[ContentUnderstanding] Add to_llm_input helper for converting analysis results to LLM-friendly text

[chienyuanchang]

Description

Adds the to_llm_input() public helper function to the azure-ai-contentunderstanding package. This function converts a CU AnalysisResult into a formatted text string (YAML front matter + markdown body) suitable for injecting into LLM prompts, storing in vector databases, or returning as tool output in agentic workflows.

Key features:

• Renders all content types: documents (with page markers), audio/video (with time ranges for multi-segment), and classification hierarchies (parent auto-skipped, children rendered with category labels)
• _resolve_fields() recursively flattens all 9 ContentField subtypes (StringField, NumberField, ObjectField, ArrayField, etc.) into plain Python dicts
• Span-based page markers from pages[].spans offsets, with <!-- PageBreak --> fallback for older content
• Minimal built-in YAML serializer (no external dependency) with proper quoting for dates, booleans, and YAML-special characters
• RAI warnings always included in output regardless of include_fields/include_markdown flags
• Single AV content omits timeRange; only multi-segment AV includes timeRange per segment (per design spec)
• Configurable via include_fields, include_markdown, and metadata keyword arguments

Files changed

File	Change
azure/ai/contentunderstanding/_helpers.py	New file — to_llm_input(), _resolve_fields(), and supporting internal functions
azure/ai/contentunderstanding/_patch.py	Import and re-export to_llm_input in __all__
tests/test_to_llm_input.py	New file — 60 unit tests across 10 categories (public API, error handling, field resolution, documents, AV, multi-segment, classification, parameters, front matter, edge cases, real CU patterns)

How to verify

cd sdk/contentunderstanding/azure-ai-contentunderstanding
pip install -e .
python -m pytest tests/test_to_llm_input.py -v

[Copilot]

Pull request overview

Adds a public to_llm_input() helper to azure-ai-contentunderstanding to convert AnalysisResult objects into LLM-friendly text (YAML front matter + Markdown), along with unit tests and version/docs updates.

Changes:

• Introduces to_llm_input() and supporting field/page/YAML rendering helpers.
• Re-exports to_llm_input from the package public surface and bumps version to 1.1.0.
• Adds a comprehensive test suite covering content types, rendering rules, and edge cases.

Reviewed changes

Copilot reviewed 6 out of 6 changed files in this pull request and generated 2 comments.

Show a summary per file

File	Description
azure/ai/contentunderstanding/_helpers.py	New helper implementation, including minimal YAML serializer and content rendering logic.
azure/ai/contentunderstanding/_patch.py	Re-exports to_llm_input via __all__ for package-level import.
tests/test_to_llm_input.py	New unit tests validating helper behavior across documents, AV, classification, and flags.
azure/ai/contentunderstanding/_version.py	Version bump to 1.1.0.
README.md	Adds 1.1.0 to the SDK-to-service-version table.
CHANGELOG.md	Adds release notes entry for 1.1.0.