How to fix Lepiter loading errors

Errors can happen during the loading of a Lepiter knowledgebase, usualy because of duplicated pages or corrupted JSON files.

When an error happens during startup a notification is shows detailing the error. There are two types of notifications for Lepiter errors with the following notification message:

Fatal error during Lepiter reload - This is an error that causes the reload to stop completely. The database is most likely in an inconsistent state and you should fix the problem before contiuning.

Content error during Lepiter reload - Indicates that some content in the database (most likely a page) could not be loaded. Only the affectet content is not loaded, the rest of the database should be ok.

For both types of errors do a manual reload using the snippet below to get a live debugger on the reload.

LeDatabasesRegistry defaultLogicalDatabase reloadWithErrorDebugging

As an alternative, the following snippet does the reload and opens each error in a forked process.

LeDatabasesRegistry defaultLogicalDatabase reloadWithForkedDebugging

To get the same behavior that happens on startup run the reload by only showing notifications.

LeDatabasesRegistry defaultLogicalDatabase reloadWithNotificationOnError

Disable Lepiter reload

In case of more complicated errors, it can be useful to avoid the automatic reload of Lepiter databases on startup, to check the state of those databases in the image.

In that case open Glamorous Toolkit from a terminal and pass the --noLepiterReload command line option (LeDatabaseStore>>#noLepiterReload noLepiterReload "Answer a boolean indicating whether the image has been started with the --noLepiterReload option" ^ Smalltalk commandLine argumentsInclude: '--noLepiterReload' )