docs(nitrite): refresh query support boundaries

Update the Nitrite guide to match the current TCK-backed repository and criteria support.

Document common Jakarta Data query semantics, sort/page/limit precedence, LIKE and regex support, property comparisons, logical predicates, and runtime-evaluated computed expressions. Remove stale limitations that listed lower and upper as unsupported now that those expressions are handled.

Replace the stale Jackson mapper section with the current object-mapping contract: Nitrite uses Micronaut Data mapping over Nitrite collections, does not require Nitrite's Jackson mapper module, and treats a Micronaut Serde ObjectMapper as optional.

Author: sfkamath

src/main/docs/guide/nitrite.adoc

@@ -7,7 +7,8 @@ It supports common Micronaut Data repository patterns, including:
 - <<nitriteMapping, Entity Mapping>>
 - <<nitriteAssociationFetching, Association Fetching>>
 - <<nitriteCriteriaSpecifications, Criteria API>>
+- Jakarta Data repository query semantics for common document queries
 - <<nitriteTransactions, Transactions>> (via Micronaut Data TX infrastructure)
 - <<nitriteSpatial, Spatial Queries>>
 
-Nitrite is intended for practical embedded use-cases that need Micronaut Data repositories without running an external datastore. It supports the core repository, mapping, association, criteria, transaction, optimistic locking, and optional spatial features described in this chapter. It is not a feature-equivalent MongoDB replacement; see <<nitriteLimitations>>.
+Nitrite is intended for practical embedded use-cases that need Micronaut Data repositories without running an external datastore. It supports the core repository, mapping, association, criteria, projection, sorting, pagination, transaction, optimistic locking, and optional spatial features described in this chapter. It is not a feature-equivalent MongoDB replacement; see <<nitriteLimitations>>.

src/main/docs/guide/nitrite/nitriteConfiguration.adoc

@@ -64,11 +64,9 @@ The spatial module enables:
 
 See <<nitriteSpatial>> for detailed usage. See the https://nitrite.dizitart.com/java-sdk/modules/spatial/index.html[Nitrite Spatial Module documentation] and https://nitrite.dizitart.com/java-sdk/filter/index.html#spatial-filters[Spatial Filters documentation] for more details.
 
-=== Jackson Mapper
+=== Object Mapping
 
-The Jackson mapper is included by default and provides JSON serialization support. No additional configuration is required.
-
-See the https://nitrite.dizitart.com/java-sdk/modules/jackson/index.html#jackson-mapper[Nitrite Jackson Mapper documentation] for more details.
+Micronaut Data Nitrite uses its own Micronaut Data mapping layer with Nitrite collections. It does not require Nitrite's Jackson mapper module, and a Micronaut Serde `ObjectMapper` bean is optional. If Micronaut Serde is present, the mapper can use it for non-simple nested values; otherwise it relies on Micronaut introspection and the built-in conversion service.
 
 == Batch Operations
 

src/main/docs/guide/nitrite/nitriteCriteriaSpecifications/nitriteCriteriaExecuteQuery.adoc

@@ -25,6 +25,24 @@ The `findAll(Specification, Pageable)` method returns a `Page<T>` with paginatio
 
 snippet::example.PersonRepositoryExample[project-base="doc-examples/nitrite-example", source="main", tags="sorting-pagination-usage"]
 
+Method-name `OrderBy` clauses, dynamic `Sort`, `Pageable` sorting, `Limit`, and offset pagination are supported for repository and criteria queries. When both a static method sort and a dynamic sort are present, Micronaut Data's standard ordering precedence is preserved.
+
+== Predicate and Expression Support
+
+Nitrite supports the common Micronaut Data and Jakarta Data predicate set for document repositories, including:
+
+* equality and inequality
+* comparison operators (`greaterThan`, `greaterThanOrEqualTo`, `lessThan`, `lessThanOrEqualTo`)
+* `between`
+* `in` and `not in`, including empty collections
+* `isNull`, `isNotNull`, `isTrue`, `isFalse`, empty and not-empty string checks
+* nested `and`, `or`, and `not` predicates
+* property-to-property comparisons
+* `like`, `not like`, case-insensitive like, regex predicates, and custom LIKE escape characters
+* string and computed expressions including `length`, `lower`, `upper`, `concat`, `right`, and multiplication (`prod`)
+
+Computed expression predicates are evaluated by the Nitrite runtime against candidate documents when Nitrite has no equivalent native indexed operator. They are useful for repository correctness and Jakarta Data compatibility, but they should not be treated as indexed predicates for large collections.
+
 == Association Queries
 
 Nitrite supports Criteria queries on associations:

src/main/docs/guide/nitrite/nitriteLimitations.adoc

@@ -15,7 +15,7 @@ It is still a Nitrite-backed embedded store, not a MongoDB compatibility layer.
 Micronaut Data Nitrite supports the following features for embedded document storage:
 
 * <<nitriteRepositories, Repository patterns>> with derived query methods
-* <<nitriteCriteriaSpecifications, Criteria API>> with sorting and pagination
+* <<nitriteCriteriaSpecifications, Criteria API>> with sorting, pagination, projection, logical predicates, LIKE/regex matching, and common computed expressions
 * <<nitriteAssociationMapping, Association mapping>> with cascade operations
 * <<nitriteAssociationFetching, Association fetching>> with `@Join` annotation
 * <<nitriteOptimisticLocking, Optimistic locking>> with `@Version` annotation
@@ -30,6 +30,7 @@ For details and examples, see the respective sections above.
 The following boundaries remain:
 
 * `@Join` uses Nitrite's default fetch behavior; custom join types such as `LEFT_FETCH` and `INNER_FETCH` are not currently distinguished.
-* Criteria subqueries (`cb.exists(subquery)`), arithmetic transforms such as `sum` and `diff`, and string transforms such as `lower` and `upper` are not supported. These are shared limitations across the Micronaut Data document-family criteria implementation (Mongo included), not something specific to Nitrite. String-length (`cb.length(...)`) and multiplication (`cb.prod(...)`) expressions *are* supported in Nitrite, but evaluated in application code against each candidate document rather than pushed down (Nitrite has no `$strLenCP`/`$multiply` equivalent), so they are unindexed and scan the collection.
+* Criteria subqueries (`cb.exists(subquery)`) are not supported by Nitrite.
+* Criteria `sum`, `diff`, and unsupported string transforms such as `trim` are rejected. Supported computed expressions such as `length`, `lower`, `upper`, `concat`, `right`, and `prod` are evaluated in application code against candidate documents rather than pushed down to Nitrite indexes, so use them with care on large collections.
 * Advanced Mongo-style aggregation pipelines are out of scope. Use repository aggregate methods, including property-scoped distinct counts, and Criteria API support where applicable.
 * Embedded/nested POJO values are persisted and round-trip correctly, but derived nested-property queries against arbitrary embedded object fields depend on annotation processor support.