Reference only; verify against the target FastGPT release and deployment documentation.
What this category of data looks like
Typical data sources include meteorological warning systems, hydrological monitoring systems, emergency command platforms, grassroots disaster reporting systems, and video surveillance. The update frequency is daily. Daily documents include duty logs, warning notices, consultation minutes, rapid disaster reports, emergency response plans, and material allocation slips. Document structures contain hierarchical headings, nested paragraphs, professional numerical tables, and on-site images. Fields cover rainfall and water level data, danger level, material quantity, number of people transferred. Units include millimeters, cubic meters, number of people, and other professional measurement standards.
Constraints on document parsing and chunking
Multi-source and heterogeneous data formats require the parsing module to support multiple document types and structured data. Daily updated batch documents require stable parsing time to avoid disrupting emergency workflows. Documents containing professional values and units must avoid splitting values from their corresponding units to ensure data readability. Documents with embedded images and tables must retain the contextual association between text, images and tables to avoid loss of critical disaster information. Differences in document hierarchies across cross-departmental collaboration require chunking logic to adapt to the structural characteristics of different documents.
Configuration Settings
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
chunkSize | 800–1200 characters | Adapts to documents containing professional values and units, avoiding damage to data integrity |
chunkSplitter | Split by paragraph + heading | Adapts to the hierarchical structure of consultation minutes and rapid disaster reports, retaining contextual association |
PARSE_FILE_TIMEOUT_SECONDS | 600 seconds | Adapts to the parsing time of daily batch documents, avoiding timeout interruptions |
UPLOAD_FILE_MAX_SIZE | 1000 MB | Accommodates large documents such as emergency response plans with multi-page images and text |
customPdfParse | Enabled | Parses PDF documents with embedded tables and images, avoiding format loss |
chunkSettingMode | Manual configuration | Adapts the chunking logic for different types of emergency documents to optimize parsing results |
Three Common Errors
- Uploading a docx file with images returns an
Invalid image fileerror. The root cause is that thecustomPdfParseconfiguration is not enabled, so the system cannot recognize and associate content related to embedded images. - A
slow operation xxxxmslog appears during batch parsing of daily updated documents. The root cause is thatPARSE_FILE_TIMEOUT_SECONDSis set too short, failing to match the overall parsing time of batch documents. - Markdown output of rapid disaster reports is truncated. The root cause is that
chunkSizeis set too small, forcing splitting of long paragraphs and causing professional data to be split apart.
How to Confirm Proper Configuration
- Upload an emergency response plan document with embedded images, and verify that the parsed result fully displays text descriptions associated with the images.
- Review parsing logs to confirm that no timeout or
slow operationrelated error messages appear. - Select a rapid disaster report document containing professional values and units, and check that the chunking result does not split values from their corresponding units.
- Upload multiple copies of the same type of daily updated document, and confirm that the parsing process does not experience abnormal interruptions.
The values listed here are common starting points and should be evaluated against the user's own samples.
Question material comes from public community discussions. Configuration values are common starting points and should be measured against your own samples. Verified on 2026-09-28.