feat(dimensional): replace course_order placeholder in bridge_program_course

[blarghmatey]

Context

PR #1868 added bridge_program_course mapping programs to their constituent courses. The course_order column is hardcoded to 1 for all rows because source systems (MITx Online, xPro, MicroMasters) do not expose a canonical ordering of courses within a program.

A SQL comment documents this:

course_order is set to 1 as a placeholder. The source systems do not expose a canonical ordering of courses within a program, so a meaningful sort order cannot be derived without additional business logic or manual curation.

What

Investigate whether any source system or intermediate model exposes course ordering within a program:

• int__mitxonline__coursesinprogram — check for any position/order field
• int__mitxpro__coursesinprogram — same
• int__micromasters__* — check for course sequence data
• If ordering is available, populate course_order from source
• If not, consider adding a manual curation seed table (CSV) mapping program + course to order

Acceptance Criteria

• Source system ordering capability assessed and documented
• course_order either populated from source data or from a curated seed
• bridge_program_course updated and tested

[rachellougee]

@blarghmatey This issue seems outdated. I don't see the hardcoded course_order in bridge_program_course and the reference model int__mitxonline__coursesinprogram (int__mitxonline__program_requirements) . But we still need to validate bridge_program_course. Is that what this issue is about?

[blarghmatey]

Closing as implemented. As noted in the comment from @rachellougee, course_order is no longer hardcoded to 1. The current bridge_program_course.sql sources it from:

• MITx Online: course_order_in_program from int__mitxonline__program_requirements
• xPro: course_order from int__mitxpro__coursesinprogram
• MicroMasters: course_position_in_program from the requirements CTE

The original issue's acceptance criteria are met. If there are validation gaps around the bridge table's correctness, those are being tracked under the grain audit findings in #2073 (specifically G1 covers the MicroMasters elective-set fan-out risk).