Reference only; verify against the target FastGPT release and deployment documentation.
What this category of data looks like
Data related to airport service inquiries and responses mainly comes from official terminal guidance manuals, quarterly flight schedules, temporary flight adjustment announcements, passenger service notices and other documents. There are two update cadences: fixed-cycle updates for quarterly flight schedules and annual service manuals, and real-time updates for temporary flight changes and temporary control notices. Document structures include mixed formats: structured Excel spreadsheets with multiple sheets, PDF service guides with title hierarchies, and Word documents for service process descriptions. Fields include flight number, departure and arrival time, terminal number, baggage claim area number, service inquiry phone number and others. Units cover multiple types such as minutes, terminal numbers and phone numbers.
Constraints Imposed on Document Parsing and Chunking
Airport service inquiry documents have mixed characteristics of structured data and unstructured text, plus special data forms such as multi-sheet Excel schedules and real-time updated temporary notices. If only the default sheet is parsed for multi-sheet Excel files, a large amount of cross-zone flight and ground crew information will be lost. If real-time updated temporary notices are chunked into excessively large segments, precise matching of latest flight change information during retrieval will fail. If a single parsing rule is used uniformly for mixed-format documents, column associations of structured data and title hierarchies of unstructured text will be lost. Additionally, single files for airport documents may be large, requiring adaptation to large file upload and parsing time limits.
Configuration Settings
| Config Item | Recommended Value | Rationale |
|---|---|---|
chunkSize | 800–1200 characters | Airport service documents include structured flight data and unstructured service descriptions. This range balances information integrity and retrieval accuracy |
chunkSplitter | Split by paragraph + title | Airport documents have clear service process chapters and flight zone titles. Splitting by title preserves context association |
customPdfParse | Enable text layer-first parsing | Airport PDF service manuals are often a mix of scanned documents and text. Parsing the text layer first preserves retrievable content |
UPLOAD_FILE_MAX_SIZE | 1000 MB | Airports may upload complete quarterly flight schedule Excel packages. This upper limit covers conventional bulk upload needs |
PARSE_FILE_TIMEOUT_SECONDS | 600 seconds | Parsing large multi-sheet Excel files takes significant time. This duration prevents timeout interruptions |
chunkSettingMode | Manual configuration | Chunking rules need adjustment for different document types. Manual configuration adapts to the mixed structure of airport documents |
Three Common Errors
- Phenomenon: After parsing an Excel file, sheet names cannot be obtained, and the parsing result only contains the content of the default sheet. Cause: Custom parsing configuration is not enabled, or multi-sheet recognition rules are not specified.
- Phenomenon: After importing a PPT or DOC document, the PDF enhancement function is enabled, and no valid parsing results are generated. Cause: The PDF enhancement function only applies to PDF format documents. Non-PDF documents require parsing modes corresponding to their format.
- Phenomenon: When importing an Excel file into the knowledge base, only two columns of data are recognized, and the content of the remaining columns is lost. Cause: CSV parsing mode is enabled by default. This mode only supports reading single-column or two-column structured data.
How to Verify Correct Configuration
- Upload a single-sheet Excel flight schedule, check whether the parsed result contains complete multi-column data and the sheet name.
- Upload a non-PDF service process document, confirm that normally retrievable chunked content is generated when the PDF enhancement function is not enabled.
- Upload a PDF of a large airport service manual under 100 MB, check that the parsing process does not trigger a timeout prompt.
- View the chunked document fragments, confirm that the title hierarchy and context association of the original document are preserved.
The values given are common starting points and should be measured against the reader'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.