feat: add custom seed path to config

[avallete]

What kind of change does this PR introduce?

Feature enhancement: Improved seed data handling and configuration

What is the current behavior?

Currently, the system uses a single, fixed seed file path (utils.SeedDataPath) for database seeding. This limits flexibility in organizing and applying seed data.

What is the new behavior?

1. Introduces a new configuration option in config.toml for specifying multiple seed file paths using glob patterns:

[db.seed]
enabled = true
sql_paths = ['./seed.sql', './seeds/*.sql']

2. Implements a new function GetSeedFiles that:

   • Matches files based on the configured glob patterns
   • Deduplicates matched files to avoid duplicate seeding
   • Sorts the files to ensure consistent application order

3. Updates various parts of the codebase to use the new seeding mechanism:

   • Modifies the SeedDatabase function to use the new GetSeedFiles function
   • Updates error handling and logging related to seed file operations
   • Adjusts test cases to accommodate the new seeding behavior

4. Change the logging to display the seed file paths being used

Additional context

This change allows for more flexible organization of seed data, enabling users to split seed data across multiple files and use patterns to include them. It maintains backwards compatibility by defaulting to the previous behavior if no custom configuration is provided.

I'm just wondering about how this should work with supabase "branch deployments" and so on. Locally we'll always find the seeds files at the right location. I wonder how the branching approach work with that. From my understanding it does "clone" the user repository so it should work the same as locally but I might need some help to get that behavior tested.

[coveralls]

Pull Request Test Coverage Report for Build 11027502555

Details

• 20 of 26 (76.92%) changed or added relevant lines in 5 files are covered.
• 8 unchanged lines in 3 files lost coverage.
• Overall coverage decreased (-0.05%) to 60.002%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
cmd/start.go	0	1	0.0%
internal/migration/apply/apply.go	3	4	75.0%
internal/db/push/push.go	3	5	60.0%
internal/utils/misc.go	13	15	86.67%

Files with Coverage Reduction	New Missed Lines	%
internal/migration/apply/apply.go	1	83.33%
internal/storage/rm/rm.go	2	89.53%
internal/gen/keys/keys.go	5	12.9%

Totals
Change from base Build 11014121556:	-0.05%
Covered Lines:	6458
Relevant Lines:	10763

---

💛 - Coveralls

[avallete]

With the new files searching behavior we need to ensure the file actually exist on the filesystem. Otherwise, the "Glob" filter will just not return anything, the seeds will be skipped and the file will never be attempted to open.

[jacob-8]

This broke my seeding ability on Windows. No matter if I use the default supabase/seed.sql location or use something like sql_paths = ['./seeds/*.sql'] I always get the response No seed files matched pattern: supabase\seed.sql (or whatever my pattern happens to be). Is anyone else having this problem?

[avallete]

This broke my seeding ability on Windows. No matter if I use the default supabase/seed.sql location or use something like sql_paths = ['./seeds/*.sql'] I always get the response No seed files matched pattern: supabase\seed.sql (or whatever my pattern happens to be). Is anyone else having this problem?

Hey thanks for reporting, this is already handled and should be fixed by #2735 in the next upcoming release sorry about that. Our testing pipeline for windows is still flawky we want to improve it that's in our roadmap.

Meanwhile, you should be able to test this fix by using the beta channel via npx supabase@beta start.

Please use github issues to report such cases as it has a much better discoverability for the team than comments on closed PR 👍