Do not swallow non-JSON http error responses

[Strift]

Pull Request

Related issue

Fixes #129

What does this PR do?

• Use raw response text when an API error body is not valid JSON

AI disclosure

Cursor with gpt-5.4-low and codex-5.3-low

PR checklist

Please check if your PR fulfills the following requirements:

• Did you use any AI tool while implementing this PR (code, tests, docs, etc.)? If yes, disclose it in the PR description and describe what it was used for. AI usage is allowed when it is disclosed.
• Does this PR fix an existing issue, or have you listed the changes applied in the PR description (and why they are needed)?
• Have you read the contributing guidelines?
• Have you made sure that the title is accurate and descriptive of the changes?

Thank you so much for contributing to Meilisearch!

Summary by CodeRabbit

Release Notes

• Bug Fixes

  • Improved error handling for API failures when the response body is malformed or not a JSON object. The error message now reliably falls back to the raw response text instead of leaving it unset or causing parsing-related issues.

• Tests

  • Added regression tests covering non-JSON (e.g., HTML) and JSON-but-not-object (e.g., array) error responses.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

Run ID: da84df69-2c2a-45ef-8adf-62fb13802157

📥 Commits

Reviewing files that changed from the base of the PR and between 5e2f14e and ef693f6.

📒 Files selected for processing (2)

• meilisearch/errors.py
• tests/errors/test_api_error_meilisearch.py

🚧 Files skipped from review as they are similar to previous changes (1)

• meilisearch/errors.py

---

📝 Walkthrough

Walkthrough

MeilisearchApiError.__init__ now wraps json.loads(request.text) in a try/except json.JSONDecodeError and validates the parsed JSON is a dict before extracting message, code, link, and type. If parsing fails or the result is not a dict, it falls back to raw request.text. Two regression tests cover non-JSON (HTML) and non-object (JSON array) response bodies.

Changes

Non-JSON error response fallback

Layer / File(s)	Summary
JSON parse fallback with dict type validation meilisearch/errors.py	MeilisearchApiError.__init__ wraps json.loads in try/except json.JSONDecodeError, validates the parsed JSON is a dict before extracting fields, and sets self.message to raw request.text on parse failure or non-dict payload. When a dict message field is absent or falsy, also falls back to request.text.
Regression tests for non-JSON and non-dict JSON responses tests/errors/test_api_error_meilisearch.py	Two new tests mock HTTP 502 responses: one with an HTML body (non-JSON), another with a valid JSON array body (non-dict). Both assert MeilisearchApiError preserves status_code=502, sets message to the raw response text, and leaves code, link, and type as None.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~8 minutes

Poem

🐇 A gateway threw HTML my way,
JSON was nowhere to be found that day.
I caught the error, checked the type,
And kept the text when parsing's not right.
Now arrays, HTML, and missing fields all play! 🎉

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 60.00% which is insufficient. The required threshold is 80.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The PR title 'Do not swallow non-JSON http error responses' directly and clearly describes the main objective of the changeset: improving error handling to gracefully handle non-JSON HTTP error responses instead of crashing.
Linked Issues check	✅ Passed	The PR fully addresses issue #129 by implementing a try/except mechanism to parse JSON error responses and falling back to raw response text when parsing fails, resolving the JSONDecodeError crash.
Out of Scope Changes check	✅ Passed	All changes are directly scoped to the error handling issue. The modifications to MeilisearchApiError and corresponding regression tests are tightly focused on handling non-JSON HTTP error responses.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch python-1209-create_index-and-get_index-giving-json-error

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[coderabbitai]

Actionable comments posted: 1

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@meilisearch/errors.py`:
- Around line 40-43: The json.loads(request.text) call may return valid JSON
that is not a dictionary (such as a list or string), and calling .get() methods
on non-dict types will raise an AttributeError. Add a type check to verify that
json_data is actually a dictionary before attempting to call .get() on it. If
json_data is not a dictionary, fall back to using request.text as the message
value. Apply this guard to all the .get() calls on json_data for the message,
code, link, and type attributes.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

Run ID: 41cd5f41-99d1-4306-a125-2728b3a2f325

📥 Commits

Reviewing files that changed from the base of the PR and between 1464823 and 5e2f14e.

📒 Files selected for processing (2)

• meilisearch/errors.py
• tests/errors/test_api_error_meilisearch.py