Beyond SQLite & Core Data NoSQL for Mobile Sam Ritchie
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Beyond SQLite & Core Data
NoSQL for Mobile
Sam Ritchie
App Data Storage
iOS Data Storage
0%
5%
10%
15%
20%
Core Data SQLite FMDB
Android Data Storage
0%
5%
10%
15%
20%
SQLite ormlite greenDAO
So What?
Logical model
Arbitrary queries
Concurrency
Tool support
Existing skills
!
!
!
!
Replication
Full Text indexing
Spatial indexing
Schema upgrades
Object-Relational impedance
Relational DatabasesDatabases
NoSQL
Order
Order
Order
Customer
Customer
Customer
Order Line
Order Line
Order Line
Product
Product
Product
Order
Customer
Order Line
Product
Order Line
Product
Order Line
Product
Order
Order
Order
Customer
Customer
Customer
Order Line
Order Line
Order Line
Product
Product
Product
Order
Order
Order
Customer
Customer
Customer
Order Line
Order Line
Order Line
Product
Product
Product
Order Line
OrderOrder
Customer
Customer
Product
Order +
Order
Order Line
Order Line
Order
Order Line
Product
Order Line
Product
Order Line
Product
Customer
Order
Customer
Order Line
Product
Order Line
Product
Order Line
ProductProduct
Customer
!
Options?
The File System
YapDatabase
Benchmarks
{ title : …, body : …, createdDate : …, ownerId : “users/1234”, ownerDisplayName : “Sam”, … answers : [ { score : …, body : …, ownerId : “users/4567”, ownerDisplayName : “Jon Skeet”, comments : [] }], comments : [ { score : …, text : …, ownerId : “users/1234”, ownerDisplayName : “Sam” }], tags : [ “objective-c”, “iOS” ], duplicates : [ “questions/456” ], related : [] }
questions/123 users/1234{ displayName : “Sam” reputation : 1337, profileImageURL : …, websiteURL : …, age : …, location : …, creationDate : …, lastAccessDate : … }
Fetch Tag Counts Full Text Search LOC
Core Data CouchBase Lite YapDatabase LevelDB
Awesome
Crap
Cost of Abstraction
CBLQuery *query = [[self.database viewNamed:@"tags"] createQuery]; query.groupLevel = 1; for (CBLQueryRow *row in [query run:NULL]) NSLog(@"%@ (%@)", row.key, row.value);
Couchbase Lite
__block NSString *currentTag = nil; __block int count = 0; [database enumerateKeysWithPrefix:@"i/tags/" block:^(NSString *key, BOOL *stop) { NSString *tag = [[key stringByDeletingLastPathComponent] stringByReplacingOccurrencesOfString:@"i/tags/" withString:@""]; if (!currentTag || ![currentTag isEqualToString:tag]) { NSLog(@"%@ (%d)", currentTag, count); count = 0; currentTag = tag; } count++; }]; NSLog(@"%@ (%d)", currentTag, count);
LevelDB
Cost of Abstraction
select name, count(1) from Tags group by name
NSFetchRequest* fetch = [NSFetchRequest fetchRequestWithEntityName:@"Tag"]; NSEntityDescription* entity = [NSEntityDescription entityForName:@"Tag" inManagedObjectContext:self.context]; NSAttributeDescription* nameDesc = [entity.attributesByName objectForKey:@"name"]; NSExpression *keyPathExpression = [NSExpression expressionForKeyPath: @"name"]; NSExpression *countExpression = [NSExpression expressionForFunction: @"count:" arguments: [NSArray arrayWithObject:keyPathExpression]]; NSExpressionDescription *expressionDescription = [[NSExpressionDescription alloc] init]; [expressionDescription setName: @"count"]; [expressionDescription setExpression: countExpression]; [expressionDescription setExpressionResultType: NSInteger32AttributeType]; [fetch setPropertiesToFetch:[NSArray arrayWithObjects:nameDesc, expressionDescription, nil]]; [fetch setPropertiesToGroupBy:[NSArray arrayWithObject:nameDesc]]; [fetch setResultType:NSDictionaryResultType]; NSError* error = nil; NSArray *results = [self.context executeFetchRequest:fetch error:&error]; for (NSDictionary *result in results) NSLog(@"%@ (%@)", result[@"name"], result[@"count"]);
Core Data
Live Coding Demo
Related Documents
![T FS: Enhancing Metadata Efficiency in the Local File System · employed in our experiments: Ext4, XFS, BTRFS and LevelDB. 2.1 Local File System Structures Ext4[26] is the fourth](https://static.cupdf.com/doc/110x72/6021f741a6d11b4e4d0f33bf/t-fs-enhancing-metadata-eficiency-in-the-local-file-employed-in-our-experiments.jpg)
![SILK: Preventing Latency Spikes in Log-Structured Merge Key … · 2019-07-10 · structured merge key-value stores (LSM KVs), such as RocksDB [18], LevelDB [14] and Cassandra [30],](https://static.cupdf.com/doc/110x72/5f831932c58eab702e4e843e/silk-preventing-latency-spikes-in-log-structured-merge-key-2019-07-10-structured.jpg)
