com.zeoflow.depot.solver.query.result.PositionalDataSourceQueryResultBinder.kt Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of depot-compiler Show documentation
Show all versions of depot-compiler Show documentation
The Depot persistence library provides an abstraction layer over SQLite to allow for more robust database access while using the full power of SQLite.
The newest version!
/*
* Copyright (C) 2021 ZeoFlow SRL
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package com.zeoflow.depot.solver.query.result
import com.zeoflow.depot.ext.AndroidTypeNames
import com.zeoflow.depot.ext.CommonTypeNames
import com.zeoflow.depot.ext.L
import com.zeoflow.depot.ext.N
import com.zeoflow.depot.ext.DepotTypeNames
import com.zeoflow.depot.solver.CodeGenScope
import com.squareup.javapoet.FieldSpec
import com.squareup.javapoet.MethodSpec
import com.squareup.javapoet.ParameterSpec
import com.squareup.javapoet.ParameterizedTypeName
import com.squareup.javapoet.TypeName
import com.squareup.javapoet.TypeSpec
import javax.lang.model.element.Modifier
/**
* Used by both Paging2 and Paging3 pipelines.
*
* Paging 3 might create data source on the main thread and does not have any strict
* invalidation rules (e.g. query does not need to return as many records as it initially
* counted).
* Paging 2 always creates data source on a background thread and does not let data source
* return less values then it counted. To achieve that, LimitOffsetDataSource always
* synchronized invalidation trackers to the database, which is not necessary for Paging3.
*
* As a result, we change behavior based on whether we create the data source for paging 3 or 2.
* In practice, [forPaging3] parameter controls whether LimitOffsetDataSource registers its observer
* immediately (paging2) or not (paging3).
*/
class PositionalDataSourceQueryResultBinder(
val listAdapter: ListQueryResultAdapter?,
val tableNames: Set,
val forPaging3: Boolean,
) : QueryResultBinder(listAdapter) {
val itemTypeName: TypeName = listAdapter?.rowAdapter?.out?.typeName ?: TypeName.OBJECT
val typeName: ParameterizedTypeName = ParameterizedTypeName.get(
DepotTypeNames.LIMIT_OFFSET_DATA_SOURCE, itemTypeName
)
override fun convertAndReturn(
depotSQLiteQueryVar: String,
canReleaseQuery: Boolean,
dbField: FieldSpec,
inTransaction: Boolean,
scope: CodeGenScope
) {
// first comma for table names comes from the string since it might be empty in which case
// we don't need a comma. If list is empty, this prevents generating bad code (it is still
// an error to have empty list but that is already reported while item is processed)
val tableNamesList = tableNames.joinToString("") { ", \"$it\"" }
val spec = TypeSpec.anonymousClassBuilder(
"$N, $L, $L, $L $L",
dbField, depotSQLiteQueryVar, inTransaction, !forPaging3, tableNamesList
).apply {
superclass(typeName)
addMethod(createConvertRowsMethod(scope))
}.build()
scope.builder().apply {
addStatement("return $L", spec)
}
}
private fun createConvertRowsMethod(scope: CodeGenScope): MethodSpec =
MethodSpec.methodBuilder("convertRows").apply {
addAnnotation(Override::class.java)
addModifiers(Modifier.PROTECTED)
returns(ParameterizedTypeName.get(CommonTypeNames.LIST, itemTypeName))
val cursorParam = ParameterSpec.builder(AndroidTypeNames.CURSOR, "cursor")
.build()
addParameter(cursorParam)
val resultVar = scope.getTmpVar("_res")
val rowsScope = scope.fork()
listAdapter?.convert(resultVar, cursorParam.name, rowsScope)
addCode(rowsScope.builder().build())
addStatement("return $L", resultVar)
}.build()
}