Hi everyone! I'm trying to apply ibatis-migrations to manage our database schema and I found a semantic problem with it. (we are using postgres if it does matter.)
My plan is to maintain ALL changes of database with "..\migrate", including changes in stored procedures. Here is my use-case: 1. ..\migrate new create new stored procedure demo_stored_procedure_1 2. fill-up migration script with stored procedure body (and undo with drop procedure) 3. ..\migrate up looks fine so far, but "..\migrate" preprocessing my SQL code to make it single line, so body of my stored procedure (which is multiline in my script) is passed to database as single-line statement. So I can't get my stored procedure code back from database because it is required to reformat it back to multiline code. It's also hard to find errors in code, and set-up a breakpoint (if it is possible). I see next solutions for this problem: 1. special options could be added to "..\migrate" environment to force this tool to keep SQL code formatting as is. i.e. do_not_concatenate_lines=true and do_not_strip_comments=true 2. I can put all my stored procedures code in separate SQL file and apply them through "psql". I prefer 1-st plan, but 2-nd is also fine. May be I'm using "..\migrate" tool in wrong way, please suggest any better solution or explain why 1-st or 2-nd approach is good from your point of view. -- Dmtiry
