python(feat): add data import api to sift_client

[wei-qlu]

What was changed

• Added data_import resource with support for CSV and Parquet file imports
• Added low-level gRPC wrapper for DataImportService (DetectConfig, CreateDataImportFromUpload)
• detect_config auto-detects column structure, time columns, and data types from a file sample
• import_from_path handles the full flow: detect config, create import, upload file, and poll for completion via the jobs API
• Parquet supports both flat dataset and single-channel-per-row layouts, with automatic footer extraction
• Added Pydantic models in sift_types for all CSV and Parquet config types
• Import results are returned as Job objects, consistent with existing services

Verification

• Unit tests for config mutability, get_column (helper), time column validation, extension mapping, and run precedence (Integration tests will not be included for now)
• Full import cycle tested locally for both CSV and Parquet files

[wei-qlu]

data_imports.proto was updated to support tdms and some additions to CSV, will be making updates accordingly

[alexluck-sift]

Great work!