Improve formatting of logged indexing processors and reindexing resolvers

Description

There are multiple problems:

  1. The format, inspired from JSON, is not very human-readable. Something similar to YAML would be better.

  2. The node types are mentioned in a "class" attribute, but wouldn't just prepending them before the "{" be better? E.g. "myType { ... }" instead of "{class = myType, ...}"

  3. The node types, in the "class" attributes, are very technical and verbose, such as PojoIndexingProcessorPropertyNode (just "property" would be enough).

  4. Some types included in the tree don't have a toString() method: MappedIndexManagerImpl, RoutingKeyProvider$$Lambda$578/1859227008, SimpleBeanHolder, PropertyIdentifierMapping, ...

  5. Some types included in the tree are much too verbose in their toString() method. For example LuceneIndexSchemaFieldNode includes the whole parent while it should only include the parent path, and includes the various predicate/sort/projection factories which are not necessary (and they don't have a toString() method anyway).

  6. And maybe more: let's have a closer look.

Example:

Environment

None

Assignee

Yoann Rodière

Reporter

Yoann Rodière

Labels

None

Suitable for new contributors

None

Feedback Requested

None

Fix versions

Priority

Minor
Configure