ViewableTree Interface
A tree from which a TreeView can be created.
For more information about our API support guarantees, see here.
Signature
/** @sealed */
export interface ViewableTree
Methods
| Method | Alerts | Return Type | Description |
|---|---|---|---|
| viewWith(config) | System |
TreeView<TRoot> | Returns a TreeView using the provided schema. If the stored schema is compatible with the view schema specified by config, the returned TreeView will expose the root with a schema-aware API based on the provided view schema. If the provided schema is incompatible with the stored schema, the view will instead expose a status indicating the incompatibility. |
Method Details
viewWith
Returns a TreeView using the provided schema. If the stored schema is compatible with the view schema specified by config, the returned TreeView will expose the root with a schema-aware API based on the provided view schema. If the provided schema is incompatible with the stored schema, the view will instead expose a status indicating the incompatibility.
For more information about our API support guarantees, see here.
Signature
viewWith<TRoot extends ImplicitFieldSchema>(config: TreeViewConfiguration<TRoot>): TreeView<TRoot>;
Type Parameters
| Parameter | Constraint | Description |
|---|---|---|
| TRoot | ImplicitFieldSchema |
Remarks
If the tree is uninitialized (has no schema and no content), use initialize(content) on the returned view to set the schema and content together. Using viewWith followed by upgradeSchema() to initialize only the schema for a document is technically valid when the schema permits trees with no content.
Note that other clients can modify the document at any time, causing the view to change its compatibility status: see events for how to handle invalidation in these cases.
Only one schematized view may exist for a given ITree at a time. If creating a second, the first must be disposed before calling viewWith again.
TODO: Provide a way to make a generic view schema for any document. TODO: Support adapters for handling out-of-schema data.
Doing initialization here allows a small API that is hard to use incorrectly. Other approaches tend to have easy-to-make mistakes. For example, having a separate initialization function means apps can forget to call it, making an app that can only open existing documents, or call it unconditionally leaving an app that can only create new documents. It also would require the schema to be passed into separate places and could cause issues if they didn't match. Since the initialization function couldn't return a typed tree, the type checking wouldn't help catch that. Also, if an app manages to create a document, but the initialization fails to get persisted, an app that only calls the initialization function on the create code-path (for example how a schematized factory might do it), would leave the document in an unusable state which could not be repaired when it is reopened (by the same or other clients). Additionally, once out of schema content adapters are properly supported (with lazy document updates), this initialization could become just another out of schema content adapter and this initialization is no longer a special case.
Parameters
| Parameter | Type | Description |
|---|---|---|
| config | TreeViewConfiguration<TRoot> |
Returns
Return type: TreeView<TRoot>