fix: accumulate all text blocks in Anthropic non-streaming response

[AmirF194]

ProviderAnthropic._query (the non-streaming path in anthropic_source.py) sets llm_response.completion_text inside a loop over completion.content, once per text content block, instead of appending. When a Message carries more than one text block, only the last one survives; every earlier text block is silently dropped from the reply and from conversation history. This can happen with the Citations API and with extended-thinking flows, both of which can split a response into multiple text blocks. The streaming sibling, _query_stream, already accumulates correctly with final_text += event.delta.text.

Modifications / 改动点

Collect each text block into a list during the loop and join them once after the loop, instead of overwriting completion_text per block. thinking and tool_use handling are unchanged.

• This is NOT a breaking change. / 这不是一个破坏性变更。

Screenshots or Test Results / 运行截图或测试结果

Added test_query_accumulates_multiple_text_blocks, which feeds ProviderAnthropic._query a fake Message with two text blocks. It fails on unmodified master (AssertionError, only the second block's text survives) and passes with this fix. The full tests/test_anthropic_kimi_code_provider.py file: 28 passed, 1 pre-existing failure unrelated to this change (test_create_http_client_uses_anthropic_httpx_module, an SDK version drift already being fixed in open PR #9769). ruff format --check and ruff check are clean on both changed files.

Not verified: a live call to the provider's API. This was reproduced and tested against the SDK's Message/content-block shape directly, not against a real multi-block response from the API.

---

Checklist / 检查清单

• If there are new features added in the PR, I have discussed it with the authors through issues/emails, etc.
  / 如果 PR 中有新加入的功能，已经通过 Issue / 邮件等方式和作者讨论过。

• My changes have been well-tested, and "Verification Steps" and "Screenshots" have been provided above.
  / 我的更改经过了良好的测试，并已在上方提供了"验证步骤"和"运行截图"。

• I have ensured that no new dependencies are introduced, OR if new dependencies are introduced, they have been added to the appropriate locations in requirements.txt and pyproject.toml.
  / 我确保没有引入新依赖库，或者引入了新依赖库的同时将其添加到 requirements.txt 和 pyproject.toml 文件相应位置。

• My changes do not introduce malicious code.
  / 我的更改没有引入恶意代码。

Summary by Sourcery

Accumulate every text block returned by Anthropic’s non-streaming API so complete responses and conversation history are preserved.

Bug Fixes:

• Preserve all text content blocks in Anthropic non-streaming responses instead of returning only the final block.

Tests:

• Add coverage verifying that multiple Anthropic text blocks are concatenated in the response.

[sourcery-ai]

Hey - I've reviewed your changes and they look great!

Sourcery assessment

Approved.

---

Sourcery is free for open source - if you like our reviews please consider sharing them ✨

• X
• Mastodon
• LinkedIn
• Facebook

_{Help me be more useful! Please click 👍 or 👎 on each comment and I'll use the feedback to improve your reviews.}