Data ingestion function app bug-fix

[placerda]

This pull request includes several changes to improve and fix the document chunking function app, focusing on refining the chunking process for various file types and enhancing logging and configuration settings. The most important changes include updates to the chunker classes, adjustments to the configuration settings, and improvements to logging.

Chunker Class Updates:

• chunking/chunkers/langchain_chunker.py: Modified methods to download blob data and decode it into text before chunking. Added a text parameter to _chunk_content method. (F5d12d29L70R80, F5d12d29L96R100, F5d12d29L133R137)
• chunking/chunkers/spreadsheet_chunker.py: Added environment variable support for max_chunk_size and improved logging for chunk processing. (F7853bb3L55R54, F7853bb3L71R69, F7853bb3L96R89)
• chunking/chunkers/transcription_chunker.py: Updated methods to pass text parameter for chunking and improved logging. (chunking/chunkers/transcription_chunker.pyL63-R68, F420b885L78R79, F420b885L98R99, F420b885L107R108)

Configuration and Logging Enhancements:

• local.settings.json.template: Added SPREADSHEET_NUM_TOKENS configuration setting.
• tools/aoai.py: Increased MAX_RETRIES for rate limit errors from 3 to 10.
• function_app.py: Removed documentContent from logging and request schema. [1] [2]

Documentation Updates:

• README.md: Updated supported formats and removed outdated information about Document Intelligence API 4.0. [1] [2] F0524c14L89R90)

Minor Changes:

• setup.py: Changed dataToExtract configuration to allMetadata. (Fb1524a3L663R667)
• chunking/document_chunking.py: Corrected logging message format in _error_message method.