Support @file syntax for JMESPath queries

[joshrotenberg]

Summary

Add support for loading JMESPath queries from files using @filename syntax, similar to how --data @file.json works.

Proposed

# Load query from file
redisctl cloud cost-report download $ID -q @queries/top-resources.jmespath

# Still support inline queries
redisctl cloud cost-report download $ID -q 'sum([].BilledCost)'

Use case: Canned FOCUS queries

Ship a set of pre-built queries for common cost analysis patterns:

queries/focus/
├── top-n-resources.jmespath
├── cost-by-region.jmespath
├── cost-by-resource-name.jmespath
├── cost-by-charge-category.jmespath
└── zero-cost-resources.jmespath

Example cost-by-region.jmespath:

group_by(@, \`RegionName\`) | items(@) | [*].{region: [0], total: sum([1][*].BilledCost)} | sort_by(@, &total) | reverse(@)

Users get powerful analysis without memorizing complex JMESPath syntax.

References

• jmespath-extensions CLI query files: https://joshrotenberg.com/jmespath-extensions/cli/query-files.html
• Related: Add one-shot cost-report export command #592 (one-shot export), Add cost report documentation for FinOps use cases #588 (cost report docs)

[joshrotenberg]

Now tracked under #726 (JMESPath epic). Removed from #680 (Cloud epic) since it's a cross-cutting JMESPath concern.