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
Class boundaries are now detected by brace matching, so top-level code
between or after classes (functions, constants, extensions) is no longer
swallowed or destroyed during conversion.
Conversions are applied through an undoable WorkspaceEdit and respect
unsaved buffer changes, instead of writing directly to disk.
Existing methods, getters and custom fromJson factories are preserved; the
converter only injects annotations, the id field, the constructor and a fromJson when missing.
Field detection is brace-depth aware: statements inside method bodies,
getters/setters and static members are no longer mistaken for fields.
Class headers using $ (e.g. with _$PhormXMixin) and generic classes are
parsed correctly.
Added
Settings: phorm.generateFullService, phorm.timestamps, phorm.paranoid, phorm.addFromJson, phorm.enableCodeLens — emitted as the matching @Schema options.
CodeLens "⚡ To PHORM Model" above plain classes, plus conversion of just the
class under the cursor.