What data for this category looks like
Data sources for this category primarily originate from the national public resource trading platform, natural resource trading systems, property right exchange systems, and state-owned asset supervision systems, with updates occurring daily. Typical documents include transfer announcements, trading notices, bid application documents, margin vouchers, transaction confirmation letters, listing result announcements, property right transfer specifications, and similar materials. Most document structures include core fields such as subject ownership descriptions, transaction restriction conditions, margin rules, and transaction process nodes. Field units commonly use identifiers for rights or amounts such as square meters, ten thousand yuan, and parcels.
What constraints do these characteristics impose on the document parsing and chunking link
Data sources for this category are scattered and updated frequently each day. This requires the document parsing and chunking link to adapt to standardized and non-standard document formats exported from multiple systems. Documents contain cross-page related content such as ownership descriptions and transaction conditions. Chunking must avoid splitting core information context. Fields with precise units such as amounts, areas, and parcel counts must not have their complete expressions truncated by chunk boundaries. Documents subject to compliance review, such as bid application documents and transaction confirmation letters, must retain sufficiently long context to support subsequent compliance verification links.
Configuration Settings
Reference only; verify against the target FastGPT release and deployment documentation.
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
chunkSize | 800–1200 characters | Adapts to long paragraphs of ownership descriptions and trading rule content in documents of this category, avoiding truncation of core cross-page related information |
chunkSplitter | Split by paragraph + title | Matches the structured format with title-based sections in documents such as transfer announcements and transaction confirmation letters, retaining logical relevance of information |
customPdfParse | Enabled | Adapts to PDF documents exported from multiple systems, retaining structured content such as tables and annotations from the original document |
PARSE_FILE_TIMEOUT_SECONDS | 600 seconds | Addresses scenarios where documents of this category include multi-page trading materials and attachments, avoiding parsing timeout interruptions |
UPLOAD_FILE_MAX_SIZE | 100 MB | Compatible with large-volume documents such as batch uploaded listing file packages and historical trading archives |
chunkSettingMode | Manual configuration | Allows adjustment of chunking parameters for different document types, such as transfer announcements vs bid applications, to adapt to diverse business scenarios |
Three Common Misconfigurations
- Uploading HTML interface documents generated by javadoc or non-standard format files with modified extensions results in no parsing results. The cause is failure to enable custom parsing mode or configure parsing rules for corresponding formats, with only preset standard document formats recognized.
- Using
chunkmode to call the pushdata API to upload trading documents leaves the task in indexing status for an extended period. The cause ischunkSizeset outside the reasonable range orPARSE_FILE_TIMEOUT_SECONDSconfiguration being insufficient, leading to parsing task backlog that cannot be completed. - Uploading a PDF trading document containing images fails to extract key information such as ownership numbers and margin amounts from the images. The cause is failure to enable the image OCR function of
customPdfParse, with only text content parsed while business information carried by images ignored.
How to Verify Correct Configuration
- Upload a standard transfer announcement PDF document, check if parsed chunked content retains cross-page ownership descriptions and trading rules to verify the configuration effects of
chunkSizeandchunkSplitter. - Call the pushdata API to upload a test file, check if the task status completes within a reasonable time to verify if the
PARSE_FILE_TIMEOUT_SECONDSconfiguration adapts to the current document size. - Upload a transaction confirmation letter PDF containing images, check if parsed results include key business information from the images to verify the enabled status of
customPdfParse. - Batch upload multiple different types of trading documents, confirm uploaded file sizes do not exceed the
UPLOAD_FILE_MAX_SIZElimit to verify configuration compatibility.
The values provided here 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.