fix(azure): Strip operation suffixes from deployment URLs to prevent 404 errors - #70
Merged
josemaria-vilaplana merged 1 commit intoJan 26, 2026
Conversation
…404 errors When api_base is configured with full deployment paths like .../openai/deployments/gpt-4o/chat/completions, the Azure SDK appends the operation path again, resulting in duplicate paths and 404 errors. This fix updates select_azure_base_url_or_endpoint to strip operation suffixes (/chat/completions, /completions, /embeddings, /audio/speech, /audio/transcriptions, /images/generations) before passing the URL to the Azure SDK. This is the counterpart to PR #61 which fixed the same issue for the Responses API. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
josemaria-vilaplana
changed the title
mateo-di
approved these changes
Jan 23, 2026
josemaria-vilaplana
deleted the
fix/azure-chat-completions-url-sanitization
branch
January 26, 2026 08:42
mateo-di
added a commit
that referenced
this pull request
Jan 29, 2026
…ution Replace simple PR listing with intelligent context building: - Fetch ALL merged PRs to carto/main (up to 1000, no arbitrary limit) - For each CARTO-modified conflicted file, show: - Which PRs modified it - PR title (what the change does) - Author and lines changed (+/-) - Summary from PR description (first 150 chars) - Provide full PR reference table sorted by recency Example context for Claude: ``` ### `litellm/llms/azure/chat/gpt_transformation.py` **2 CARTO PR(s) modified this file:** - PR #70: fix(azure): Strip operation suffixes from deployment URLs Author: @josemaria-vilaplana | +189/-1 lines - PR #61: fix: Azure Responses API URL construction Author: @josemaria-vilaplana | +77/-3 lines ``` This gives Claude full understanding of WHY each file was modified, enabling smarter conflict resolution decisions.
mateo-di
added a commit
that referenced
this pull request
Jan 29, 2026
* feat: Implement CARTO-First resolution strategy for upstream syncs Add CARTO-FIRST conflict resolution priority that mandates preserving CARTO customizations made by CartoDB org members during upstream syncs. Key changes: - Resolver: Add pre-analysis step to identify CARTO-modified files - Resolver: Update prompt with CARTO-FIRST priority table (6 priorities) - Resolver: Add known CARTO customizations list (Azure, Snowflake, etc.) - CI Fixer: Add FORBIDDEN ACTIONS (never revert CARTO code) - CI Fixer: Add REQUIRED ACTIONS (fix tests, not CARTO code) - Docs: Update CARTO_CLAUDE.md with new resolution priorities The key insight is using `git log` to identify files with commits from @carto.com or @cartodb.com authors, then prioritizing those files. [sc-XXXXX] * fix: Use GitHub org membership check instead of email domains Update CARTO file detection to use GitHub API org membership check, matching the approach in calculate_carto_version.sh (PR #72). Changes: - Resolver: Fetch CartoDB org members via `gh api orgs/CartoDB/members` - Resolver: Check commit authors against org member list - CI Fixer: Update docs to reference "CartoDB GitHub org members" - Docs: Update CARTO_CLAUDE.md with org membership reference This is more reliable than email domain matching because: - GitHub usernames are authoritative - Works for users with non-CARTO emails - Consistent with version calculator logic * feat: Build rich CARTO context from ALL merged PRs for conflict resolution Replace simple PR listing with intelligent context building: - Fetch ALL merged PRs to carto/main (up to 1000, no arbitrary limit) - For each CARTO-modified conflicted file, show: - Which PRs modified it - PR title (what the change does) - Author and lines changed (+/-) - Summary from PR description (first 150 chars) - Provide full PR reference table sorted by recency Example context for Claude: ``` ### `litellm/llms/azure/chat/gpt_transformation.py` **2 CARTO PR(s) modified this file:** - PR #70: fix(azure): Strip operation suffixes from deployment URLs Author: @josemaria-vilaplana | +189/-1 lines - PR #61: fix: Azure Responses API URL construction Author: @josemaria-vilaplana | +77/-3 lines ``` This gives Claude full understanding of WHY each file was modified, enabling smarter conflict resolution decisions.
mateo-di
added a commit
that referenced
this pull request
Feb 4, 2026
…text Simplifies the resolver for a purer goal-driven approach: - Remove CARTO-MODIFIED vs UPSTREAM-ONLY classification - Keep only PR feature summaries and file-to-PR mapping - Claude decides per-file based on context, not rigid labels - Replace complex verification steps with simple sanity check - Reduce workflow from 1284 to 1078 lines Philosophy: "This file was touched by PR #70 (Azure fix). You decide." Instead of: "This file is CARTO-MODIFIED, keep it."
3 tasks
Cartofante
pushed a commit
that referenced
this pull request
Feb 4, 2026
Features restored: - [PR #70] Azure URL suffix stripping - Strip operation suffixes (/chat/completions, /embeddings, etc.) from deployment URLs to prevent 404 errors when Azure SDK appends them again - [PR #54] JSON repair for streaming tool calls - Use JSONDecoder.raw_decode() to handle malformed JSON from Gemini streaming that sends duplicate/overlapping chunks Integration approach: - PR #70: Restored regex-based stripping in select_azure_base_url_or_endpoint() with docstring explaining the fix. Added 12 unit tests. - PR #54: Restored _validate_and_repair_tool_arguments() function and integrated it into get_combined_tool_content() and get_combined_function_call_content(). Added 19 unit tests. Verified: syntax checks, ruff linting passed Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
mateo-di
added a commit
that referenced
this pull request
Feb 10, 2026
Static registry of critical CARTO customizations that must survive upstream syncs. The resolver should verify all patterns exist after every sync to prevent silent feature regressions (like the OCI Gemini and Snowflake tool calling drops in v1.81.0). Tracks: OCI Gemini (#68), Snowflake streaming/tool calling (#38, #58), Azure URL stripping (#70), JSON repair (#54), Redis sessions (#16).
3 tasks
3 tasks
This was referenced Apr 10, 2026
3 tasks
3 tasks
3 tasks
3 tasks
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Title
fix(azure): Strip operation suffixes from deployment URLs to prevent 404 errors
Summary
When
api_baseis configured with full deployment paths like:The Azure SDK appends the operation path again, resulting in duplicate paths:
This causes 404 errors from Azure.
Solution
This fix updates
select_azure_base_url_or_endpointinlitellm/llms/azure/common_utils.pyto strip operation suffixes before passing the URL to the Azure SDK:/chat/completions/completions/embeddings/audio/speech/audio/transcriptions/images/generationsThis is the counterpart to PR #61 which fixed the same issue for the Responses API.
Relevant issues
Counterpart to #61 (Azure Responses API URL fix)
Pre-Submission checklist
tests/litellm/directory, Adding at least 1 test is a hard requirement - see detailsmake test-unitType
🐛 Bug Fix
Changes
litellm/llms/azure/common_utils.py: Updatedselect_azure_base_url_or_endpointto strip operation-specific suffixes from deployment URLstests/test_litellm/llms/azure/test_azure_common_utils.py: Added 12 unit tests for URL sanitizationTest plan
litellm-azure-url-fix)🤖 Generated with Claude Code