Kotlin Legacy→DQL Migration Guide

Overview

This guide will help you successfully migrate your Ditto Kotlin Android application from the legacy query builder APIs to the modern DQL (Ditto Query Language). After reviewing this documentation, you’ll understand how to convert method chaining patterns to DQL syntax and systematically update your data operations.

AI Agent Prompt

Use this prompt when working with an AI coding assistant to migrate your Ditto Kotlin Android app from legacy query builder to DQL.

Copy AI Migration Prompt (Click to Expand)

I need help migrating a Ditto Kotlin Android application from the legacy query builder APIs to modern DQL (Ditto Query Language). This migration involves converting method chaining patterns to SQL-like DQL syntax.

CRITICAL RULES:
1. All query builder method chains (.collection().find()) must be replaced with ditto.store.execute() using DQL
2. Use parameterized queries with :paramName syntax - NEVER string interpolation
3. Counter operations must use PN_INCREMENT BY in APPLY clause - do NOT initialize counter fields
4. Sync subscriptions must use ditto.sync.registerSubscription() instead of .find().subscribe()
5. observeLocal must be replaced with the diff-aware registerObserver overload (`(result, diff)` lambda) — do NOT construct DittoDiffer directly (its constructor is internal in v5)

---

CORE MIGRATION AREAS:

1. QUERY SYNTAX MIGRATION

BEFORE (Legacy Query Builder):
```
ditto.store.collection("cars")
    .find("color == $0", "red")
    .exec()
```

AFTER (DQL):
```
ditto.store.execute(
    "SELECT * FROM cars WHERE color = :color",
    mapOf("color" to "red")
)
```

2. INSERT OPERATIONS

BEFORE (Legacy Query Builder):
```
ditto.store.collection("cars")
    .upsert(mapOf("_id" to id, "color" to "blue"))
```

AFTER (DQL):
```
ditto.store.execute(
    "INSERT INTO cars DOCUMENTS (:car)",
    mapOf("car" to mapOf("_id" to id, "color" to "blue"))
)
```

3. UPDATE OPERATIONS

BEFORE (Legacy Query Builder):
```
ditto.store.collection("cars")
    .findById(id)
    .update { doc -> doc["color"].set("green") }
```

AFTER (DQL):
```
ditto.store.execute(
    "UPDATE cars SET color = :color WHERE _id = :id",
    mapOf("color" to "green", "id" to id)
)
```

4. DELETE OPERATIONS

BEFORE (Legacy Query Builder):
```
ditto.store.collection("cars").findById(id).remove()
```

AFTER (DQL):
```
ditto.store.execute(
    "DELETE FROM cars WHERE _id = :id",
    mapOf("id" to id)
)
```

5. EVICTION OPERATIONS

BEFORE (Legacy Query Builder):
```
ditto.store.collection("cars").findById(id).evict()
```

AFTER (DQL):
```
ditto.store.execute(
    "EVICT FROM cars WHERE _id = :id",
    mapOf("id" to id)
)
```

6. COUNTER OPERATIONS (PN_COUNTER)

BEFORE (Legacy Query Builder):
```
ditto.store.collection("cars")
    .findById(id)
    .update { doc ->
        doc["numUpdates"].counter?.increment(amount = 1.0)
    }
```

AFTER (DQL with PN_INCREMENT):
```
ditto.store.execute(
    "UPDATE cars APPLY numUpdates PN_INCREMENT BY :increment WHERE _id = :id",
    mapOf("increment" to 1, "id" to id)
)
```

IMPORTANT: Do NOT initialize counter fields in documents:
```
// WRONG - Creates a register, not a counter
mapOf("counter" to 0)

// CORRECT - Omit counter field, it's created on first PN_INCREMENT
mapOf("_id" to id, "color" to "blue")
```

7. DOCUMENT FIELD ACCESS MIGRATION

BEFORE (Legacy Query Builder):
```
val document: DittoDocument = collection.findById(res.id).exec()
val color = document.value["color"] as String
```

AFTER (DQL):
```
// store.execute(query, args) returns Unit. Use the handler form to inspect the
// result (auto-closes), or store.executeRaw() and close it yourself.
val color: String = ditto.store.execute(
    "SELECT color FROM cars WHERE _id = :id",
    mapOf("id" to id),
) { result ->
    var item = result.items.first()
    // item.value is DittoCborSerializable.Dictionary, not Map<String, Any>.
    // Use typed accessors — `.string`, `.stringOrNull`, `.int`, `.boolean`, etc.
    item.value["color"].string
}
```

8. LIVE QUERY MIGRATION (observeLocal → registerObserver)

BEFORE (Legacy observeLocal):
```
liveQuery = collection.find("_id.locationId == '${Constants.locationId}'")
    .observeLocal { docs, event ->
        when (event) {
            is DittoLiveQueryEvent.Update -> {
                adapter.delete(event.deletions)
                adapter.insert(event.insertions)
                adapter.update(event.updates)
            }
            is DittoLiveQueryEvent.Initial -> {
                adapter.setInitialCars(docs)
            }
        }
    }
```

AFTER (DQL with the diff-aware `registerObserver` overload):
```
import com.ditto.kotlin.Ditto
import com.ditto.kotlin.DittoStoreObserver

// In v5 the differ is internal — the observer owns it and closes it when the
// observer is closed. Use the `(result, diff)` overload of registerObserver.
class MainFragment : Fragment() {
    private var observer: DittoStoreObserver? = null
    private var previousDocumentIds: MutableList<String> = mutableListOf()

    private fun startLiveQuery() {
        observer = ditto.store.registerObserver(
            query = "SELECT * FROM cars WHERE _id.locationId = :locationId",
            arguments = mapOf("locationId" to Constants.locationId),
        ) { result, diff ->
            // Extract IDs as plain Kotlin strings inside the lambda — never
            // return or store the items themselves (use-after-free).
            val currentDocumentIds = result.items.map { item ->
                val id = item.value["_id"].stringOrNull ?: "unknown"
                item.dematerialize() // release the materialized value
                id
            }

            // Handle deletions using stored IDs from the previous emission
            diff.deletions.forEach { index ->
                previousDocumentIds.getOrNull(index)?.let { deletedId ->
                    // Handle deletion
                }
            }

            // Handle insertions using current IDs
            diff.insertions.forEach { index ->
                currentDocumentIds.getOrNull(index)?.let { insertedId ->
                    // Handle insertion
                }
            }

            // Handle updates using current IDs
            diff.updates.forEach { index ->
                currentDocumentIds.getOrNull(index)?.let { updatedId ->
                    // Handle update
                }
            }

            previousDocumentIds = currentDocumentIds.toMutableList()
        }
    }
}
```

9. SYNC SUBSCRIPTIONS MIGRATION

BEFORE (Legacy Query Builder):
```
val subscription = ditto.store.collection("cars")
    .find("color == $0", "red")
    .subscribe()
```

AFTER (DQL):
```
val subscription = ditto.sync.registerSubscription(
    "SELECT * FROM cars WHERE color = :color",
    mapOf("color" to "red")
)
```

---

COMMON PITFALLS TO AVOID:

1. DQL Syntax Errors
   - Use :paramName for parameters, not $0 or string interpolation

2. Missing Parameter Binding
   - NEVER use string interpolation in queries
   - Always use parameterized queries with mapOf()

3. Counter Type Errors
   - Do NOT initialize counter fields with DittoCounter() or numbers
   - Use PN_INCREMENT BY in APPLY clause
   - Pass negative values for decrements

4. Memory Leaks in Observers
   - Always call item.dematerialize() after extracting data
   - Store only IDs, not full query items

5. Attachment Handling
   - Use ATTACHMENT annotation: "(image ATTACHMENT)"
   - Create attachments with ditto.store.newAttachment()

---

MIGRATION CHECKLIST:

Search for these legacy patterns and replace:
- [ ] .collection(" → ditto.store.execute("SELECT * FROM
- [ ] .find( → Convert to DQL WHERE clause with parameters
- [ ] .findById( → Convert to DQL WHERE _id = :id
- [ ] .upsert( → Convert to DQL INSERT INTO
- [ ] .update( → Convert to DQL UPDATE SET
- [ ] .remove( → Convert to DQL DELETE FROM
- [ ] .evict( → Convert to DQL EVICT FROM
- [ ] .counter?.increment( → Convert to PN_INCREMENT BY in APPLY clause
- [ ] DittoCounter() → Remove initialization, use PN_INCREMENT
- [ ] .observeLocal( → Convert to registerObserver (use the `(result, diff)` overload if you need diffs)
- [ ] .subscribe() → Convert to ditto.sync.registerSubscription()
- [ ] DittoDocument → DittoQueryResultItem
- [ ] DittoSubscription → DittoSyncSubscription
- [ ] item.value["x"] as String → item.value["x"].string (or .stringOrNull); .intOrNull / .booleanOrNull for other types
- [ ] store.execute(query, args) used as a value → use the handler form `store.execute(q, a) { result -> ... }` or `store.executeRaw(q, a)`

---

Please help me convert all legacy query builder patterns in my codebase to DQL syntax. Focus on:
1. Maintaining the same functionality
2. Using proper parameterized queries
3. Handling counter operations correctly with PN_INCREMENT
4. Implementing proper memory management in observers
5. Converting all sync subscriptions to DQL

Start by identifying all uses of .collection() in my codebase and systematically converting each one to the appropriate DQL pattern.

Key API Changes Reference

Query Syntax Migration

Legacy Query Builder → DQL Query Syntax

ditto.store.execute(
    query = "SELECT * FROM cars WHERE color = :color",
    arguments = mapOf("color" to "red")
)

ditto.store.collection("cars")
    .find("color == $0", "red")
    .exec()

Data Operations Migration

Legacy Query Builder → DQL Insert Operations

ditto.store.execute(
    query = "INSERT INTO cars DOCUMENTS (:car)",
    arguments = mapOf("car" to mapOf("_id" to id, "color" to "blue"))
)

ditto.store.collection("cars")
    .upsert(mapOf("_id" to id, "color" to "blue"))

Legacy Query Builder → DQL Update Operations

ditto.store.execute(
    query = "UPDATE cars SET color = :color WHERE _id = :id",
    arguments = mapOf("color" to "green", "id" to id)
)

// Upsert document
ditto.store.collection("cars")
    .upsert(mapOf("_id" to id, "color" to "green"))

// Update with closure
ditto.store.collection("cars")
    .findById(id)
    .update { doc -> doc["color"].set("green") }

Legacy Query Builder → DQL Delete Operations

ditto.store.execute(
    query = "DELETE FROM cars WHERE _id = :id",
    arguments = mapOf("id" to id)
)

ditto.store.collection("cars").findById(id).remove()

Legacy Query Builder → DQL Eviction Operations

// Evict by ID
ditto.store.execute(
    query = "EVICT FROM cars WHERE _id = :id",
    arguments = mapOf("id" to id)
)

// Evict all matching documents
ditto.store.execute(
    query = "EVICT FROM cars WHERE color = :color",
    arguments = mapOf("color" to "red")
)

// Evict by ID
ditto.store.collection("cars").findById(id).evict()

// Evict all matching documents
ditto.store.collection("cars").findAll().evict()

Document Field Access Migration

Legacy Query Builder → Modern Field Access

// store.execute(query, args) returns Unit in v5. Pass a handler to inspect the
// result (auto-closed when the lambda returns), or call store.executeRaw() and
// close the result yourself.
val color: String = ditto.store.execute(
    query = "SELECT color FROM cars WHERE _id = :id",
    arguments = mapOf("id" to id),
) { result ->
    // item.value is DittoCborSerializable.Dictionary. Use typed accessors —
    // e.g. .string / .stringOrNull / .intOrNull — not Kotlin casts.
    result.items.first().value["color"].string
}

val document: DittoDocument = collection.findById(res.id).exec()
val color = document.value["color"] as String

Legacy Query Builder → Modern Document Conversion

import kotlinx.serialization.Serializable
import kotlinx.serialization.json.Json

@Serializable
data class Car(val _id: String, val color: String)

private val carJson = Json { ignoreUnknownKeys = true }

// Inside the handler (which auto-closes the result), decode each item to plain
// Kotlin data via item.jsonString(). Do NOT return DittoQueryResultItem out of
// the lambda — its native lifetime ends when the handler returns.
val cars: List<Car> = ditto.store.execute(
    "SELECT * FROM cars",
) { result ->
    result.items.map { carJson.decodeFromString<Car>(it.jsonString()) }
}

fun documentToCar(doc: DittoDocument): Car {
    return Car(
        id = doc.value["_id"] as String,
        color = doc.value["color"] as String
    )
}

Observer Migration

Legacy Query Builder → DQL Store Observer Migration

import com.ditto.kotlin.DittoStoreObserver
import kotlinx.coroutines.Dispatchers
import kotlinx.coroutines.withContext

// In v5, the differ is internal — the observer owns it. Use the `(result, diff)`
// overload of registerObserver; closing the observer frees the differ.
class MainFragment : Fragment() {
    private var observer: DittoStoreObserver? = null
    private var previousDocumentIds: List<String> = emptyList()

    private fun startLiveQuery() {
        observer?.close()

        observer = ditto.store.registerObserver(
            query = "SELECT * FROM cars WHERE _id.locationId = :locationId",
            arguments = mapOf("locationId" to Constants.locationId),
        ) { result, diff ->
            // Extract IDs as plain Kotlin strings inside the lambda — never store
            // result.items or DittoQueryResultItem instances; they are closed as
            // soon as this lambda returns (use-after-free otherwise).
            val currentDocumentIds = result.items.map { item ->
                val id = item.value["_id"].stringOrNull ?: "unknown"
                item.dematerialize()
                id
            }

            for (index in diff.deletions) {
                previousDocumentIds.getOrNull(index)?.let { deletedId ->
                    println("Deleted car with ID: $deletedId")
                }
            }
            for (index in diff.insertions) {
                currentDocumentIds.getOrNull(index)?.let { insertedId ->
                    println("Inserted car with ID: $insertedId")
                }
            }
            for (index in diff.updates) {
                currentDocumentIds.getOrNull(index)?.let { updatedId ->
                    println("Updated car with ID: $updatedId")
                }
            }

            previousDocumentIds = currentDocumentIds

            // The handler is suspend — use withContext, not runOnUiThread.
            withContext(Dispatchers.Main) {
                // Update your adapter / UI here using `currentDocumentIds` + diff.
            }
        }
    }

    override fun onDestroyView() {
        super.onDestroyView()
        observer?.close()
    }
}

liveQuery = CarsApplication.carsCollection?.find("_id.locationId == '${Constants.locationId}'")?.observeLocal { docs, event ->
    println("Live query handler called. Event: $event Docs count: ${docs.count()}")
    when (event) {
        is DittoLiveQueryEvent.Update -> {
            requireActivity().runOnUiThread {
                val adapter = (viewAdapter as CarsAdapter)
                adapter.set(docs)
                adapter.delete(event.deletions)
                adapter.insert(event.insertions)
                adapter.update(event.updates)
            }
        }
        is DittoLiveQueryEvent.Initial -> {
            requireActivity().runOnUiThread {
                val carTotal = (viewAdapter as CarsAdapter).setInitialCars(docs)
                if (carTotal > 0) {
                    recyclerView.scrollToPosition(carTotal - 1)
                }
            }
        }
    }
}

Performance Consideration: DQL observers provide more advanced return results including aggregates and projections. This requires more database full scans to ensure consistent results compared to the legacy query builder.Use indexes on query fields to maintain and improve observer performance. Indexes ensure your observers remain functional with optimal query performance.

Best Practice: Create Indexes for Observer Queries

// Create index on frequently queried fields
ditto.store.execute("""
    CREATE INDEX idx_cars_locationId
    ON cars (_id.locationId)
""")

// Then register observer - queries will use the index
val observer = ditto.store.registerObserver(
    query = "SELECT * FROM cars WHERE _id.locationId = :locationId",
    arguments = mapOf("locationId" to Constants.locationId),
) { result ->
    // Process result.items inside the lambda; extract plain Kotlin data here
    // — do NOT return result.items or DittoQueryResultItem instances.
}

For more information on creating and managing indexes, see the DQL Indexing documentation.

Sync Subscriptions Migration

Legacy Query Builder → DQL Sync Subscriptions Subscribe with Query

val subscription = ditto.sync.registerSubscription(
    query = "SELECT * FROM cars WHERE color = :color",
    arguments = mapOf("color" to "red")
)

val subscription = ditto.store.collection("cars")
    .find("color == $0", "red")
    .subscribe()

Subscribe with Parameters

val subscription = ditto.sync.registerSubscription(
    query = "SELECT * FROM cars WHERE _id.locationId = :locationId",
    arguments = mapOf("locationId" to Constants.locationId)
)

val subscription = ditto.store.collection("cars")
    .find("_id.locationId == '${Constants.locationId}'")
    .subscribe()

Multiple Subscriptions

val subscriptions = mutableListOf<DittoSyncSubscription>()
subscriptions.add(
    ditto.sync.registerSubscription(
        query = "SELECT * FROM cars WHERE color = :color",
        arguments = mapOf("color" to "red")
    )
)
subscriptions.add(
    ditto.sync.registerSubscription(
        query = "SELECT * FROM cars WHERE year > :year",
        arguments = mapOf("year" to 2020)
    )
)

val subscriptions = mutableListOf<DittoSubscription>()
subscriptions.add(
    ditto.store.collection("cars")
        .find("color == 'red'")
        .subscribe()
)
subscriptions.add(
    ditto.store.collection("cars")
        .find("year > 2020")
        .subscribe()
)

Cancel Subscription

subscription.cancel()

subscription.cancel()

Subscribe to All Documents

val subscription = ditto.sync.registerSubscription(
    "SELECT * FROM cars"
)

val subscription = ditto.store.collection("cars")
    .findAll()
    .subscribe()

Counter Type Migration

PN_COUNTER is the DQL equivalent of the legacy DittoCounter type. When migrating counter operations from the legacy query builder’s .counter?.increment() method, use PN_INCREMENT BY in the APPLY clause. This maintains full compatibility with existing counter data created by DittoCounter.

Counter Increment

ditto.store.execute(
    query = "UPDATE cars APPLY numUpdates PN_INCREMENT BY :increment WHERE _id = :id",
    arguments = mapOf("increment" to 1, "id" to id)
)

ditto.store.collection("cars")
    .findById(id)
    .update { doc ->
        doc["numUpdates"].counter?.increment(amount = 1.0)
    }

Counter Decrement

ditto.store.execute(
    query = "UPDATE cars APPLY viewCount PN_INCREMENT BY :decrement WHERE _id = :id",
    arguments = mapOf("decrement" to -1, "id" to id)
)

ditto.store.collection("cars")
    .findById(id)
    .update { doc ->
        doc["viewCount"].counter?.increment(amount = -1.0)
    }

Initialize Counter in Document

// Counter fields are automatically created on first PN_INCREMENT use
ditto.store.execute(
    query = "INSERT INTO cars DOCUMENTS (:car)",
    arguments = mapOf("car" to mapOf(
        "_id" to id,
        "color" to "blue"
        // Do NOT initialize counter fields - they are created on first PN_INCREMENT
    ))
)

// Then use PN_INCREMENT with APPLY clause to create and increment the counter
ditto.store.execute(
    query = "UPDATE cars APPLY numUpdates PN_INCREMENT BY 1 WHERE _id = :id",
    arguments = mapOf("id" to id)
)

val carData = mapOf(
    "_id" to id,
    "color" to "blue",
    "numUpdates" to DittoCounter()
)
ditto.store.collection("cars").upsert(carData)

Multiple Counter Operations

ditto.store.execute(
    """UPDATE cars
       APPLY likes PN_INCREMENT BY :likeIncrement,
             dislikes PN_INCREMENT BY :dislikeDecrement,
             views PN_INCREMENT BY :viewIncrement
       WHERE _id = :id""",
    mapOf(
        "likeIncrement" to 1,
        "dislikeDecrement" to -1,
        "viewIncrement" to 1,
        "id" to id
    )
)

ditto.store.collection("cars")
    .findById(id)
    .update { doc ->
        doc["likes"].counter?.increment(amount = 1.0)
        doc["dislikes"].counter?.increment(amount = -1.0)
        doc["views"].counter?.increment(amount = 1.0)
    }

Attachment Operations with DQL

Attachment Creation and Storage

// Create attachment using store
val attachment = ditto.store.newAttachment(
    inputStream = fileInputStream,
    metadata = metadata,
)

// Store attachment with DQL — build the document inline as a plain map,
// with the attachment as the value of the ATTACHMENT-typed column.
ditto.store.execute(
    query = "INSERT INTO COLLECTION cars (image ATTACHMENT) DOCUMENTS (:doc)",
    arguments = mapOf("doc" to mapOf("_id" to "car-1", "image" to attachment)),
)

Attachment Fetching

// Fetch attachment with progress callback
val fetchResult = ditto.store.fetchAttachment(
    token = attachmentToken,
    onFetchProgress = { progress ->
        updateProgress(progress.downloadedBytes, progress.totalBytes)
    }
)

Performance Enhancements

Indexes for Improved Query Performance

DQL observers and queries benefit significantly from proper indexing. When migrating from the legacy query builder to DQL, creating indexes on frequently queried fields is essential for maintaining optimal performance. Why Indexes Matter for DQL:

DQL observers support advanced features like aggregates and projections
These advanced features require full database scans to ensure consistent results
Indexes dramatically reduce query execution time by avoiding full scans
Combining indexes with observers provides better performance than legacy query builder

Creating Indexes:

// Create index on single field
ditto.store.execute("""
    CREATE INDEX idx_cars_color
    ON cars (color)
""")

// Create compound index on multiple fields
ditto.store.execute("""
    CREATE INDEX idx_cars_color_year
    ON cars (color, year)
""")

// Create index on nested field
ditto.store.execute("""
    CREATE INDEX idx_cars_location
    ON cars (_id.locationId)
""")

Best Practices:

Create indexes on fields used in WHERE clauses
Create indexes before registering observers for those queries
Use compound indexes for queries with multiple filter conditions
Monitor query performance and add indexes as needed

For comprehensive information on indexing strategies, syntax, and best practices, see the DQL Indexing documentation.

Common Pitfalls to Avoid

1. DQL Syntax Errors

Use :paramName for parameters, not $0 or string interpolation.

// ❌ Wrong: String interpolation
val color = "red"
ditto.store.execute("SELECT * FROM cars WHERE color = '$color'")

// ✅ Correct: Using :paramName with mapOf()
ditto.store.execute(
    query = "SELECT * FROM cars WHERE color = :color",
    arguments = mapOf("color" to "red")
)

2. Missing Parameter Binding

NEVER use string interpolation in queries. Always use parameterized queries with Map<String, Any>.

// ❌ Wrong: Direct string interpolation
val locationId = "loc_123"
ditto.store.execute(
    "SELECT * FROM cars WHERE _id.locationId = '$locationId'"
)

// ✅ Correct: Parameterized query
ditto.store.execute(
    query = "SELECT * FROM cars WHERE _id.locationId = :locationId",
    arguments = mapOf("locationId" to locationId)
)

3. Counter Type Errors

Use COUNTER annotation in collection definitions. Do NOT use SET with COUNTER fields. Use APPLY with PN_INCREMENT BY. Pass negative values for decrements.

// ❌ Wrong: Initializing counter with a number (creates REGISTER, not COUNTER)
ditto.store.execute(
    query = "INSERT INTO items DOCUMENTS (:doc)",
    arguments = mapOf("doc" to mapOf("counter" to 0, "_id" to id))
)

// ❌ Wrong: Using SET on counter field
ditto.store.execute(
    query = "UPDATE items SET counter = 5 WHERE _id = :id",
    arguments = mapOf("id" to id)
)

// ✅ Correct: Use PN_INCREMENT BY with APPLY clause (creates counter on first use)
ditto.store.execute(
    query = "UPDATE COLLECTION items (counter COUNTER) APPLY counter PN_INCREMENT BY :value WHERE _id = :id",
    arguments = mapOf("value" to 1, "id" to id)
)

// ✅ Correct: Decrement by passing negative value
ditto.store.execute(
    query = "UPDATE items APPLY counter PN_INCREMENT BY :value WHERE _id = :id",
    arguments = mapOf("value" to -1, "id" to id)
)

4. Memory Management with Observers

The DittoQueryResult passed to your handler — and every DittoQueryResultItem it owns — is closed as soon as your lambda returns. Never return result.items out of the lambda or store individual items in a field; doing so will throw IllegalStateException: Query result item has already been closed. Use the (result, diff) overload of registerObserver when you need a diff — the differ is managed by the observer in v5; you do not construct one yourself.

// ❌ Wrong: storing DittoQueryResultItem instances outside the lambda
var items: List<DittoQueryResultItem> = emptyList()
val observer = ditto.store.registerObserver(
    query = "SELECT * FROM cars"
) { result ->
    items = result.items  // CRASH on next emission: items are already closed
}

// ✅ Correct: extract plain Kotlin data from *only* the items pointed to by
// the diff, and maintain your own mirror across emissions. Never return
// DittoQueryResultItem instances out of the lambda. Close the observer in onDestroy
class MainFragment : Fragment() {
    // Local mirror of the observed list — updated incrementally via diffs.
    private val documentIds: MutableList<String> = mutableListOf()
    private var observer: DittoStoreObserver? = null

    private fun startLiveQuery() {
        observer = ditto.store.registerObserver(
            query = "SELECT * FROM cars",
        ) { result, diff ->
            // Only touch items pointed to by the diff
            diff.deletions.sortedDescending().forEach { documentIds.removeAt(it) }
            diff.insertions.sorted().forEach { i ->
                val item = result.items[i]
                documentIds.add(i, item.value["_id"].stringOrNull ?: "unknown")
                item.dematerialize()
            }
            diff.updates.forEach { i ->
                val item = result.items[i]
                documentIds[i] = item.value["_id"].stringOrNull ?: "unknown"
                item.dematerialize()
            }
        }
    }

    override fun onDestroyView() {
        super.onDestroyView()
        observer?.close()
    }
}

5. Attachment Handling

Use ATTACHMENT annotation in collection definitions. Create attachments with ditto.store.newAttachment().

// ❌ Wrong: Missing ATTACHMENT annotation — the attachment is inserted as
// an opaque nested map, not an attachment-typed column.
ditto.store.execute(
    query = "INSERT INTO cars DOCUMENTS (:doc)",
    arguments = mapOf("doc" to mapOf("_id" to id, "image" to attachment))
)

// ✅ Correct: Use ATTACHMENT annotation in COLLECTION definition
val attachment = ditto.store.newAttachment(
    inputStream = fileInputStream,
    metadata = metadata
)

ditto.store.execute(
    query = "INSERT INTO COLLECTION cars (image ATTACHMENT) DOCUMENTS (:doc)",
    arguments = mapOf("doc" to mapOf(
        "_id" to id,
        "image" to attachment
    ))
)

Overview & Setup

SDKs

Accessing Data

Syncing Data

Authentication and Authorization

Managing the Mesh

Deploying your App

Kotlin Legacy→DQL Migration Guide

Overview

AI Agent Prompt

Key API Changes Reference

Query Syntax Migration

Data Operations Migration

Document Field Access Migration

Observer Migration

Sync Subscriptions Migration

Counter Type Migration

Attachment Operations with DQL

Performance Enhancements

Indexes for Improved Query Performance

Common Pitfalls to Avoid

1. DQL Syntax Errors

2. Missing Parameter Binding

3. Counter Type Errors

4. Memory Management with Observers

5. Attachment Handling

​Overview

​AI Agent Prompt

​Key API Changes Reference

​Query Syntax Migration

​Data Operations Migration

​Document Field Access Migration

​Observer Migration

​Sync Subscriptions Migration

​Counter Type Migration

​Attachment Operations with DQL

​Performance Enhancements

​Indexes for Improved Query Performance

​Common Pitfalls to Avoid

​1. DQL Syntax Errors

​2. Missing Parameter Binding

​3. Counter Type Errors

​4. Memory Management with Observers

​5. Attachment Handling

Overview

AI Agent Prompt

Key API Changes Reference

Query Syntax Migration

Data Operations Migration

Document Field Access Migration

Observer Migration

Sync Subscriptions Migration

Counter Type Migration

Attachment Operations with DQL

Performance Enhancements

Indexes for Improved Query Performance

Common Pitfalls to Avoid

1. DQL Syntax Errors

2. Missing Parameter Binding

3. Counter Type Errors

4. Memory Management with Observers

5. Attachment Handling