Paper

Paper's aim is to provide a simple yet fast object storage option for Android. It allows to use Java/Kotlin classes as is: without annotations, factory methods, mandatory class extensions etc. Moreover adding or removing fields to data classes is no longer a pain – all data structure changes are handled automatically.

What's new in 2.6

New API:

Paper.bookOn(path) to set custom storage location;
book.getPath() or book.getPath(key) to get path for content of book or key.

Improvements:

simultaneous read/write for different keys, up to 97% performance gain per thread.
name change: use book.contains(key) instead of deprecated book.exist(key)

Thanks @hiperioncn and @cezar-carneiro for your contribution!

Add dependency

compile 'io.paperdb:paperdb:2.6'

RxJava wrapper for Paper is available as a separate lib RxPaper2. Thanks @pakoito for it!

Initialize Paper

Should be initialized once in Application.onCreate():

Paper.init(context);

Threading

Paper.init() should be called in UI thread;
All other APIs (write, read etc.) are thread-safe and obviously must be called outside of UI thread. Reading/writing for different keys can be done in parallel.

Save

Save any object, Map, List, HashMap etc. including all internal objects. Use your existing data classes as is.

List<Person> contacts = ...
Paper.book().write("contacts", contacts);

Read

Read data objects is as easy as

List<Person> = Paper.book().read("contacts");

the instantiated class is exactly the one used to save data. Limited changes to the class structure are handled automatically. See Handle data class changes.

Use default values if object doesn't exist in the storage.

List<Person> = Paper.book().read("contacts", new ArrayList<>());

Delete

Delete data for one key.

Paper.book().delete("contacts");

Remove all keys for the given Book. Paper.init() must be called prior calling destroy().

Paper.book().destroy();

Use custom book

You can create custom Book with separate storage using

Paper.book("for-user-1").write("contacts", contacts);
Paper.book("for-user-2").write("contacts", contacts);

Each book is located in a separate file folder.

Get all keys

Returns all keys for objects in the book.

List<String> allKeys = Paper.book().getAllKeys();

Handle data structure changes

You can add or remove fields to the class. Then on next read attempt of a new class:

Newly added fields will have their default values.
Removed field will be ignored.

Note: field type changes are not supported.

For example, if you have following data class saved in Paper storage:

class Volcano {
    public String name;
    public boolean isActive;
}

And then you realized you need to change the class like:

class Volcano {
    public String name;
    // public boolean isActive; removed field
    public Location location; // New field
}

the isActive field will be ignored on next read and new location field will have its default value as null.

Exclude fields

Use transient keyword for fields which you want to exclude from saving process.

public transient String tempId = "default"; // Won't be saved

Set storage location for Book instances

By default, all the Paper data files are located with all files belonging to your app, at ../you-app-package-name/files. To save data on SDCard or at any other location you can use new API:

Paper.bookOn("/path/to/the/new/location")
or Paper.bookOn("path/to/the/new/location", "book-for-user-1") to create custom book.

Export/Import

Use Paper.book().getPath() to get path for a folder containing all *.pt files for a given book.
Use Paper.book().getPath(key) to get path for a particular *.pt file containing saved object for a given key. Feel free to copy/rewrite those files for export/import purposes. It's your responsibility to finalize file's export/import operations prior accessing data over Paper API.

Proguard config

Keep your data classes from modification by Proguard:

-keep class your.app.data.model.** { *; }

also you can implement Serializable for all your data classes and keep all of them using:

-keep class * implements java.io.Serializable { *; }

How it works

Paper is based on the following assumptions:

Datasets on mobile devices are small and usually don't have relations in between;
Random file access on flash storage is very fast;

Paper saves each object for given key in a separate file and every write/read operations write/read the whole file.

The Kryo is used for object graph serialization and to provide data compatibility support.

Benchmark results

Running Benchmark on Nexus 4, in ms:

Benchmark	Paper	Hawk
Read/write 500 contacts	187	447
Write 500 contacts	108	221
Read 500 contacts	79	155

Limitations

Circular references are not supported

Apps using Paper

AppDialer – Paper initially has been developed as internal lib to reduce start up time for AppDialer. Currently AppDialer has the best start up time in its class. And simple no-sql-pain data storage layer like a bonus.
Busmap - This application provide all things you need for travelling by bus in Ho Chi Minh city, Vietnam. While the source code is not opened, it is found that the application use Paper internally to manange the bus stop data, route data, time data,... and more.

License

Copyright 2015 Aleksey Masny

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.

yunusemrecetin/Paper