); + const createResponse = await fetch(`/ui/${ingestGraphName}/create_ingest`, { + method: "POST", + headers: { + "Content-Type": "application/json", + Authorization: `Basic ${creds}`, + }, + body: JSON.stringify(createIngestConfig), + }); - if (!createResponse.ok) { - const errorData = await createResponse.json(); - throw new Error(errorData.detail || `Failed to create ingest job: ${createResponse.statusText}`); - } + if (!createResponse.ok) { + const errorData = await createResponse.json(); + throw new Error(errorData.detail || `Failed to create ingest job: ${createResponse.statusText}`); + } - const createData = await createResponse.json(); - //console.log("Create ingest response:", createData); + const createData = await createResponse.json(); + console.log("Create ingest response:", createData); - // Step 2: Run ingest - loadingInfo = { - load_job_id: createData.load_job_id, - data_source_id: createData.data_source_id, - file_path: outputBucket, - }; + // Step 2: Run ingest + setIngestMessage("Step 2/2: Running document ingest..."); - const filesToIngest = createData.data_source_id.bda_jobs.map((job: any) => job.jobId.split("/")[-1]); - setIngestMessage(`Step 2/2: Running document ingest for ${filesToIngest.length} files in ${outputBucket}...`); + // Determine file path based on format + let filePath = ""; + if (fileFormat === "multi") { + filePath = outputBucket; // For multi format, use output bucket + } else if (fileFormat === "json") { + filePath = dataPath; // For json format, use the provided data path } + const loadingInfo = { + load_job_id: createData.load_job_id, + data_source_id: createData.data_source_id, + file_path: filePath, + }; + const ingestResponse = await fetch(`/ui/${ingestGraphName}/ingest`, { method: "POST", headers: { @@ -629,13 +609,15 @@ const Setup = () => { } const ingestData = await ingestResponse.json(); - //console.log("Ingest response:", ingestData); - const filesIngested = ingestData.summary.map((file: any) => file.file_path); - - setIngestMessage(`✅ Document ingestion completed successfully! Ingested ${filesIngested.length} into your knowledge graph.`); + console.log("Ingest response:", ingestData); + if (fileFormat === "multi") { + setIngestMessage(`✅ Data ingested successfully! AWS Bedrock BDA processed documents from ${inputBucket} and loaded results from ${outputBucket}.`); + } else { + setIngestMessage(`✅ Data ingested successfully! Processed documents from ${dataPath}.`); + } } catch (error: any) { - console.error("Error ingesting files:", error); + console.error("Error ingesting S3 data:", error); setIngestMessage(`❌ Error: ${error.message}`); } finally { setIsIngesting(false); @@ -1121,8 +1103,8 @@ const Setup = () => { - + @@ -1139,35 +1121,32 @@ const Setup = () => { )} + {ingestGraphName && ( +

+ Files will be uploaded to: uploads/{ingestGraphName}/ +

+ )} - { - // Block tab switching when ingesting - if (!isIngesting) { - setActiveTab(value); - } - }} className="w-full"> + - + Upload Files - + Download from Cloud - - - Use Amazon BDA + + + Amazon BDA Configuration {/* Upload Data Tab */}
-

- Upload local files to the server and ingest them into your knowledge graph. -

@@ -1295,9 +1274,6 @@ const Setup = () => { {/* Download from Cloud Storage Tab */}
-

- Download files from cloud storage and ingest them into your knowledge graph. -

)} - {ingestGraphName && ( -

- Download destination: downloaded_files_cloud/{ingestGraphName}/ -

- )}
+

+ Files will be downloaded to: downloaded_files_cloud/{ingestGraphName}/ +

- {/* Amazon BDA Configuration Tab */} - -
-

- Process multimodal documents stored in S3 with Amazon Bedrock Data Automation and ingest them into your knowledge graph. -

+ {/* S3 Bedrock Configuration Tab */} + +
+
+ + +
{/* Common fields */}
@@ -1625,7 +1610,6 @@ const Setup = () => { onChange={(e) => setAwsAccessKey(e.target.value)} placeholder="Enter AWS access key" className="dark:border-[#3D3D3D] dark:bg-shadeA" - disabled={isIngesting} />
@@ -1639,74 +1623,76 @@ const Setup = () => { onChange={(e) => setAwsSecretKey(e.target.value)} placeholder="Enter AWS secret key" className="dark:border-[#3D3D3D] dark:bg-shadeA" - disabled={isIngesting} />
-
-
- -
- -
- - setOutputBucket(e.target.value)} - placeholder="Enter output bucket name" - className="dark:border-[#3D3D3D] dark:bg-shadeA" - disabled={isIngesting} - /> -
+ ) : ( + <> +
+ + setInputBucket(e.target.value)} + placeholder="Enter input bucket name" + className="dark:border-[#3D3D3D] dark:bg-shadeA" + /> +
-
- - setRegionName(e.target.value)} - placeholder="e.g., us-east-1" - className="dark:border-[#3D3D3D] dark:bg-shadeA" - disabled={isIngesting} - /> -
+
+ + setOutputBucket(e.target.value)} + placeholder="Enter output bucket name" + className="dark:border-[#3D3D3D] dark:bg-shadeA" + /> +
- {ingestGraphName && ( -

- Processing destination: Input bucket ({inputBucket || "not specified"}) → Output bucket ({outputBucket || "not specified"}) → Knowledge graph ({ingestGraphName}) -

+
+ + setRegionName(e.target.value)} + placeholder="e.g., us-east-1" + className="dark:border-[#3D3D3D] dark:bg-shadeA" + /> +
+ )} - {/* Ingest S3 Files with Amazon BDA Section */} + {/* Ingest S3 Bedrock Data Section */}
+

+ Ingest S3 Data into Knowledge Graph +

+

+ Process S3 data and add it to the knowledge graph using AWS Bedrock BDA for multimodal documents +

@@ -1771,7 +1757,7 @@ const Setup = () => { Refresh Knowledge Graph - Rebuild the graph content and rerun community detection for your knowledge graph + Rebuild the graph content of your knowledge graph @@ -1780,8 +1766,8 @@ const Setup = () => { - + @@ -1805,7 +1791,7 @@ const Setup = () => { ⚠️ Warning

- This operation will process new documents and rerun community detection that will interrupt related queries. + This operation will rebuild the graph content that will interrupt related queries. Please confirm to proceed.

From b752b8606a85c84273c491df3a8d4d2174fe366c Mon Sep 17 00:00:00 2001 From: Prins Kumar Date: Tue, 18 Nov 2025 18:27:24 +0530 Subject: [PATCH 02/15] Update README for OpenAI and Bedrock config, add pymupdf4llm license --- README.md | 78 ++-- licenses/pymupdf4llm-AGPL-3.0.txt | 661 ++++++++++++++++++++++++++++++ 2 files changed, 704 insertions(+), 35 deletions(-) create mode 100644 licenses/pymupdf4llm-AGPL-3.0.txt diff --git a/README.md b/README.md index 8c38f6c..13c88b3 100644 --- a/README.md +++ b/README.md @@ -103,24 +103,23 @@ Organizing the data as a knowledge graph allows a chatbot to access accurate, fa ### Quick Start #### Use TigerGraph Docker-Based Instance -Set your LLM Provider (supported `openai` or `gemini`) api key as environment varabiel LLM_API_KEY and use the following command for a one-step quick deployment with TigerGraph Community Edition and default configurations: +Set your OpenAI api key as environment varabiel OPENAI_API_KEY and use the following command for a one-step quick deployment with TigerGraph Community Edition and default configurations: ``` -curl -k https://raw.githubusercontent.com/tigergraph/graphrag/refs/heads/main/docs/tutorials/setup_graphrag.sh | bash +curl -k https://raw.githubusercontent.com/tigergraph/graphrag/refs/heads/main/docs/tutorials/setup_graphrag.sh | sh ``` The GraphRAG instances will be deployed at `./graphrag` folder and TigerGraph instance will be available at `http://localhost:14240`. -To change installation folder, use `bash -s -- ` instead of `bash` at the end of the above command. - -> Note: for other LLM providers, manually update `configs/server_config.json` accordingly and re-run `docker compose up -d` +To change installation folder, use `sh -s -- ` instead of `sh` at the end of the above command. #### Use Pre-Installed TigerGraph Instance -Similar to the above setup, and use the following command for a one-step quick deployment connecting to a pre-installed TigerGraph with default configurations: + +Using the following command for a one-step quick deployment with TigerGraph Community Edition and default configurations: ``` -curl -k https://raw.githubusercontent.com/tigergraph/graphrag/refs/heads/main/docs/tutorials/setup_graphrag_tg.sh | bash +curl -k https://raw.githubusercontent.com/tigergraph/graphrag/refs/heads/main/docs/tutorials/setup_graphrag_tg.sh | sh ``` The GraphRAG instances will be deployed at `./graphrag` folder and connect to TigerGraph instance at `http://localhost:14240` by default. -To change installation folder, TigerGraph instance location or username/password, use `bash -s -- ` instead of `bash` at the end of the above command. +To change installation folder, TigerGraph instance location or username/password, use `sh -s -- ` instead of `sh` at the end of the above command. [Go back to top](#top) @@ -152,7 +151,7 @@ Here’s what the folder structure looks like: ##### Step 3: Adjust configurations -Edit `llm_config` section of `configs/server_config.json` and replace `` to your own LLM_API_KEY for the LLM provider. +Edit `llm_config` section of `configs/server_config.json` and replace `` to your own OPENAI_API_KEY. > If desired, you can also change the model to be used for the embedding service and completion service to your preferred models to adjust the output from the LLM service. @@ -470,23 +469,27 @@ In addition to the `OPENAI_API_KEY`, `llm_model` and `model_name` can be edited ```json { "llm_config": { + "authentication_configuration": { + "OPENAI_API_KEY": "YOUR_OPENAI_API_KEY_HERE" + }, "embedding_service": { - "embedding_model_service": "openai", "model_name": "text-embedding-3-small", - "authentication_configuration": { - "OPENAI_API_KEY": "YOUR_OPENAI_API_KEY_HERE" - } + "embedding_model_service": "openai" }, "completion_service": { "llm_service": "openai", "llm_model": "gpt-4.1-mini", - "authentication_configuration": { - "OPENAI_API_KEY": "YOUR_OPENAI_API_KEY_HERE" - }, "model_kwargs": { "temperature": 0 }, "prompt_path": "./common/prompts/openai_gpt4/" + }, + "multimodal_service": { + "llm_service": "openai", + "llm_model": "gpt-4o-mini", + "model_kwargs": { + "temperature": 0 + } } } } @@ -546,7 +549,7 @@ And your JSON config should follow as: "model_kwargs": { "temperature": 0 }, - "prompt_path": "./common/prompts/gcp_vertexai_palm/" + "prompt_path": "./app/prompts/gcp_vertexai_palm/" } } } @@ -583,7 +586,7 @@ In addition to the `AZURE_OPENAI_ENDPOINT`, `AZURE_OPENAI_API_KEY`, and `azure_d "model_kwargs": { "temperature": 0 }, - "prompt_path": "./common/prompts/azure_open_ai_gpt35_turbo_instruct/" + "prompt_path": "./app/prompts/azure_open_ai_gpt35_turbo_instruct/" } } } @@ -594,27 +597,32 @@ In addition to the `AZURE_OPENAI_ENDPOINT`, `AZURE_OPENAI_API_KEY`, and `azure_d ```json { "llm_config": { + "authentication_configuration": { + "AWS_ACCESS_KEY_ID": "YOUR_AWS_ACCESS_KEY", + "AWS_SECRET_ACCESS_KEY": "YOUR_AWS_SECRET_KEY", + "AWS_REGION_NAME": "us-west-2" + }, "embedding_service": { + "model_name": "amazon.titan-embed-text-v1", "embedding_model_service": "bedrock", - "model_name":"amazon.titan-embed-text-v2", - "region_name":"us-west-2", - "authentication_configuration": { - "AWS_ACCESS_KEY_ID": "ACCESS_KEY", - "AWS_SECRET_ACCESS_KEY": "SECRET" - } + "dimensions": 1536 }, "completion_service": { "llm_service": "bedrock", - "llm_model": "us.anthropic.claude-3-7-sonnet-20250219-v1:0", - "region_name":"us-west-2", - "authentication_configuration": { - "AWS_ACCESS_KEY_ID": "ACCESS_KEY", - "AWS_SECRET_ACCESS_KEY": "SECRET" - }, + "llm_model": "anthropic.claude-3-5-sonnet-20240620-v1:0", "model_kwargs": { "temperature": 0, + "max_tokens": 4096 }, - "prompt_path": "./common/prompts/aws_bedrock_claude3haiku/" + "prompt_path": "./common/prompts/openai_gpt4/" + }, + "multimodal_service": { + "llm_service": "bedrock", + "llm_model": "anthropic.claude-3-5-sonnet-20240620-v1:0", + "model_kwargs": { + "temperature": 0, + "max_tokens": 4096 + } } } } @@ -640,7 +648,7 @@ In addition to the `AZURE_OPENAI_ENDPOINT`, `AZURE_OPENAI_API_KEY`, and `azure_d "model_kwargs": { "temperature": 0.0000001 }, - "prompt_path": "./common/prompts/openai_gpt4/" + "prompt_path": "./app/prompts/openai_gpt4/" } } } @@ -670,7 +678,7 @@ Example configuration for a model on Hugging Face with a dedicated endpoint is s "model_kwargs": { "temperature": 0.1 }, - "prompt_path": "./common/prompts/openai_gpt4/" + "prompt_path": "./app/prompts/openai_gpt4/" } } } @@ -697,7 +705,7 @@ Example configuration for a model on Hugging Face with a serverless endpoint is "model_kwargs": { "temperature": 0.1 }, - "prompt_path": "./common/prompts/llama_70b/" + "prompt_path": "./app/prompts/llama_70b/" } } } @@ -724,7 +732,7 @@ Example configuration for a model on Hugging Face with a serverless endpoint is "model_kwargs": { "temperature": 0.1 }, - "prompt_path": "./common/prompts/openai_gpt4/" + "prompt_path": "./app/prompts/openai_gpt4/" } } } diff --git a/licenses/pymupdf4llm-AGPL-3.0.txt b/licenses/pymupdf4llm-AGPL-3.0.txt new file mode 100644 index 0000000..0ad25db --- /dev/null +++ b/licenses/pymupdf4llm-AGPL-3.0.txt @@ -0,0 +1,661 @@ + GNU AFFERO GENERAL PUBLIC LICENSE + Version 3, 19 November 2007 + + Copyright (C) 2007 Free Software Foundation, Inc. + Everyone is permitted to copy and distribute verbatim copies + of this license document, but changing it is not allowed. + + Preamble + + The GNU Affero General Public License is a free, copyleft license for +software and other kinds of works, specifically designed to ensure +cooperation with the community in the case of network server software. + + The licenses for most software and other practical works are designed +to take away your freedom to share and change the works. By contrast, +our General Public Licenses are intended to guarantee your freedom to +share and change all versions of a program--to make sure it remains free +software for all its users. + + When we speak of free software, we are referring to freedom, not +price. Our General Public Licenses are designed to make sure that you +have the freedom to distribute copies of free software (and charge for +them if you wish), that you receive source code or can get it if you +want it, that you can change the software or use pieces of it in new +free programs, and that you know you can do these things. + + Developers that use our General Public Licenses protect your rights +with two steps: (1) assert copyright on the software, and (2) offer +you this License which gives you legal permission to copy, distribute +and/or modify the software. + + A secondary benefit of defending all users' freedom is that +improvements made in alternate versions of the program, if they +receive widespread use, become available for other developers to +incorporate. Many developers of free software are heartened and +encouraged by the resulting cooperation. However, in the case of +software used on network servers, this result may fail to come about. +The GNU General Public License permits making a modified version and +letting the public access it on a server without ever releasing its +source code to the public. + + The GNU Affero General Public License is designed specifically to +ensure that, in such cases, the modified source code becomes available +to the community. It requires the operator of a network server to +provide the source code of the modified version running there to the +users of that server. Therefore, public use of a modified version, on +a publicly accessible server, gives the public access to the source +code of the modified version. + + An older license, called the Affero General Public License and +published by Affero, was designed to accomplish similar goals. This is +a different license, not a version of the Affero GPL, but Affero has +released a new version of the Affero GPL which permits relicensing under +this license. + + The precise terms and conditions for copying, distribution and +modification follow. + + TERMS AND CONDITIONS + + 0. Definitions. + + "This License" refers to version 3 of the GNU Affero General Public License. + + "Copyright" also means copyright-like laws that apply to other kinds of +works, such as semiconductor masks. + + "The Program" refers to any copyrightable work licensed under this +License. Each licensee is addressed as "you". "Licensees" and +"recipients" may be individuals or organizations. + + To "modify" a work means to copy from or adapt all or part of the work +in a fashion requiring copyright permission, other than the making of an +exact copy. The resulting work is called a "modified version" of the +earlier work or a work "based on" the earlier work. + + A "covered work" means either the unmodified Program or a work based +on the Program. + + To "propagate" a work means to do anything with it that, without +permission, would make you directly or secondarily liable for +infringement under applicable copyright law, except executing it on a +computer or modifying a private copy. Propagation includes copying, +distribution (with or without modification), making available to the +public, and in some countries other activities as well. + + To "convey" a work means any kind of propagation that enables other +parties to make or receive copies. Mere interaction with a user through +a computer network, with no transfer of a copy, is not conveying. + + An interactive user interface displays "Appropriate Legal Notices" +to the extent that it includes a convenient and prominently visible +feature that (1) displays an appropriate copyright notice, and (2) +tells the user that there is no warranty for the work (except to the +extent that warranties are provided), that licensees may convey the +work under this License, and how to view a copy of this License. If +the interface presents a list of user commands or options, such as a +menu, a prominent item in the list meets this criterion. + + 1. Source Code. + + The "source code" for a work means the preferred form of the work +for making modifications to it. "Object code" means any non-source +form of a work. + + A "Standard Interface" means an interface that either is an official +standard defined by a recognized standards body, or, in the case of +interfaces specified for a particular programming language, one that +is widely used among developers working in that language. + + The "System Libraries" of an executable work include anything, other +than the work as a whole, that (a) is included in the normal form of +packaging a Major Component, but which is not part of that Major +Component, and (b) serves only to enable use of the work with that +Major Component, or to implement a Standard Interface for which an +implementation is available to the public in source code form. A +"Major Component", in this context, means a major essential component +(kernel, window system, and so on) of the specific operating system +(if any) on which the executable work runs, or a compiler used to +produce the work, or an object code interpreter used to run it. + + The "Corresponding Source" for a work in object code form means all +the source code needed to generate, install, and (for an executable +work) run the object code and to modify the work, including scripts to +control those activities. However, it does not include the work's +System Libraries, or general-purpose tools or generally available free +programs which are used unmodified in performing those activities but +which are not part of the work. For example, Corresponding Source +includes interface definition files associated with source files for +the work, and the source code for shared libraries and dynamically +linked subprograms that the work is specifically designed to require, +such as by intimate data communication or control flow between those +subprograms and other parts of the work. + + The Corresponding Source need not include anything that users +can regenerate automatically from other parts of the Corresponding +Source. + + The Corresponding Source for a work in source code form is that +same work. + + 2. Basic Permissions. + + All rights granted under this License are granted for the term of +copyright on the Program, and are irrevocable provided the stated +conditions are met. This License explicitly affirms your unlimited +permission to run the unmodified Program. The output from running a +covered work is covered by this License only if the output, given its +content, constitutes a covered work. This License acknowledges your +rights of fair use or other equivalent, as provided by copyright law. + + You may make, run and propagate covered works that you do not +convey, without conditions so long as your license otherwise remains +in force. You may convey covered works to others for the sole purpose +of having them make modifications exclusively for you, or provide you +with facilities for running those works, provided that you comply with +the terms of this License in conveying all material for which you do +not control copyright. Those thus making or running the covered works +for you must do so exclusively on your behalf, under your direction +and control, on terms that prohibit them from making any copies of +your copyrighted material outside their relationship with you. + + Conveying under any other circumstances is permitted solely under +the conditions stated below. Sublicensing is not allowed; section 10 +makes it unnecessary. + + 3. Protecting Users' Legal Rights From Anti-Circumvention Law. + + No covered work shall be deemed part of an effective technological +measure under any applicable law fulfilling obligations under article +11 of the WIPO copyright treaty adopted on 20 December 1996, or +similar laws prohibiting or restricting circumvention of such +measures. + + When you convey a covered work, you waive any legal power to forbid +circumvention of technological measures to the extent such circumvention +is effected by exercising rights under this License with respect to +the covered work, and you disclaim any intention to limit operation or +modification of the work as a means of enforcing, against the work's +users, your or third parties' legal rights to forbid circumvention of +technological measures. + + 4. Conveying Verbatim Copies. + + You may convey verbatim copies of the Program's source code as you +receive it, in any medium, provided that you conspicuously and +appropriately publish on each copy an appropriate copyright notice; +keep intact all notices stating that this License and any +non-permissive terms added in accord with section 7 apply to the code; +keep intact all notices of the absence of any warranty; and give all +recipients a copy of this License along with the Program. + + You may charge any price or no price for each copy that you convey, +and you may offer support or warranty protection for a fee. + + 5. Conveying Modified Source Versions. + + You may convey a work based on the Program, or the modifications to +produce it from the Program, in the form of source code under the +terms of section 4, provided that you also meet all of these conditions: + + a) The work must carry prominent notices stating that you modified + it, and giving a relevant date. + + b) The work must carry prominent notices stating that it is + released under this License and any conditions added under section + 7. This requirement modifies the requirement in section 4 to + "keep intact all notices". + + c) You must license the entire work, as a whole, under this + License to anyone who comes into possession of a copy. This + License will therefore apply, along with any applicable section 7 + additional terms, to the whole of the work, and all its parts, + regardless of how they are packaged. This License gives no + permission to license the work in any other way, but it does not + invalidate such permission if you have separately received it. + + d) If the work has interactive user interfaces, each must display + Appropriate Legal Notices; however, if the Program has interactive + interfaces that do not display Appropriate Legal Notices, your + work need not make them do so. + + A compilation of a covered work with other separate and independent +works, which are not by their nature extensions of the covered work, +and which are not combined with it such as to form a larger program, +in or on a volume of a storage or distribution medium, is called an +"aggregate" if the compilation and its resulting copyright are not +used to limit the access or legal rights of the compilation's users +beyond what the individual works permit. Inclusion of a covered work +in an aggregate does not cause this License to apply to the other +parts of the aggregate. + + 6. Conveying Non-Source Forms. + + You may convey a covered work in object code form under the terms +of sections 4 and 5, provided that you also convey the +machine-readable Corresponding Source under the terms of this License, +in one of these ways: + + a) Convey the object code in, or embodied in, a physical product + (including a physical distribution medium), accompanied by the + Corresponding Source fixed on a durable physical medium + customarily used for software interchange. + + b) Convey the object code in, or embodied in, a physical product + (including a physical distribution medium), accompanied by a + written offer, valid for at least three years and valid for as + long as you offer spare parts or customer support for that product + model, to give anyone who possesses the object code either (1) a + copy of the Corresponding Source for all the software in the + product that is covered by this License, on a durable physical + medium customarily used for software interchange, for a price no + more than your reasonable cost of physically performing this + conveying of source, or (2) access to copy the + Corresponding Source from a network server at no charge. + + c) Convey individual copies of the object code with a copy of the + written offer to provide the Corresponding Source. This + alternative is allowed only occasionally and noncommercially, and + only if you received the object code with such an offer, in accord + with subsection 6b. + + d) Convey the object code by offering access from a designated + place (gratis or for a charge), and offer equivalent access to the + Corresponding Source in the same way through the same place at no + further charge. You need not require recipients to copy the + Corresponding Source along with the object code. If the place to + copy the object code is a network server, the Corresponding Source + may be on a different server (operated by you or a third party) + that supports equivalent copying facilities, provided you maintain + clear directions next to the object code saying where to find the + Corresponding Source. Regardless of what server hosts the + Corresponding Source, you remain obligated to ensure that it is + available for as long as needed to satisfy these requirements. + + e) Convey the object code using peer-to-peer transmission, provided + you inform other peers where the object code and Corresponding + Source of the work are being offered to the general public at no + charge under subsection 6d. + + A separable portion of the object code, whose source code is excluded +from the Corresponding Source as a System Library, need not be +included in conveying the object code work. + + A "User Product" is either (1) a "consumer product", which means any +tangible personal property which is normally used for personal, family, +or household purposes, or (2) anything designed or sold for incorporation +into a dwelling. In determining whether a product is a consumer product, +doubtful cases shall be resolved in favor of coverage. For a particular +product received by a particular user, "normally used" refers to a +typical or common use of that class of product, regardless of the status +of the particular user or of the way in which the particular user +actually uses, or expects or is expected to use, the product. A product +is a consumer product regardless of whether the product has substantial +commercial, industrial or non-consumer uses, unless such uses represent +the only significant mode of use of the product. + + "Installation Information" for a User Product means any methods, +procedures, authorization keys, or other information required to install +and execute modified versions of a covered work in that User Product from +a modified version of its Corresponding Source. The information must +suffice to ensure that the continued functioning of the modified object +code is in no case prevented or interfered with solely because +modification has been made. + + If you convey an object code work under this section in, or with, or +specifically for use in, a User Product, and the conveying occurs as +part of a transaction in which the right of possession and use of the +User Product is transferred to the recipient in perpetuity or for a +fixed term (regardless of how the transaction is characterized), the +Corresponding Source conveyed under this section must be accompanied +by the Installation Information. But this requirement does not apply +if neither you nor any third party retains the ability to install +modified object code on the User Product (for example, the work has +been installed in ROM). + + The requirement to provide Installation Information does not include a +requirement to continue to provide support service, warranty, or updates +for a work that has been modified or installed by the recipient, or for +the User Product in which it has been modified or installed. Access to a +network may be denied when the modification itself materially and +adversely affects the operation of the network or violates the rules and +protocols for communication across the network. + + Corresponding Source conveyed, and Installation Information provided, +in accord with this section must be in a format that is publicly +documented (and with an implementation available to the public in +source code form), and must require no special password or key for +unpacking, reading or copying. + + 7. Additional Terms. + + "Additional permissions" are terms that supplement the terms of this +License by making exceptions from one or more of its conditions. +Additional permissions that are applicable to the entire Program shall +be treated as though they were included in this License, to the extent +that they are valid under applicable law. If additional permissions +apply only to part of the Program, that part may be used separately +under those permissions, but the entire Program remains governed by +this License without regard to the additional permissions. + + When you convey a copy of a covered work, you may at your option +remove any additional permissions from that copy, or from any part of +it. (Additional permissions may be written to require their own +removal in certain cases when you modify the work.) You may place +additional permissions on material, added by you to a covered work, +for which you have or can give appropriate copyright permission. + + Notwithstanding any other provision of this License, for material you +add to a covered work, you may (if authorized by the copyright holders of +that material) supplement the terms of this License with terms: + + a) Disclaiming warranty or limiting liability differently from the + terms of sections 15 and 16 of this License; or + + b) Requiring preservation of specified reasonable legal notices or + author attributions in that material or in the Appropriate Legal + Notices displayed by works containing it; or + + c) Prohibiting misrepresentation of the origin of that material, or + requiring that modified versions of such material be marked in + reasonable ways as different from the original version; or + + d) Limiting the use for publicity purposes of names of licensors or + authors of the material; or + + e) Declining to grant rights under trademark law for use of some + trade names, trademarks, or service marks; or + + f) Requiring indemnification of licensors and authors of that + material by anyone who conveys the material (or modified versions of + it) with contractual assumptions of liability to the recipient, for + any liability that these contractual assumptions directly impose on + those licensors and authors. + + All other non-permissive additional terms are considered "further +restrictions" within the meaning of section 10. If the Program as you +received it, or any part of it, contains a notice stating that it is +governed by this License along with a term that is a further +restriction, you may remove that term. If a license document contains +a further restriction but permits relicensing or conveying under this +License, you may add to a covered work material governed by the terms +of that license document, provided that the further restriction does +not survive such relicensing or conveying. + + If you add terms to a covered work in accord with this section, you +must place, in the relevant source files, a statement of the +additional terms that apply to those files, or a notice indicating +where to find the applicable terms. + + Additional terms, permissive or non-permissive, may be stated in the +form of a separately written license, or stated as exceptions; +the above requirements apply either way. + + 8. Termination. + + You may not propagate or modify a covered work except as expressly +provided under this License. Any attempt otherwise to propagate or +modify it is void, and will automatically terminate your rights under +this License (including any patent licenses granted under the third +paragraph of section 11). + + However, if you cease all violation of this License, then your +license from a particular copyright holder is reinstated (a) +provisionally, unless and until the copyright holder explicitly and +finally terminates your license, and (b) permanently, if the copyright +holder fails to notify you of the violation by some reasonable means +prior to 60 days after the cessation. + + Moreover, your license from a particular copyright holder is +reinstated permanently if the copyright holder notifies you of the +violation by some reasonable means, this is the first time you have +received notice of violation of this License (for any work) from that +copyright holder, and you cure the violation prior to 30 days after +your receipt of the notice. + + Termination of your rights under this section does not terminate the +licenses of parties who have received copies or rights from you under +this License. If your rights have been terminated and not permanently +reinstated, you do not qualify to receive new licenses for the same +material under section 10. + + 9. Acceptance Not Required for Having Copies. + + You are not required to accept this License in order to receive or +run a copy of the Program. Ancillary propagation of a covered work +occurring solely as a consequence of using peer-to-peer transmission +to receive a copy likewise does not require acceptance. However, +nothing other than this License grants you permission to propagate or +modify any covered work. These actions infringe copyright if you do +not accept this License. Therefore, by modifying or propagating a +covered work, you indicate your acceptance of this License to do so. + + 10. Automatic Licensing of Downstream Recipients. + + Each time you convey a covered work, the recipient automatically +receives a license from the original licensors, to run, modify and +propagate that work, subject to this License. You are not responsible +for enforcing compliance by third parties with this License. + + An "entity transaction" is a transaction transferring control of an +organization, or substantially all assets of one, or subdividing an +organization, or merging organizations. If propagation of a covered +work results from an entity transaction, each party to that +transaction who receives a copy of the work also receives whatever +licenses to the work the party's predecessor in interest had or could +give under the previous paragraph, plus a right to possession of the +Corresponding Source of the work from the predecessor in interest, if +the predecessor has it or can get it with reasonable efforts. + + You may not impose any further restrictions on the exercise of the +rights granted or affirmed under this License. For example, you may +not impose a license fee, royalty, or other charge for exercise of +rights granted under this License, and you may not initiate litigation +(including a cross-claim or counterclaim in a lawsuit) alleging that +any patent claim is infringed by making, using, selling, offering for +sale, or importing the Program or any portion of it. + + 11. Patents. + + A "contributor" is a copyright holder who authorizes use under this +License of the Program or a work on which the Program is based. The +work thus licensed is called the contributor's "contributor version". + + A contributor's "essential patent claims" are all patent claims +owned or controlled by the contributor, whether already acquired or +hereafter acquired, that would be infringed by some manner, permitted +by this License, of making, using, or selling its contributor version, +but do not include claims that would be infringed only as a +consequence of further modification of the contributor version. For +purposes of this definition, "control" includes the right to grant +patent sublicenses in a manner consistent with the requirements of +this License. + + Each contributor grants you a non-exclusive, worldwide, royalty-free +patent license under the contributor's essential patent claims, to +make, use, sell, offer for sale, import and otherwise run, modify and +propagate the contents of its contributor version. + + In the following three paragraphs, a "patent license" is any express +agreement or commitment, however denominated, not to enforce a patent +(such as an express permission to practice a patent or covenant not to +sue for patent infringement). To "grant" such a patent license to a +party means to make such an agreement or commitment not to enforce a +patent against the party. + + If you convey a covered work, knowingly relying on a patent license, +and the Corresponding Source of the work is not available for anyone +to copy, free of charge and under the terms of this License, through a +publicly available network server or other readily accessible means, +then you must either (1) cause the Corresponding Source to be so +available, or (2) arrange to deprive yourself of the benefit of the +patent license for this particular work, or (3) arrange, in a manner +consistent with the requirements of this License, to extend the patent +license to downstream recipients. "Knowingly relying" means you have +actual knowledge that, but for the patent license, your conveying the +covered work in a country, or your recipient's use of the covered work +in a country, would infringe one or more identifiable patents in that +country that you have reason to believe are valid. + + If, pursuant to or in connection with a single transaction or +arrangement, you convey, or propagate by procuring conveyance of, a +covered work, and grant a patent license to some of the parties +receiving the covered work authorizing them to use, propagate, modify +or convey a specific copy of the covered work, then the patent license +you grant is automatically extended to all recipients of the covered +work and works based on it. + + A patent license is "discriminatory" if it does not include within +the scope of its coverage, prohibits the exercise of, or is +conditioned on the non-exercise of one or more of the rights that are +specifically granted under this License. You may not convey a covered +work if you are a party to an arrangement with a third party that is +in the business of distributing software, under which you make payment +to the third party based on the extent of your activity of conveying +the work, and under which the third party grants, to any of the +parties who would receive the covered work from you, a discriminatory +patent license (a) in connection with copies of the covered work +conveyed by you (or copies made from those copies), or (b) primarily +for and in connection with specific products or compilations that +contain the covered work, unless you entered into that arrangement, +or that patent license was granted, prior to 28 March 2007. + + Nothing in this License shall be construed as excluding or limiting +any implied license or other defenses to infringement that may +otherwise be available to you under applicable patent law. + + 12. No Surrender of Others' Freedom. + + If conditions are imposed on you (whether by court order, agreement or +otherwise) that contradict the conditions of this License, they do not +excuse you from the conditions of this License. If you cannot convey a +covered work so as to satisfy simultaneously your obligations under this +License and any other pertinent obligations, then as a consequence you may +not convey it at all. For example, if you agree to terms that obligate you +to collect a royalty for further conveying from those to whom you convey +the Program, the only way you could satisfy both those terms and this +License would be to refrain entirely from conveying the Program. + + 13. Remote Network Interaction; Use with the GNU General Public License. + + Notwithstanding any other provision of this License, if you modify the +Program, your modified version must prominently offer all users +interacting with it remotely through a computer network (if your version +supports such interaction) an opportunity to receive the Corresponding +Source of your version by providing access to the Corresponding Source +from a network server at no charge, through some standard or customary +means of facilitating copying of software. This Corresponding Source +shall include the Corresponding Source for any work covered by version 3 +of the GNU General Public License that is incorporated pursuant to the +following paragraph. + + Notwithstanding any other provision of this License, you have +permission to link or combine any covered work with a work licensed +under version 3 of the GNU General Public License into a single +combined work, and to convey the resulting work. The terms of this +License will continue to apply to the part which is the covered work, +but the work with which it is combined will remain governed by version +3 of the GNU General Public License. + + 14. Revised Versions of this License. + + The Free Software Foundation may publish revised and/or new versions of +the GNU Affero General Public License from time to time. Such new versions +will be similar in spirit to the present version, but may differ in detail to +address new problems or concerns. + + Each version is given a distinguishing version number. If the +Program specifies that a certain numbered version of the GNU Affero General +Public License "or any later version" applies to it, you have the +option of following the terms and conditions either of that numbered +version or of any later version published by the Free Software +Foundation. If the Program does not specify a version number of the +GNU Affero General Public License, you may choose any version ever published +by the Free Software Foundation. + + If the Program specifies that a proxy can decide which future +versions of the GNU Affero General Public License can be used, that proxy's +public statement of acceptance of a version permanently authorizes you +to choose that version for the Program. + + Later license versions may give you additional or different +permissions. However, no additional obligations are imposed on any +author or copyright holder as a result of your choosing to follow a +later version. + + 15. Disclaimer of Warranty. + + THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY +APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT +HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY +OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, +THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR +PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM +IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF +ALL NECESSARY SERVICING, REPAIR OR CORRECTION. + + 16. Limitation of Liability. + + IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING +WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS +THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY +GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE +USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF +DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD +PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), +EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF +SUCH DAMAGES. + + 17. Interpretation of Sections 15 and 16. + + If the disclaimer of warranty and limitation of liability provided +above cannot be given local legal effect according to their terms, +reviewing courts shall apply local law that most closely approximates +an absolute waiver of all civil liability in connection with the +Program, unless a warranty or assumption of liability accompanies a +copy of the Program in return for a fee. + + END OF TERMS AND CONDITIONS + + How to Apply These Terms to Your New Programs + + If you develop a new program, and you want it to be of the greatest +possible use to the public, the best way to achieve this is to make it +free software which everyone can redistribute and change under these terms. + + To do so, attach the following notices to the program. It is safest +to attach them to the start of each source file to most effectively +state the exclusion of warranty; and each file should have at least +the "copyright" line and a pointer to where the full notice is found. + + + Copyright (C) + + This program is free software: you can redistribute it and/or modify + it under the terms of the GNU Affero General Public License as published + by the Free Software Foundation, either version 3 of the License, or + (at your option) any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU Affero General Public License for more details. + + You should have received a copy of the GNU Affero General Public License + along with this program. If not, see . + +Also add information on how to contact you by electronic and paper mail. + + If your software can interact with users remotely through a computer +network, you should also make sure that it provides a way for users to +get its source. For example, if your program is a web application, its +interface could display a "Source" link that leads users to an archive +of the code. There are many ways you could offer source, and different +solutions will be better for different programs; see section 13 for the +specific requirements. + + You should also get your employer (if you work as a programmer) or school, +if any, to sign a "copyright disclaimer" for the program, if necessary. +For more information on this, and how to apply and follow the GNU AGPL, see +. From a73de5d4a39759c7dd657d12ab1a27d77e06777f Mon Sep 17 00:00:00 2001 From: Prins Kumar Date: Tue, 18 Nov 2025 18:32:55 +0530 Subject: [PATCH 03/15] Update README for OpenAI and Bedrock config, add pymupdf4llm license --- README.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/README.md b/README.md index 13c88b3..9469ad6 100644 --- a/README.md +++ b/README.md @@ -482,7 +482,7 @@ In addition to the `OPENAI_API_KEY`, `llm_model` and `model_name` can be edited "model_kwargs": { "temperature": 0 }, - "prompt_path": "./common/prompts/openai_gpt4/" + "prompt_path": "./app/prompts/openai_gpt4/" }, "multimodal_service": { "llm_service": "openai", @@ -614,7 +614,7 @@ In addition to the `AZURE_OPENAI_ENDPOINT`, `AZURE_OPENAI_API_KEY`, and `azure_d "temperature": 0, "max_tokens": 4096 }, - "prompt_path": "./common/prompts/openai_gpt4/" + "prompt_path": "./app/prompts/aws_bedrock_claude3haiku/" }, "multimodal_service": { "llm_service": "bedrock", From 5b18648869eb1b2251444a1a81774ea753d1bfa0 Mon Sep 17 00:00:00 2001 From: Prins Kumar Date: Tue, 18 Nov 2025 22:55:22 +0530 Subject: [PATCH 04/15] Fix prompt_path to use ./common/prompts/ for OpenAI and Bedrock --- README.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/README.md b/README.md index 9469ad6..13c88b3 100644 --- a/README.md +++ b/README.md @@ -482,7 +482,7 @@ In addition to the `OPENAI_API_KEY`, `llm_model` and `model_name` can be edited "model_kwargs": { "temperature": 0 }, - "prompt_path": "./app/prompts/openai_gpt4/" + "prompt_path": "./common/prompts/openai_gpt4/" }, "multimodal_service": { "llm_service": "openai", @@ -614,7 +614,7 @@ In addition to the `AZURE_OPENAI_ENDPOINT`, `AZURE_OPENAI_API_KEY`, and `azure_d "temperature": 0, "max_tokens": 4096 }, - "prompt_path": "./app/prompts/aws_bedrock_claude3haiku/" + "prompt_path": "./common/prompts/openai_gpt4/" }, "multimodal_service": { "llm_service": "bedrock", From 380aa7c72ecfe0ab6314b9d75b40de9c8ea936fe Mon Sep 17 00:00:00 2001 From: Prins Kumar Date: Fri, 21 Nov 2025 20:57:53 +0530 Subject: [PATCH 05/15] bug fixes --- graphrag/app/routers/ui.py | 1 + graphrag/app/supportai/supportai.py | 18 ++++++++++++------ 2 files changed, 13 insertions(+), 6 deletions(-) diff --git a/graphrag/app/routers/ui.py b/graphrag/app/routers/ui.py index 9637347..114b489 100644 --- a/graphrag/app/routers/ui.py +++ b/graphrag/app/routers/ui.py @@ -395,6 +395,7 @@ async def serve_image_from_vertex( LogWriter.info(f"Serving image {image_id} from graph {graphname}") # Fetch the Image vertex by ID + # TigerGraph loading job uses gsql_lower() so all IDs are stored in lowercase image_vertices = conn.getVerticesById('Image', [image_id.lower()]) if not image_vertices: diff --git a/graphrag/app/supportai/supportai.py b/graphrag/app/supportai/supportai.py index d2efe8a..6b93df0 100644 --- a/graphrag/app/supportai/supportai.py +++ b/graphrag/app/supportai/supportai.py @@ -337,9 +337,9 @@ def create_ingest( conn: TigerGraphConnection, ): # Check for invalid combination of multi format and non-s3 data source - if ingest_config.data_source.lower() in ["bda", "server"] and ingest_config.get("file_format", "").lower() != "multi": - logger.warning(f"File format {ingest_config.get('file_format', '').lower()} is not supported for data source {ingest_config.data_source.lower()}") - ingest_config["file_format"] = "multi" + if ingest_config.data_source.lower() in ["bda", "server"] and ingest_config.file_format.lower() != "multi": + logger.warning(f"File format {ingest_config.file_format.lower()} is not supported for data source {ingest_config.data_source.lower()}") + ingest_config.file_format = "multi" res_ingest_config = {"data_source": ingest_config.data_source.lower()} res_ingest_config["file_format"] = ingest_config.file_format.lower() @@ -481,9 +481,9 @@ def create_ingest( except Exception as e: raise Exception(f"Error during Amazon BDA preprocessing: {e}") elif ingest_config.data_source.lower() == "server": - data_path = ingest_config.data_source_config.get("data_path", None) + data_path = ingest_config.data_source_config.get("folder_path", None) if data_path is None: - raise Exception("Data path not provided for server processing") + raise Exception("Folder path not provided for server processing") try: extractor = TextExtractor() server_processing_result = extractor.process_folder(data_path, graphname=graphname) @@ -652,7 +652,10 @@ def ingest( data_source_id = ingest_config.get("data_source_id", "DocumentContent") if ingest_config.get("server_jobs"): for doc_data in ingest_config.get("server_jobs"): - if not doc_data.get("doc_id") or not doc_data.get("content"): + if not doc_data.get("doc_id"): + continue + # Skip documents with neither content nor image_data + if not doc_data.get("content") and not doc_data.get("image_data"): continue if doc_data.get("image_data"): payload = { @@ -660,8 +663,11 @@ def ingest( "doc_type": "image", "image_data": doc_data.get("image_data", ""), "image_format": doc_data.get("image_format", "jpg"), + "image_description": doc_data.get("image_description", ""), "parent_doc": doc_data.get("parent_doc", ""), "page_number": doc_data.get("page_number", 0), + "width": doc_data.get("width", 0), + "height": doc_data.get("height", 0), "position": doc_data.get("position", 0), "content": "" } From 16c8721bc25d9fe064df88321e9bbeb677dd4ea0 Mon Sep 17 00:00:00 2001 From: Prins Kumar Date: Mon, 24 Nov 2025 20:00:28 +0530 Subject: [PATCH 06/15] Add local temp file storage for ingestion review --- common/utils/text_extractors.py | 59 +++-- graphrag-ui/src/pages/Setup.tsx | 319 +++++++++++++++++++++++++--- graphrag/app/routers/ui.py | 136 ++++++++++++ graphrag/app/supportai/supportai.py | 70 +++++- 4 files changed, 530 insertions(+), 54 deletions(-) diff --git a/common/utils/text_extractors.py b/common/utils/text_extractors.py index b900cae..ec5b140 100644 --- a/common/utils/text_extractors.py +++ b/common/utils/text_extractors.py @@ -8,6 +8,7 @@ import uuid import base64 import io +import threading from pathlib import Path import shutil import asyncio @@ -15,6 +16,9 @@ logger = logging.getLogger(__name__) +# Global lock for pymupdf4llm calls (not thread-safe) +_pymupdf4llm_lock = threading.Lock() + class TextExtractor: """Class for handling text extraction from various file formats and cleanup.""" @@ -202,26 +206,39 @@ def _extract_pdf_with_images_as_docs(file_path, base_doc_id, graphname=None): shutil.rmtree(image_output_folder, ignore_errors=True) # Convert PDF to markdown with extracted image files - try: - markdown_content = pymupdf4llm.to_markdown( - file_path, - write_images=True, - image_path=str(image_output_folder), # unique folder per PDF - force_text=False, - margins=0, - image_size_limit=0.08, - ) - except Exception as e: - logger.error(f"pymupdf4llm failed for {file_path}: {e}") - # Cleanup folder if it was created - if image_output_folder.exists(): - shutil.rmtree(image_output_folder, ignore_errors=True) - return [{ - "doc_id": base_doc_id, - "doc_type": "markdown", - "content": f"[PDF extraction failed: {e}]", - "position": 0 - }] + # Use lock because pymupdf4llm's table extraction is not thread-safe + # See: https://github.com/pymupdf/PyMuPDF/issues/3241 + with _pymupdf4llm_lock: + try: + markdown_content = pymupdf4llm.to_markdown( + file_path, + write_images=True, + image_path=str(image_output_folder), # unique folder per PDF + margins=0, + image_size_limit=0.08, + ) + except Exception: + # Retry with table_strategy="lines" if first attempt fails + try: + markdown_content = pymupdf4llm.to_markdown( + file_path, + write_images=True, + image_path=str(image_output_folder), # unique folder per PDF + margins=0, + image_size_limit=0.08, + table_strategy="lines", + ) + except Exception as e: + logger.error(f"pymupdf4llm failed for {file_path}: {e}") + # Cleanup folder if it was created + if image_output_folder.exists(): + shutil.rmtree(image_output_folder, ignore_errors=True) + return [{ + "doc_id": base_doc_id, + "doc_type": "markdown", + "content": f"[PDF extraction failed: {e}]", + "position": 0 + }] if not markdown_content or not markdown_content.strip(): logger.warning(f"No content extracted from PDF: {file_path}") @@ -461,4 +478,4 @@ def get_supported_extensions(): def is_supported_file(file_path): """Check if a file is supported for text extraction.""" extension = Path(file_path).suffix.lower() - return extension in get_supported_extensions() + return extension in get_supported_extensions() \ No newline at end of file diff --git a/graphrag-ui/src/pages/Setup.tsx b/graphrag-ui/src/pages/Setup.tsx index 2aaee99..c844896 100644 --- a/graphrag-ui/src/pages/Setup.tsx +++ b/graphrag-ui/src/pages/Setup.tsx @@ -56,6 +56,12 @@ const Setup = () => { const [uploadMessage, setUploadMessage] = useState(""); const [isIngesting, setIsIngesting] = useState(false); const [ingestMessage, setIngestMessage] = useState(""); + + // Ingestion temp files state + const [tempSessionId, setTempSessionId] = useState(null); + const [tempFiles, setTempFiles] = useState([]); + const [showTempFiles, setShowTempFiles] = useState(false); + const [ingestJobData, setIngestJobData] = useState(null); // Refresh state const [refreshOpen, setRefreshOpen] = useState(false); @@ -416,6 +422,125 @@ const Setup = () => { } }; + // Fetch temp processed files + const fetchTempFiles = async (sessionId: string) => { + if (!ingestGraphName || !sessionId) return; + + try { + const creds = localStorage.getItem("creds"); + const response = await fetch(`/ui/${ingestGraphName}/ingestion_temp/list?session_id=${sessionId}`, { + headers: { Authorization: `Basic ${creds}` }, + }); + const data = await response.json(); + if (data.status === "success" && data.sessions.length > 0) { + setTempFiles(data.sessions[0].files || []); + setShowTempFiles(true); + } + } catch (error) { + console.error("Error fetching temp files:", error); + } + }; + + // Delete a specific temp file + const handleDeleteTempFile = async (filename: string) => { + if (!ingestGraphName || !tempSessionId) return; + + try { + const creds = localStorage.getItem("creds"); + const response = await fetch( + `/ui/${ingestGraphName}/ingestion_temp/delete?session_id=${tempSessionId}&filename=${encodeURIComponent(filename)}`, + { + method: "DELETE", + headers: { Authorization: `Basic ${creds}` }, + } + ); + const data = await response.json(); + if (data.status === "success") { + setIngestMessage(`✅ ${data.message}`); + // Refresh the temp files list + await fetchTempFiles(tempSessionId); + } + } catch (error: any) { + setIngestMessage(`❌ Error: ${error.message}`); + } + }; + + // Delete all temp files for session + const handleDeleteAllTempFiles = async () => { + if (!ingestGraphName || !tempSessionId) return; + + try { + const creds = localStorage.getItem("creds"); + const response = await fetch( + `/ui/${ingestGraphName}/ingestion_temp/delete?session_id=${tempSessionId}`, + { + method: "DELETE", + headers: { Authorization: `Basic ${creds}` }, + } + ); + const data = await response.json(); + if (data.status === "success") { + setIngestMessage(`✅ ${data.message}`); + setTempFiles([]); + setShowTempFiles(false); + setTempSessionId(null); + } + } catch (error: any) { + setIngestMessage(`❌ Error: ${error.message}`); + } + }; + + // Run final ingest after user reviews temp files + const handleRunIngest = async () => { + if (!ingestJobData) { + setIngestMessage("❌ No ingest job data available"); + return; + } + + setIsIngesting(true); + setIngestMessage("Running final document ingest..."); + + try { + const creds = localStorage.getItem("creds"); + + const loadingInfo = { + load_job_id: ingestJobData.load_job_id, + data_source_id: ingestJobData.data_source_id, + file_path: ingestJobData.data_path, + }; + + const ingestResponse = await fetch(`/ui/${ingestGraphName}/ingest`, { + method: "POST", + headers: { + "Content-Type": "application/json", + Authorization: `Basic ${creds}`, + }, + body: JSON.stringify(loadingInfo), + }); + + if (!ingestResponse.ok) { + const errorData = await ingestResponse.json(); + throw new Error(errorData.detail || `Failed to run ingest: ${ingestResponse.statusText}`); + } + + const ingestData = await ingestResponse.json(); + console.log("Ingest response:", ingestData); + + setIngestMessage(`✅ Data ingested successfully! Processed ${tempFiles.length} documents.`); + + // Clear temp state + setTempFiles([]); + setShowTempFiles(false); + setTempSessionId(null); + setIngestJobData(null); + } catch (error: any) { + console.error("Error running ingest:", error); + setIngestMessage(`❌ Error: ${error.message}`); + } finally { + setIsIngesting(false); + } + }; + // Ingest files into knowledge graph (uploaded or downloaded) const handleIngestDocuments = async (sourceType: "uploaded" | "downloaded" = "uploaded") => { if (!ingestGraphName) { @@ -460,37 +585,53 @@ const Setup = () => { const createData = await createResponse.json(); console.log("Create ingest response:", createData); - // Step 2: Run ingest - setIngestMessage("Step 2/2: Running document ingest..."); - - const loadingInfo = { - load_job_id: createData.load_job_id, - data_source_id: createData.data_source_id, - file_path: createData.data_path || createData.file_path, // Handle both field names - }; + // Check if temp files were created (for server data source) + const sessionId = createData.data_source_id?.temp_session_id; + + if (sessionId) { + // Files are saved to temp storage - show them for review + setTempSessionId(sessionId); + setIngestJobData({ + load_job_id: createData.load_job_id, + data_source_id: createData.data_source_id, + data_path: createData.data_path || createData.file_path, + }); + setIngestMessage(`✅ Processed ${createData.data_source_id.file_count} files. Review them below before ingesting.`); + await fetchTempFiles(sessionId); + setIsIngesting(false); + } else { + // No temp files (e.g., S3 Bedrock) - proceed directly to ingest + setIngestMessage("Step 2/2: Running document ingest..."); - const ingestResponse = await fetch(`/ui/${ingestGraphName}/ingest`, { - method: "POST", - headers: { - "Content-Type": "application/json", - Authorization: `Basic ${creds}`, - }, - body: JSON.stringify(loadingInfo), - }); + const loadingInfo = { + load_job_id: createData.load_job_id, + data_source_id: createData.data_source_id, + file_path: createData.data_path || createData.file_path, + }; - if (!ingestResponse.ok) { - const errorData = await ingestResponse.json(); - throw new Error(errorData.detail || `Failed to run ingest: ${ingestResponse.statusText}`); - } + const ingestResponse = await fetch(`/ui/${ingestGraphName}/ingest`, { + method: "POST", + headers: { + "Content-Type": "application/json", + Authorization: `Basic ${creds}`, + }, + body: JSON.stringify(loadingInfo), + }); + + if (!ingestResponse.ok) { + const errorData = await ingestResponse.json(); + throw new Error(errorData.detail || `Failed to run ingest: ${ingestResponse.statusText}`); + } - const ingestData = await ingestResponse.json(); - console.log("Ingest response:", ingestData); + const ingestData = await ingestResponse.json(); + console.log("Ingest response:", ingestData); - setIngestMessage(`✅ Data ingested successfully! Processed documents from ${folderPath}/`); + setIngestMessage(`✅ Data ingested successfully! Processed documents from ${folderPath}/`); + setIsIngesting(false); + } } catch (error: any) { console.error("Error ingesting data:", error); setIngestMessage(`❌ Error: ${error.message}`); - } finally { setIsIngesting(false); } }; @@ -1237,6 +1378,71 @@ const Setup = () => { {ingestMessage}
)} + + {/* Processed Temp Files - Review before ingesting */} + {showTempFiles && tempFiles.length > 0 && ( +
+
+

+ Processed Files ({tempFiles.length}) +

+ +
+

+ Review the processed files below. You can delete any file before ingesting. +

+
+ {tempFiles.map((file, index) => ( +
+
+

+ {file.doc_id} +

+

+ {(file.size / 1024).toFixed(2)} KB +

+
+ +
+ ))} +
+ +
+ )}
)} @@ -1576,6 +1782,71 @@ const Setup = () => { {ingestMessage}
)} + + {/* Processed Temp Files - Review before ingesting */} + {showTempFiles && tempFiles.length > 0 && ( +
+
+

+ Processed Files ({tempFiles.length}) +

+ +
+

+ Review the processed files below. You can delete any file before ingesting. +

+
+ {tempFiles.map((file, index) => ( +
+
+

+ {file.doc_id} +

+

+ {(file.size / 1024).toFixed(2)} KB +

+
+ +
+ ))} +
+ +
+ )}
)}
diff --git a/graphrag/app/routers/ui.py b/graphrag/app/routers/ui.py index 114b489..9b012ec 100644 --- a/graphrag/app/routers/ui.py +++ b/graphrag/app/routers/ui.py @@ -1380,3 +1380,139 @@ async def delete_cloud_downloads( logger.debug_pii(f"Delete error trace:\n{exc}") raise HTTPException(status_code=500, detail=f"Error deleting files: {str(e)}") + +# Ingestion Temp Files Endpoints + +@router.get(route_prefix + "/{graphname}/ingestion_temp/list") +async def list_ingestion_temp_files( + graphname: str, + credentials: Annotated[HTTPBase, Depends(security)], + session_id: str = None, +): + """ + List processed files in the ingestion temp folder for a specific graph. + """ + try: + base_temp_dir = os.path.join("uploads", "ingestion_temp", graphname) + + if not os.path.exists(base_temp_dir): + return { + "status": "success", + "graphname": graphname, + "sessions": [], + "total_files": 0, + } + + sessions = [] + total_files = 0 + + # If session_id provided, list only that session + if session_id: + session_dir = os.path.join(base_temp_dir, session_id) + if os.path.exists(session_dir) and os.path.isdir(session_dir): + files = [] + for filename in os.listdir(session_dir): + filepath = os.path.join(session_dir, filename) + if os.path.isfile(filepath) and filename.endswith('.json'): + file_stat = os.stat(filepath) + # Read doc_id from file + try: + with open(filepath, 'r', encoding='utf-8') as f: + doc_data = json.load(f) + doc_id = doc_data.get('doc_id', 'unknown') + except: + doc_id = 'unknown' + + files.append({ + "filename": filename, + "doc_id": doc_id, + "size": file_stat.st_size, + "modified": file_stat.st_mtime, + }) + sessions.append({ + "session_id": session_id, + "files": files, + "file_count": len(files), + }) + total_files = len(files) + + return { + "status": "success", + "graphname": graphname, + "sessions": sessions, + "total_files": total_files, + } + + except Exception as e: + exc = traceback.format_exc() + logger.error(f"Error listing ingestion temp files for graph {graphname}: {e}") + logger.debug_pii(f"List error trace:\n{exc}") + raise HTTPException(status_code=500, detail=f"Error listing temp files: {str(e)}") + + +@router.delete(route_prefix + "/{graphname}/ingestion_temp/delete") +async def delete_ingestion_temp_files( + graphname: str, + credentials: Annotated[HTTPBase, Depends(security)], + session_id: str = None, + filename: str = None, +): + """ + Delete files from ingestion temp folder. + """ + try: + base_temp_dir = os.path.join("uploads", "ingestion_temp", graphname) + + if not session_id: + raise HTTPException(status_code=400, detail="session_id is required") + + session_dir = os.path.join(base_temp_dir, session_id) + + if not os.path.exists(session_dir): + return { + "status": "success", + "message": f"No temp files found for session {session_id}", + "deleted_files": [], + } + + deleted_files = [] + + if filename: + # Delete specific file + file_path = os.path.join(session_dir, filename) + if os.path.exists(file_path) and os.path.isfile(file_path): + os.remove(file_path) + deleted_files.append(filename) + logger.info(f"Deleted temp file {filename} from session {session_id}") + + # If session folder is now empty, remove it + if not os.listdir(session_dir): + os.rmdir(session_dir) + logger.info(f"Removed empty session folder {session_id}") + else: + raise HTTPException(status_code=404, detail=f"File {filename} not found") + else: + # Delete entire session folder + import shutil + for filename in os.listdir(session_dir): + if os.path.isfile(os.path.join(session_dir, filename)): + deleted_files.append(filename) + + shutil.rmtree(session_dir) + logger.info(f"Deleted session folder {session_id} for graph {graphname}") + + return { + "status": "success", + "message": f"Successfully deleted {len(deleted_files)} file(s)", + "deleted_files": deleted_files, + "session_id": session_id, + } + + except HTTPException: + raise + except Exception as e: + exc = traceback.format_exc() + logger.error(f"Error deleting ingestion temp files for graph {graphname}: {e}") + logger.debug_pii(f"Delete error trace:\n{exc}") + raise HTTPException(status_code=500, detail=f"Error deleting temp files: {str(e)}") + diff --git a/graphrag/app/supportai/supportai.py b/graphrag/app/supportai/supportai.py index 6b93df0..88542dc 100644 --- a/graphrag/app/supportai/supportai.py +++ b/graphrag/app/supportai/supportai.py @@ -489,14 +489,37 @@ def create_ingest( server_processing_result = extractor.process_folder(data_path, graphname=graphname) if server_processing_result.get("statusCode") != 200: raise Exception(f"Server folder processing failed: {server_processing_result}") - else: - logger.info(f"Server folder processing completed successfully: {server_processing_result}") - - res_ingest_config["server_jobs"] = server_processing_result.get("documents", []) + + # Log only summary, NOT the full documents to avoid memory logging + logger.info(f"Server folder processing completed: {server_processing_result.get('message')}") + + # Save processed documents to temporary folder instead of keeping in memory + temp_session_id = str(uuid.uuid4()) + temp_folder = os.path.join("uploads", "ingestion_temp", graphname, temp_session_id) + os.makedirs(temp_folder, exist_ok=True) + + documents = server_processing_result.get("documents", []) + doc_count = len(documents) + + # Save each document as a separate JSON file + for idx, doc_data in enumerate(documents): + doc_filename = f"doc_{idx}_{doc_data.get('doc_id', 'unknown')}.json" + doc_filepath = os.path.join(temp_folder, doc_filename) + with open(doc_filepath, 'w', encoding='utf-8') as f: + json.dump(doc_data, f, ensure_ascii=False, indent=2) + + # Clear documents from memory immediately after saving + documents.clear() + server_processing_result.clear() + + logger.info(f"Saved {doc_count} processed documents to {temp_folder}") + + res_ingest_config["temp_session_id"] = temp_session_id + res_ingest_config["temp_folder"] = temp_folder + res_ingest_config["file_count"] = doc_count res_ingest_config["data_source_id"] = "DocumentContent" - # Use a placeholder path that doesn't start with "/" to avoid pyTigerGraph treating it as a file - # The actual folder path is stored in server_jobs, this is just for the API call - res["data_path"] = "in_response" + # Use a placeholder path to indicate temp storage + res["data_path"] = "in_temp_storage" res["data_source_id"] = res_ingest_config except Exception as e: raise Exception(f"Error during server folder processing: {e}") @@ -650,13 +673,30 @@ def ingest( try: processed_files = [] data_source_id = ingest_config.get("data_source_id", "DocumentContent") - if ingest_config.get("server_jobs"): - for doc_data in ingest_config.get("server_jobs"): + + # Read from temporary folder + temp_folder = ingest_config.get("temp_folder") + if not temp_folder or not os.path.exists(temp_folder): + raise Exception(f"Temporary folder not found: {temp_folder}") + + # Read all JSON files from temp folder + json_files = [f for f in os.listdir(temp_folder) if f.endswith('.json')] + logger.info(f"Reading {len(json_files)} documents from {temp_folder}") + + for json_filename in json_files: + json_filepath = os.path.join(temp_folder, json_filename) + try: + with open(json_filepath, 'r', encoding='utf-8') as f: + doc_data = json.load(f) + if not doc_data.get("doc_id"): + logger.warning(f"Skipping invalid document: {json_filename}") continue # Skip documents with neither content nor image_data if not doc_data.get("content") and not doc_data.get("image_data"): + logger.warning(f"Skipping document with no content: {json_filename}") continue + if doc_data.get("image_data"): payload = { "doc_id": doc_data.get("doc_id", ""), @@ -684,6 +724,18 @@ def ingest( 'parent_doc': doc_data.get("parent_doc", ""), }) logger.info(f"Data uploading done for doc_id: {doc_data.get('doc_id', 'unknown')}") + except Exception as file_error: + logger.error(f"Error processing file {json_filename}: {file_error}") + continue + + # Clean up temp folder after successful ingestion + try: + import shutil + shutil.rmtree(temp_folder) + logger.info(f"Cleaned up temporary folder: {temp_folder}") + except Exception as cleanup_error: + logger.warning(f"Failed to cleanup temp folder {temp_folder}: {cleanup_error}") + except Exception as e: raise Exception(f"Error during server markdown extraction and TigerGraph loading: {e}") return { From e25ba1eea8467f9bbc879ed554fac007ef21fa99 Mon Sep 17 00:00:00 2001 From: Prins Kumar Date: Mon, 24 Nov 2025 20:31:36 +0530 Subject: [PATCH 07/15] Add direct ingestion option with checkbox to skip file review --- graphrag-ui/src/pages/Setup.tsx | 37 ++++++++++++++++++++++++++++++--- 1 file changed, 34 insertions(+), 3 deletions(-) diff --git a/graphrag-ui/src/pages/Setup.tsx b/graphrag-ui/src/pages/Setup.tsx index c844896..e86eefb 100644 --- a/graphrag-ui/src/pages/Setup.tsx +++ b/graphrag-ui/src/pages/Setup.tsx @@ -62,6 +62,7 @@ const Setup = () => { const [tempFiles, setTempFiles] = useState([]); const [showTempFiles, setShowTempFiles] = useState(false); const [ingestJobData, setIngestJobData] = useState(null); + const [directIngestion, setDirectIngestion] = useState(false); // Refresh state const [refreshOpen, setRefreshOpen] = useState(false); @@ -588,8 +589,8 @@ const Setup = () => { // Check if temp files were created (for server data source) const sessionId = createData.data_source_id?.temp_session_id; - if (sessionId) { - // Files are saved to temp storage - show them for review + if (sessionId && !directIngestion) { + // Files are saved to temp storage - show them for review (only if not direct ingestion) setTempSessionId(sessionId); setIngestJobData({ load_job_id: createData.load_job_id, @@ -600,7 +601,7 @@ const Setup = () => { await fetchTempFiles(sessionId); setIsIngesting(false); } else { - // No temp files (e.g., S3 Bedrock) - proceed directly to ingest + // No temp files (e.g., S3 Bedrock) OR direct ingestion enabled - proceed directly to ingest setIngestMessage("Step 2/2: Running document ingest..."); const loadingInfo = { @@ -1350,6 +1351,21 @@ const Setup = () => {

Process uploaded files and add them to the knowledge graph

+ + {/* Direct Ingestion Checkbox */} +
+ setDirectIngestion(e.target.checked)} + className="mr-2 h-4 w-4 rounded border-gray-300 text-blue-600 focus:ring-blue-500" + /> + +
+ - -

- Review the processed files below. You can delete any file before ingesting. -

-
- {tempFiles.map((file, index) => ( -
-
-

- {file.doc_id} -

-

- {(file.size / 1024).toFixed(2)} KB -

-
- -
- ))} -
- - - )} )} @@ -1771,23 +1882,9 @@ const Setup = () => { Process downloaded files and add them to the knowledge graph

- {/* Direct Ingestion Checkbox */} -
- setDirectIngestion(e.target.checked)} - className="mr-2 h-4 w-4 rounded border-gray-300 text-blue-600 focus:ring-blue-500" - /> - -
- - -

- Review the processed files below. You can delete any file before ingesting. -

-
- {tempFiles.map((file, index) => ( -
-
-

- {file.doc_id} -

-

- {(file.size / 1024).toFixed(2)} KB -

-
- -
- ))} -
- - - )} )} From 37ccf7089b0acff64cf9aba7827061df097af4ac Mon Sep 17 00:00:00 2001 From: Prins Kumar Date: Tue, 18 Nov 2025 18:32:55 +0530 Subject: [PATCH 09/15] Update README for OpenAI and Bedrock config, add pymupdf4llm license --- README.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/README.md b/README.md index 13c88b3..9469ad6 100644 --- a/README.md +++ b/README.md @@ -482,7 +482,7 @@ In addition to the `OPENAI_API_KEY`, `llm_model` and `model_name` can be edited "model_kwargs": { "temperature": 0 }, - "prompt_path": "./common/prompts/openai_gpt4/" + "prompt_path": "./app/prompts/openai_gpt4/" }, "multimodal_service": { "llm_service": "openai", @@ -614,7 +614,7 @@ In addition to the `AZURE_OPENAI_ENDPOINT`, `AZURE_OPENAI_API_KEY`, and `azure_d "temperature": 0, "max_tokens": 4096 }, - "prompt_path": "./common/prompts/openai_gpt4/" + "prompt_path": "./app/prompts/aws_bedrock_claude3haiku/" }, "multimodal_service": { "llm_service": "bedrock", From 74ce839bc08266433ba26d867eb39cd882a17f69 Mon Sep 17 00:00:00 2001 From: Prins Kumar Date: Tue, 18 Nov 2025 22:55:22 +0530 Subject: [PATCH 10/15] Fix prompt_path to use ./common/prompts/ for OpenAI and Bedrock --- README.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/README.md b/README.md index 9469ad6..13c88b3 100644 --- a/README.md +++ b/README.md @@ -482,7 +482,7 @@ In addition to the `OPENAI_API_KEY`, `llm_model` and `model_name` can be edited "model_kwargs": { "temperature": 0 }, - "prompt_path": "./app/prompts/openai_gpt4/" + "prompt_path": "./common/prompts/openai_gpt4/" }, "multimodal_service": { "llm_service": "openai", @@ -614,7 +614,7 @@ In addition to the `AZURE_OPENAI_ENDPOINT`, `AZURE_OPENAI_API_KEY`, and `azure_d "temperature": 0, "max_tokens": 4096 }, - "prompt_path": "./app/prompts/aws_bedrock_claude3haiku/" + "prompt_path": "./common/prompts/openai_gpt4/" }, "multimodal_service": { "llm_service": "bedrock", From 5090a71ded00a672bf248e3af5143ce8842bbe18 Mon Sep 17 00:00:00 2001 From: Prins Kumar Date: Fri, 21 Nov 2025 20:57:53 +0530 Subject: [PATCH 11/15] bug fixes --- graphrag/app/supportai/supportai.py | 79 +++++++++-------------------- 1 file changed, 23 insertions(+), 56 deletions(-) diff --git a/graphrag/app/supportai/supportai.py b/graphrag/app/supportai/supportai.py index 88542dc..8dea43a 100644 --- a/graphrag/app/supportai/supportai.py +++ b/graphrag/app/supportai/supportai.py @@ -501,18 +501,17 @@ def create_ingest( documents = server_processing_result.get("documents", []) doc_count = len(documents) - # Save each document as a separate JSON file - for idx, doc_data in enumerate(documents): - doc_filename = f"doc_{idx}_{doc_data.get('doc_id', 'unknown')}.json" - doc_filepath = os.path.join(temp_folder, doc_filename) - with open(doc_filepath, 'w', encoding='utf-8') as f: - json.dump(doc_data, f, ensure_ascii=False, indent=2) + # Save all documents to a single JSONL file (our new logic) + jsonl_filepath = os.path.join(temp_folder, "processed_documents.jsonl") + with open(jsonl_filepath, 'w', encoding='utf-8') as f: + for doc_data in documents: + f.write(json.dumps(doc_data, ensure_ascii=False) + '\n') # Clear documents from memory immediately after saving documents.clear() server_processing_result.clear() - logger.info(f"Saved {doc_count} processed documents to {temp_folder}") + logger.info(f"Saved {doc_count} processed documents to {jsonl_filepath}") res_ingest_config["temp_session_id"] = temp_session_id res_ingest_config["temp_folder"] = temp_folder @@ -671,7 +670,6 @@ def ingest( } elif ingest_config.get("data_source") == "server": try: - processed_files = [] data_source_id = ingest_config.get("data_source_id", "DocumentContent") # Read from temporary folder @@ -679,54 +677,23 @@ def ingest( if not temp_folder or not os.path.exists(temp_folder): raise Exception(f"Temporary folder not found: {temp_folder}") - # Read all JSON files from temp folder - json_files = [f for f in os.listdir(temp_folder) if f.endswith('.json')] - logger.info(f"Reading {len(json_files)} documents from {temp_folder}") + # Read the processed_documents.jsonl file (our new logic) + jsonl_file = os.path.join(temp_folder, "processed_documents.jsonl") + if not os.path.exists(jsonl_file): + raise Exception(f"Processed documents file not found: {jsonl_file}") - for json_filename in json_files: - json_filepath = os.path.join(temp_folder, json_filename) - try: - with open(json_filepath, 'r', encoding='utf-8') as f: - doc_data = json.load(f) - - if not doc_data.get("doc_id"): - logger.warning(f"Skipping invalid document: {json_filename}") - continue - # Skip documents with neither content nor image_data - if not doc_data.get("content") and not doc_data.get("image_data"): - logger.warning(f"Skipping document with no content: {json_filename}") - continue - - if doc_data.get("image_data"): - payload = { - "doc_id": doc_data.get("doc_id", ""), - "doc_type": "image", - "image_data": doc_data.get("image_data", ""), - "image_format": doc_data.get("image_format", "jpg"), - "image_description": doc_data.get("image_description", ""), - "parent_doc": doc_data.get("parent_doc", ""), - "page_number": doc_data.get("page_number", 0), - "width": doc_data.get("width", 0), - "height": doc_data.get("height", 0), - "position": doc_data.get("position", 0), - "content": "" - } - else: - payload = { - "doc_id": doc_data.get("doc_id", ""), - "doc_type": doc_data.get("doc_type", "markdown"), - "content": doc_data.get("content", "") - } - payload_json = json.dumps(payload) - conn.runLoadingJobWithData(payload_json, data_source_id, loader_info.load_job_id) - processed_files.append({ - 'file_path': doc_data.get("doc_id", ""), - 'parent_doc': doc_data.get("parent_doc", ""), - }) - logger.info(f"Data uploading done for doc_id: {doc_data.get('doc_id', 'unknown')}") - except Exception as file_error: - logger.error(f"Error processing file {json_filename}: {file_error}") - continue + # Read entire JSONL content as a single string + with open(jsonl_file, 'r', encoding='utf-8') as f: + jsonl_content = f.read() + + # Count documents for logging + document_count = jsonl_content.count('\n') if jsonl_content.strip() else 0 + logger.info(f"Ingesting {document_count} documents from {jsonl_file}") + + # Pass entire JSONL content in ONE call (efficient!) + conn.runLoadingJobWithData(jsonl_content, data_source_id, loader_info.load_job_id) + + logger.info(f"Successfully ingested {document_count} documents") # Clean up temp folder after successful ingestion try: @@ -740,7 +707,7 @@ def ingest( raise Exception(f"Error during server markdown extraction and TigerGraph loading: {e}") return { "job_name": loader_info.load_job_id, - "summary": processed_files + "summary": f"Data ingestion successful - processed {document_count} documents" } else: raise Exception("Data source and file format combination not implemented") From 5417f8872fd55ef7d2598b95314b6e7d05eb9fd0 Mon Sep 17 00:00:00 2001 From: Prins Kumar Date: Mon, 1 Dec 2025 17:16:28 +0530 Subject: [PATCH 12/15] Merge latest main and consolidate markdown_parsing.py into text_extractors.py --- common/requirements.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/common/requirements.txt b/common/requirements.txt index f0022f3..3bbd096 100644 --- a/common/requirements.txt +++ b/common/requirements.txt @@ -108,7 +108,7 @@ ordered-set==4.1.0 orjson==3.10.18 packaging==24.2 pandas==2.2.3 -#pathtools==0.1.2 +pathtools==0.1.2 pillow==11.2.1 #PyMuPDF==1.26.4 pymupdf4llm==0.2.0 From eefcd67d012fd787a2555ba2f7825c585ae0a211 Mon Sep 17 00:00:00 2001 From: Prins Kumar Date: Wed, 3 Dec 2025 15:23:46 +0530 Subject: [PATCH 13/15] Redesign temp file storage: save immediately during file processing instead of after --- common/utils/text_extractors.py | 77 ++++++++++++++++++++++++----- graphrag/app/supportai/supportai.py | 14 +++++- 2 files changed, 77 insertions(+), 14 deletions(-) diff --git a/common/utils/text_extractors.py b/common/utils/text_extractors.py index ec5b140..4f22822 100644 --- a/common/utils/text_extractors.py +++ b/common/utils/text_extractors.py @@ -42,10 +42,11 @@ def __init__(self): '.jpg': 'image/jpeg' } - async def _process_file_async(self, file_path, folder_path_obj, graphname): + async def _process_file_async(self, file_path, folder_path_obj, graphname, temp_folder=None, file_counter=None): """ Async helper to process a single file. Runs in thread pool to avoid blocking on I/O operations. + If temp_folder is provided, saves documents immediately and returns metadata only. """ try: loop = asyncio.get_event_loop() @@ -57,6 +58,27 @@ async def _process_file_async(self, file_path, folder_path_obj, graphname): graphname ) + # If temp_folder provided, save immediately and return metadata only + if temp_folder and doc_entries: + saved_files = [] + for idx, doc_data in enumerate(doc_entries): + # Use file_counter for unique naming across all files + counter_val = next(file_counter) if file_counter else idx + doc_filename = f"doc_{counter_val}_{doc_data.get('doc_id', 'unknown')}.json" + doc_filepath = os.path.join(temp_folder, doc_filename) + with open(doc_filepath, 'w', encoding='utf-8') as f: + json.dump(doc_data, f, ensure_ascii=False, indent=2) + saved_files.append(doc_filename) + + # Return metadata only, not full documents (memory efficient) + return { + 'success': True, + 'file_path': str(file_path), + 'saved_files': saved_files, + 'num_documents': len(doc_entries) + } + + # No temp_folder - return documents in memory (legacy behavior) return { 'success': True, 'file_path': str(file_path), @@ -72,10 +94,11 @@ async def _process_file_async(self, file_path, folder_path_obj, graphname): logger.warning(f"Failed to process file {file_path}: {e}") return {'success': False, 'file_path': str(file_path), 'error': str(e)} - async def _process_folder_async(self, folder_path, graphname=None, max_concurrent=10): + async def _process_folder_async(self, folder_path, graphname=None, max_concurrent=10, temp_folder=None): """ Async version of process_folder for parallel file processing. This prevents conflicts when multiple users process folders simultaneously. + If temp_folder is provided, saves documents immediately to disk instead of holding in memory. """ logger.info(f"Processing local folder ASYNC: {folder_path} for graph: {graphname} (max_concurrent={max_concurrent})") @@ -87,6 +110,11 @@ async def _process_folder_async(self, folder_path, graphname=None, max_concurren if not folder_path_obj.is_dir(): raise Exception(f"Path is not a directory: {folder_path}") + # Create temp folder if provided + if temp_folder: + os.makedirs(temp_folder, exist_ok=True) + logger.info(f"Saving processed documents to: {temp_folder}") + def safe_walk(path): try: for item in path.iterdir(): @@ -111,16 +139,20 @@ def safe_walk(path): logger.info(f"Found {len(files_to_process)} files to process") semaphore = asyncio.Semaphore(max_concurrent) + + # Thread-safe counter for unique file naming + file_counter = iter(range(100000)) if temp_folder else None async def process_with_semaphore(file_path): async with semaphore: - return await self._process_file_async(file_path, folder_path_obj, graphname) + return await self._process_file_async(file_path, folder_path_obj, graphname, temp_folder, file_counter) tasks = [process_with_semaphore(fp) for fp in files_to_process] results = await asyncio.gather(*tasks, return_exceptions=True) all_documents = [] processed_files_info = [] + total_saved_files = [] for result in results: if isinstance(result, Exception): @@ -128,10 +160,15 @@ async def process_with_semaphore(file_path): continue if result.get('success'): - all_documents.extend(result.get('documents', [])) + # If temp_folder was used, documents are saved to disk + if temp_folder: + total_saved_files.extend(result.get('saved_files', [])) + else: + all_documents.extend(result.get('documents', [])) + processed_files_info.append({ 'file_path': result['file_path'], - 'num_documents': result.get('num_documents', len(result.get('documents', []))), + 'num_documents': result.get('num_documents', 0), 'status': 'success' }) else: @@ -141,23 +178,39 @@ async def process_with_semaphore(file_path): 'error': result.get('error', 'Unknown error') }) - logger.info(f"Processed {len(processed_files_info)} files, extracted {len(all_documents)} total documents") + total_docs = len(total_saved_files) if temp_folder else len(all_documents) + logger.info(f"Processed {len(processed_files_info)} files, extracted {total_docs} total documents") - return { + response = { 'statusCode': 200, - 'message': f'Processed {len(processed_files_info)} files, {len(all_documents)} documents', - 'documents': all_documents, + 'message': f'Processed {len(processed_files_info)} files, {total_docs} documents', 'files': processed_files_info, - 'num_documents': len(all_documents) + 'num_documents': total_docs } + + # Only include documents in response if NOT saving to temp_folder + if temp_folder: + response['saved_to_temp'] = True + response['temp_folder'] = temp_folder + response['saved_files'] = total_saved_files + else: + response['documents'] = all_documents + + return response - def process_folder(self, folder_path, graphname=None): + def process_folder(self, folder_path, graphname=None, temp_folder=None): """ Process local folder with multiple file formats and extract text content. Uses async processing internally for parallel file handling. + + Args: + folder_path: Path to the folder containing files to process + graphname: Name of the graph (for context) + temp_folder: Optional path to save processed documents immediately. + If provided, documents are saved to disk instead of returned in memory. """ logger.info(f"Processing local folder: {folder_path} for graph: {graphname}") - return asyncio.run(self._process_folder_async(folder_path, graphname)) + return asyncio.run(self._process_folder_async(folder_path, graphname, temp_folder=temp_folder)) def extract_text_from_file_with_images_as_docs(file_path, graphname=None): diff --git a/graphrag/app/supportai/supportai.py b/graphrag/app/supportai/supportai.py index 8dea43a..63cfa3d 100644 --- a/graphrag/app/supportai/supportai.py +++ b/graphrag/app/supportai/supportai.py @@ -485,12 +485,22 @@ def create_ingest( if data_path is None: raise Exception("Folder path not provided for server processing") try: + # Create temp folder BEFORE processing so extractor can save directly + temp_session_id = str(uuid.uuid4()) + temp_folder = os.path.join("uploads", "ingestion_temp", graphname, temp_session_id) + + # Process files and save immediately to temp folder (memory efficient) extractor = TextExtractor() - server_processing_result = extractor.process_folder(data_path, graphname=graphname) + server_processing_result = extractor.process_folder( + data_path, + graphname=graphname, + temp_folder=temp_folder # Extractor saves files as it processes + ) + if server_processing_result.get("statusCode") != 200: raise Exception(f"Server folder processing failed: {server_processing_result}") - # Log only summary, NOT the full documents to avoid memory logging + doc_count = server_processing_result.get("num_documents", 0) logger.info(f"Server folder processing completed: {server_processing_result.get('message')}") # Save processed documents to temporary folder instead of keeping in memory From 555a500f3d24bec8fafd556847c324b5f8f5761f Mon Sep 17 00:00:00 2001 From: Prins Kumar Date: Thu, 4 Dec 2025 21:02:50 +0530 Subject: [PATCH 14/15] files processing with processed file json creation then using single job to insert --- common/utils/text_extractors.py | 191 +++++++++++++++++++--------- graphrag-ui/src/pages/Setup.tsx | 107 ++++++++-------- graphrag/app/routers/ui.py | 140 +++++++++++++++----- graphrag/app/supportai/supportai.py | 43 ++----- 4 files changed, 310 insertions(+), 171 deletions(-) diff --git a/common/utils/text_extractors.py b/common/utils/text_extractors.py index 4f22822..9b5b652 100644 --- a/common/utils/text_extractors.py +++ b/common/utils/text_extractors.py @@ -42,11 +42,11 @@ def __init__(self): '.jpg': 'image/jpeg' } - async def _process_file_async(self, file_path, folder_path_obj, graphname, temp_folder=None, file_counter=None): + async def _process_file_async(self, file_path, folder_path_obj, graphname, temp_folder, jsonl_file, jsonl_lock): """ Async helper to process a single file. Runs in thread pool to avoid blocking on I/O operations. - If temp_folder is provided, saves documents immediately and returns metadata only. + Appends documents immediately to JSONL file. """ try: loop = asyncio.get_event_loop() @@ -58,31 +58,21 @@ async def _process_file_async(self, file_path, folder_path_obj, graphname, temp_ graphname ) - # If temp_folder provided, save immediately and return metadata only - if temp_folder and doc_entries: - saved_files = [] - for idx, doc_data in enumerate(doc_entries): - # Use file_counter for unique naming across all files - counter_val = next(file_counter) if file_counter else idx - doc_filename = f"doc_{counter_val}_{doc_data.get('doc_id', 'unknown')}.json" - doc_filepath = os.path.join(temp_folder, doc_filename) - with open(doc_filepath, 'w', encoding='utf-8') as f: - json.dump(doc_data, f, ensure_ascii=False, indent=2) - saved_files.append(doc_filename) - - # Return metadata only, not full documents (memory efficient) - return { - 'success': True, - 'file_path': str(file_path), - 'saved_files': saved_files, - 'num_documents': len(doc_entries) - } + # Append each document to JSONL file immediately + if doc_entries: + # Use lock to ensure thread-safe writing to JSONL file + async with jsonl_lock: + await loop.run_in_executor( + None, + self._append_to_jsonl, + jsonl_file, + doc_entries + ) - # No temp_folder - return documents in memory (legacy behavior) + # Return metadata only, documents already saved to JSONL return { 'success': True, 'file_path': str(file_path), - 'documents': doc_entries, 'num_documents': len(doc_entries) } @@ -93,12 +83,21 @@ async def _process_file_async(self, file_path, folder_path_obj, graphname, temp_ except Exception as e: logger.warning(f"Failed to process file {file_path}: {e}") return {'success': False, 'file_path': str(file_path), 'error': str(e)} + + def _append_to_jsonl(self, jsonl_file, doc_entries): + """ + Append document entries to JSONL file. + Each document is written as a separate line. + """ + with open(jsonl_file, 'a', encoding='utf-8') as f: + for doc_data in doc_entries: + json_line = json.dumps(doc_data, ensure_ascii=False) + f.write(json_line + '\n') - async def _process_folder_async(self, folder_path, graphname=None, max_concurrent=10, temp_folder=None): + async def _process_folder_async(self, folder_path, graphname, temp_folder, max_concurrent=10): """ Async version of process_folder for parallel file processing. - This prevents conflicts when multiple users process folders simultaneously. - If temp_folder is provided, saves documents immediately to disk instead of holding in memory. + Saves all documents immediately to a single JSONL file as they are processed. """ logger.info(f"Processing local folder ASYNC: {folder_path} for graph: {graphname} (max_concurrent={max_concurrent})") @@ -110,10 +109,12 @@ async def _process_folder_async(self, folder_path, graphname=None, max_concurren if not folder_path_obj.is_dir(): raise Exception(f"Path is not a directory: {folder_path}") - # Create temp folder if provided - if temp_folder: - os.makedirs(temp_folder, exist_ok=True) - logger.info(f"Saving processed documents to: {temp_folder}") + # Create temp folder and JSONL file + os.makedirs(temp_folder, exist_ok=True) + jsonl_file = os.path.join(temp_folder, "processed_documents.jsonl") + # Create async lock for thread-safe JSONL writing + jsonl_lock = asyncio.Lock() + logger.info(f"Saving processed documents to: {jsonl_file}") def safe_walk(path): try: @@ -139,20 +140,16 @@ def safe_walk(path): logger.info(f"Found {len(files_to_process)} files to process") semaphore = asyncio.Semaphore(max_concurrent) - - # Thread-safe counter for unique file naming - file_counter = iter(range(100000)) if temp_folder else None async def process_with_semaphore(file_path): async with semaphore: - return await self._process_file_async(file_path, folder_path_obj, graphname, temp_folder, file_counter) + return await self._process_file_async(file_path, folder_path_obj, graphname, temp_folder, jsonl_file, jsonl_lock) tasks = [process_with_semaphore(fp) for fp in files_to_process] results = await asyncio.gather(*tasks, return_exceptions=True) - all_documents = [] processed_files_info = [] - total_saved_files = [] + total_docs = 0 for result in results: if isinstance(result, Exception): @@ -160,15 +157,12 @@ async def process_with_semaphore(file_path): continue if result.get('success'): - # If temp_folder was used, documents are saved to disk - if temp_folder: - total_saved_files.extend(result.get('saved_files', [])) - else: - all_documents.extend(result.get('documents', [])) + num_docs = result.get('num_documents', 0) + total_docs += num_docs processed_files_info.append({ 'file_path': result['file_path'], - 'num_documents': result.get('num_documents', 0), + 'num_documents': num_docs, 'status': 'success' }) else: @@ -178,39 +172,118 @@ async def process_with_semaphore(file_path): 'error': result.get('error', 'Unknown error') }) - total_docs = len(total_saved_files) if temp_folder else len(all_documents) logger.info(f"Processed {len(processed_files_info)} files, extracted {total_docs} total documents") - response = { + return { 'statusCode': 200, 'message': f'Processed {len(processed_files_info)} files, {total_docs} documents', 'files': processed_files_info, - 'num_documents': total_docs + 'num_documents': total_docs, + 'temp_folder': temp_folder, + 'jsonl_file': jsonl_file } - - # Only include documents in response if NOT saving to temp_folder - if temp_folder: - response['saved_to_temp'] = True - response['temp_folder'] = temp_folder - response['saved_files'] = total_saved_files - else: - response['documents'] = all_documents - - return response - def process_folder(self, folder_path, graphname=None, temp_folder=None): + def process_folder(self, folder_path, graphname, temp_folder): """ Process local folder with multiple file formats and extract text content. Uses async processing internally for parallel file handling. + Saves all documents to JSONL file immediately as they are processed. Args: folder_path: Path to the folder containing files to process graphname: Name of the graph (for context) - temp_folder: Optional path to save processed documents immediately. - If provided, documents are saved to disk instead of returned in memory. + temp_folder: Path to save processed documents as JSONL file """ logger.info(f"Processing local folder: {folder_path} for graph: {graphname}") - return asyncio.run(self._process_folder_async(folder_path, graphname, temp_folder=temp_folder)) + return asyncio.run(self._process_folder_async(folder_path, graphname, temp_folder)) + + def delete_file_from_jsonl(self, temp_folder, filename): + """ + Delete all documents related to a specific file from the JSONL file. + + Args: + temp_folder: Path to the temp folder containing processed_documents.jsonl + filename: Original filename (e.g., "report.pdf", "stock_gs200.jpg") + + Returns: + dict with status and number of documents removed + """ + jsonl_file = os.path.join(temp_folder, "processed_documents.jsonl") + + if not os.path.exists(jsonl_file): + logger.warning(f"JSONL file not found: {jsonl_file}") + return {'success': False, 'error': 'JSONL file not found'} + + # Get base name without extension to match doc_id + base_name = Path(filename).stem + logger.info(f"Deleting documents for file: {filename} (base_name: '{base_name}')") + + # Read all lines and filter out ones matching this file + remaining_lines = [] + removed_count = 0 + removed_doc_ids = [] + + try: + with open(jsonl_file, 'r', encoding='utf-8') as f: + for line_num, line in enumerate(f, 1): + line = line.strip() + if not line: + continue + + try: + doc_data = json.loads(line) + doc_id = doc_data.get('doc_id', '') + + # Check if doc_id matches the base_name or starts with base_name_ + # Handles: "stock_gs200" == "stock_gs200" or "stock_gs200_image_1".startswith("stock_gs200_") + if doc_id == base_name or doc_id.startswith(f"{base_name}_"): + removed_count += 1 + removed_doc_ids.append(doc_id) + logger.info(f"Removing document: {doc_id}") + else: + remaining_lines.append(line) + except json.JSONDecodeError as e: + logger.warning(f"Skipping invalid JSON at line {line_num}: {e}") + # Keep invalid lines in case they're important + remaining_lines.append(line) + + if removed_count == 0: + logger.warning(f"No documents found matching base_name: '{base_name}'") + return { + 'success': False, + 'error': f'No documents found for {filename}', + 'removed_count': 0 + } + + # If no lines remain, delete the entire temp folder + if not remaining_lines: + logger.info(f"No documents remaining, deleting temp folder: {temp_folder}") + import shutil + shutil.rmtree(temp_folder, ignore_errors=True) + return { + 'success': True, + 'removed_count': removed_count, + 'removed_doc_ids': removed_doc_ids, + 'temp_folder_deleted': True + } + + # Write remaining lines back to JSONL + with open(jsonl_file, 'w', encoding='utf-8') as f: + for line in remaining_lines: + f.write(line + '\n') + + logger.info(f"Removed {removed_count} documents ({', '.join(removed_doc_ids)}), {len(remaining_lines)} remaining") + return { + 'success': True, + 'removed_count': removed_count, + 'removed_doc_ids': removed_doc_ids, + 'remaining_count': len(remaining_lines), + 'temp_folder_deleted': False + } + + except Exception as e: + logger.error(f"Error deleting from JSONL: {e}") + return {'success': False, 'error': str(e)} def extract_text_from_file_with_images_as_docs(file_path, graphname=None): diff --git a/graphrag-ui/src/pages/Setup.tsx b/graphrag-ui/src/pages/Setup.tsx index 17f952c..c3e4689 100644 --- a/graphrag-ui/src/pages/Setup.tsx +++ b/graphrag-ui/src/pages/Setup.tsx @@ -398,11 +398,18 @@ const Setup = () => { const data = await response.json(); if (data.status === "success") { - setDownloadMessage(`✅ ${data.message}. Processing...`); + setDownloadMessage(`✅ ${data.message}. Processed ${data.doc_count || 0} document(s)`); await fetchDownloadedFiles(); - // Step 2: Call create_ingest to process downloaded files - await handleCreateIngestAfterUpload("downloaded"); + // Save session ID from automatic processing + if (data.temp_session_id) { + setTempSessionId(data.temp_session_id); + await fetchTempFiles(data.temp_session_id); + } + } else if (data.status === "partial_success") { + setDownloadMessage(`⚠️ ${data.message}`); + await fetchDownloadedFiles(); + // Don't call create_ingest if processing already attempted } else if (data.status === "warning") { setDownloadMessage(`⚠️ ${data.message}`); } else { @@ -421,22 +428,26 @@ const Setup = () => { try { const creds = localStorage.getItem("creds"); - const response = await fetch( - `/ui/${ingestGraphName}/cloud/delete?filename=${encodeURIComponent(filename)}`, - { - method: "DELETE", - headers: { Authorization: `Basic ${creds}` }, - } - ); - const data = await response.json(); - // Also delete corresponding temp files if session exists + // Build URL with session_id if available + let deleteUrl = `/ui/${ingestGraphName}/cloud/delete?filename=${encodeURIComponent(filename)}`; if (tempSessionId) { - await handleDeleteTempFilesForOriginal(filename); + deleteUrl += `&session_id=${encodeURIComponent(tempSessionId)}`; } + const response = await fetch(deleteUrl, { + method: "DELETE", + headers: { Authorization: `Basic ${creds}` }, + }); + const data = await response.json(); + setDownloadMessage(`✅ ${data.message}`); await fetchDownloadedFiles(); + + // Refresh temp files list if session exists + if (tempSessionId) { + await fetchTempFiles(tempSessionId); + } } catch (error: any) { setDownloadMessage(`❌ Error: ${error.message}`); } @@ -451,13 +462,26 @@ const Setup = () => { try { const creds = localStorage.getItem("creds"); - const response = await fetch(`/ui/${ingestGraphName}/cloud/delete`, { + + // Build URL with session_id if available + let deleteUrl = `/ui/${ingestGraphName}/cloud/delete`; + if (tempSessionId) { + deleteUrl += `?session_id=${encodeURIComponent(tempSessionId)}`; + } + + const response = await fetch(deleteUrl, { method: "DELETE", headers: { Authorization: `Basic ${creds}` }, }); const data = await response.json(); setDownloadMessage(`✅ ${data.message}`); await fetchDownloadedFiles(); + + // Clear session ID and refresh temp files + if (tempSessionId) { + setTempSessionId(null); + setTempFiles([]); + } } catch (error: any) { setDownloadMessage(`❌ Error: ${error.message}`); } @@ -541,44 +565,24 @@ const Setup = () => { } try { - // Extract base name without extension (e.g., "document.pdf" -> "document") - const baseName = originalFilename.replace(/\.[^/.]+$/, ""); - console.log("Base name:", baseName); - const creds = localStorage.getItem("creds"); - // Fetch temp files to find matches - const response = await fetch(`/ui/${ingestGraphName}/ingestion_temp/list?session_id=${tempSessionId}`, { - headers: { Authorization: `Basic ${creds}` }, - }); - const data = await response.json(); - console.log("Temp files list response:", data); - - if (data.status === "success" && data.sessions.length > 0) { - const files = data.sessions[0].files || []; - console.log("All temp files:", files.map((f: any) => f.filename)); - - // Find temp files matching pattern: doc_{idx}_{baseName}*.json - const matchingFiles = files.filter((f: any) => f.filename.includes(`_${baseName}`)); - console.log("Matching files to delete:", matchingFiles.map((f: any) => f.filename)); - - // Delete each matching file - for (const file of matchingFiles) { - console.log("Deleting temp file:", file.filename); - const deleteResponse = await fetch( - `/ui/${ingestGraphName}/ingestion_temp/delete?session_id=${tempSessionId}&filename=${encodeURIComponent(file.filename)}`, - { - method: "DELETE", - headers: { Authorization: `Basic ${creds}` }, - } - ); - const deleteData = await deleteResponse.json(); - console.log("Delete response:", deleteData); + // Call the delete endpoint with the original filename + // The backend will handle removing all related documents from the JSONL file + const deleteResponse = await fetch( + `/ui/${ingestGraphName}/ingestion_temp/delete?session_id=${tempSessionId}&filename=${encodeURIComponent(originalFilename)}`, + { + method: "DELETE", + headers: { Authorization: `Basic ${creds}` }, } - - console.log(`Successfully deleted ${matchingFiles.length} temp file(s)`); + ); + const deleteData = await deleteResponse.json(); + console.log("Delete temp files response:", deleteData); + + if (deleteData.status === "success") { + console.log(`Successfully deleted processed documents for ${originalFilename}`); } else { - console.log("No temp files found or empty sessions"); + console.error("Failed to delete temp files:", deleteData); } } catch (error: any) { console.error("Error deleting temp files:", error); @@ -621,7 +625,8 @@ const Setup = () => { const ingestData = await ingestResponse.json(); console.log("Ingest response:", ingestData); - setIngestMessage(`✅ Data ingested successfully! Processed ${tempFiles.length} documents.`); + const docCount = ingestData.document_count || tempFiles.length; + setIngestMessage(`✅ Data ingested successfully! Processed ${docCount} document(s).`); // Clear temp state setTempFiles([]); @@ -691,7 +696,7 @@ const Setup = () => { data_source_id: createData.data_source_id, data_path: createData.data_path || createData.file_path, }); - setIngestMessage(`✅ Processed ${createData.data_source_id.file_count} files. Review them below before ingesting.`); + setIngestMessage(`✅ Files processed successfully. Review them below before ingesting.`); await fetchTempFiles(sessionId); setIsIngesting(false); } else { @@ -802,7 +807,7 @@ const Setup = () => { await handleRunIngest(); } else { // Save for later - files ready for ingestion - setUploadMessage(`✅ Successfully processed ${createData.data_source_id.file_count} files. Ready for ingestion.`); + setUploadMessage(`✅ Files processed successfully. Ready for ingestion.`); } } else { console.warn("No session ID returned from create_ingest"); diff --git a/graphrag/app/routers/ui.py b/graphrag/app/routers/ui.py index 9b012ec..9d8cc33 100644 --- a/graphrag/app/routers/ui.py +++ b/graphrag/app/routers/ui.py @@ -1058,7 +1058,8 @@ async def download_from_cloud( request_body: dict = Body(...), ): """ - Download files from cloud storage (S3, GCS, or Azure) to local directory. + Download files from cloud storage (S3, GCS, or Azure) to local directory + and automatically process them to create JSONL files for ingestion. Parameters: - graphname: The graph name to associate downloaded files with @@ -1252,14 +1253,49 @@ async def download_from_cloud( logger.info(f"Downloaded {len(downloaded_files)} file(s) from {provider} for graph {graphname}") - return { - "status": "success", - "message": f"Successfully downloaded {len(downloaded_files)} file(s) from {provider}", - "graphname": graphname, - "provider": provider, - "downloaded_files": downloaded_files, - "local_path": download_dir, - } + # Automatically process downloaded files to create JSONL + from common.utils.text_extractors import TextExtractor + temp_session_id = str(uuid.uuid4()) + temp_folder = os.path.join("uploads", "ingestion_temp", graphname, temp_session_id) + + try: + extractor = TextExtractor() + processing_result = extractor.process_folder( + download_dir, + graphname=graphname, + temp_folder=temp_folder + ) + + if processing_result.get("statusCode") != 200: + logger.error(f"Cloud file processing failed: {processing_result}") + raise Exception(f"Failed to process downloaded files: {processing_result}") + + doc_count = processing_result.get("num_documents", 0) + logger.info(f"Processed {doc_count} documents from downloaded files") + + return { + "status": "success", + "message": f"Successfully downloaded and processed {len(downloaded_files)} file(s) from {provider}", + "graphname": graphname, + "provider": provider, + "downloaded_files": downloaded_files, + "local_path": download_dir, + "temp_session_id": temp_session_id, + "temp_folder": temp_folder, + "doc_count": doc_count, + } + except Exception as e: + logger.error(f"Error processing downloaded files: {e}") + # Return success for download but warn about processing failure + return { + "status": "partial_success", + "message": f"Downloaded {len(downloaded_files)} file(s) but processing failed: {str(e)}", + "graphname": graphname, + "provider": provider, + "downloaded_files": downloaded_files, + "local_path": download_dir, + "processing_error": str(e), + } except HTTPException: raise @@ -1322,13 +1358,16 @@ async def delete_cloud_downloads( graphname: str, credentials: Annotated[HTTPBase, Depends(security)], filename: str = None, + session_id: str = None, ): """ Delete downloaded cloud files for a specific graph. + Also deletes corresponding processed documents from the JSONL file. Parameters: - graphname: The graph name whose downloaded files to clear - filename: If provided, only delete this specific file. Otherwise, delete all files. + - session_id: The session ID for the temp folder containing processed JSONL """ try: download_dir = os.path.join("downloaded_files_cloud", graphname) @@ -1343,9 +1382,26 @@ async def delete_cloud_downloads( deleted_files = [] if filename: - # Delete specific file + # Delete specific file AND its processed documents from JSONL file_path = os.path.join(download_dir, filename) if os.path.exists(file_path) and os.path.isfile(file_path): + # If session_id provided, also delete from JSONL + if session_id: + from common.utils.text_extractors import TextExtractor + extractor = TextExtractor() + + temp_folder = os.path.join("uploads", "ingestion_temp", graphname, session_id) + if os.path.exists(temp_folder): + # Delete from JSONL first + delete_result = extractor.delete_file_from_jsonl(temp_folder, filename) + logger.info(f"JSONL delete result for {filename}: {delete_result}") + + # If JSONL delete failed (and JSONL exists), warn but continue with file deletion + if not delete_result.get('success'): + logger.warning(f"Failed to delete from JSONL: {delete_result.get('error')}") + # Continue with file deletion even if JSONL deletion failed + + # Delete the original downloaded file os.remove(file_path) deleted_files.append(filename) logger.info(f"Deleted cloud download {filename} for graph {graphname}") @@ -1353,13 +1409,21 @@ async def delete_cloud_downloads( raise HTTPException(status_code=404, detail=f"File {filename} not found") else: # Delete all files in the directory - for filename in os.listdir(download_dir): - file_path = os.path.join(download_dir, filename) + for fname in os.listdir(download_dir): + file_path = os.path.join(download_dir, fname) if os.path.isfile(file_path): os.remove(file_path) - deleted_files.append(filename) + deleted_files.append(fname) - # Remove the directory if it's empty + # If session_id provided, delete the entire temp folder + if session_id: + temp_folder = os.path.join("uploads", "ingestion_temp", graphname, session_id) + if os.path.exists(temp_folder): + import shutil + shutil.rmtree(temp_folder, ignore_errors=True) + logger.info(f"Deleted temp folder for session {session_id}") + + # Remove the download directory if it's empty if not os.listdir(download_dir): os.rmdir(download_dir) @@ -1478,25 +1542,41 @@ async def delete_ingestion_temp_files( deleted_files = [] if filename: - # Delete specific file - file_path = os.path.join(session_dir, filename) - if os.path.exists(file_path) and os.path.isfile(file_path): - os.remove(file_path) - deleted_files.append(filename) - logger.info(f"Deleted temp file {filename} from session {session_id}") - - # If session folder is now empty, remove it - if not os.listdir(session_dir): - os.rmdir(session_dir) - logger.info(f"Removed empty session folder {session_id}") + # Delete processed documents from JSONL for this original filename + # Note: Original files are NOT in temp folder, only processed_documents.jsonl is here + from common.utils.text_extractors import TextExtractor + extractor = TextExtractor() + + # Delete from JSONL - MUST succeed + delete_result = extractor.delete_file_from_jsonl(session_dir, filename) + logger.info(f"JSONL delete result for {filename}: {delete_result}") + + # If JSONL delete failed, return error + if not delete_result.get('success'): + error_msg = delete_result.get('error', 'Unknown error') + logger.error(f"Failed to delete from JSONL: {error_msg}") + raise HTTPException(status_code=500, detail=f"Failed to delete processed documents: {error_msg}") + + deleted_files.append(filename) + logger.info(f"Deleted {delete_result.get('removed_count', 0)} processed documents for {filename} from JSONL") + + # Check if temp folder was deleted by JSONL cleanup + if delete_result.get('temp_folder_deleted'): + logger.info(f"Session folder {session_id} was automatically deleted (no documents remaining)") + elif not os.path.exists(session_dir): + logger.info(f"Session folder {session_id} was deleted") + elif not os.listdir(session_dir): + # Clean up empty session folder + os.rmdir(session_dir) + logger.info(f"Removed empty session folder {session_id}") else: - raise HTTPException(status_code=404, detail=f"File {filename} not found") + logger.info(f"Removed {delete_result.get('removed_count', 0)} documents from JSONL, {delete_result.get('remaining_count', 0)} remaining") else: - # Delete entire session folder + # Delete entire session folder (including JSONL) import shutil - for filename in os.listdir(session_dir): - if os.path.isfile(os.path.join(session_dir, filename)): - deleted_files.append(filename) + for fname in os.listdir(session_dir): + if os.path.isfile(os.path.join(session_dir, fname)): + deleted_files.append(fname) shutil.rmtree(session_dir) logger.info(f"Deleted session folder {session_id} for graph {graphname}") diff --git a/graphrag/app/supportai/supportai.py b/graphrag/app/supportai/supportai.py index 63cfa3d..a9dbbe0 100644 --- a/graphrag/app/supportai/supportai.py +++ b/graphrag/app/supportai/supportai.py @@ -502,26 +502,6 @@ def create_ingest( doc_count = server_processing_result.get("num_documents", 0) logger.info(f"Server folder processing completed: {server_processing_result.get('message')}") - - # Save processed documents to temporary folder instead of keeping in memory - temp_session_id = str(uuid.uuid4()) - temp_folder = os.path.join("uploads", "ingestion_temp", graphname, temp_session_id) - os.makedirs(temp_folder, exist_ok=True) - - documents = server_processing_result.get("documents", []) - doc_count = len(documents) - - # Save all documents to a single JSONL file (our new logic) - jsonl_filepath = os.path.join(temp_folder, "processed_documents.jsonl") - with open(jsonl_filepath, 'w', encoding='utf-8') as f: - for doc_data in documents: - f.write(json.dumps(doc_data, ensure_ascii=False) + '\n') - - # Clear documents from memory immediately after saving - documents.clear() - server_processing_result.clear() - - logger.info(f"Saved {doc_count} processed documents to {jsonl_filepath}") res_ingest_config["temp_session_id"] = temp_session_id res_ingest_config["temp_folder"] = temp_folder @@ -682,28 +662,28 @@ def ingest( try: data_source_id = ingest_config.get("data_source_id", "DocumentContent") - # Read from temporary folder + # Read from temporary folder's JSONL file temp_folder = ingest_config.get("temp_folder") if not temp_folder or not os.path.exists(temp_folder): raise Exception(f"Temporary folder not found: {temp_folder}") - # Read the processed_documents.jsonl file (our new logic) + # Read the entire JSONL file as a string jsonl_file = os.path.join(temp_folder, "processed_documents.jsonl") if not os.path.exists(jsonl_file): - raise Exception(f"Processed documents file not found: {jsonl_file}") + raise Exception(f"JSONL file not found: {jsonl_file}") - # Read entire JSONL content as a single string + logger.info(f"Reading JSONL file: {jsonl_file}") + + # Read entire JSONL content with open(jsonl_file, 'r', encoding='utf-8') as f: jsonl_content = f.read() - # Count documents for logging - document_count = jsonl_content.count('\n') if jsonl_content.strip() else 0 - logger.info(f"Ingesting {document_count} documents from {jsonl_file}") - - # Pass entire JSONL content in ONE call (efficient!) + # Load all documents in one call - runLoadingJobWithData supports JSONL format conn.runLoadingJobWithData(jsonl_content, data_source_id, loader_info.load_job_id) - logger.info(f"Successfully ingested {document_count} documents") + # Count documents for reporting + doc_count = sum(1 for line in jsonl_content.strip().split('\n') if line.strip()) + logger.info(f"Successfully ingested {doc_count} documents from JSONL") # Clean up temp folder after successful ingestion try: @@ -717,7 +697,8 @@ def ingest( raise Exception(f"Error during server markdown extraction and TigerGraph loading: {e}") return { "job_name": loader_info.load_job_id, - "summary": f"Data ingestion successful - processed {document_count} documents" + "summary": f"Successfully ingested {doc_count} documents from JSONL", + "document_count": doc_count } else: raise Exception("Data source and file format combination not implemented") From 2d54b02a766ab209c65b8d27bd0c2741f906efb4 Mon Sep 17 00:00:00 2001 From: Prins Kumar Date: Fri, 5 Dec 2025 21:52:01 +0530 Subject: [PATCH 15/15] Fix: Delete processed content from JSONL before deleting original files --- graphrag-ui/src/pages/Setup.tsx | 364 ++++++++++++++++---------------- graphrag/app/routers/ui.py | 170 ++++++--------- 2 files changed, 238 insertions(+), 296 deletions(-) diff --git a/graphrag-ui/src/pages/Setup.tsx b/graphrag-ui/src/pages/Setup.tsx index c3e4689..0e15939 100644 --- a/graphrag-ui/src/pages/Setup.tsx +++ b/graphrag-ui/src/pages/Setup.tsx @@ -40,7 +40,7 @@ const Setup = () => { const navigate = useNavigate(); const [confirm, confirmDialog, isConfirmDialogOpen] = useConfirm(); const [availableGraphs, setAvailableGraphs] = useState([]); - + const [initializeGraphOpen, setInitializeGraphOpen] = useState(false); const [graphName, setGraphName] = useState(""); const [isInitializing, setIsInitializing] = useState(false); @@ -71,7 +71,7 @@ const Setup = () => { const [refreshGraphName, setRefreshGraphName] = useState(""); const [isRebuildRunning, setIsRebuildRunning] = useState(false); const [isCheckingStatus, setIsCheckingStatus] = useState(false); - + // S3 state const [fileFormat, setFileFormat] = useState<"json" | "multi">("json"); const [awsAccessKey, setAwsAccessKey] = useState(""); @@ -129,7 +129,7 @@ const Setup = () => { } const filesArray = Array.from(selectedFiles); - + // Check if any single file exceeds the server limit const oversizedFiles = filesArray.filter((file) => file.size > MAX_UPLOAD_SIZE_BYTES); if (oversizedFiles.length > 0) { @@ -169,7 +169,7 @@ const Setup = () => { setUploadMessage(`✅ ${data.message} Processing...`); setSelectedFiles(null); await fetchUploadedFiles(); - + // Step 2: Call create_ingest to process uploaded files console.log("Calling handleCreateIngestAfterUpload from main upload..."); await handleCreateIngestAfterUpload("uploaded"); @@ -200,9 +200,9 @@ const Setup = () => { for (let i = 0; i < filesArray.length; i++) { const file = filesArray[i]; const fileNumber = i + 1; - + setUploadMessage(`Uploading file ${fileNumber}/${totalFiles}: ${file.name} (${formatBytes(file.size)})...`); - + const formData = new FormData(); formData.append("files", file); @@ -236,10 +236,10 @@ const Setup = () => { } else { setUploadMessage(`⚠️ Uploaded ${uploadedCount} files successfully, ${failedCount} failed. Processing...`); } - + setSelectedFiles(null); await fetchUploadedFiles(); - + // Step 2: Call create_ingest to process uploaded files console.log("Calling handleCreateIngestAfterUpload..."); await handleCreateIngestAfterUpload("uploaded"); @@ -261,25 +261,25 @@ const Setup = () => { try { const creds = localStorage.getItem("creds"); - - // Also delete corresponding temp files FIRST if session exists - if (tempSessionId) { - console.log("Calling handleDeleteTempFilesForOriginal..."); - await handleDeleteTempFilesForOriginal(filename); - } - - // Then delete original file - const response = await fetch( - `/ui/${ingestGraphName}/uploads?filename=${encodeURIComponent(filename)}`, - { - method: "DELETE", - headers: { Authorization: `Basic ${creds}` }, - } - ); + + // Delete original file (backend will also delete processed content from JSONL if session_id is provided) + const url = tempSessionId + ? `/ui/${ingestGraphName}/uploads?filename=${encodeURIComponent(filename)}&session_id=${tempSessionId}` + : `/ui/${ingestGraphName}/uploads?filename=${encodeURIComponent(filename)}`; + + const response = await fetch(url, { + method: "DELETE", + headers: { Authorization: `Basic ${creds}` }, + }); const data = await response.json(); - + setUploadMessage(`✅ ${data.message}`); await fetchUploadedFiles(); + + // Refresh temp files list if session exists + if (tempSessionId) { + await fetchTempFiles(tempSessionId); + } } catch (error: any) { console.error("Delete error:", error); setUploadMessage(`❌ Error: ${error.message}`); @@ -300,12 +300,12 @@ const Setup = () => { headers: { Authorization: `Basic ${creds}` }, }); const data = await response.json(); - + // Also clear temp session if (tempSessionId) { await handleDeleteAllTempFiles(); } - + setUploadMessage(`✅ ${data.message}`); await fetchUploadedFiles(); } catch (error: any) { @@ -341,7 +341,7 @@ const Setup = () => { try { const creds = localStorage.getItem("creds"); - + // Prepare request body based on provider let requestBody: any = { provider: cloudProvider }; @@ -398,18 +398,11 @@ const Setup = () => { const data = await response.json(); if (data.status === "success") { - setDownloadMessage(`✅ ${data.message}. Processed ${data.doc_count || 0} document(s)`); + setDownloadMessage(`✅ ${data.message}. Processing...`); await fetchDownloadedFiles(); - - // Save session ID from automatic processing - if (data.temp_session_id) { - setTempSessionId(data.temp_session_id); - await fetchTempFiles(data.temp_session_id); - } - } else if (data.status === "partial_success") { - setDownloadMessage(`⚠️ ${data.message}`); - await fetchDownloadedFiles(); - // Don't call create_ingest if processing already attempted + + // Step 2: Call create_ingest to process downloaded files + await handleCreateIngestAfterUpload("downloaded"); } else if (data.status === "warning") { setDownloadMessage(`⚠️ ${data.message}`); } else { @@ -428,22 +421,21 @@ const Setup = () => { try { const creds = localStorage.getItem("creds"); - - // Build URL with session_id if available - let deleteUrl = `/ui/${ingestGraphName}/cloud/delete?filename=${encodeURIComponent(filename)}`; - if (tempSessionId) { - deleteUrl += `&session_id=${encodeURIComponent(tempSessionId)}`; - } - - const response = await fetch(deleteUrl, { + + // Delete original file (backend will also delete processed content from JSONL if session_id is provided) + const url = tempSessionId + ? `/ui/${ingestGraphName}/cloud/delete?filename=${encodeURIComponent(filename)}&session_id=${tempSessionId}` + : `/ui/${ingestGraphName}/cloud/delete?filename=${encodeURIComponent(filename)}`; + + const response = await fetch(url, { method: "DELETE", headers: { Authorization: `Basic ${creds}` }, }); const data = await response.json(); - + setDownloadMessage(`✅ ${data.message}`); await fetchDownloadedFiles(); - + // Refresh temp files list if session exists if (tempSessionId) { await fetchTempFiles(tempSessionId); @@ -462,26 +454,13 @@ const Setup = () => { try { const creds = localStorage.getItem("creds"); - - // Build URL with session_id if available - let deleteUrl = `/ui/${ingestGraphName}/cloud/delete`; - if (tempSessionId) { - deleteUrl += `?session_id=${encodeURIComponent(tempSessionId)}`; - } - - const response = await fetch(deleteUrl, { + const response = await fetch(`/ui/${ingestGraphName}/cloud/delete`, { method: "DELETE", headers: { Authorization: `Basic ${creds}` }, }); const data = await response.json(); setDownloadMessage(`✅ ${data.message}`); await fetchDownloadedFiles(); - - // Clear session ID and refresh temp files - if (tempSessionId) { - setTempSessionId(null); - setTempFiles([]); - } } catch (error: any) { setDownloadMessage(`❌ Error: ${error.message}`); } @@ -558,31 +537,51 @@ const Setup = () => { // Delete temp files matching original filename const handleDeleteTempFilesForOriginal = async (originalFilename: string) => { console.log("handleDeleteTempFilesForOriginal called with:", originalFilename); - + if (!ingestGraphName || !tempSessionId) { console.log("No graph name or session ID, returning"); return; } try { + // Extract base name without extension (e.g., "document.pdf" -> "document") + const baseName = originalFilename.replace(/\.[^/.]+$/, ""); + console.log("Base name:", baseName); + const creds = localStorage.getItem("creds"); - - // Call the delete endpoint with the original filename - // The backend will handle removing all related documents from the JSONL file - const deleteResponse = await fetch( - `/ui/${ingestGraphName}/ingestion_temp/delete?session_id=${tempSessionId}&filename=${encodeURIComponent(originalFilename)}`, - { - method: "DELETE", - headers: { Authorization: `Basic ${creds}` }, + + // Fetch temp files to find matches + const response = await fetch(`/ui/${ingestGraphName}/ingestion_temp/list?session_id=${tempSessionId}`, { + headers: { Authorization: `Basic ${creds}` }, + }); + const data = await response.json(); + console.log("Temp files list response:", data); + + if (data.status === "success" && data.sessions.length > 0) { + const files = data.sessions[0].files || []; + console.log("All temp files:", files.map((f: any) => f.filename)); + + // Find temp files matching pattern: doc_{idx}_{baseName}*.json + const matchingFiles = files.filter((f: any) => f.filename.includes(`_${baseName}`)); + console.log("Matching files to delete:", matchingFiles.map((f: any) => f.filename)); + + // Delete each matching file + for (const file of matchingFiles) { + console.log("Deleting temp file:", file.filename); + const deleteResponse = await fetch( + `/ui/${ingestGraphName}/ingestion_temp/delete?session_id=${tempSessionId}&filename=${encodeURIComponent(file.filename)}`, + { + method: "DELETE", + headers: { Authorization: `Basic ${creds}` }, + } + ); + const deleteData = await deleteResponse.json(); + console.log("Delete response:", deleteData); } - ); - const deleteData = await deleteResponse.json(); - console.log("Delete temp files response:", deleteData); - - if (deleteData.status === "success") { - console.log(`Successfully deleted processed documents for ${originalFilename}`); + + console.log(`Successfully deleted ${matchingFiles.length} temp file(s)`); } else { - console.error("Failed to delete temp files:", deleteData); + console.log("No temp files found or empty sessions"); } } catch (error: any) { console.error("Error deleting temp files:", error); @@ -625,9 +624,8 @@ const Setup = () => { const ingestData = await ingestResponse.json(); console.log("Ingest response:", ingestData); - const docCount = ingestData.document_count || tempFiles.length; - setIngestMessage(`✅ Data ingested successfully! Processed ${docCount} document(s).`); - + setIngestMessage(`✅ Data ingested successfully! Processed ${tempFiles.length} documents.`); + // Clear temp state setTempFiles([]); setShowTempFiles(false); @@ -648,7 +646,7 @@ const Setup = () => { return; } - const folderPath = sourceType === "uploaded" + const folderPath = sourceType === "uploaded" ? `uploads/${ingestGraphName}` : `downloaded_files_cloud/${ingestGraphName}`; @@ -687,7 +685,7 @@ const Setup = () => { // Check if temp files were created (for server data source) const sessionId = createData.data_source_id?.temp_session_id; - + if (sessionId && !directIngestion) { // Files are saved to temp storage - show them for review (only if not direct ingestion) setTempSessionId(sessionId); @@ -696,37 +694,37 @@ const Setup = () => { data_source_id: createData.data_source_id, data_path: createData.data_path || createData.file_path, }); - setIngestMessage(`✅ Files processed successfully. Review them below before ingesting.`); + setIngestMessage(`✅ Processed ${createData.data_source_id.file_count} files. Review them below before ingesting.`); await fetchTempFiles(sessionId); setIsIngesting(false); } else { // No temp files (e.g., S3 Bedrock) OR direct ingestion enabled - proceed directly to ingest - setIngestMessage("Step 2/2: Running document ingest..."); + setIngestMessage("Step 2/2: Running document ingest..."); - const loadingInfo = { - load_job_id: createData.load_job_id, - data_source_id: createData.data_source_id, + const loadingInfo = { + load_job_id: createData.load_job_id, + data_source_id: createData.data_source_id, file_path: createData.data_path || createData.file_path, - }; + }; - const ingestResponse = await fetch(`/ui/${ingestGraphName}/ingest`, { - method: "POST", - headers: { - "Content-Type": "application/json", - Authorization: `Basic ${creds}`, - }, - body: JSON.stringify(loadingInfo), - }); + const ingestResponse = await fetch(`/ui/${ingestGraphName}/ingest`, { + method: "POST", + headers: { + "Content-Type": "application/json", + Authorization: `Basic ${creds}`, + }, + body: JSON.stringify(loadingInfo), + }); - if (!ingestResponse.ok) { - const errorData = await ingestResponse.json(); - throw new Error(errorData.detail || `Failed to run ingest: ${ingestResponse.statusText}`); - } + if (!ingestResponse.ok) { + const errorData = await ingestResponse.json(); + throw new Error(errorData.detail || `Failed to run ingest: ${ingestResponse.statusText}`); + } - const ingestData = await ingestResponse.json(); - console.log("Ingest response:", ingestData); + const ingestData = await ingestResponse.json(); + console.log("Ingest response:", ingestData); - setIngestMessage(`✅ Data ingested successfully! Processed documents from ${folderPath}/`); + setIngestMessage(`✅ Data ingested successfully! Processed documents from ${folderPath}/`); setIsIngesting(false); } } catch (error: any) { @@ -740,16 +738,16 @@ const Setup = () => { const handleCreateIngestAfterUpload = async (sourceType: "uploaded" | "downloaded" = "uploaded") => { console.log("handleCreateIngestAfterUpload called with sourceType:", sourceType); console.log("ingestGraphName:", ingestGraphName); - + if (!ingestGraphName) { console.log("No graph name, returning early"); return; } - const folderPath = sourceType === "uploaded" + const folderPath = sourceType === "uploaded" ? `uploads/${ingestGraphName}` : `downloaded_files_cloud/${ingestGraphName}`; - + console.log("folderPath:", folderPath); try { @@ -764,7 +762,7 @@ const Setup = () => { loader_config: {}, file_format: "multi" }; - + console.log("Calling create_ingest with config:", createIngestConfig); const createResponse = await fetch(`/ui/${ingestGraphName}/create_ingest`, { @@ -775,7 +773,7 @@ const Setup = () => { }, body: JSON.stringify(createIngestConfig), }); - + console.log("create_ingest response status:", createResponse.status); if (!createResponse.ok) { @@ -786,10 +784,10 @@ const Setup = () => { const createData = await createResponse.json(); console.log("create_ingest response data:", createData); - + const sessionId = createData.data_source_id?.temp_session_id; console.log("Session ID:", sessionId); - + if (sessionId) { // Save session ID for later ingest setTempSessionId(sessionId); @@ -798,16 +796,16 @@ const Setup = () => { data_source_id: createData.data_source_id, data_path: createData.data_path || createData.file_path, }); - + console.log("Direct ingestion enabled:", directIngestion); - + if (directIngestion) { // Direct ingestion - proceed to ingest immediately setUploadMessage("Running direct ingestion..."); await handleRunIngest(); } else { // Save for later - files ready for ingestion - setUploadMessage(`✅ Files processed successfully. Ready for ingestion.`); + setUploadMessage(`✅ Successfully processed ${createData.data_source_id.file_count} files. Ready for ingestion.`); } } else { console.warn("No session ID returned from create_ingest"); @@ -969,9 +967,9 @@ const Setup = () => { const statusData = await statusResponse.json(); const wasRunning = isRebuildRunning; const isCurrentlyRunning = statusData.is_running || false; - + setIsRebuildRunning(isCurrentlyRunning); - + if (isCurrentlyRunning) { const startTime = statusData.started_at ? new Date(statusData.started_at * 1000).toLocaleString() : "unknown time"; setRefreshMessage(`⚠️ A rebuild is already in progress for "${graphName}" (started at ${startTime}). Please wait for it to complete.`); @@ -1027,7 +1025,7 @@ const Setup = () => { try { const creds = localStorage.getItem("creds"); - + const response = await fetch(`/ui/${refreshGraphName}/rebuild_graph`, { method: "POST", headers: { @@ -1058,12 +1056,12 @@ const Setup = () => { if (refreshOpen && refreshGraphName) { // Check status immediately when dialog opens checkRebuildStatus(refreshGraphName, true); - + // Set up polling to check status every 5 seconds while dialog is open const intervalId = setInterval(() => { checkRebuildStatus(refreshGraphName, false); }, 5000); - + return () => clearInterval(intervalId); } }, [refreshOpen, refreshGraphName]); @@ -1163,10 +1161,10 @@ const Setup = () => { setIsInitializing(false); return; } - + setStatusMessage(`✅ Graph "${graphName}" created and initialized successfully! You can now close this dialog.`); setStatusType("success"); - + // Add the new graph to the available graphs list const newGraph = graphName; setAvailableGraphs(prev => { @@ -1180,7 +1178,7 @@ const Setup = () => { } return prev; }); - + // Set the newly created graph as selected for ingestion setIngestGraphName(graphName); setRefreshGraphName(graphName); @@ -1216,7 +1214,7 @@ const Setup = () => { {/* Three cards displayed horizontally */}
- + {/* Section 1: Initialize Knowledge Graph */}
@@ -1231,7 +1229,7 @@ const Setup = () => {

-
-
-
{/* Initialize Graph Dialog */} - { // Prevent closing if confirm dialog is open @@ -1302,7 +1300,7 @@ const Setup = () => { setInitializeGraphOpen(open); }} > - e.preventDefault()} > @@ -1312,7 +1310,7 @@ const Setup = () => { Enter the name of your knowledge graph. The system will create it if necessary and initialize it with the GraphRAG schema. - +
{/* Data Ingest Dialog */} - { // Prevent closing if confirm dialog is open if (!open && isConfirmDialogOpen) { @@ -1410,7 +1407,7 @@ const Setup = () => { setIngestOpen(open); }} > - e.preventDefault()} > @@ -1481,23 +1478,23 @@ const Setup = () => { disabled={isUploading} className="dark:border-[#3D3D3D] dark:bg-shadeA" /> -

- Maximum upload per request: {MAX_UPLOAD_SIZE_MB} MB. -

- - {/* Direct Ingestion Checkbox */} -
- setDirectIngestion(e.target.checked)} - className="mr-2 h-4 w-4 rounded border-gray-300 text-blue-600 focus:ring-blue-500" - /> - -
+

+ Maximum upload per request: {MAX_UPLOAD_SIZE_MB} MB. +

+ + {/* Direct Ingestion Checkbox */} +
+ setDirectIngestion(e.target.checked)} + className="mr-2 h-4 w-4 rounded border-gray-300 text-blue-600 focus:ring-blue-500" + /> + +
@@ -1546,7 +1543,7 @@ const Setup = () => {

Process uploaded files and add them to the knowledge graph

- + {ingestMessage && ( -
+ }`}> {ingestMessage}
)} @@ -1806,7 +1802,7 @@ const Setup = () => {

Files will be downloaded to: downloaded_files_cloud/{ingestGraphName}/

-
{downloadMessage && ( -
+ }`}> {downloadMessage}
)} @@ -1886,7 +1881,7 @@ const Setup = () => {

Process downloaded files and add them to the knowledge graph

- + {ingestMessage && ( -
+ }`}> {ingestMessage}
)} @@ -2048,13 +2042,12 @@ const Setup = () => { )} {ingestMessage && ( -
+ }`}> {ingestMessage}
)} @@ -2079,8 +2072,8 @@ const Setup = () => { {/* Refresh Graph Dialog */} - { // Prevent closing if confirm dialog is open if (!open && isConfirmDialogOpen) { @@ -2089,7 +2082,7 @@ const Setup = () => { setRefreshOpen(open); }} > - e.preventDefault()} > @@ -2130,19 +2123,18 @@ const Setup = () => { ⚠️ Warning

- This operation will rebuild the graph content that will interrupt related queries. + This operation will rebuild the graph content that will interrupt related queries. Please confirm to proceed.

{refreshMessage && ( -
+ }`}> {refreshMessage}
)} diff --git a/graphrag/app/routers/ui.py b/graphrag/app/routers/ui.py index 9d8cc33..b69afae 100644 --- a/graphrag/app/routers/ui.py +++ b/graphrag/app/routers/ui.py @@ -52,6 +52,7 @@ from common.logs.logwriter import LogWriter from common.metrics.prometheus_metrics import metrics as pmetrics from supportai import supportai +from common.utils.text_extractors import TextExtractor from common.py_schemas.schemas import ( AgentProgess, CreateIngestConfig, @@ -990,6 +991,7 @@ async def clear_uploaded_files( graphname: str, creds: Annotated[tuple[list[str], HTTPBasicCredentials], Depends(ui_basic_auth)], filename: str | None = None, + session_id: str | None = None, ): """ Clear uploaded files for a specific graphname. @@ -997,6 +999,7 @@ async def clear_uploaded_files( Parameters: - graphname: The graph name whose files to clear - filename: If provided, only delete this specific file. Otherwise, delete all files. + - session_id: Optional session ID to delete processed content from temp folder """ try: upload_dir = os.path.join("uploads", graphname) @@ -1009,9 +1012,21 @@ async def clear_uploaded_files( } deleted_files = [] + text_extractor = TextExtractor() if filename: - # Delete specific file + # Delete processed content from JSONL FIRST if session_id provided + if session_id: + temp_folder = os.path.join("uploads", "ingestion_temp", graphname, session_id) + if os.path.exists(temp_folder): + logger.info(f"Deleting processed content for {filename} from temp folder") + result = text_extractor.delete_file_from_jsonl(temp_folder, filename) + if result.get('success'): + logger.info(f"Removed {result.get('removed_count', 0)} processed documents for {filename}") + else: + logger.warning(f"Failed to remove processed content: {result.get('error', 'Unknown error')}") + + # Then delete the original file file_path = os.path.join(upload_dir, filename) if os.path.exists(file_path) and os.path.isfile(file_path): os.remove(file_path) @@ -1049,6 +1064,7 @@ async def clear_uploaded_files( raise HTTPException(status_code=500, detail=f"Error deleting files: {str(e)}") + # Cloud Storage Download Endpoints @router.post(route_prefix + "/{graphname}/cloud/download") @@ -1058,8 +1074,7 @@ async def download_from_cloud( request_body: dict = Body(...), ): """ - Download files from cloud storage (S3, GCS, or Azure) to local directory - and automatically process them to create JSONL files for ingestion. + Download files from cloud storage (S3, GCS, or Azure) to local directory. Parameters: - graphname: The graph name to associate downloaded files with @@ -1253,49 +1268,14 @@ async def download_from_cloud( logger.info(f"Downloaded {len(downloaded_files)} file(s) from {provider} for graph {graphname}") - # Automatically process downloaded files to create JSONL - from common.utils.text_extractors import TextExtractor - temp_session_id = str(uuid.uuid4()) - temp_folder = os.path.join("uploads", "ingestion_temp", graphname, temp_session_id) - - try: - extractor = TextExtractor() - processing_result = extractor.process_folder( - download_dir, - graphname=graphname, - temp_folder=temp_folder - ) - - if processing_result.get("statusCode") != 200: - logger.error(f"Cloud file processing failed: {processing_result}") - raise Exception(f"Failed to process downloaded files: {processing_result}") - - doc_count = processing_result.get("num_documents", 0) - logger.info(f"Processed {doc_count} documents from downloaded files") - - return { - "status": "success", - "message": f"Successfully downloaded and processed {len(downloaded_files)} file(s) from {provider}", - "graphname": graphname, - "provider": provider, - "downloaded_files": downloaded_files, - "local_path": download_dir, - "temp_session_id": temp_session_id, - "temp_folder": temp_folder, - "doc_count": doc_count, - } - except Exception as e: - logger.error(f"Error processing downloaded files: {e}") - # Return success for download but warn about processing failure - return { - "status": "partial_success", - "message": f"Downloaded {len(downloaded_files)} file(s) but processing failed: {str(e)}", - "graphname": graphname, - "provider": provider, - "downloaded_files": downloaded_files, - "local_path": download_dir, - "processing_error": str(e), - } + return { + "status": "success", + "message": f"Successfully downloaded {len(downloaded_files)} file(s) from {provider}", + "graphname": graphname, + "provider": provider, + "downloaded_files": downloaded_files, + "local_path": download_dir, + } except HTTPException: raise @@ -1362,12 +1342,11 @@ async def delete_cloud_downloads( ): """ Delete downloaded cloud files for a specific graph. - Also deletes corresponding processed documents from the JSONL file. Parameters: - graphname: The graph name whose downloaded files to clear - filename: If provided, only delete this specific file. Otherwise, delete all files. - - session_id: The session ID for the temp folder containing processed JSONL + - session_id: Optional session ID to delete processed content from temp folder """ try: download_dir = os.path.join("downloaded_files_cloud", graphname) @@ -1380,28 +1359,23 @@ async def delete_cloud_downloads( } deleted_files = [] + text_extractor = TextExtractor() if filename: - # Delete specific file AND its processed documents from JSONL + # Delete processed content from JSONL FIRST if session_id provided + if session_id: + temp_folder = os.path.join("uploads", "ingestion_temp", graphname, session_id) + if os.path.exists(temp_folder): + logger.info(f"Deleting processed content for {filename} from temp folder") + result = text_extractor.delete_file_from_jsonl(temp_folder, filename) + if result.get('success'): + logger.info(f"Removed {result.get('removed_count', 0)} processed documents for {filename}") + else: + logger.warning(f"Failed to remove processed content: {result.get('error', 'Unknown error')}") + + # Then delete the original file file_path = os.path.join(download_dir, filename) if os.path.exists(file_path) and os.path.isfile(file_path): - # If session_id provided, also delete from JSONL - if session_id: - from common.utils.text_extractors import TextExtractor - extractor = TextExtractor() - - temp_folder = os.path.join("uploads", "ingestion_temp", graphname, session_id) - if os.path.exists(temp_folder): - # Delete from JSONL first - delete_result = extractor.delete_file_from_jsonl(temp_folder, filename) - logger.info(f"JSONL delete result for {filename}: {delete_result}") - - # If JSONL delete failed (and JSONL exists), warn but continue with file deletion - if not delete_result.get('success'): - logger.warning(f"Failed to delete from JSONL: {delete_result.get('error')}") - # Continue with file deletion even if JSONL deletion failed - - # Delete the original downloaded file os.remove(file_path) deleted_files.append(filename) logger.info(f"Deleted cloud download {filename} for graph {graphname}") @@ -1409,21 +1383,13 @@ async def delete_cloud_downloads( raise HTTPException(status_code=404, detail=f"File {filename} not found") else: # Delete all files in the directory - for fname in os.listdir(download_dir): - file_path = os.path.join(download_dir, fname) + for filename in os.listdir(download_dir): + file_path = os.path.join(download_dir, filename) if os.path.isfile(file_path): os.remove(file_path) - deleted_files.append(fname) - - # If session_id provided, delete the entire temp folder - if session_id: - temp_folder = os.path.join("uploads", "ingestion_temp", graphname, session_id) - if os.path.exists(temp_folder): - import shutil - shutil.rmtree(temp_folder, ignore_errors=True) - logger.info(f"Deleted temp folder for session {session_id}") + deleted_files.append(filename) - # Remove the download directory if it's empty + # Remove the directory if it's empty if not os.listdir(download_dir): os.rmdir(download_dir) @@ -1542,41 +1508,25 @@ async def delete_ingestion_temp_files( deleted_files = [] if filename: - # Delete processed documents from JSONL for this original filename - # Note: Original files are NOT in temp folder, only processed_documents.jsonl is here - from common.utils.text_extractors import TextExtractor - extractor = TextExtractor() - - # Delete from JSONL - MUST succeed - delete_result = extractor.delete_file_from_jsonl(session_dir, filename) - logger.info(f"JSONL delete result for {filename}: {delete_result}") - - # If JSONL delete failed, return error - if not delete_result.get('success'): - error_msg = delete_result.get('error', 'Unknown error') - logger.error(f"Failed to delete from JSONL: {error_msg}") - raise HTTPException(status_code=500, detail=f"Failed to delete processed documents: {error_msg}") - - deleted_files.append(filename) - logger.info(f"Deleted {delete_result.get('removed_count', 0)} processed documents for {filename} from JSONL") - - # Check if temp folder was deleted by JSONL cleanup - if delete_result.get('temp_folder_deleted'): - logger.info(f"Session folder {session_id} was automatically deleted (no documents remaining)") - elif not os.path.exists(session_dir): - logger.info(f"Session folder {session_id} was deleted") - elif not os.listdir(session_dir): - # Clean up empty session folder - os.rmdir(session_dir) - logger.info(f"Removed empty session folder {session_id}") + # Delete specific file + file_path = os.path.join(session_dir, filename) + if os.path.exists(file_path) and os.path.isfile(file_path): + os.remove(file_path) + deleted_files.append(filename) + logger.info(f"Deleted temp file {filename} from session {session_id}") + + # If session folder is now empty, remove it + if not os.listdir(session_dir): + os.rmdir(session_dir) + logger.info(f"Removed empty session folder {session_id}") else: - logger.info(f"Removed {delete_result.get('removed_count', 0)} documents from JSONL, {delete_result.get('remaining_count', 0)} remaining") + raise HTTPException(status_code=404, detail=f"File {filename} not found") else: - # Delete entire session folder (including JSONL) + # Delete entire session folder import shutil - for fname in os.listdir(session_dir): - if os.path.isfile(os.path.join(session_dir, fname)): - deleted_files.append(fname) + for filename in os.listdir(session_dir): + if os.path.isfile(os.path.join(session_dir, filename)): + deleted_files.append(filename) shutil.rmtree(session_dir) logger.info(f"Deleted session folder {session_id} for graph {graphname}")