You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
keep a history of generated types/schemata, each in their own folder. simplest way: make it part of e.g. build:schema, with the output path being src/migration/history/{semver}
put this in place now to lay groundwork for later
programmatically check for relevant schema changes (change name, type, or properties, not description), throw during package build if theres no migration available.
should happen for even trivial renames. sometimes thatll mean an empty function, but it needs to be a conscious choice
id migrations should include a regex pattern + a replacement template string using the regex capture groups
consider ways to make this useful for non-js platforms.
id regexes could be presented as a JSON dictionary object (with its own schema). key is the regex string, value is the replacement string. (validation for this: check capture groups by running the replacer function, and seeing if theres any unfilled ones by matching \$\d
The text was updated successfully, but these errors were encountered:
rsek
added
the
epic
A big task that's broken down into constituent tasks.
label
Jun 16, 2024
build:schema
, with the output path beingsrc/migration/history/{semver}
\$\d
The text was updated successfully, but these errors were encountered: