Oracle package deploy/revert suggestion

[fluca1978]

I'm seeking for suggestions about a best practice approach when dealing with the (stupid) way Oracle handles packages, that means the whole package has to be rebuilt even if a small function changes in it.

Imagine I've changes like 01.sql, 02.sql, 03.sql where the revert script of 02.sql is the deployment of 01.sql and so on.
01 is deployed, 02 and 03 are ongoing work.

A bug is found in the deployment of 01.sql, so I sqitch revert, fix the bug and sqtich deploy. So far so good, but now I have to change manually all the revert scripts of subsequent changes.

I cannot implement a subsequent bug fix as in 04.sql because changes 02 an 03 could not be ready for deployment yet.

Playing with git branches could help, but requires a little extra effort.

How do you deal with a situation like this?

[theory]

Me personally I've just manually changed all the intervening scripts. It hasn't come up often enough for it to be much of an issue.

There is logic to do something like this built into the rework command. The command itself requires a tag and special file naming, but the basic pattern for generating revert scripts from prior deploys is there. Maybe there's some way to create a script that follows the same basic logic for this use case?