### What problem does this PR solve?
The Upstage Go driver landed in #14817 with chat, list models, and check
connection. `Embed` was left as a stub that returns `"not implemented"`.
This PR fills the gap.
Upstage exposes an OpenAI-compatible embeddings endpoint at
`https://api.upstage.ai/v1/solar/embeddings` via the
`solar-embedding-1-large` family (`solar-embedding-1-large-query` for
queries, `solar-embedding-1-large-passage` for passages), and the Python
side has had `UpstageEmbed(OpenAIEmbed)` in `rag/llm/embedding_model.py`
for a long time targeting this same path. The existing
`conf/models/upstage.json` did not list any embedding model out of the
box, so a tenant who wanted to use Upstage end to end could not run an
embedding call. This PR fills the gap.
### What this PR includes
- `conf/models/upstage.json`: add `"embedding": "embeddings"` under
`url_suffix` so the driver can build the URL from config (matches the
`URLSuffix.Embedding` field already used by openai, mistral,
siliconflow, zhipu-ai), and add `solar-embedding-1-large-query` and
`solar-embedding-1-large-passage` entries under `models`.
- `internal/entity/models/upstage.go`: replace the `Embed` stub with a
real implementation that POSTs to `/v1/solar/embeddings`. Adds local
response types `upstageEmbeddingData` and `upstageEmbeddingResponse`.
No factory change. No interface change.
### How the implementation works
- Validate `apiConfig`, the API key, and the model name. Use the
existing `baseURLForRegion` helper so an unknown region fails fast with
a clear error.
- Wrap the request with `context.WithTimeout(nonStreamCallTimeout)` so
the call has a clear deadline. Same pattern as `ChatWithMessages` and
`ListModels` already use in this file.
- Send all input texts in one request. The Upstage API accepts the
`input` field as an array.
- Parse `data[*].embedding` and copy each slice into a `[]EmbeddingData`
indexed by `data[*].index` so the output order matches the input order
even if the API returns items in a different order.
- An empty input slice returns `[]EmbeddingData{}` with no HTTP call.
- Non-200 responses propagate the upstream status line and body.
- A final pass checks that every input slot got a vector. If any slot is
still empty, return a clear error so the caller does not silently use a
zero vector.
### Note on stacking
This PR builds on #14817 (the Upstage driver). Until #14817 merges, this
PR's diff on GitHub will include both that PR's commits and this one.
After #14817 lands on `main`, GitHub will auto-reduce this PR to only
the `Embed` changes (one commit, ~119 line diff in `upstage.go` plus ~15
lines in `upstage.json`).
### Type of change
- [x] New Feature (non-breaking change which adds functionality)
### How was this tested?
- `go build ./internal/entity/models/...` returns exit 0 on go 1.25 (the
`go.mod` minimum).
- The full method set on `UpstageModel` still matches the `ModelDriver`
interface.
- Pattern parity with the existing Mistral Embed
(`internal/entity/models/mistral.go`) and OpenAI Embed
(`internal/entity/models/openai.go`) implementations.
Closes#14818
Depends on #14817
Tracking: #14736
---------
Co-authored-by: Jin Hai <haijin.chn@gmail.com>
### What problem does this PR solve?
This PR completes the Baichuan provider
**The following functionalities are now supported:**
**Baichuan:**
- [x] Chat / Stream Chat
- [x] Embedding
- [ ] ~~Rerank~~
- [ ] ~~Model listing~~
- [ ] ~~Provider connection checking~~
- [ ] ~~Balance~~
**Verified examples from the CLI:**
```plaintext
# Baichuan
RAGFlow(user)> embed text 'walkerwhat' 'jumperwho' with 'Baichuan-Text-Embedding@test@baichuan' dimension 16;
+-----------+-------+
| dimension | index |
+-----------+-------+
| 1024 | 0 |
| 1024 | 1 |
+-----------+-------+
AGFlow(user)> chat with 'Baichuan-M2@test@baichuan' message 'who r u'
Answer: I'm BaiChuan, a helpful AI assistant created by Baichuan-AI. I'm designed to be a knowledgeable, friendly, and reliable assistant for various tasks like answering questions, explaining concepts, writing content, and more. Feel free to ask me anything! 😊
Time: 1.637975
RAGFlow(user)> stream chat with 'Baichuan-M2@test@baichuan' message 'who r u'
Answer: I'm BaiChuan-m2, an AI assistant developed by Baichuan-AI. My purpose is to help you with a wide range of tasks by providing information, answering questions, solving problems, and assisting with creative projects. Think of me as a helpful digital companion! If you have any questions or need assistance, just let me know.😊
Time: 1.692321
```
### Type of change
- [x] New Feature (non-breaking change which adds functionality)
- [x] Refactoring
This PR adds focused unit tests for aggregate_by_field in OceanBase
memory utilities to improve behavior coverage for real-world input
shapes.
- Adds test coverage for list-valued aggregation fields, including
whitespace trimming and skipping invalid list entries.
- Adds test coverage for scalar field values to ensure blank/non-string
values are ignored.
- Confirms aggregation output remains correct and stable for
mixed-quality message payloads.
### Why this helps
It strengthens regression protection for aggregation logic used by
memory retrieval flows, with no production code changes and minimal
review risk.
### What problem does this PR solve?
fix:
update null checks to use 'is None' for better clarity
replace RAGFlowSelect with SelectWithSearch in DebugContent
add max height and overflow to DialogContent in ParameterDialog
remove unused types from DataOperationsForm
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
Fixes#13817
### What problem does this PR solve?
The "knowledge graph construction" link on line 21 of
`docs/guides/dataset/run_retrieval_test.md` points to
`./construct_knowledge_graph.md`, which doesn't exist. The actual file
is at `./advanced/construct_knowledge_graph.md`.
### Type of change
- [x] Documentation Update
Signed-off-by: majiayu000 <1835304752@qq.com>
### What problem does this PR solve?
Add a Go driver for StepFun (阶跃星辰), one of the unchecked providers on
the umbrella tracking issue #14736.
Until this PR, a tenant who configured `stepfun` as a model provider in
the Go layer fell through to the default branch of
`internal/entity/models/factory.go` and got the dummy driver. Chat, list
models, and check connection all returned `"not implemented"` instead of
reaching the StepFun API.
The Python side has had StepFun registered in `rag/llm/__init__.py` as a
`SupportedLiteLLMProvider` with base URL `https://api.stepfun.com/v1`,
plus `StepFunCV` for vision and `StepFunSeq2txt` for ASR, but no Go
path. StepFun's chat API is OpenAI-compatible, so the implementation
pattern is the same as the merged Moonshot driver (#14433) and OpenAI
driver (#14605).
### What this PR includes
- New file `internal/entity/models/stepfun.go` with a `StepFunModel`
that implements the `ModelDriver` interface.
- `factory.go`: route the `"stepfun"` provider name to
`NewStepFunModel`.
- New `conf/models/stepfun.json` with the public StepFun chat models
(step-2-16k, step-1 family in 8k/32k/128k/256k context lengths,
step-1-flash, and the step-1v / step-1o vision models) and `url_suffix`
entries for `chat` and `models`.
### How the driver works
- StepFun exposes the OpenAI-compatible API at
`https://api.stepfun.com/v1`.
- `ChatWithMessages` and `ChatStreamlyWithSender` post to
`/chat/completions` in the same shape as the merged moonshot,
openrouter, and openai drivers.
- `ListModels` and `CheckConnection` call `/models` to list available
ids and confirm the API key works.
- `Embed` is left as `"not implemented"`. StepFun has not advertised a
public embeddings endpoint in the API reference linked from the umbrella
issue
(`https://platform.stepfun.com/docs/en/api-reference/chat/chat-completion-create`
is the chat endpoint), so any real implementation belongs in a separate
follow-up only after the endpoint is verified.
- `Rerank` and `Balance` return `"no such method"` because StepFun does
not expose either.
### Type of change
- [x] New Feature (non-breaking change which adds functionality)
### How was this tested?
- `go build ./internal/entity/models/...` returns exit 0 with no errors
on go 1.25 (the `go.mod` minimum).
- Method set of `StepFunModel` matches the `ModelDriver` interface:
`NewInstance`, `Name`, `ChatWithMessages`, `ChatStreamlyWithSender`,
`Embed`, `Rerank`, `ListModels`, `Balance`, `CheckConnection`.
- Pattern parity with the merged moonshot (#14433), openai (#14605),
openrouter (#14652), and xai (#14550) drivers.
Closes#14814
Tracking: #14736
Bumps [urllib3](https://github.com/urllib3/urllib3) from 2.6.3 to 2.7.0.
<details>
<summary>Release notes</summary>
<p><em>Sourced from <a
href="https://github.com/urllib3/urllib3/releases">urllib3's
releases</a>.</em></p>
<blockquote>
<h2>2.7.0</h2>
<h2>🚀 urllib3 is fundraising for HTTP/2 support</h2>
<p><a
href="https://sethmlarson.dev/urllib3-is-fundraising-for-http2-support">urllib3
is raising ~$40,000 USD</a> to release HTTP/2 support and ensure
long-term sustainable maintenance of the project after a sharp decline
in financial support. If your company or organization uses Python and
would benefit from HTTP/2 support in Requests, pip, cloud SDKs, and
thousands of other projects <a
href="https://opencollective.com/urllib3">please consider contributing
financially</a> to ensure HTTP/2 support is developed sustainably and
maintained for the long-haul.</p>
<p>Thank you for your support.</p>
<h2>Security</h2>
<p>Addressed high-severity security issues. Impact was limited to
specific use cases detailed in the accompanying advisories; overall user
exposure was estimated to be marginal.</p>
<ul>
<li>
<p>Decompression-bomb safeguards of the streaming API were bypassed:</p>
<ol>
<li>When <code>HTTPResponse.drain_conn()</code> was called after the
response had been read and decompressed partially. (Reported by <a
href="https://github.com/Cycloctane"><code>@Cycloctane</code></a>)</li>
<li>During the second <code>HTTPResponse.read(amt=N)</code> or
<code>HTTPResponse.stream(amt=N)</code> call when the response was
decompressed using the official <a
href="https://pypi.org/project/brotli/">Brotli</a> library. (Reported by
<a
href="https://github.com/kimkou2024"><code>@kimkou2024</code></a>)</li>
</ol>
<p>See GHSA-mf9v-mfxr-j63j for details.</p>
</li>
<li>
<p>HTTP pools created using
<code>ProxyManager.connection_from_url</code> did not strip sensitive
headers specified in <code>Retry.remove_headers_on_redirect</code> when
redirecting to a different host. (GHSA-qccp-gfcp-xxvc reported by <a
href="https://github.com/christos-spearbit"><code>@christos-spearbit</code></a>)</p>
</li>
</ul>
<h2>Deprecations and Removals</h2>
<ul>
<li>Used <code>FutureWarning</code> instead of
<code>DeprecationWarning</code> for better visibility of existing
deprecation notices. Rescheduled the removal of deprecated features to
version 3.0. (<a
href="https://redirect.github.com/urllib3/urllib3/issues/3763">urllib3/urllib3#3763</a>)</li>
<li>Removed support for end-of-life Python 3.9. (<a
href="https://redirect.github.com/urllib3/urllib3/issues/3720">urllib3/urllib3#3720</a>)</li>
<li>Removed support for end-of-life PyPy3.10. (<a
href="https://redirect.github.com/urllib3/urllib3/issues/4979">urllib3/urllib3#4979</a>)</li>
<li>Bumped the minimum supported pyOpenSSL version to 19.0.0. (<a
href="https://redirect.github.com/urllib3/urllib3/issues/3777">urllib3/urllib3#3777</a>)</li>
</ul>
<h2>Bugfixes</h2>
<ul>
<li>Fixed a bug where <code>HTTPResponse.read(amt=None)</code> was
ignoring decompressed data buffered from previous partial reads. (<a
href="https://redirect.github.com/urllib3/urllib3/issues/3636">urllib3/urllib3#3636</a>)</li>
<li>Fixed a bug where <code>HTTPResponse.read()</code> could cache only
part of the response after a partial read when
<code>cache_content=True</code>. (<a
href="https://redirect.github.com/urllib3/urllib3/issues/4967">urllib3/urllib3#4967</a>)</li>
<li>Fixed <code>HTTPResponse.stream()</code> and
<code>HTTPResponse.read_chunked()</code> to handle <code>amt=0</code>.
(<a
href="https://redirect.github.com/urllib3/urllib3/issues/3793">urllib3/urllib3#3793</a>)</li>
<li>Updated <code>_TYPE_BODY</code> type alias to include missing
<code>Iterable[str]</code>, matching the documented and runtime behavior
of chunked request bodies. (<a
href="https://redirect.github.com/urllib3/urllib3/issues/3798">urllib3/urllib3#3798</a>)</li>
<li>Fixed <code>LocationParseError</code> when paths resembling
schemeless URIs were passed to
<code>HTTPConnectionPool.urlopen()</code>. (<a
href="https://redirect.github.com/urllib3/urllib3/issues/3352">urllib3/urllib3#3352</a>)</li>
<li>Fixed <code>BaseHTTPResponse.readinto()</code> type annotation to
accept <code>memoryview</code> in addition to <code>bytearray</code>,
matching the <code>io.RawIOBase.readinto</code> contract and enabling
use with <code>io.BufferedReader</code> without type errors. (<a
href="https://redirect.github.com/urllib3/urllib3/issues/3764">urllib3/urllib3#3764</a>)</li>
</ul>
</blockquote>
</details>
<details>
<summary>Changelog</summary>
<p><em>Sourced from <a
href="https://github.com/urllib3/urllib3/blob/main/CHANGES.rst">urllib3's
changelog</a>.</em></p>
<blockquote>
<h1>2.7.0 (2026-05-07)</h1>
<h2>Security</h2>
<p>Addressed high-severity security issues.
Impact was limited to specific use cases detailed in the accompanying
advisories; overall user exposure was estimated to be marginal.</p>
<ul>
<li>
<p>Decompression-bomb safeguards of the streaming API were bypassed:</p>
<ol>
<li>When <code>HTTPResponse.drain_conn()</code> was called after the
response had been
read and decompressed partially.</li>
<li>During the second <code>HTTPResponse.read(amt=N)</code> or
<code>HTTPResponse.stream(amt=N)</code> call when the response was
decompressed
using the official <code>Brotli
<https://pypi.org/project/brotli/></code>__ library.</li>
</ol>
<p>See <code>GHSA-mf9v-mfxr-j63j
<https://github.com/urllib3/urllib3/security/advisories/GHSA-mf9v-mfxr-j63j></code>__
for details.</p>
</li>
<li>
<p>HTTP pools created using
<code>ProxyManager.connection_from_url</code> did not strip
sensitive headers specified in
<code>Retry.remove_headers_on_redirect</code> when
redirecting to a different host.
(<code>GHSA-qccp-gfcp-xxvc
<https://github.com/urllib3/urllib3/security/advisories/GHSA-qccp-gfcp-xxvc></code>__)</p>
</li>
</ul>
<h2>Deprecations and Removals</h2>
<ul>
<li>Used <code>FutureWarning</code> instead of
<code>DeprecationWarning</code> for better
visibility of existing deprecation notices. Rescheduled the removal of
deprecated features to version 3.0.
(<code>[#3763](https://github.com/urllib3/urllib3/issues/3763)
<https://github.com/urllib3/urllib3/issues/3763></code>__)</li>
<li>Removed support for end-of-life Python 3.9.
(<code>[#3720](https://github.com/urllib3/urllib3/issues/3720)
<https://github.com/urllib3/urllib3/issues/3720></code>__)</li>
<li>Removed support for end-of-life PyPy3.10.
(<code>[#4979](https://github.com/urllib3/urllib3/issues/4979)
<https://github.com/urllib3/urllib3/issues/4979></code>__)</li>
<li>Bumped the minimum supported pyOpenSSL version to 19.0.0.
(<code>[#3777](https://github.com/urllib3/urllib3/issues/3777)
<https://github.com/urllib3/urllib3/issues/3777></code>__)</li>
</ul>
<h2>Bugfixes</h2>
<ul>
<li>Fixed a bug where <code>HTTPResponse.read(amt=None)</code> was
ignoring decompressed
data buffered from previous partial reads.
(<code>[#3636](https://github.com/urllib3/urllib3/issues/3636)
<https://github.com/urllib3/urllib3/issues/3636></code>__)</li>
<li>Fixed a bug where <code>HTTPResponse.read()</code> could cache only
part of the
response after a partial read when <code>cache_content=True</code>.</li>
</ul>
<!-- raw HTML omitted -->
</blockquote>
<p>... (truncated)</p>
</details>
<details>
<summary>Commits</summary>
<ul>
<li><a
href="9a950b92d9"><code>9a950b9</code></a>
Release 2.7.0</li>
<li><a
href="5ec0de499b"><code>5ec0de4</code></a>
Merge commit from fork</li>
<li><a
href="2bdcc44d1e"><code>2bdcc44</code></a>
Merge commit from fork</li>
<li><a
href="f45b0df09d"><code>f45b0df</code></a>
Fix a misleading example for <code>ProxyManager</code> (<a
href="https://redirect.github.com/urllib3/urllib3/issues/4970">#4970</a>)</li>
<li><a
href="577193ca02"><code>577193c</code></a>
Switch to nightly PyPy3.11 in CI for now (<a
href="https://redirect.github.com/urllib3/urllib3/issues/4984">#4984</a>)</li>
<li><a
href="e90af45bb0"><code>e90af45</code></a>
Avoid infinite loop in <code>HTTPResponse.read_chunked</code> when
<code>amt=0</code> (<a
href="https://redirect.github.com/urllib3/urllib3/issues/4974">#4974</a>)</li>
<li><a
href="67ed74fdae"><code>67ed74f</code></a>
Bump dev dependencies (<a
href="https://redirect.github.com/urllib3/urllib3/issues/4972">#4972</a>)</li>
<li><a
href="3abd481097"><code>3abd481</code></a>
Upgrade mypy to version 1.20.2 (<a
href="https://redirect.github.com/urllib3/urllib3/issues/4978">#4978</a>)</li>
<li><a
href="2b8725dfca"><code>2b8725d</code></a>
Drop support for EOL PyPy3.10 (<a
href="https://redirect.github.com/urllib3/urllib3/issues/4979">#4979</a>)</li>
<li><a
href="2944b2a0a6"><code>2944b2a</code></a>
Upgrade <code>setup-chrome</code> and <code>setup-firefox</code> to fix
warnings (<a
href="https://redirect.github.com/urllib3/urllib3/issues/4973">#4973</a>)</li>
<li>Additional commits viewable in <a
href="https://github.com/urllib3/urllib3/compare/2.6.3...2.7.0">compare
view</a></li>
</ul>
</details>
<br />
[](https://docs.github.com/en/github/managing-security-vulnerabilities/about-dependabot-security-updates#about-compatibility-scores)
Dependabot will resolve any conflicts with this PR as long as you don't
alter it yourself. You can also trigger a rebase manually by commenting
`@dependabot rebase`.
[//]: # (dependabot-automerge-start)
[//]: # (dependabot-automerge-end)
---
<details>
<summary>Dependabot commands and options</summary>
<br />
You can trigger Dependabot actions by commenting on this PR:
- `@dependabot rebase` will rebase this PR
- `@dependabot recreate` will recreate this PR, overwriting any edits
that have been made to it
- `@dependabot show <dependency name> ignore conditions` will show all
of the ignore conditions of the specified dependency
- `@dependabot ignore this major version` will close this PR and stop
Dependabot creating any more for this major version (unless you reopen
the PR or upgrade to it yourself)
- `@dependabot ignore this minor version` will close this PR and stop
Dependabot creating any more for this minor version (unless you reopen
the PR or upgrade to it yourself)
- `@dependabot ignore this dependency` will close this PR and stop
Dependabot creating any more for this dependency (unless you reopen the
PR or upgrade to it yourself)
You can disable automated security fix PRs for this repo from the
[Security Alerts
page](https://github.com/infiniflow/ragflow/network/alerts).
</details>
Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
### What problem does this PR solve?
Closes#14674.
This PR improves RAPTOR configuration and tree construction while
preserving the existing RAPTOR behavior as the default.
RAPTOR currently builds summary layers with the original UMAP + GMM
clustering path. This PR keeps that default path, and adds:
- A hidden backend tree-builder option:
- `tree_builder="raptor"`: default, existing RAPTOR behavior.
- `tree_builder="psi"`: rank-aware Psi-style tree builder using original
embedding-space cosine ranking.
- A user-facing clustering method option for the default RAPTOR builder:
- `clustering_method="gmm"`: existing default.
- `clustering_method="ahc"`: agglomerative hierarchical clustering path.
- A RAPTOR UI setting for `Clustering method` and `Max cluster`.
### What changed
#### Backend
- Added `tree_builder` support for RAPTOR/Psi.
- Added `clustering_method` support for GMM/AHC.
- Kept existing RAPTOR + GMM as the default.
- Added Psi tree building from original-space cosine similarity.
- Added bucketed Psi building controls for large inputs:
- `raptor.ext.psi_exact_max_leaves`
- `raptor.ext.psi_bucket_size`
- Added method-aware RAPTOR summary metadata using existing
`extra.raptor_method`.
- Avoided adding a dedicated DB schema field for experimental method
tracking.
- Added cleanup/migration logic to avoid mixing stale RAPTOR summary
trees.
- Added defensive checks for Psi tree construction and summary failures.
#### Frontend/UI
- Added `Clustering method` in RAPTOR settings with `GMM` and `AHC`.
- Added/kept `Max cluster` in RAPTOR settings.
- Enlarged max cluster UI limit to `1024`, matching backend validation.
- Kept AHC editable even when a RAPTOR task has already finished.
- Fixed the UI save payload so `clustering_method` and `tree_builder`
are serialized through `parser_config.raptor.ext`, avoiding backend
validation errors for extra top-level RAPTOR fields.
Example saved RAPTOR config:
```json
{
"raptor": {
"max_cluster": 317,
"ext": {
"clustering_method": "ahc",
"tree_builder": "raptor"
}
}
}
Co-authored-by: CaptainTimon <CaptainTimon@users.noreply.github.com>
## Summary
- Add GET method handler to `/api/v1/dify/retrieval` endpoint for Dify
external knowledge base connectivity verification
- GET requests return a simple success response; POST requests retain
existing retrieval logic unchanged
## Problem
When Dify integrates with RAGFlow as an external knowledge base, it
sends periodic GET requests to the retrieval endpoint for
health/connectivity checks. The endpoint only accepted POST, causing
werkzeug to return `405 Method Not Allowed`. After several successful
POST retrievals, the failing GET health checks trigger Dify's circuit
breaker, causing all subsequent requests to fail.
Traceback from the issue:
```
werkzeug.exceptions.MethodNotAllowed: 405 Method Not Allowed: The method is not allowed for the requested URL.
```
## Changes
- `api/apps/sdk/dify_retrieval.py`: Added a separate GET route handler
(`retrieval_health_check`) that returns `get_json_result(data=True)`
## Test plan
- [ ] Verify `GET /api/v1/dify/retrieval` returns `{"code": 0,
"message": "success", "data": true}`
- [ ] Verify `POST /api/v1/dify/retrieval` with valid API key and body
still works as before
- [ ] Verify Dify external knowledge base integration no longer returns
405 errors
Closes#13788🤖 Generated with [Claude Code](https://claude.com/claude-code)
---------
Co-authored-by: Asksksn <Asksksn@noreply.gitcode.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: Kevin Hu <kevinhu.sh@gmail.com>
### What problem does this PR solve?
1. Add region check in zhipu-ai embed method
2. Fix retrieval test
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
Signed-off-by: Jin Hai <haijin.chn@gmail.com>
### What problem does this PR solve?
This PR completes the Cohere provider integration (upgrading to the new
Cohere V2 API) and enhances the Fish Audio provider in RAGFlow.
**The following functionalities are now supported:**
**Cohere:**
- [x] Chat / Think Chat / Stream Chat / Stream Think Chat
- [x] Embedding
- [x] Rerank
- [x] Model listing
- [x] Provider connection checking
- [ ] Balance
**Fish Audio:**
- [x] Model listing (`ListModels`)
- [x] Balance (`Balance`)
-----
**Verified examples from the CLI:**
```plaintext
# Cohere
RAGFlow(user)> think chat with 'command-a-reasoning-08-2025@test3@cohere' message 'jumperwho'
Thinking: Okay, the user wrote "jumperwho". Let me try to figure out what they might be asking. First, I'll check if it's a misspelling. "Jumper" ...... Hmm. Since the query is unclear, the best approach is to ask the user to provide more context or correct any possible typos.
Answer: It seems there might be a typo or missing context in your query "jumperwho." Could you clarify what you're referring to? For example:
- Are you asking about a **jumper** (a type of sweater, a person who jumps, or a component in electronics)?
- Is this related to a specific context, like a movie (e.g., the 2008 film *Jumper*) or a game?
- Did you mean to ask about a person ("who") associated with jumping (e.g., a parachutist)?
Let me know so I can provide a helpful response! 😊
Time: 6.710331
RAGFlow(user)> stream think chat with 'command-a-reasoning-08-2025@test3@cohere' message 'jumperwho'
Thinking: , the user mentioned "jumperwho". Let me try to figure out what they're referring to. First, I'll check if it's a misspelling. "Jumper" could be a typo for "jumper" or maybe a username. Alternatively, it might be a combination of words like "jumper who",....... the best approach is to inform the user that I don't recognize the term and ask if they can provide more context or clarify what they mean by "jumperwho". That way, I can assist them better once I have more information.
Answer: seems "jumperwho" isn't a widely recognized term, proper noun, or acronym in common usage. Could you provide more context or clarify what you mean by "jumperwho"? This will help me understand your question or request better!
Time: 4.513596
RAGFlow(user)> embed text 'walkerwhat' 'jumperwho' with 'embed-v4.0@test3@cohere' dimension 16;
+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+-------+
| embedding | index |
+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+-------+
| [-0.016643638 -0.001957038 0.0055713872 0.009027058 0.05275187 -0.024542313 -0.044006906 0.024119169 0.0014192933 0.006558722 0.0019129605 -0.021016119 -0.026516981 -0.017489925 0.021298215 0.017772019 0.04569948 0.008886009 0.012059584 -0.0014721862 0.... | 0 |
| [0.018778935 -0.0063459855 -0.0006839742 0.0046623563 0.0067668925 -0.018001877 -0.03963003 0.035744734 -0.014246088 -0.0020721585 -0.006313608 0.025124922 -0.010749322 0.01217393 -0.010231283 -0.025254432 0.021498645 -0.028880708 0.019167464 -0.0058279... | 1 |
+------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+-------+
RAGFlow(user)> rerank query 'what is rag' document 'rag is retrieval augment generation' 'rag need llm' 'famous rag project includes ragflow' with 'rerank-v4.0-pro@test@cohere' top 3;
+-------+-----------------+
| index | relevance_score |
+-------+-----------------+
| 0 | 0.91744334 |
| 1 | 0.7458429 |
| 2 | 0.68729424 |
+-------+-----------------+
RAGFlow(user)> list supported models from 'cohere' 'test'
+-------------------------------------+
| model_name |
+-------------------------------------+
| c4ai-aya-expanse-32b |
| c4ai-aya-vision-32b |
| cohere-transcribe-03-2026 |
| command-a-03-2025 |
| command-a-reasoning-08-2025 |
| command-a-translate-08-2025 |
| command-a-vision-07-2025 |
| command-r-08-2024 |
| command-r-plus-08-2024 |
| command-r7b-12-2024 |
| command-r7b-arabic-02-2025 |
| embed-english-light-v3.0 |
| embed-english-light-v3.0-image |
| embed-english-v3.0 |
| embed-english-v3.0-image |
| embed-multilingual-light-v3.0 |
| embed-multilingual-light-v3.0-image |
| embed-multilingual-v3.0 |
| embed-multilingual-v3.0-image |
| embed-v4.0 |
+-------------------------------------+
RAGFlow(user)> check instance 'test' from 'cohere'
SUCCESS
# FishAudio
RAGFlow(user)> list supported models from 'fishaudio' 'test'
+----------------------------------------+
| model_name |
+----------------------------------------+
| Valentino Narración Biblica Fer |
| Super Smash Bros. 4/Ultimate Announcer |
| Farid Dieck |
| عصام الشوالي |
| ALEX_CHIKNA |
| Energetic Male |
| voz de locutor k |
| يي |
| ELITE |
| Mortal Kombat |
+----------------------------------------+
RAGFlow(user)> show balance from 'fishaudio' 'test'
+----------------------------------+-----------------------------+--------+-----------------+------------------+-----------------------------+----------------------------------+
| _id | created_at | credit | has_free_credit | has_phone_sha256 | updated_at | user_id |
+----------------------------------+-----------------------------+--------+-----------------+------------------+-----------------------------+----------------------------------+
| 82ffec12cf984d88a30ec504d7909812 | 2026-05-09T07:52:16.119000Z | 0 | | false | 2026-05-09T07:52:16.119000Z | 2578ab1126804d6eaa630552400d7ff3 |
+----------------------------------+-----------------------------+--------+-----------------+------------------+-----------------------------+----------------------------------+
```
### Type of change
- [x] New Feature (non-breaking change which adds functionality)
- [x] Refactoring
## Summary
- Replaces the `"no such method"` stub on `NvidiaModel.Rerank`
(`internal/entity/models/nvidia.go`) with a real implementation against
NVIDIA NIM's `/ranking` endpoint.
- Mirrors the existing Python `NvidiaRerank` class at
`rag/llm/rerank_model.py:149-190` for behavior parity: same
`passages`/`query.text`/`logit` payload shape; `top_n` set to
`len(documents)` so every input gets a score returned in original order
(the issue body's spec omitted `top_n`, which would cause silent data
loss).
- Adds the `"rerank": "ranking"` URL suffix and two NIM rerank model
entries (`nvidia/nv-rerankqa-mistral-4b-v3`,
`nvidia/llama-3.2-nv-rerankqa-1b-v2`) to `conf/models/nvidia.json` so
the picker exposes them.
- Follows the same shape as the recently merged Aliyun (#14676), Gitee
(#14656), and ZhipuAI (#14608) Rerank implementations: lowercase
per-driver request/response types, conversion to the project-wide
`RerankResponse{Data: []RerankResult}`, per-call `context.WithTimeout`
of 30s.
Closes#14720
## Test plan
- [x] `gofmt -l internal/entity/models/nvidia.go` — clean
- [x] `go vet ./internal/entity/models/...` — no new errors introduced
(the two pre-existing vet errors in `baidu.go:642` and
`openrouter.go:566` are unrelated to this PR)
- [x] `go build ./internal/entity/models/...` — succeeds
- [x] `python3 -c "import json;
json.load(open('conf/models/nvidia.json'))"` — JSON valid
- [ ] Live smoke test against NVIDIA NIM with a real API key (requires
reviewer with NIM credentials)
## Notes for reviewers
- The issue body suggested omitting `top_n`. The Python reference
includes it (`top_n: len(texts)`), and without it NVIDIA returns only
the default top-K rankings rather than scores for every input. This PR
follows the Python.
- The URL host is `integrate.api.nvidia.com` (kept consistent with the
existing chat/embeddings BaseURL in `nvidia.go`), not the legacy
`ai.api.nvidia.com` host the Python uses. NIM's unified endpoint accepts
the model names as-is, so no per-model URL transform is needed.
### What problem does this PR solve?
As the title suggests.
### Type of change
- [x] Documentation Update
Signed-off-by: Jin Hai <haijin.chn@gmail.com>
### What problem does this PR solve?
Fixes#14570. On OpenSearch backends (`DOC_ENGINE=opensearch`) every
document-metadata write failed with `'OSConnection' object has no
attribute 'create_doc_meta_idx'`, so both `PATCH
/api/v1/datasets/{ds}/documents/{doc}` with `meta_fields` and `POST
/api/v1/datasets/{ds}/metadata/update` were unusable while every other
document operation (retrieval, parsing, name update, chunk management)
worked correctly on the same OpenSearch cluster.
The bug runs deeper than the missing method name in the error message
suggests. `DocMetadataService` also reached into
`settings.docStoreConn.es.*` directly for the index refresh, the
scripted partial update, and the count call, which means that even after
adding `create_doc_meta_idx` to `OSConnection` the very next call in the
same metadata flow would still raise `AttributeError` because
`OSConnection` exposes `self.os` rather than `self.es`. Fixing only the
reported symptom would have moved the failure one line down without
restoring the feature.
This PR adds a uniform document-metadata dispatch surface to both
connection classes so they present the same abstract API, and routes the
service layer through that surface via `getattr` guards instead of
poking at backend-specific attributes. The four new methods on
`OSConnection` and `ESConnectionBase` are `create_doc_meta_idx`,
`refresh_idx`, `count_idx`, and `replace_meta_fields`.
`OSConnection.create_doc_meta_idx` reuses the existing
`conf/doc_meta_es_mapping.json` schema in the OpenSearch `body=` form
because OpenSearch and Elasticsearch share the same index-creation
payload, and `replace_meta_fields` emits a full scripted assignment
(`ctx._source.meta_fields = params.meta_fields`) on both backends so
removed keys actually disappear instead of being preserved by deep-merge
semantics.
The `getattr`-guarded dispatch in `DocMetadataService` keeps the
existing fall-through paths intact for Infinity and OceanBase, which
continue to rely on their search-based count fallback and on the
delete-then-insert metadata replacement they used before, so this change
is strictly additive for those two backends.
Verification: `pytest
test/unit_test/rag/utils/test_opensearch_doc_meta.py` runs 16 new unit
tests that pass locally and pin the `OSConnection` dispatch surface, the
`create_doc_meta_idx` short-circuit when the index already exists, the
mapping-file payload routing, the `IndicesClient.create` failure path,
the `refresh_idx` and `count_idx` success and error sentinels, and the
full-assignment script emitted by `replace_meta_fields`. The test module
stubs `common.settings` and `rag.nlp` at import time so the suite runs
without the heavy backend SDKs that the rest of the repository pulls in
transitively.
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
---------
Co-authored-by: tmimmanuel <tmimmanuel@users.noreply.github.com>
### What problem does this PR solve?
fix some comments to improve readability
### Type of change
- [x] Documentation Update
---------
Signed-off-by: box4wangjing <box4wangjing@outlook.com>
Fixes#13851
## Problem
`OCR.detect()` in `deepdoc/vision/ocr.py` returns `None, None,
time_dict` (a truthy 3-tuple) when the text detector fails or receives a
`None` image. However, the caller in `pdf_parser.py:__ocr()` checks:
```python
bxs = self.ocr.detect(np.array(img), device_id)
if not bxs: # False! (None, None, time_dict) is a non-empty tuple → truthy
self.boxes.append([])
return
bxs = [(line[0], line[1][0]) for line in bxs] # iterates (None, None, time_dict)
# line = None → None[0] → TypeError: 'NoneType' object is not subscriptable
```
This causes the `NoneType object is not subscriptable` error that
appears after "OCR started" in the chunking pipeline when using PDF +
General parser.
## Solution
Simplified `OCR.detect()` to return `None` (falsy) instead of `None,
None, time_dict` on failure. The `time_dict` was unused by the only
caller of this method. The early-return guard `if not bxs:` in
`pdf_parser.py` then correctly catches it.
## Testing
- The method's only caller (`pdf_parser.py:__ocr`) already has a `if not
bxs:` guard that handles the `None` return correctly.
- No other callers of `OCR.detect()` exist in the codebase.
<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->
## Summary by CodeRabbit
* **Refactor**
* Modified OCR detection function return behavior to streamline output.
The function now returns detection results only, without timing
metadata. Error cases now return `None` instead of empty tuple values.
<!-- end of auto-generated comment: release notes by coderabbit.ai -->
This PR fixes a UI issue where the .docx document preview was displayed
incompletely when clicking on a citation/reference link during a
knowledge base conversation.
### What problem does this PR solve?
The Issue:
In the chat interface, when a user clicks the source citation at the end
of an answer, the DocPreviewer opens. However, for .docx files, if the
content exceeded the window height, it was truncated and unscrollable,
preventing users from reading the full referenced text.
Changes:
web/src/components/document-preview/doc-preview.tsx: Added the
overflow-auto Tailwind class to the DocPreviewer root container to
ensure scrollbars appear automatically when content overflows.
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
Co-authored-by: nie.weiyang <nie.weiyang@embedway.com>
### What problem does this PR solve?
The document parse status was set to DONE before the document chunks
were actually retrievable from Elasticsearch/Opensearch because it did
not wait for the index refresh. This meant that it was possible that the
document parse status returned by the API was DONE but when trying to
retrieve chunks there were none. Since the index refreshes every 1
second this was quite likely to happen when wait for document parsing by
polling with a short interval and then immediately trying to retrieve
chunks once the status was DONE.
I fixed this bug and added a test case that would have caught it.
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
### What problem does this PR solve?
Added a private helper _visibility_and_status_filter(joined_tenant_ids,
user_id) that returns the Peewee condition: visible to user (team or
own) and status is VALID.
### Type of change
- [x] Refactoring
---------
Co-authored-by: Serobabov Aleksandr <40SerobabovAS@region.cbr.ru>
Co-authored-by: Yingfeng <yingfeng.zhang@gmail.com>
### What problem does this PR solve?
Addresses event-loop blocking under high concurrency reported in #13825.
When multiple requests hit the API simultaneously, synchronous DB/Redis
calls block the async event loop, preventing Quart from handling other
requests and causing cascading 502/504 timeouts.
This PR wraps all remaining blocking DB/Redis calls in `canvas_app.py`,
`chat_api.py`, `session.py`, and `canvas_service.py` with `await
thread_pool_exec()`
- Offload all synchronous `Service.*`, `REDIS_CONN.*`, and
`APIToken.query` calls to the thread pool
- Convert sync endpoint handlers (`list_chats`, `get_chat`, `templates`,
`sessions`, etc.) to `async def`
- Convert sync helper functions (`_ensure_owned_chat`,
`_validate_llm_id`, `_validate_dataset_ids`, etc.) to async - no
duplicate sync/async pairs
- Wrap `CanvasReplicaService` Redis IO calls (`bootstrap`,
`replace_for_set`, `commit_after_run`)
- Use `asyncio.gather()` for concurrent file uploads and chat response
building
**Note:** This fixes the code-level event-loop blocking, which is a
prerequisite for handling concurrent requests. For the full "30
concurrent requests without 502/504" goal described in the issue, users
should also tune deployment config:
- `WS=4` or higher (HTTP worker processes, default 1)
- `MAX_CONCURRENT_CHATS=50` (default 10)
- `SANDBOX_EXECUTOR_MANAGER_POOL_SIZE` for workflow-heavy workloads
### Performance verification
Reviewer asked for a before-vs-after comparison
([comment](https://github.com/infiniflow/ragflow/pull/13941#issuecomment-4393667231)).
I built a self-contained microbenchmark that reproduces the exact
failure mode this PR targets: an async handler that performs blocking
DB/Redis-style calls (50 ms each, 3 per request, 30 concurrent requests)
is run twice — once with the pre-PR pattern (sync call directly inside
the async handler) and once with the post-PR pattern (`await
thread_pool_exec(...)`). The benchmark imports nothing from RAGFlow
except `thread_pool_exec` itself, so it is hermetic and reproducible
(`THREAD_POOL_MAX_WORKERS=128`, Python 3.13.12).
**Throughput — wall-clock for 30 concurrent requests (lower is better)**
| flavour | wall(s) | p50(s) | p95(s) | max(s) |
|---|---:|---:|---:|---:|
| before | 4.986 | 0.158 | 0.207 | 0.269 |
| after | 0.248 | 0.181 | 0.230 | 0.231 |
The pre-PR handler serializes the entire load on the event-loop thread,
so 30 × 3 × 50 ms ≈ 4.5 s shows up as the wall time. The post-PR handler
parallelizes the blocking work across the thread pool and finishes the
same load in 248 ms — a **~20× speedup** on this workload.
**Event-loop responsiveness — latency of an unrelated probe coroutine
while the 30 slow requests are running (lower is better)**
| flavour | samples | probe p50 (ms) | probe p95 (ms) | probe max (ms) |
|---|---:|---:|---:|---:|
| before | 1 | 5442.26 | 5442.26 | 5442.26 |
| after | 28 | 0.88 | 11.53 | 98.02 |
This is the metric that maps directly to "the API still answers other
requests while one is busy". A 5 ms-interval probe was scheduled while
the 30 slow handlers ran. With the pre-PR code the event loop was frozen
for the entire duration of the blocking work, so only one probe sample
was ever picked up and it waited **5,442 ms**. After the PR, 28 probe
samples landed with **p50 0.88 ms / p95 11.53 ms**, meaning unrelated
requests are no longer starved by the slow ones. That is the regression
mode behind the cascading 502/504s reported in #13825.
<details>
<summary>Raw benchmark output</summary>
```
config: 30 concurrent requests, 3 blocking calls of 50ms each per request, THREAD_POOL_MAX_WORKERS=128
=== Throughput (lower wall is better) ===
flavour wall(s) p50(s) p95(s) max(s)
before 4.986 0.158 0.207 0.269
after 0.248 0.181 0.230 0.231
=== Event-loop responsiveness (lower probe latency is better) ===
flavour samples probe p50(ms) probe p95(ms) probe max(ms)
before 1 5442.26 5442.26 5442.26
after 28 0.88 11.53 98.02
```
</details>
The benchmark script is included as a comment on the PR for
reproducibility.
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
- [x] Performance Improvement
Closes [#13825](https://github.com/infiniflow/ragflow/issues/13825)
---------
Co-authored-by: tmimmanuel <tmimmanuel@users.noreply.github.com>
Co-authored-by: Kevin Hu <kevinhu.sh@gmail.com>
### What problem does this PR solve?
- Moved if not all([email, new_pwd, new_pwd2]) guard to the top, before
any decryption that could crash on None value
- Removed the redundant REDIS_CONN.get() call — one call is sufficient
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
- [x] Refactoring
### What problem does this PR solve?
Provide embedding index according to the input text
### Type of change
- [x] Refactoring
---------
Signed-off-by: Jin Hai <haijin.chn@gmail.com>
## Summary
- Wrap 2 `ThreadPoolExecutor` instances in `file_service.py` with `with`
statement
- Ensures threads are properly shut down after all futures complete
## Problem
`parse_docs()` (line 532) and the file processing method (line 694)
create `ThreadPoolExecutor` instances that are never shut down. In a
long-running server process, this leaks thread resources on every
invocation — threads remain alive consuming memory even after all
submitted work is complete.
## Fix
Replace bare `ThreadPoolExecutor()` with `with ThreadPoolExecutor() as
exe:` context manager, which calls `executor.shutdown(wait=True)` on
exit.
## Test plan
- [x] Verified both call sites use `with` statement after fix
- [x] No remaining bare `ThreadPoolExecutor` in `file_service.py`
- [x] `document_service.py:1066` is a module-level executor (different
pattern, not changed in this PR)
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-authored-by: Kevin Hu <kevinhu.sh@gmail.com>
### What problem does this PR solve?
issue: https://github.com/infiniflow/ragflow/issues/14748
change: dataset search rerank id type
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
### Related issues
Closes#14744
### What problem does this PR solve?
The Memory REST endpoint `POST /api/v1/messages` previously persisted
whatever `user_id` the client sent in the JSON body. Memory rows were
therefore attributed to an arbitrary string, even when the caller
authenticated as a normal workspace user via JWT (browser/session-style
bearer token decoded into an access token). That broke attribution and
audit semantics for shared memories (team visibility): any authorized
writer could spoof another subject id.
The Python SDK already sends an optional `user_id` for integrations
using **API keys** (`APIToken`) to tag an external subject distinct from
the tenant owner user.
### Solution
- Record **`g.auth_via_api_token`** in `_load_user`
(`api/apps/__init__.py`): set `True` only when authentication resolves
via `APIToken`, otherwise `False` after JWT-based login succeeds.
- In **`POST /messages`** (`memory_api.add_message`): if the request was
authenticated with an API key, keep accepting optional `user_id` from
the body (default empty string). For JWT-authenticated users, **always**
set stored `user_id` to **`current_user.id`** and ignore the client
field.
- Guard reads of `g` with **`RuntimeError`** handling so isolated
imports or tests without a Quart application context do not fail when
resolving `user_id`.
- Document on **`RAGFlow.add_message`** that `user_id` is only
meaningful for API-key authentication.
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
- [ ] New Feature (non-breaking change which adds functionality)
- [ ] Documentation Update
- [ ] Refactoring
- [ ] Performance Improvement
- [ ] Other (please describe):
### Testing
- `python -m py_compile` on modified modules (`api/apps/__init__.py`,
`api/apps/restful_apis/memory_api.py`).
- Recommended: run web/SDK memory message tests (`test_add_message`,
`test_message_routes_unit`) against a full environment with `quart` and
configured services.
### Notes for reviewers
- Behavior change **only** for callers using JWT-style authorization on
`POST /messages`; API-key callers keep prior optional `user_id`
semantics.
Co-authored-by: jony376 <jony376@gmail.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
## What problem does this PR solve?
The Dify-compatible `/dify/retrieval` endpoint recently gained stricter
parsing and validation for its request payload, including:
- Normalized `retrieval_setting.top_k` and
`retrieval_setting.score_threshold` types.
- Clear separation between malformed arguments vs missing required
fields.
Previously, there was no unit test explicitly guarding the exact error
code and message contract for these cases.
## What does this PR change?
- **Add guard-style unit test** in `test_dify_retrieval_routes_unit.py`:
- `test_retrieval_argument_error_messages`:
- Sends a request with malformed numeric options:
- `retrieval_setting = {"top_k": "not-int", "score_threshold":
"not-float"}`
- Asserts `code == RetCode.ARGUMENT_ERROR` and message contains
`"invalid or malformed arguments:"`.
- Sends a request with required fields missing:
- Empty payload (`{}`)
- Asserts `code == RetCode.ARGUMENT_ERROR` and message contains
`"required arguments are missing:"`.
This test encodes the intended behavior of the Dify retrieval API so
future refactors cannot silently regress error handling.
## Type of change
- [x] Tests (add coverage and guardrails for existing behavior)
Co-authored-by: Kevin Hu <kevinhu.sh@gmail.com>
pending_cell_images should be scoped by sheet
### What problem does this PR solve?
_Briefly describe what this PR aims to solve. Include background context
that will help reviewers understand the purpose of the PR._
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
### What problem does this PR solve?
GraphRAG feature - Part 1 - add spacy to extract entity and relation
<img width="1621" height="1288" alt="image"
src="https://github.com/user-attachments/assets/aadeddad-94da-46c6-adad-9c3784181f61"
/>
### Type of change
- [x] New Feature (non-breaking change which adds functionality)
## Summary
- Wrap the `ThreadPoolExecutor` instances in `FileService.parse_docs`
and `FileService.get_files` with `with ... as exe:` blocks for
deterministic cleanup
- Replace the `concurrent.futures.ThreadPoolExecutor` in
`do_handle_task` with `asyncio.create_task(asyncio.to_thread(build_TOC,
...))`, preserving the existing parallelism with chunk insertion while
leveraging the surrounding async context
- Drop the now-unused `import concurrent` and the
`executor.shutdown(wait=False)` call in the `finally` block
Closes#14622.
No behavioral change, no public API change. Net diff: ~19 insertions /
25 deletions across two files.
## Test plan
- [ ] `uv run ruff check api/db/services/file_service.py
rag/svr/task_executor.py` passes
- [ ] Upload a multi-file batch through the chat/file endpoint and
confirm `FileService.parse_docs` still returns combined parsed text
- [ ] Trigger `FileService.get_files` via the chat reference flow with a
mix of image and non-image files; verify both `raw=True` and `raw=False`
paths return correctly
- [ ] Run a `naive`-parser document task with `toc_extraction: true` and
confirm the TOC chunk is generated and inserted exactly as before
- [ ] Run a `naive`-parser document task with `toc_extraction: false`
and confirm the path with `toc_thread = None` is unaffected
- [ ] Cancel a running task to exercise the `finally` block and confirm
cleanup still works without the executor shutdown call
---------
Co-authored-by: web-dev0521 <jasonpette1783@gmail.com>
Co-authored-by: Wang Qi <wangq8@outlook.com>
### What problem does this PR solve?
The OpenRouter `Encode` method silently swallowed malformed responses.
If a `data[]` item from the API was missing a field (`index`,
`embedding`, or unexpected shape), the loop did `continue` instead of
returning an error — leaving `nil` entries in the result slice. Callers
got back partial results with no indication anything went wrong, which
then crashes downstream consumers when they try to use a `nil` vector.
There were three concrete gaps:
- No count-mismatch check between `data` length and input texts (only
checked for empty)
- No duplicate-index detection (a duplicate would silently overwrite)
- Parse failures on individual items returned partial slices instead of
erroring
This PR replaces `map[string]interface{}` parsing with a typed
`openrouterEmbeddingResponse` struct and applies the same 3-layer
validation used in the other drivers (count mismatch → out-of-range
index → duplicate index), so any malformed response produces a clear
error instead of corrupted data.
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
### What problem does this PR solve?
The LM Studio Go driver shipped with a stub \`Encode\` method that
returned \`no such method\`, even though LM Studio is one of the most
common local LLM runners on macOS and Windows and exposes an
OpenAI-compatible embeddings endpoint at \`/v1/embeddings\`.
LM Studio users routinely load local embedding models such as
\`nomic-ai/nomic-embed-text-v1.5\`,
\`mixedbread-ai/mxbai-embed-large-v1\`, or \`BAAI/bge-m3\`. They run on
the same \`/v1\` namespace as chat. The existing \`ListModels\` already
discovers them, but because \`Encode\` was a stub, a tenant who picked
one of these models in the Go layer could not actually run an embedding
call.
This finishes the local-LLM trio: Ollama Encode (#14664) and vLLM Encode
(#14688) are already in flight, both using the
same OpenAI-compatible \`/embeddings\` shape.
### What this PR includes
- \`conf/models/lmstudio.json\`: add \`\"embedding\": \"embeddings\"\`
under \`url_suffix\` so the driver can build the URL from config.
- \`internal/entity/models/lmstudio.go\`: replace the \`Encode\` stub
with a real implementation. Adds a small local response type that
matches the OpenAI-compatible shape.
No factory change. No interface change.
### How the driver works
- Validate the model name. The API key is optional for local LM Studio,
so the Authorization header is only set when both \`apiConfig\` and
\`ApiKey\` are non-nil and non-empty, the same pattern the recently
merged CheckConnection PR (#14614) uses.
- Resolve the region with a default fallback. Return a clear "missing
base URL" error when the user has not configured
the local access address yet.
- Use a per-call \`context.WithTimeout(30s)\` and
\`http.NewRequestWithContext\`, the same pattern the merged
Aliyun Encode (#14647) and the in-flight Ollama Encode (#14664) and vLLM
Encode (#14688) use.
- Send \`{model, input: [texts]}\` in one request.
- Parse \`data[*].embedding\` and copy each slice into a \`[][]float64\`
indexed by \`data[*].index\`, so the output
order matches the input order.
- Handle both \`float64\` and \`float32\` element types.
- Empty input returns \`[][]float64{}\` with no HTTP call.
- Length mismatch between input and result, out-of-range index, and any
missing slot all return clear errors instead
of silent zero vectors.
### Type of change
- [x] New Feature (non-breaking change which adds functionality)
### How was this tested?
- \`go build ./internal/entity/models/...\` in a clean go 1.25 image
returns exit 0.
- The full method set on \`LmStudioModel\` still matches the
\`ModelDriver\` interface.
- Pattern parity with the merged Aliyun Encode (#14647), the in-flight
Ollama Encode (#14664) and vLLM Encode (#14688), and the existing
SiliconFlow Encode.
Closes#14693
### What problem does this PR solve?
The SiliconFlow `Encode` method sent one HTTP request per text, which is
wasteful and slow when indexing many documents (e.g., 100 docs = 100
round-trips).
SiliconFlow's `/v1/embeddings` is OpenAI-compatible and accepts an array
of strings in `input` (officially documented at
https://docs.siliconflow.cn/en/api-reference/embeddings/create-embeddings,
with a documented max array size of 32). This PR batches the requests up
to that limit, reducing 100 docs to ~4 round-trips, and replaces
`map[string]interface{}` parsing with a typed struct using the same
3-layer validation (count mismatch, out-of-range index, duplicate index)
used in the other drivers.
### Type of change
- [x] Performance Improvement
### What problem does this PR solve?
The NVIDIA Go driver in `internal/entity/models/nvidia.go` shipped with
a stub `Encode`
method that returned `no such method`. `conf/models/nvidia.json` already
lists
`nvidia/llama-3.2-nemoretriever-1b-vlm-embed-v1` as an embedding model,
but the conf had
no `embedding` URL suffix, so the picker had nothing wired even if
`Encode` worked.
A tenant who wanted to use NVIDIA NIM for chat (already working) and
embeddings from a
single provider could not, even though the upstream endpoint is public
at
`https://integrate.api.nvidia.com/v1/embeddings` and uses an
OpenAI-compatible request
body extended with the NVIDIA-specific `input_type` and `truncate`
fields. Several other
Go drivers already implement `Encode` (siliconflow, zhipu-ai, aliyun),
so the interface
and the pattern are well-established.
This PR fills the gap.
### What this PR includes
* `conf/models/nvidia.json`: declare the `embedding` URL suffix
alongside the existing
`chat` and `models` entries. The embedding model entry was already
present, so no
model addition is needed.
* `internal/entity/models/nvidia.go`: replace the `Encode` stub with a
real
implementation. Adds a small local response type that matches the
OpenAI-compatible
shape NVIDIA NIM returns.
No factory change. No interface change.
### How the driver works
* Validates `apiConfig` and the API key, validates the model name,
resolves the region
with a default fallback (matching the pattern the merged `ListModels`
and
`CheckConnection` paths in this driver already use), and builds the URL
from
`BaseURL[region] + URLSuffix.Embedding`.
* Sends all input texts in one request as the `input` array, with the
NVIDIA-specific `input_type: "query"`, `encoding_format: "float"`, and
`truncate: "END"`
fields, mirroring the Python `NvidiaEmbed` reference.
* Parses `data[*].embedding` and copies each slice into `[][]float64`
indexed by
`data[*].index` so the output order matches the input order even if the
API returns
items in a different order.
* Handles both `float64` and `float32` element types.
* Empty input returns `[][]float64{}` with no HTTP call.
* Non-200 responses propagate the upstream status line and body.
* A final pass checks every input slot got a vector and returns a clear
error if any
slot is still nil.
* Per-call 30s context deadline so a slow call cannot block forever.
### Type of change
- [x] New Feature (non-breaking change which adds functionality)
### How was this tested?
* `go build ./internal/entity/models/...` returns exit 0.
* `go vet ./internal/entity/models/...` is clean.
* `gofmt -l internal/entity/models/nvidia.go` is clean.
* The full method set on `NvidiaModel` still matches the `ModelDriver`
interface.
* Pattern parity with the just-merged Aliyun `Encode` (#14647).
Closes#14699
### What problem does this PR solve?
The Ollama Go driver shipped with a stub \`Encode\` method that returned
\`no such method\`, even though Ollama is one of the most common local
LLM runners and exposes an OpenAI-compatible embeddings endpoint at
\`/v1/embeddings\`.
Ollama users routinely run local embedding models such as
\`nomic-embed-text\`, \`mxbai-embed-large\`, or \`bge-m3\`.
Pulled with \`ollama pull <model>\` and served on the same \`/v1\`
namespace as chat. The existing \`ListModels\` already
discovers them, but because \`Encode\` was a stub, a tenant who picked
one of these models in the Go layer could not
actually run an embedding call.
### What this PR includes
- \`conf/models/ollama.json\`: add \`\"embedding\": \"embeddings\"\`
under \`url_suffix\` so the
driver can build the URL from config.
- \`internal/entity/models/ollama.go\`: replace the \`Encode\` stub with
a real implementation. Adds a small local response
type that matches the OpenAI-compatible shape.
No factory change. No interface change.
### How the driver works
- Validate the model name. The API key is optional for local Ollama, so
the Authorization header is only set when both
\`apiConfig\` and \`ApiKey\` are non-nil and non-empty, the same pattern
the recently merged CheckConnection PR (#14614) uses.
- Resolve the region with a default fallback. Return a clear "missing
base URL" error when the user has not configured
the local access address yet.
- Use a per-call \`context.WithTimeout(30s)\` and
\`http.NewRequestWithContext\`, the same pattern the merged
Aliyun Encode (#14647) uses.
- Send \`{model, input: [texts]}\` in one request.
- Parse \`data[*].embedding\` and copy each slice into a \`[][]float64\`
indexed by \`data[*].index\`, so the output
order matches the input order.
- Handle both \`float64\` and \`float32\` element types.
- Empty input returns \`[][]float64{}\` with no HTTP call.
- Length mismatch between input and result, out-of-range index, and any
missing slot all return clear errors instead
of silent zero vectors.
### Type of change
- [x] New Feature (non-breaking change which adds functionality)
### How was this tested?
- \`go build ./internal/entity/models/...\` in a clean go 1.25 image
returns exit 0.
- The full method set on \`OllamaModel\` still matches the
\`ModelDriver\` interface.
- Pattern parity with the merged Aliyun Encode (#14647) and the existing
SiliconFlow Encode.
Closes#14662
## Summary
This PR fixes the `message_fit_in()` truncation bug reported in #13607.
Changes:
- fix the user-message truncation branch to reserve room for the system
prompt token budget
- guard the zero-token edge case to avoid dividing by zero in the
truncation ratio check
- add focused regression tests covering both the user-dominant
truncation path and the zero-token boundary case
## Validation
```bash
pytest -q --noconftest test/unit_test/rag/prompts/test_generator_message_fit_in.py
```
Result: `2 passed`
Closes#13607
## Summary
This PR fully addresses all CodeRabbit review feedback and enhances the
robustness of the reranking module with 100% backward compatibility.
## Key Fixes
1. Fixed JinaRerank hardcoded base_url to support subclass endpoint
overrides
2. Corrected GPUStackRerank exception handling to use proper requests
exceptions and preserve stack traces
3. Added 30s timeout to all API calls to prevent service hanging
4. Added empty input validation for all rerank providers
5. Replaced direct dict key access with .get() to eliminate KeyError
crashes
6. Fixed _normalize_rank edge case for empty arrays
7. Implemented missing functionality for Ai302Rerank
8. Standardized type hints and fixed typo issues
## Compatibility
- No breaking changes to any existing functionality
- All rerank providers work as originally intended
- Fully compatible with existing configurations and workflows
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
- [x] Refactoring
---------
Co-authored-by: Kevin Hu <kevinhu.sh@gmail.com>
### What problem does this PR solve?
The vLLM Go driver shipped with a stub \`Encode\` method that returned
\`not implemented\`, even though vLLM is one of the most common
production-grade self-hosted inference servers and exposes an
OpenAI-compatible embeddings endpoint at \`/v1/embeddings\`.
Users who self-host \`BAAI/bge-m3\`, \`Qwen3-Embedding-*\`,
\`NV-Embed-v2\`, or similar models on vLLM could not run an embedding
call through the Go layer. The existing \`ListModels\` already discovers
the loaded models, but the embedding path failed because \`Encode\` was
a stub.
### What this PR includes
- \`conf/models/vllm.json\`: add \`\"embedding\": \"embeddings\"\` under
\`url_suffix\` so the driver can build the URL from config.
- \`internal/entity/models/vllm.go\`: replace the \`Encode\` stub with a
real implementation. Adds a small local response
type that matches the OpenAI-compatible shape.
No factory change. No interface change.
### How the driver works
- Validate the model name. The API key is optional for self-hosted vLLM,
so the Authorization header is only set when both \`apiConfig\` and
\`ApiKey\` are non-nil and non-empty, the same pattern the recently
merged CheckConnection PR (#14614) uses.
- Resolve the region with a default fallback. Return a clear "missing
base URL" error when the user has not configured
the local access address yet.
- Use a per-call \`context.WithTimeout(30s)\` and
\`http.NewRequestWithContext\`, the same pattern the merged
Aliyun Encode (#14647) and in-flight Ollama Encode (#14664) use.
- Send \`{model, input: [texts]}\` in one request.
- Parse \`data[*].embedding\` and copy each slice into a \`[][]float64\`
indexed by \`data[*].index\`, so the output
order matches the input order.
- Handle both \`float64\` and \`float32\` element types.
- Empty input returns \`[][]float64{}\` with no HTTP call.
- Length mismatch between input and result, out-of-range index, and any
missing slot all return clear errors instead
of silent zero vectors.
### Type of change
- [x] New Feature (non-breaking change which adds functionality)
### How was this tested?
- \`go build ./internal/entity/models/...\` in a clean go 1.25 image
returns exit 0.
- The full method set on \`VllmModel\` still matches the \`ModelDriver\`
interface.
- Pattern parity with the merged Aliyun Encode (#14647), the in-flight
Ollama Encode (#14664), and the existing
SiliconFlow Encode.
Closes#14687
## What
Widen the keyword delimiter in `rag/svr/task_executor.py`:
both `build_chunks` (LLM `keyword_extraction` cache parsing) and
`run_dataflow` (chunk-level `keywords` ingestion) now split on
`, ， ; ； 、 \r \n` instead of only ASCII comma.
## Why
`rag/prompts/keyword_prompt.md` instructs the LLM:
> The keywords are delimited by ENGLISH COMMA.
In practice, Chinese-leaning models (Qwen / Tongyi-Qianwen, GLM,
etc.) frequently ignore this instruction when the source content is
Chinese and emit Chinese commas (`，`) instead. Result:
`cached.split(",")` sees the full LLM output as a *single* keyword.
Repro: `auto_keywords>=4` + Chinese docs + `qwen-plus@Tongyi-Qianwen`.
We observed entries in `important_kwd` like
`"功能介绍，配置说明，参数详解，问题排查"` — one bucket instead of four.
## Impact
- Silent data-quality bug; no exception thrown.
- BM25 `important_kwd^30` boost effectively stops firing — the
indexed term is the whole list, never matches user query tokens.
- Any downstream aggregating `important_kwd` (tagging, analytics,
candidate-keyword review UIs) sees garbage.
## Compatibility
- Pure widening of the splitter; ASCII-comma-only outputs continue
to work identically.
- No schema / API change.
## Test plan
Manually verified against `qwen-plus@Tongyi-Qianwen` with
`auto_keywords=10` on Chinese .txt files:
- Before: `important_kwd` contains one element per chunk that is the
full LLM string with `，`-separated phrases inside.
- After: `important_kwd` contains N elements, one per phrase, as the
LLM intended.
### What problem does this PR solve?
The Gitee AI Go driver in `internal/entity/models/gitee.go` shipped with
a stub `Encode` method that returned `gitee, no such method`, even
though `conf/models/gitee.json` already wires the `embedding` URL
suffix. The conf also listed no embedding models, so the picker had
nothing to select.
This blocked any tenant who wanted to use Gitee AI for chat, rerank
(already working, see #14656), and embeddings from a single provider.
This PR fills the gap, mirroring the just-merged Aliyun `Encode`
(#14647):
- `internal/entity/models/gitee.go`: replace the `Encode` stub with a
real implementation.
Validates inputs, resolves the region with a default fallback, POSTs the
standard OpenAI-compatible `{"model", "input": [...]}` body to
`BaseURL[region] + URLSuffix.Embedding`, parses `data[*].embedding`
indexed by `data[*].index` so output order matches input order, handles
both `float64` and `float32` element types, and uses a 30s per-call
context deadline matching the merged `Rerank`.
- `conf/models/gitee.json`: add `BAAI/bge-m3` so the embedding picker
has something to select.
No factory change. No interface change. No URL suffix change.
Verified with `go build`, `go vet`, and `gofmt -l` : all clean.
Closes#14697
### Type of change
- [x] New Feature (non-breaking change which adds functionality)
### What problem does this PR solve?
`retrieval_by_children()` in `rag/nlp/search.py` crashes with a
`TypeError: 'NoneType' object is not subscriptable` when a parent
("mom") chunk referenced by child chunks is missing from the index.
This happens when the index is in an inconsistent state — for example
after a partial re-index, a document deletion that didn't clean up all
children, or a race condition during ingestion. `dataStore.get()`
returns `None` for the missing parent, and the subsequent access to
`chunk["content_with_weight"]` raises a `TypeError`.
**Stack trace:**
```
TypeError: 'NoneType' object is not subscriptable
File "rag/nlp/search.py", line 792, in retrieval_by_children
"content_with_weight": chunk["content_with_weight"],
```
### Type of change
- [x] Bug Fix
### Fix
When `dataStore.get()` returns `None` for a parent chunk, fall back to
using the child chunks directly and continue processing the remaining
parents. This preserves retrieval results for all other chunks rather
than aborting the entire query with an exception.
```python
chunk = self.dataStore.get(id, idx_nms[0], [ck["kb_id"] for ck in cks])
if chunk is None:
chunks.extend(cks)
continue
```
---------
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
## Summary
Two bypass vectors in the sandbox code security analyzer allowed
malicious code to pass the safety check undetected and reach the Docker
executor.
### 1. JavaScript: template-literal bypass of `require()` block
The `SecureJavaScriptAnalyzer` regex patterns used `['"]` to match
module names, covering only single and double quotes. An attacker could
use ES6 template literals to bypass all three `require` checks:
`javascript
const cp = require(`child_process`);
async function main() {
return cp.execSync('cat /etc/passwd').toString();
}
`
The same bypass applied to `fs` and `worker_threads`.
**Fix:** Updated all three `require` patterns from `['"]` to `['"\]` to
also match backtick template literals.
### 2. Python: `builtins` not blocked + attribute-call blind spot in
`visit_Call`
`visit_Call` only checked `ast.Name` nodes, so attribute-style calls
like `module.func()` were invisible to the analyzer. Additionally,
`builtins` was absent from `DANGEROUS_IMPORTS`. Combined, this allowed:
`python
import builtins
def main():
builtins.exec('import os; os.system("id")')
`
Neither the import nor the exec call triggered any flag.
**Fix:** Added `builtins` to `DANGEROUS_IMPORTS` and added an
`ast.Attribute` branch to `visit_Call` so that `module.dangerous_func()`
style calls are caught alongside bare `dangerous_func()` calls.
## Tests
Added four regression tests covering each new bypass vector:
- `test_javascript_child_process_template_literal_is_rejected`
- `test_javascript_fs_template_literal_is_rejected`
- `test_python_builtins_import_is_rejected`
- `test_python_attribute_eval_call_is_rejected`
---------
Co-authored-by: bounty-hunter <bounty@hunter.local>
### What problem does this PR solve?
Two bugs in the Aliyun Go driver:
1. **`Name()` returns `"siliconflow"`** — a copy-paste bug from when the
driver was created. `Name()` is used in error messages and log output,
so every Aliyun error incorrectly attributed itself to SiliconFlow.
2. **Silent empty URL for unknown regions in `ChatWithMessages`,
`ChatStreamlyWithSender`, and `ListModels`** — all three methods
construct the request URL as `z.BaseURL[region]` without checking
whether the key exists. For an unrecognised region this returns `""`,
producing a malformed URL like `"/chat/completions"` that the HTTP
transport rejects with a confusing error. `Encode` and `Rerank` (already
merged) correctly fall back to `"default"` and return a clear error.
This PR applies the same pattern to the remaining three methods.
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
### What problem does this PR solve?
Closes#14703
`GoogleModel.CheckConnection` currently returns a hardcoded `no such
method` error even though the Google Go driver already supports
`ListModels`. This makes provider connection checks fail regardless of
whether the configured API key can list Google models.
This PR makes `CheckConnection` call `ListModels`, adds a small API-key
guard for nil, empty, and whitespace-only keys, and keeps `ListModels`
useful by following paginated Google model responses.
### What stays unchanged
* Google model listing still uses the Google GenAI SDK with
`genai.BackendGeminiAPI`.
* Model names still come from `models.Items[*].Name`.
* `Balance`, `Encode`, chat, streaming, provider config, and factory
wiring are unchanged.
### Tests and validation
Added focused unit coverage for:
* `CheckConnection` delegating to `ListModels` and returning its error
* nil, missing, empty, and whitespace-only API key validation
* model-name passthrough from the list-models adapter
* paginated model listing, empty-result preservation, and next-page
error propagation
Validated current PR head `17ceef43515ba8c46c254dd349b9085bf26dcbea`
locally with Go 1.25.0:
* `go test ./internal/entity/models -run
'TestGoogleModel|TestCollectGoogleModelNames' -count=1 -v` - PASS
* `go test ./internal/entity/models -count=1` - PASS
* `go test -race ./internal/entity/models -count=1` - PASS
* `gofmt -w internal/entity/models/google.go
internal/entity/models/google_test.go` - PASS, no diff
* `git diff --check` - PASS
### Type of change
* [x] Bug Fix (non-breaking change which fixes an issue)
Co-authored-by: Jin Hai <haijin.chn@gmail.com>
