zlei6/ragflow
1
0
Fork 0
mirror of https://github.com/infiniflow/ragflow.git synced 2026-07-18 21:57:27 +08:00
Code Issues Packages Projects Releases Wiki Activity
7,451 Commits 2 Branches 75 Tags
982b9c7b25205b91cedf2f07008d6cfa212bdc87
Commit Graph

697 Commits

Author SHA1 Message Date
deadtrickster
982b9c7b25 fix(deepdoc): recover word boundaries for non-Latin scripts; skip OCR fallback the recogniser can't serve (#16958) 2026-07-18 00:36:25 +08:00
Jin Hai
8ebdc02cf6 Go: add LLM usage (#17049) 
### Summary

```
RAGFlow(api/default)> CHAT WITH 'glm-4-flash@new_test@zhipu-ai' MESSAGE '30 words describes LLM';
Answer: Hello! I'm ChatGLM, an AI assistant. Feel free to ask me any questions or request help with any tasks.
Input tokens: 5
Output tokens: 28
Time: 12.748241
```

Signed-off-by: Jin Hai <haijin.chn@gmail.com>
 2026-07-17 21:28:43 +08:00
euvre
6da5085073 feat: add chat-level TTS and ASR endpoints for Go API server (#17036) 2026-07-17 17:22:57 +08:00
Haruko386
344477f3f0 fix: loop component usage issue (#17028) 
### Summary

As title

<img width="3718" height="2029" alt="image"
src="https://github.com/user-attachments/assets/227bcd8b-e5b8-465e-983c-c0c0650a4ea5"
/>

---------

Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>
 2026-07-17 16:40:42 +08:00
Haruko386
ddd7fc02ee feat: support yahoo_finance search (#17022) 
### Summary

As title

<img width="3116" height="1949" alt="image"
src="https://github.com/user-attachments/assets/13692cb2-06d6-470b-b13e-182773c8225a"
/>
 2026-07-17 16:05:19 +08:00
Hz_
e55bc969f9 fix(go-agent): filter SSE done sentinel from model stream (#17034) 
## Summary

- Filter the `[DONE]` transport sentinel from Eino model streams
- Add regression coverage for stream messages and callbacks

## Testing

- `bash build.sh --test ./internal/entity/models`
- `bash build.sh --test ./internal/agent/component`
- `bash build.sh --test -run
'Test(AgentChatCompletions_Stream|RunAgent_Stream)' ./internal/handler`
 2026-07-17 15:55:00 +08:00
Hz_
1b77e3ebcd fix(go-agent): preserve canvas system state across turns (#17010) 
## Summary

- Preserve request-scoped system variables such as files and user IDs
during Canvas execution.
- Persist conversation history, turn counts, and tool memory in the
session DSL across turns.
- Parse agent uploads into `sys.files` and align system variable
rendering with Python.

## Testing

- `bash build.sh --test ./internal/agent/...`
- `bash build.sh --test ./internal/service/...`

<img width="1896" height="1232" alt="image"
src="https://github.com/user-attachments/assets/b420cd97-53c3-470f-a3e1-d39cea26a213"
/>
 2026-07-17 15:53:58 +08:00
Lynn
7336c27814 Fix: set Builtin models as default model (#17029) 2026-07-17 15:53:45 +08:00
qinling0210
995e405e8c Support pipeline DSL modification through dataset configuration (backend) (#16991) 
…end)

### Summary

Support pipeline DSL modification through dataset configuration
(backend)

Key modification: knowledgebase.parser_config

---------

Co-authored-by: yzc <yuzhichang@gmail.com>
 2026-07-17 14:40:09 +08:00
Jin Hai
cf71c7193a Go: update server config (#17027) 
### Summary

```
RAGFlow(admin)> list services;
+--------+--------------+----+---------------------+------+---------------+-----------+
| enable | host         | id | name                | port | service_type  | status    |
+--------+--------------+----+---------------------+------+---------------+-----------+
|        | localhost    | 0  | redis               | 6379 | cache         | alive     |
|        | localhost    | 1  | minio               | 9000 | file_store    | alive     |
|        | localhost    | 2  | elasticsearch       | 1200 | retrieval     | alive     |
|        | localhost    | 3  | mysql               | 3306 | meta_data     | alive     |
| false  | localhost    | 4  | jaeger              | 4318 | tracing       | unknown   |
|        | localhost    | 5  | clickhouse          | 9900 | olap          | unknown   |
|        | localhost    | 6  | nats                | 4222 | message_queue | CONNECTED |
|        | 192.168.1.68 | 7  | ragflow-server-9384 | 9384 | api_server    | alive     |
+--------+--------------+----+---------------------+------+---------------+-----------+

```

Signed-off-by: Jin Hai <haijin.chn@gmail.com>
 2026-07-17 13:54:19 +08:00
buua436
27d091b5b6 fix: make await response wait for user input (#16995) 
### What problem does this PR solve?

Await Response incorrectly consumed the initial `sys.query` as its
input, so the first user interaction was skipped.

This change makes Await Response wait for actual user input while
preserving the existing initial-query behavior for the Begin node.

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)
 2026-07-17 13:12:42 +08:00
euvre
d0ac79dc96 fix: ASR/VLM/TTS models not selectable in default model settings (#17024) 2026-07-17 13:05:22 +08:00
Jin Hai
7c698f8e4b Python: remove unused index (#17008) 
### Summary

Remove below indexes.
```
idx_tenant_langfuse_secret_key
idx_tenant_langfuse_public_key
idx_tenant_langfuse_host
```

---------

Signed-off-by: Jin Hai <haijin.chn@gmail.com>
 2026-07-16 23:00:44 +08:00
euvre
c599dc6c52 fix: remove dangling MetaFields reference and skip affected Go backend tests (#16983) 2026-07-16 20:20:29 +08:00
Jin Hai
f8474a67f5 Go: tracing framework (#17004) 
### Summary

Tracing framework

---------

Signed-off-by: Jin Hai <haijin.chn@gmail.com>
 2026-07-16 19:35:43 +08:00
maybehokori
f3a2fb7a10 fix: update somark.tech to somark.cn/somark.ai with purchase URLs (#16892) 
## Summary

Update all `somark.tech` references to `somark.cn` (default for China) /
`somark.ai` (for overseas including Taiwan, China; Hong Kong, China;
Macau, China). Users fill in `base_url` manually — default is
`somark.cn`.

### Changes

| File | Change |
|------|--------|
| `constants.py` | Default → `somark.cn` |
| `somark_parser.py` | SAAS_BASE_URL + fallback → `somark.cn` |
| `ocr_model.py` | Default → `somark.cn` |
| `pdf_parser_common.go` | Go default → `somark.cn` |
| `llm.ts` | API key URL → `somark.cn` |
| `en.ts` | Base URL descriptions + purchase URLs |
| `zh.ts` | Base URL descriptions + purchase URLs |

### Purchase URLs
- China: `https://somark.cn/workbench/purchase`
- Overseas: `https://somark.ai/studio/purchase`

---------

Co-authored-by: justinychuang <huangyicheng@soulcode.cn>
Co-authored-by: Jin Hai <haijin.chn@gmail.com>
 2026-07-16 19:18:05 +08:00
Haruko386
101155bdc7 fix: support message content (#16980) 
### Summary

As Title
<img width="3241" height="1934" alt="image"
src="https://github.com/user-attachments/assets/cb22031c-f200-462a-b871-b4a739e7a66a"
/>
 2026-07-16 19:08:36 +08:00
Haruko386
5307ecd520 Go: add tools for moonshot, baidu and minimax (#17000) 
### Summary

As title
related to #16990
 2026-07-16 19:03:59 +08:00
Jin Hai
c7a623ff81 Go: introduce clickhouse (#16996) 
### Summary

Introduce Clickhouse for data statistics.

Signed-off-by: Jin Hai <haijin.chn@gmail.com>
 2026-07-16 18:36:56 +08:00
euvre
feba9e1158 fix: custom empty_response not shown in streaming chat (Go mode) (#16989) 2026-07-16 17:02:40 +08:00
Hz_
9003584e90 fix(go-agent): add input form of ListOperationsComponent and VariableAggregatorComponent (#16987) 
## Summary
- Add `GetInputForm()` for `ListOperationsComponent` to expose `Query`
input field in debug UI
- Add `GetInputForm()` for `VariableAggregatorComponent` to expose
`Variables` input field in debug UI

## Test
- Verify input form fields render correctly for both components in the
debug UI
 2026-07-16 16:58:38 +08:00
Jin Hai
8bb2cd0fb7 Go: fix db alignment (#16964) 
### Summary

Align go db schema with python

---------

Signed-off-by: Jin Hai <haijin.chn@gmail.com>
 2026-07-16 16:00:56 +08:00
Yingfeng
5ab424c10c Revert "fix: remove MySQL document.meta_fields mapping to match Python schema" (#16981) 
Reverts infiniflow/ragflow#16929
 2026-07-16 15:17:49 +08:00
Jack
fd5487e064 feat: add /api/v1/language endpoint for Go/Python runtime detection (#16952) 
Both backends serve GET /api/v1/language. Frontend calls it once and
caches. By this way, front end can know the backend is go or python and
thus can determine which part of logic to load.

---------

Co-authored-by: Claude <noreply@anthropic.com>
 2026-07-16 15:13:43 +08:00
Wang Qi
46e14fe52b Fix doc.MetaFields undefined (#16976) 
Follow on #16929
 2026-07-16 13:27:08 +08:00
Zhichang Yu
d2f4304ff4 Fix Pipeline.Run callers and rename SetupOverrides (#16973) 
Adapt Pipeline.Run call sites to the new signature and rename the
SetupOverrides family to OverrideParams for consistency.
 2026-07-16 13:12:25 +08:00
Haruko386
f9a1fbfab9 fix: support categorize component in agent (#16966) 
<img width="3716" height="2039" alt="image"
src="https://github.com/user-attachments/assets/62709d24-60ba-4c92-81f1-48b6d23e3be1"
/>
 2026-07-16 13:10:29 +08:00
euvre
30de65991b fix: remove MySQL document.meta_fields mapping to match Python schema (#16929) 2026-07-16 13:07:49 +08:00
Jack
047c714ffb Feat: ingestion cleanup (#16953) 
### Summary

1. Remove dead code (replaced by builtin ingestion pipeline)
2. Refactor (move document parsing progress from http api into ingestion
executor)
 2026-07-16 11:05:26 +08:00
Jin Hai
2a6e210020 Go: align db schema to python (#16935) 
As title.

---------

Signed-off-by: Jin Hai <haijin.chn@gmail.com>
 2026-07-15 23:05:06 +08:00
Haruko386
ff3d566a4f fix[go]: support retrieval kb in agent chat (#16944) 
### Summary

As title
- [x] fix tool & component `Retrieval KB`
- [x] fix agent cannot use  `Retrieval KB` in agent chat

#### Main cause

Model provider do not set up:
```Go
if chatModelConfig.Tools != nil {
    reqBody["tools"] = chatModelConfig.Tools
}
```

#### Working Now
<img width="3774" height="2128" alt="image"
src="https://github.com/user-attachments/assets/400a349d-0211-43e5-a7ec-7a014acf77a6"
/>

```
 ____________________________________
< This PR takes me all day to do it. >
 ------------------------------------
  \
   \   (\__/)
       (•ㅅ•)
       /   づ                                                                                                                                                                                                    
```
 2026-07-15 21:44:28 +08:00
Hz_
a7da78d0d7 refactor(go-agent): unify tool-backed canvas components (#16912) 
## Summary

This PR consolidates Eino-backed Agent tools behind the shared
ToolBackedComponent implementation and aligns their Canvas
configuration,
runtime inputs, output conversion, validation, and registration.

## What changed

- Migrated these Canvas components to the unified tool-backed path:
    - Tavily Search and Extract
    - Execute SQL
    - Google
    - Yahoo Finance
    - Email
    - DuckDuckGo
    - Wikipedia
    - Google Scholar
    - ArXiv
    - PubMed
    - BGPT
    - GitHub
    - WenCai
    - SearXNG
    - Keenable Search

- Removed superseded component wrappers and their duplicate tests.
- Added dedicated registry builders for node-level configuration and
validation.
- Kept model-emitted runtime inputs separate from Canvas node
configuration.
- Moved Email defaults, template resolution, recipient parsing, SMTP
execution, and output conversion into the owning tool.
- Added complete ToolComponent specifications for Canvas inputs,
outputs, and input forms.
- Preserved raw upstream fields where downstream workflows may depend on
them.
- Added workflow registration coverage for all migrated component names.
- Kept HTTP Request, Docs Generator, and Browser as standalone
components because they are not Eino-backed tools.

## Testing

Passed:

bash build.sh --test ./internal/agent/tool/...
bash build.sh --test ./internal/agent/component/...
bash build.sh --test ./internal/agent/runtime/...

<img width="2057" height="1111" alt="image"
src="https://github.com/user-attachments/assets/c728d7a3-9d15-4c5c-b0eb-6b77ad0e41ac"
/>
<img width="2057" height="1111" alt="image"
src="https://github.com/user-attachments/assets/ef13119a-ef92-4b43-9061-eee9511bf492"
/>
<img width="2057" height="1111" alt="image"
src="https://github.com/user-attachments/assets/f50f2c30-eab0-463c-b0b3-0a02523219f1"
/>
 2026-07-15 21:42:08 +08:00
Jakob
d55de09b7d Updated RAGcon model provider (go) (#16950) 
### Summary
We have updated our model driver to work with go.
It is based on OpenAI-API-Compatible model provider.

Draft
#15519

Our old model provider
#13425
 2026-07-15 20:59:23 +08:00
zhifu gao
794c6b63e4 fix(agent): return DataOperations input form (#16930) 
## Summary

- return the input form collected by
`DataOperationsComponent.GetInputForm`
- add a regression test covering system variables, component outputs,
and duplicate references
 2026-07-15 20:32:59 +08:00
euvre
70a49c947d Remove unused /datasets/<id>/embedding endpoint (#16936) 2026-07-15 17:42:41 +08:00
Lynn
63c010cd04 Fix: code-rabbit review comment (#16940) 2026-07-15 15:44:08 +08:00
Jack
e543ff02c4 Feat: nats message processing refactor and support built in DSL (#16873) 2026-07-15 14:53:16 +08:00
Haruko386
14d361aa51 refactor: remove unused getInputElementsFromText function (#16927) 
### Summary

Removed unused function getInputElementsFromText and its associated
comments.
 2026-07-15 11:38:52 +08:00
Haruko386
a19af34d31 fix: unable to use agent in agent (#16910) 
### Summary

As title

#### Agent:
<img width="3707" height="2030" alt="image"
src="https://github.com/user-attachments/assets/68480004-9c79-4302-ae22-eb3375b7fac9"
/>

#### Data_operations
<img width="3716" height="2044" alt="image"
src="https://github.com/user-attachments/assets/068cc1a6-1ba5-4746-a447-a706d2149ef5"
/>
 2026-07-15 10:58:59 +08:00
Haruko386
215d291cfa fix: support email send for go backend (#16895) 
### Summary

As title

<img width="3761" height="2036" alt="image"
src="https://github.com/user-attachments/assets/775f8c2a-784e-4713-bcf0-c26bfc368dac"
/>


<img width="1256" height="2760" alt="d2aeeb50dd11cc8dc76bc4743425680e"
src="https://github.com/user-attachments/assets/80f9c572-549d-427f-8e48-35dc04cfa98a"
/>
 2026-07-15 10:56:36 +08:00
Haruko386
654279a745 fix: support code_execute for go backend (#16903) 
### Summary

As title

<img width="3764" height="2033" alt="image"
src="https://github.com/user-attachments/assets/3f5084c2-0a04-4a44-bb31-a39a59c2a91f"
/>
 2026-07-15 10:55:39 +08:00
Haruko386
3339040987 fix: add GetInputForm for string_transform (#16900) 
### Summary

As title:

#### merge:
<img width="2972" height="1473" alt="image"
src="https://github.com/user-attachments/assets/2171e31f-13f4-4799-a6d0-1b9326210658"
/>

---
#### split
<img width="858" height="416" alt="image"
src="https://github.com/user-attachments/assets/f89d39e3-da85-4349-a53c-e27a17cc9dfa"
/>
 2026-07-15 10:54:18 +08:00
Hz_
276941f7b4 fix(go-agent): add VariableAssigner input form (#16902) 
## Summary

- Add the runtime input form for VariableAssigner components
- Allow the component input-form endpoint to resolve VariableAssigner
nodes

## Testing

- `bash build.sh --test ./internal/agent/component/...`
 2026-07-15 10:47:49 +08:00
Jin Hai
b506d8ec5f Go: add more frameworks (#16915) 
### Summary

1. Add hooks for server init
2. Add hooks for router init
3. Add jaeger and otel related config

Signed-off-by: Jin Hai <haijin.chn@gmail.com>
 2026-07-14 22:20:26 +08:00
Jin Hai
c242669a9d Go: fix bugs (#16911) 
### Summary

1. Fix 'list users' in admin CLI
2. Change 3 slow test cases priority.

Signed-off-by: Jin Hai <haijin.chn@gmail.com>
 2026-07-14 21:58:33 +08:00
Lynn
bd7a3bb878 Fix: align go provider apis with python apis (#16867) 
### Summary

As title.

---------

Co-authored-by: Jin Hai <haijin.chn@gmail.com>
Co-authored-by: Wang Qi <wangq8@outlook.com>
 2026-07-14 21:06:11 +08:00
Jin Hai
26161ce844 Go: refactor (#16906) 
### Summary

remove and merge code

Signed-off-by: Jin Hai <haijin.chn@gmail.com>
 2026-07-14 19:29:34 +08:00
euvre
22dd1ad401 fix: resolve provider name case mismatch and 301 redirect in Go mode (#16889) 2026-07-14 17:05:21 +08:00
euvre
315e5725c2 Fix: Search mind map not displaying (#16866) 2026-07-14 17:04:47 +08:00
Hz_
ea18ab3ba0 fix(go-agent): align ExeSQL component inputs with Python (#16896) 
## Summary

- Resolve configured ExeSQL SQL references before tool invocation.
- Align ExeSQL parameters, defaults, JSON tags, and tool schema with
Python.
- Preserve SQL string literals and restore Canvas output fields.

## Testing

- `bash build.sh --test ./internal/agent/tool/...`
- `bash build.sh --test ./internal/agent/component/...`
- `bash build.sh --test ./internal/agent/runtime/...`

<img width="2039" height="1041" alt="image"
src="https://github.com/user-attachments/assets/9f4beca7-ca28-4641-adda-0570415fcaa1"
/>
 2026-07-14 15:02:28 +08:00