mtfyun/ragflow
1
0 Fork 0
mirror of https://github.com/infiniflow/ragflow.git synced 2026-11-11 22:01:50 +08:00
Code Issues Packages Projects Releases Wiki Activity
9,247 Commits 3 Branches 194 Tags
0a00a732905dfd6c04c2873ddafc1904a73132a2
Commit Graph
1434 Commits
Author SHA1 Message Date
buua436 0a00a73290 fix: align mindmap relations with Python (#19579) 2026-09-14 16:30:20 +08:00
xugangqiang 7c99668aee fix(parser): emit xlsx spreadsheets as typed table chunks (#19509) 2026-09-14 15:10:48 +08:00
buua436 ec5a0136de fix: split knowledge compilation logs by task type (#19548) 2026-09-14 15:02:41 +08:00
buua436 9562f0f3ab fix: merge same-name knowledge graph entities (#19525) 2026-09-14 15:00:54 +08:00
qinling0210 aa78e8d224 Remove internal cli 'insert chunks from file' and 'insert metadata from file' (#19591) 2026-09-14 14:23:04 +08:00
Harsh Kashyap fa2467d33f fix(api): intersect metadata filter results with the caller's doc scope (#19353) 
Fix #19352 — metadata filter now intersects with document_ids (was extending/widening with duplicates). Mirrors Go's constrainDocIDs()
 2026-09-14 14:16:07 +08:00
Sbaaoui Idriss d3a6aaae31 fix(storage):preserve original errors after retries (#19580) 2026-09-14 13:11:36 +08:00
mkaaad 0ff7e79c95 feat(connector): support file_extension for MySQL/PostgreSQL sync (#19466) 
Read file_extension from the connector config (default .txt) and use it
as the document extension instead of the hardcoded .txt when rows are
synced from MySQL and PostgreSQL data sources.
Python side: #15998
 2026-09-14 11:54:38 +08:00
ming1523 d8d06bcc3a fix(model): delete provider models and instances in one transaction (#19562) 2026-09-14 11:42:37 +08:00
AnonrouterandSunday Labs --labrat b69afb5005 feat: add AnonRouter model provider (#19561) 
I maintain AnonRouter and added it as a hosted OpenAI-compatible model
provider

Co-authored-by: Sunday Labs --labrat <hi.sundaylabs@gmail.com>
 2026-09-14 11:31:02 +08:00
mkaaad 3e1396d042 feat(syncer): add Feishu Wiki connector for the Go backend (#19453) 
Implements a Feishu Wiki file-node data source connector mirroring the
Python connector semantics: recursive subtree traversal with pagination,
extension/keyword/window filters, tenant token caching, size-capped
streaming downloads with Feishu JSON error-envelope protection, bounded
429/5xx retry honoring x-ogw-ratelimit-reset/Retry-After, and
checkpoint/resume via SourceID anchors. Registers the "feishu_wiki"
source in the built-in connector registry; prune remains unsupported.
Python side #19425
 2026-09-14 11:27:43 +08:00
coderdailyoneandcoderdailyone 7371255eda fix(api): authorize the dataset document download route by tenant in the Go API (#19543) 
Fix #19542 — Go document download now checks caller access (datasetService.Accessible); previously any logged-in user could download any tenant's document. Returns same "not found" envelope to avoid leaking existence. 

Co-authored-by: coderdailyone <codecookie@proton.me>
 2026-09-14 10:30:13 +08:00
Sbaaoui Idriss 9bc9c36ce6 fix(ollama): disable reasoning for vision descriptions (#19529) 2026-09-14 09:59:05 +08:00
Yingfeng 4e54fe1338 Port Agentic RAG to Golang (#19563) 2026-09-14 09:38:51 +08:00
euvre b19a10fb70 fix(wiki): version history always empty on Go API due to page-slug key mismatch (#19305) 2026-09-13 15:50:37 +08:00
Sourav Rajvi ec0ff067c1 fix(parser): handle large sparse spreadsheets, JSON BOM, and XLSX QA chunking (#19237) 2026-09-13 15:46:34 +08:00
Yingfeng 1f5d33333a Add claim/evidence to tree/pageindex compilation pipeline (#19267) 2026-09-11 22:37:33 +08:00
jay77721 868451c639 fix(ingestion): stop dropping QA pairs from multiline table cells (#19536) 2026-09-11 19:35:18 +08:00
天海蒼灆 ebbdd5028a feat(agent): support document_ids in retrieval node and tool (#16938) 2026-09-11 19:15:23 +08:00
jay77721 fbd374e92f refactor(parser): unify parser outputs to structured JSON items (#19467) 2026-09-11 18:52:15 +08:00
xugangqiang a36615c6fa Migrate cache backend from Valkey 8 to apache/kvrocks:2.16.0 (#19374) 2026-09-11 17:39:11 +08:00
buua436 c122243042 fix: update knowledge artifacts on document status changes (#19532) 2026-09-11 17:31:51 +08:00
changshenhan 78d426abda fix(chunker): one BPE pass per hardSplitPiece iteration instead of two (#18648) 2026-09-11 17:29:10 +08:00
VincentLambert 941edd1269 feat(nlp): rerank on natural chunk text, not tokenized content_ltks (#16961) 
Fix reranker using tokenized text (10–94x lower scores) instead of natural chunk text.
 2026-09-11 16:59:06 +08:00
buua436 6051b2150f refa: remove Go artifact deletion APIs (#19519) 2026-09-11 16:08:00 +08:00
qinling0210 ce34ec42f5 Port agentic RAG PRs to GO (#19527) 
### Summary

Port
https://github.com/infiniflow/ragflow/pull/19424/
https://github.com/infiniflow/ragflow/pull/19459/

Address comments from https://github.com/infiniflow/ragflow/pull/19500
 2026-09-11 15:58:42 +08:00
Aroh Maurya ea2ce6d323 fix(ssrf): unwrap IPv6 transition addresses in the URL guard (#19517) 
Fix SSRF guard bypass via IPv6 transition addresses (NAT64/6to4/Teredo wrapping private IPv4). Now unwraps and re-checks embedded address in Python + Go.
 2026-09-11 13:44:00 +08:00
Sbaaoui Idriss bcf33c6238 fix(structure-graph): discover buckets from raw rows (#19523) 2026-09-11 13:37:01 +08:00
taek105 4ad809634f fix(document): validate immutable fields on update (#19444) 
Fix #16831 — Go document endpoints now reject changes to ingestion-managed fields (progress, chunk_num, token_num, etc.), including zero-valued. PUT now persists name only, not full record.
 2026-09-11 12:39:19 +08:00
xugangqiangandmaoyifeng 2e6003cfd4 Unify ONNX Runtime native to 1.29.0 from infiniflow/ragflow-build (#19446) 
Switch Go DeepDoc's static ONNX Runtime to our own infiniflow/ragflow-build release (v1.29.0) instead of third-party. Keeps fused kernels; normalizes dir naming; added ORT version consistency check.

---------

Co-authored-by: maoyifeng <myfbee5@163.com>
 2026-09-11 12:12:34 +08:00
Sbaaoui Idriss 681f418424 fix(invoke): tolerate invalid headers across runtimes (#19481) 2026-09-11 10:21:25 +08:00
buua436 289100188d fix: avoid nested compiler pool embedding deadlock (#19486) 2026-09-11 10:17:12 +08:00
xu jiaming 3dab5db315 fix(mcp): send rerank_candidates_count so retrieval paging works past 64 results (#19018) 
Fix #19017 — MCP retrieval now sends rerank_candidates_count = max(64, page * page_size), fixing paging past 64 results.
 2026-09-10 20:12:43 +08:00
mkaaad c51f44bab3 fix(agent/tool): translate MCP tool inputSchema.properties in Info() (#19244) 
Fix MCPToolAdapter.Info() — now reads parameter names from inputSchema["properties"] instead of top-level schema keys. Model gets correct param names/types/required flags.
 2026-09-10 20:04:15 +08:00
Wang Qi 607f986664 Revert "fix(ingestion): allow re-parsing a document whose parse was canceled at progress 0" (#19503) 
Reverts infiniflow/ragflow#19258
 2026-09-10 20:03:22 +08:00
qinling0210 ef800ba5c2 Port agentic RAG to GO (#19500) 
### Summary

Port agentic RAG to GO
 2026-09-10 20:02:48 +08:00
Lynn 97029b08f1 Fix: num token usage (#19495) 2026-09-10 17:55:17 +08:00
buua436 e9a6397ab8 fix: support keyword search for dataset structure graphs (#19473) 2026-09-10 17:29:06 +08:00
euvre 7090ca834b fix(go): stop masking document preview failures as "document not found" and add tenant access check (#19348) 2026-09-10 16:56:32 +08:00
euvre 347d0827a0 fix(ingestion): allow re-parsing a document whose parse was canceled at progress 0 (#19258) 2026-09-10 16:22:43 +08:00
Wang Qi 3dbca958db Fix: NOT generate graph all entities/relations as a graph chunk (#19474) 2026-09-10 13:43:59 +08:00
Yash Raj Pandey 05168a37fa Decode a single-part email body by its Content-Transfer-Encoding (#19178) 2026-09-10 10:34:18 +08:00
Sbaaoui Idriss 11f4d11920 fix(agent): preserve Starter values beside query inputs (#19420) 2026-09-10 10:16:32 +08:00
Jin Hai d5a89c324c Go: update build.sh and development.md (#19448) 
Signed-off-by: Jin Hai <haijin.chn@gmail.com>
 2026-09-09 21:03:48 +08:00
Konstantin Romankov 4731ee3322 feat: add Hubris as a model provider (#19341) 
Added Hubris as a new OpenAI-compatible model provider for the Russian market, with per-model thinking.default_value empirically measured (not guessed), endpoint pinned (not configurable)
 2026-09-09 17:19:46 +08:00
WOLIKIMCHENG 1a99ebea38 fix(chunk): enforce document scope on updates (#19221) 
Fixed the update endpoint to enforce document-scoped chunk updates by validating doc_id before writes, preventing cross-document modifications.
 2026-09-09 17:18:40 +08:00
Wang Qi 2284c8e1ab Refactor: Follow #19404 to add RERANK_TOKEN_LIMIT_MODE=truncate|passthrough|raise_error (#19421) 2026-09-09 16:32:08 +08:00
Sbaaoui Idriss 3fa257a0da fix(compiler): resolve global built-in template scope (#19426) 2026-09-09 16:12:14 +08:00
xugangqiang 8c8253a3be chore(deps): let download_go_deps.py fetch Go DeepDoc .ort weights (#19376) 
Go DeepDoc now uses .ort weights downloaded via download_go_deps.py — no MODEL_DIR needed from repo root. Script fetches 5 model files into rag/res/deepdoc/; supports mirrors; CI updated. .onnx path unchanged.
 2026-09-09 16:07:43 +08:00
jay77721 1ed2fc26f0 refactor(ingestion): use worker-driven task pull dispatcher (#19314) 2026-09-09 14:13:07 +08:00