[FEA]: Ingest pdf with custom metadata

Is this a new feature, an improvement, or a change to existing functionality?

New Feature

How would you describe the priority of this feature request

Currently preventing usage

Please provide a clear description of problem this feature solves

Is it possible to take custom metadata along when sending the data to nv-ingest?
Example below ingest a pdf with native request. How to add custom metadata with it so in the vector db it has those fields too.

` data = {
"collection_name": collection_name,
"extraction_options": {
"extract_text": True,
"extract_tables": True,
"extract_charts": True,
"extract_images": False, # Set to True if you want to extract images, ensure the VLM model is deployed
"extract_method": "pdfium",
"text_depth": "page",
},
"split_options": {
"chunk_size": 1024,
"chunk_overlap": 150
}
}

form_data = aiohttp.FormData()
for file_path in files[1:]:
    print(f"Uploading {file_path} to {collection_name}")
    form_data.add_field("documents", open(file_path, "rb"), filename=os.path.basename(file_path), content_type="application/pdf")

form_data.add_field("data", json.dumps(data), content_type="application/json")

async with aiohttp.ClientSession() as session:
    try:
        async with session.post(f"{BASE_URL}/v1/documents", data=form_data) as response: # Replace with session.patch for reingesting
            await print_response(response)
    except aiohttp.ClientError as e:
        print(f"Error: {e}")`

Describe the feature, and optionally a solution or implementation and any alternatives

A new feature example
data = { "collection_name": collection_name, "extraction_options": { "extract_text": True, "extract_tables": True, "extract_charts": True, "extract_images": False, # Set to True if you want to extract images, ensure the VLM model is deployed "extract_method": "pdfium", "text_depth": "page", }, "split_options": { "chunk_size": 1024, "chunk_overlap": 150 }, // add custom metadata "custom_metadata": {"abc":123} }

Additional context

No response

The issue identifies the /v1/documents endpoint and a proposed custom_metadata field, but no repository files or tests. Locate the endpoint implementation and trace how document metadata is passed through ingestion; verify that supplied fields reach the vector database and add or run the relevant API tests if present.