[EPIC] Migrate Python notebooks that belong to a single job

[nfx]

Most of the Databricks Workflows are just scheduled Python notebooks. We need to migrate all relevant parts of the code in those notebooks including but not limited to:

• [EPIC] Lint all notebooks that belong to a job in scope of the assessment workflow #1085
• [FEATURE]: If a code computes a value dynamically, do value inference, at least at the state of linting #1205
• [FEATURE]: Rewrite incompatible SQL UDF declarations #1119
• [FEATURE]: Upgrade usage of s3fs library #1117
• [FEATURE]: Rewrite usages of DBFS Mount paths in Init Shell Scripts #1115
• [FEATURE]: Rewrite usages of DBFS Mount paths in Python code, including /dbfs/mnt local fs, dbfs:/, and /mnt #1109
• [FEATURE] Create mapping of directfs accesses into UC tables #392
• [FEATURE]: Regression testing harness for post-migration validation of configurations (Workload Assets) #1122
• [FEATURE]: Rewrite incompatible SQL UDF declarations #1119
• [FEATURE]: Rewrite usages of DBFS Mount paths in Init Shell Scripts #1115
• [FEATURE]: HMS Function Migration #885
• Migrate streaming checkpoints #492
• [FEATURE]: ensure Spark.sql-like code is not migrated accidentally #1213
• [FEATURE]: Mark notebooks as migrated #1327

Proposed Solution

• Create migrate-notebooks-for-job and revert-migrate-notebooks-for-job CLI commands
• Build on top of https://github.com/databrickslabs/ucx/blob/main/src/databricks/labs/ucx/source_code/notebooks.py#L20-L32
• Backup notebooks before doing code migration
• Execute code migration
• tag relevant job as uc-migrated

Additional Context

No response