datawhalechina/hello-agents · error · ValueError

Unsupported input file format: {source_path.suffix}

Error message

Unsupported input file format: {source_path.suffix}

What it means

`preview_pdf_tables` resolves the input path and immediately rejects it with ValueError unless its lowercased suffix is in SUPPORTED_DOCUMENT_SUFFIXES (the PDF-style document set). Message: 'Unsupported input file format: {suffix}'. This guards the table-preview entrypoint before any pdfplumber work starts.

Source

Thrown at Co-creation-projects/healer-666-Academic-Data-Agent/src/data_analysis_agent/document_ingestion.py:252

        "pdf_multi_table_mode": True,
        "source_pdf": source_pdf.resolve().as_posix(),
        "background_literature_context": background_literature_context,
        "text_excerpt": full_text_excerpt,
        "selected_table_id": selected_table_id,
        "candidate_tables": candidate_tables,
        "candidate_table_summaries": candidate_tables,
    }


def preview_pdf_tables(
    data_path: str | Path,
    *,
    max_pdf_pages: int = 20,
    max_candidate_tables: int = 5,
) -> PdfPreviewResult:
    source_path = Path(data_path).resolve()
    if source_path.suffix.lower() not in SUPPORTED_DOCUMENT_SUFFIXES:
        raise ValueError(f"Unsupported input file format: {source_path.suffix}")

    scratch_root = source_path.parent / ".pdf_preview_tmp"
    scratch_root.mkdir(parents=True, exist_ok=True)
    scratch_dir = Path(tempfile.mkdtemp(prefix="pdf_preview_", dir=scratch_root))
    try:
        full_text, records = _extract_pdf_payload(
            source_path,
            max_pdf_pages=max_pdf_pages,
            max_candidate_tables=max_candidate_tables,
            extracted_tables_dir=scratch_dir,
            persist_csv=False,
        )
    finally:
        shutil.rmtree(scratch_dir, ignore_errors=True)

    default_record = _select_primary_table(records)
    warnings: list[str] = []
    if records:

View on GitHub (pinned to 606a07d341)

Solutions

1. Pass a file whose extension is in SUPPORTED_DOCUMENT_SUFFIXES (check that constant; typically .pdf).
2. For .csv/.xlsx inputs, skip document preview — they are already structured tables.
3. Convert .docx to PDF (e.g. via LibreOffice) before previewing tables.
4. Ensure the path points at the original source document, not a generated artifact.

from pathlib import Path

def is_supported_document(path: str | Path, supported: set[str]) -> bool:
    return Path(path).suffix.lower() in supported

# before calling preview_pdf_tables
if not is_supported_document(pdf_path, SUPPORTED_DOCUMENT_SUFFIXES):
    raise ValueError(f"preview needs a document file, got {Path(pdf_path).suffix}")

try:
    preview = preview_pdf_tables(path)
except ValueError as e:
    if "Unsupported input file format" in str(e):
        path = convert_to_pdf(path)  # e.g. LibreOffice export
        preview = preview_pdf_tables(path)
    else:
        raise

When it happens

Trigger: Calling preview_pdf_tables on a .csv/.xlsx (tabular, not a document — use the tabular path instead), on .docx/.png/.jpg, or on a scanned PDF stored with an odd extension; also fires for files with no extension.

Common situations: Users trying to preview tables inside Word documents or images; passing the wrong path variable (a cleaned CSV output instead of the source PDF); double extensions like .pdf.bak.

Related errors

• Unsupported data file format: {data_path.suffix}
• Selected table_id '{requested_table_id}' was not found in th
• Unsupported latency_mode: {latency_mode}
• Unsupported vision_review_mode: {vision_review_mode}
• Unsupported quality_mode: {quality_mode}

AI-assisted analysis of datawhalechina/hello-agents@606a07d341 (2026-08-14). Data as JSON: /api/errors/c585d48607a88ade. Report an issue: GitHub.