What the data for this category looks like
The ledger and report data for new energy development primarily originates from project initiation documents, on-site operation and maintenance records, grid connection acceptance reports, power generation statistics reports, and similar materials. The data update rhythm is adjusted based on the project phase: files are synchronized per milestone during the project initiation and approval phases, while operational data is updated weekly or monthly during the operation and maintenance phase. Document formats are mixed, including structured Excel ledgers, PDF reports with charts and formulas, Word approval documents, and other types. Fields include project ID, installed capacity, grid connection date, cumulative power generation, operation and maintenance duration, material consumption, and more. Units involved include kilowatts, kilowatt-hours, hours, cubic meters, and others.
What constraints do these characteristics impose on the document parsing and chunking link
Mixed-format documents contain both structured tables and complex formulas, so parsing tools must support both structured content extraction and formula recognition. Document formats from multiple sources are inconsistent, so adaptation to different parsing engines is required to ensure consistent extraction results. Ledgers have numerous fields and diverse units, so the association between fields and business logic must be preserved during chunking to avoid field breakage or information loss after splitting. Irregularly updated documents must support incremental parsing and chunking to avoid reprocessing full historical data and improve processing efficiency.
Configuration Settings
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
chunkSize | 800–1200 characters | New energy development ledgers often contain long paragraphs of operation and maintenance descriptions and structured tables. This range balances context completeness and chunk granularity |
chunkSplitter | Split by paragraph + title | Ledger documents have clear project chapters and field groups. Splitting by title preserves business logic associations |
customPdfParse | Enabled | Some grid connection acceptance PDFs contain complex mathematical formulas and nested tables. Enabling custom parsing improves the accuracy of formula and structured content extraction |
PARSE_FILE_TIMEOUT_SECONDS | 600 seconds | Single large grid connection reports have many pages, so sufficient parsing time must be reserved to avoid timeouts |
UPLOAD_FILE_MAX_SIZE | 1000 MB | Project archive PDFs often contain high-definition drawings and attachments, so the upload limit must be relaxed to support full document uploads |
chunkSettingMode | Manual configuration | Chunking parameters need to be adjusted for different types of ledger documents (operation and maintenance logs / approval reports). Manual configuration allows flexible adaptation to scenarios |
Three Common Configuration Errors
Reference only; verify against the target FastGPT release and deployment documentation.
- Clicking chunk preview on a PDF in the knowledge base returns a message that the file content cannot be read. The cause is that the
customPdfParseconfiguration is not enabled, or the uploaded file exceeds theUPLOAD_FILE_MAX_SIZElimit. - After configuring the Feishu knowledge base, PPT and PDF documents cannot be synchronized. The cause is that the synchronization permission for the corresponding document types is not enabled, or the
PARSE_FILE_TIMEOUT_SECONDSsetting is too short, causing large document parsing to time out.
How to Verify Configuration Success
- Upload a grid connection acceptance PDF containing complex mathematical formulas, and check whether the formulas display normally in the parsing results.
- Upload a project archive document larger than 500 MB, and confirm that the upload and parsing processes do not trigger timeout or file size limit error messages.
- Perform a chunk preview on a structured Excel ledger document, and check whether all fields are retained and not incorrectly split.
- Adjust the
chunkSizeparameter and re-run the chunking operation, and confirm that the content length after chunking matches the preset configuration range.
The values provided are common starting points and should be evaluated 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.