http://amberframework.org - Questions? Post on StackOverflow and contribute to community knowledge! https://stackoverflow.com/questions/tagged/amber-framework - IRC? Share your chat messages across http://webchat.freenode.net/?channels=#amber
so just add the SQL to add your new column?
or if you're still developing, alter the original until you get it in a good state then can down/up it
idt any framework has that feature
(in crystal)
in case if you can afford just drop the old schema. Guess i will go and take a look at source, it doesn't sound like a complex matter
Because currently if developer need 10+ or even more columns it will be kinda unconvient to pass all of them to the
amber g model user age:int etc
others are more familiar might have thoughts/suggestions
Anyway thank you! <3
btw do you know is this?
``<<-DEPENDENCY
require "../src/models/**"
DEPENDENCY
<<- this operator
it just a fancy syntax for making multi line strings
Interesting
@Yashko at the moment there is no option to automatically modify database tables based on model fields, so (post public release) you need to create a migration to modify the table and add the column. As @Blacksmoke16 points out, prior to public release you should simply modify the migration that creates the table. It is almost an anti-pattern to create new table altering migrations prior to the first public release.
@Yashko At one point, I attempted to do what you are suggesting but it was clear this wouldn’t work. You need to support migrating from one column to another and there is no way of predicting what the developer intended to do. I tried an additive only approach but you would still need to go back and remove columns so it would require a two step process. Migrations are better in that they run in order and the database keeps track of what migration you are on so a database can be in any state and the migrations will run from that point forward.
However, It’s not as dry as your suggestion.
i suppose i should modify this block
`
class EventControllerTest < GarnetSpec::Controller::Test
getter handler : Amber::Pipe::Pipeline
def initialize
@handler = Amber::Pipe::Pipeline.new
@handler.build :web do
plug Amber::Pipe::Error.new
plug Amber::Pipe::Session.new
plug Amber::Pipe::Flash.new
end
@handler.prepare_pipelines
end
end
`
in my spec file.
but i don't know how?
er
.find_by, and find use .first under the hood
otherwise you can pass some raw sql to
.first