tuttle
tuttle copied to clipboard
lexman
Make for data
A `virtual://` that does not produce anything could help in situations where you want to do an actions after a resource is produced. For exemple, sending a mail after a...
- [ ] a spreadshit resource, with nice functionnalities in `auto`, like loading to/from database tables or csv files - [ ] a spreadshit processor, for some exotic operations ?
When a process is run, it can declare at runtime that it used a resource. This could be used for : - following changes in python modules without having to...
Bundle your workflow in a package you can distribute, to allow other people to apply the same workflow to their data without installing neither tuttle nor your project
for operation that break guarantiy of consistency between code and the data produced... _Ok, I know what I'm doing_ : - I've only added a comment, but I really know...
Implement a strict parser with a nice grammar Could be the occasion to allow list of resources an several lines, and the processor on its own line
Any suggestion welcome to make syntax easier to work with... eg : declare the processor on its own line, allow line breaks between resources lists, to avoid very long lines
- [ ] _never delete intermediate data_ : even if intermediate data doesn't seem usefull, never delete a table or a file : it can be used as a savepoint....