Getting Started

If you haven't set up your project yet, please head over to the QuickStart guide to get up and running. You can also check out our API Reference for more detailed information about our SDK.

The Parse platform provides a complete backend solution for your mobile application. Generally, our goal is to eliminate the need for managing servers and writing server-side code. Our PHP SDK is for those applications and situations where server-side code is necessary, or simply preferred.

The Parse PHP SDK is fully open source, and anyone can contribute to make it better, or make their own changes if necessary. Check out the GitHub repository for more information.

The Parse PHP SDK requires version 5.4 or greater of the PHP runtime.

On Parse, you create an App for each of your applications. Each App has its own application id and REST API key that you apply to your web app. Your account on Parse can accommodate multiple Apps. This is useful even if you have one application, since you can deploy different versions for test and production.

Want to contribute to this doc? Edit this section.
Was this section helpful? NO YES Thanks for your feedback, we’re always working to make our docs better.

Objects

ParseObject

Storing data on Parse is built around the ParseObject. Each ParseObject contains key-value pairs of JSON-compatible data. This data is schemaless, which means that you don't need to specify ahead of time what keys exist on each ParseObject. You simply set whatever key-value pairs you want, and our backend will store it.

For example, let's say you're tracking high scores for a game. A single ParseObject could contain:

score: 1337, playerName: "Sean Plott", cheatMode: false

Keys must be alphanumeric strings. Values can be strings, numbers, booleans, or even sequential arrays and associative arrays - anything that can be JSON-encoded. Note however that Arrays and Associative Arrays require separate methods to set them on a ParseObject.

Saving Objects

Let's say you want to save the GameScore described above to the Parse Cloud. The interface is similar to a our other SDKs, including the save method:

$gameScore = new ParseObject("GameScore");

$gameScore->set("score", 1337);
$gameScore->set("playerName", "Sean Plott");
$gameScore->set("cheatMode", false);

try {
  $gameScore->save();
  echo 'New object created with objectId: ' . $gameScore->getObjectId();
} catch (ParseException $ex) {  
  // Execute any logic that should take place if the save fails.
  // error is a ParseException object with an error code and message.
  echo 'Failed to create new object, with error message: ' . $ex->getMessage();
}

After this code runs, you will probably be wondering if anything really happened. To make sure the data was saved, you can look at the Data Browser in your app on Parse. You should see something like this:

objectId: "xWMyZ4YEGZ", score: 1337, playerName: "Sean Plott", cheatMode: false,
createdAt:"2011-06-10T18:33:42Z", updatedAt:"2011-06-10T18:33:42Z"

There are two things to note here. You didn't have to configure or set up a new Class called GameScore before running this code. Your Parse app lazily creates this Class for you when it first encounters it.

There are also a few fields you don't need to specify that are provided as a convenience. objectId is a unique identifier for each saved object. createdAt and updatedAt represent the time that each object was created and last modified in the cloud. Each of these fields is filled in by Parse, so they don't exist on a ParseObject until a save operation has completed.

Retrieving Objects

Saving data to the cloud is fun, but it's even more fun to get that data out again. If you have the objectId, you can retrieve the whole ParseObject using a ParseQuery:

$query = new ParseQuery("GameScore");
try {
  $gameScore = $query->get("xWMyZ4YEGZ");
  // The object was retrieved successfully.
} catch (ParseException $ex) {
  // The object was not retrieved successfully.
  // error is a ParseException with an error code and message.
}

To get the values out of the ParseObject, use the get method.

$score = $gameScore->get("score");
$playerName = $gameScore->get("playerName");
$cheatMode = $gameScore->get("cheatMode");

The three special values are provided as the result of methods:

$objectId = $gameScore->getObjectId();
$updatedAt = $gameScore->getUpdatedAt();
$createdAt = $gameScore->getCreatedAt();

If you need to refresh an object you already have with the latest data that is in the Parse Cloud, you can call the fetch method like so:

$gameScore->fetch();

Updating Objects

Updating an object is simple. Just set some new data on it and call the save method. For example:

// Create the object.
$gameScore = new ParseObject("GameScore");

$gameScore->set("score", 1337);
$gameScore->set("playerName", "Sean Plott");
$gameScore->set("cheatMode", false);
$gameScore->setArray("skills", ["pwnage", "flying"]);

$gameScore->save();
// Now let's update it with some new data. In this case, only cheatMode and score
// will get sent to the cloud. playerName hasn't changed.
$gameScore->set("cheatMode", true);
$gameScore->set("score", 1338);
$gameScore->save();

Parse automatically figures out which data has changed so only "dirty" fields will be sent to the Parse Cloud. You don't need to worry about squashing data that you didn't intend to update.

Counters

The above example contains a common use case. The "score" field is a counter that we'll need to continually update with the player's latest score. Using the above method works but it's cumbersome and can lead to problems if you have multiple clients trying to update the same counter.

To help with storing counter-type data, Parse provides methods that atomically increment (or decrement) any number field. So, the same update can be rewritten as:

$gameScore->increment("score");
$gameScore->save();

You can also increment by any amount by passing in a second argument to increment. When no amount is specified, 1 is used by default.

Arrays

To help with storing array data, there are three operations that can be used to atomically change an array associated with a given key:

  • add append the given object to the end of an array field.
  • addUnique add the given object only if it isn't already contained in an array field. The position of the insert is not guaranteed.
  • remove remove all instances of the given object from an array field.

For example, we can add items to the set-like "skills" field like so:

$gameScore->addUnique("skills", ["flying"]);
$gameScore->addUnique("skills", ["kungfu"]);
$gameScore->save();

Note that it is not currently possible to atomically add and remove items from an array in the same save. You will have to call save in between every different kind of array operation.

Destroying Objects

To delete an object from the cloud:

$gameScore->destroy();

You can delete a single field from an object with the delete method:

// After this, the playerName field will be empty
$gameScore->delete("playerName");

// Saves the field deletion to the Parse Cloud
$gameScore->save();

Relational Data

Objects may have relationships with other objects. For example, in a blogging application, a Post object may have many Comment objects. Parse supports all kind of relationships, including one-to-one, one-to-many, and many-to-many.

One-to-One and One-to-Many Relationships

One-to-one and one-to-many relationships are modeled by saving a ParseObject as a value in the other object. For example, each Comment in a blogging app might correspond to one Post.

To create a new Post with a single Comment, you could write:

// Create the post
$myPost = new ParseObject("Post");
$myPost->set("title", "I'm Hungry");
$myPost->set("content", "Where should we go for lunch?");

// Create the comment
$myComment = new ParseObject("Comment");
$myComment->set("content", "Let's do Sushirrito.");

// Add the post as a value in the comment
$myComment->set("parent", $myPost);

// This will save both myPost and myComment
$myComment->save();

Internally, the Parse framework will store the referred-to object in just one place, to maintain consistency. You can also link objects using just their objectIds like so:

$post = new ParseObject("Post", "1zEcyElZ80");

$myComment->set("parent", $post);

By default, when fetching an object, related ParseObjects are not fetched. These objects' values cannot be retrieved until they have been fetched like so:

$post = $fetchedComment->get("parent");
$post->fetch();
$title = $post->get("title");

Many-to-Many Relationships

Many-to-many relationships are modeled using ParseRelation. This works similar to storing an array of ParseObjects in a key, except that you don't need to fetch all of the objects in a relation at once. In addition, this allows ParseRelation to scale to many more objects than the array of ParseObject approach. For example, a User may have many Posts that she might like. In this case, you can store the set of Posts that a User likes using relation. In order to add a Post to the "likes" array of the User, you can do:

$user = ParseUser::getCurrentUser();
$relation = $user->getRelation("likes");
$relation->add($post);
$user->save();

You can remove a post from a ParseRelation:

$relation->remove($post);
$user->save();

You can call add and remove multiple times before calling save:

$relation->remove($post1);
$relation->remove($post2);
$user->save();

You can also pass in an array of ParseObject to add and remove:

$relation->add([$post1, $post2, $post3]);
$user->save();

By default, the array of objects in this relation are not downloaded. You can get an array of the posts that a user likes by using the ParseQuery returned by getQuery. The code looks like:

$postsLiked = $relation->getQuery()->find();
// $postsLiked contains the posts that the current user likes.

If you want only a subset of the Posts, you can add extra constraints to the ParseQuery returned by query like this:

$query = $relation->getQuery();
$query->equalTo("title", "I'm Hungry");
$postsLiked = $query->find();
// $postsLiked contains post liked by the current user which have the title "I'm Hungry".

For more details on ParseQuery, please look at the query portion of this guide. A ParseRelation behaves similar to an array of ParseObject for querying purposes, so any query you can do on an array of objects, you can do on a ParseRelation.

Data Types

So far we've used values with type String, Integer, and ParseObject. Parse also supports PHP DateTimes and null. You can nest PHP arrays and associative arrays (JSON Objects) to store more structured data within a single ParseObject. Overall, the following types are allowed for each field in your object:

  • String => String
  • Number => Integer and float
  • Bool => Boolean
  • Array => PHP arrays
  • Object => associative arrays (JSON Objects)
  • Date => DateTime
  • File => ParseFile
  • Pointer => other ParseObject
  • Relation => ParseRelation
  • Null => null

Some examples:

$number = 42;
$string = "the number is " . $number;
$date = new DateTime();
$array = [$string, $number];
$object = ["number" => $number, "string" => $string];

$bigObject = new ParseObject("BigObject");
$bigObject->set("myNumber", $number);
$bigObject->set("myString", $string);
$bigObject->set("myDate", $date);
$bigObject->setArray("myArray", $array);
$bigObject->setAssociativeArray("myObject", $object);
$bigObject->set("anyKey", null); // this value can only be saved to an existing key
$bigObject->save();

We do not recommend storing large pieces of binary data like images or documents on ParseObject. ParseObjects should not exceed 128 kilobytes in size. We recommend you use ParseFiles to store images, documents, and other types of files. You can do so by instantiating a ParseFile object and setting it on a field. See Files for more details.

For more information about how Parse handles data, check out our documentation on Data.

Subclassing ParseObject

Each ParseObject is an instance of a specific subclass with a class name that you can use to distinguish different sorts of data. For example, we could call the high score object a GameScore. We recommend that you NameYourClassesLikeThis and nameYourKeysLikeThis, just to keep your code looking pretty.

To create a new subclass, create a new class which extends the ParseObject class, add the $parseClassName static property, and call the registerSubclass method before use. Any ParseQuery will return instances of the new class for any ParseObject with the same class name.

class GameScore extends ParseObject
{
  public static $parseClassName = "GameScore";
}
// Do this once, at the start of your app, before ParseClient::initialize(...);
GameScore::registerSubclass();

// Create a new instance of that class.
$gameScore = new GameScore();

You can add additional methods and properties to your subclasses of ParseObject.

// A complex subclass of ParseObject
class Monster extends ParseObject
{
  public static $parseClassName = "Monster";

  public function hasSuperHumanStrength() {
    return this->get("strength") > 18;
  }

  public static function spawn($strength) {
    $monster = new Monster();
    $monster->set("strength", $strength);
    return $monster;
  }
}
$monster = Monster::spawn(200);
echo monster->strength();  // Displays 200.
echo monster->hasSuperHumanStrength();  // Displays true.

If you want to override the __construct method, make sure the first three params are exactly as same as the parent ParseObject constructor:

class GameScore extends ParseObject
{
  public static $parseClassName = "GameScore";

  public function __construct($className = null, $objectId = null, $isPointer = false, $another_param) {
    parent::__construct("GameScore", $objectId, $isPointer);
    // ...
  }
}
Want to contribute to this doc? Edit this section.
Was this section helpful? NO YES Thanks for your feedback, we’re always working to make our docs better.

Queries

We've already seen how a ParseQuery with get can retrieve a single ParseObject from Parse. There are many other ways to retrieve data with ParseQuery - you can retrieve many objects at once, put conditions on the objects you wish to retrieve, and more.

Basic Queries

In many cases, get isn't powerful enough to specify which objects you want to retrieve. ParseQuery offers different ways to retrieve an array of objects rather than just a single object.

The general pattern is to create a ParseQuery, put conditions on it, and then retrieve an Array of matching ParseObjects using find. For example, to retrieve of the scores with a particular playerName, use the equalTo method to constrain the value for a key.

$query = new ParseQuery("GameScore");
$query->equalTo("playerName", "Dan Stemkoski");
$results = $query->find();
echo "Successfully retrieved " . count($results) . " scores.";
// Do something with the returned ParseObject values
for ($i = 0; $i < count($results); $i++) {
  $object = $results[$i];
  echo $object->getObjectId() . ' - ' . $object->get('playerName');
}

Query Constraints

There are several ways to put constraints on the objects found by a ParseQuery. You can filter out objects with a particular key-value pair with notEqualTo:

$query->notEqualTo("playerName", "Michael Yabuti");

You can give multiple constraints, and objects will only be in the results if they match all of the constraints. In other words, it's like an AND of constraints.

$query->notEqualTo("playerName", "Michael Yabuti");
$query->greaterThan("playerAge", 18);

You can limit the number of results by setting limit. By default, results are limited to 100, but anything from 1 to 1000 is a valid limit:

$query->limit(10); // limit to at most 10 results

If you want exactly one result, a more convenient alternative may be to use first instead of using find.

$query = new ParseQuery("GameScore");
$query->equalTo("playerEmail", "dstemkoski@example.com");
$object = $query->first();

You can skip the first results by setting skip. This can be useful for pagination, though it is limited to a maximum of ten thousand:

$query->skip(10); // skip the first 10 results

For sortable types like numbers and strings, you can control the order in which results are returned:

// Sorts the results in ascending order by the score field
$query->ascending("score");

// Sorts the results in descending order by the score field
$query->descending("score");

For sortable types, you can also use comparisons in queries:

// Restricts to wins < 50
$query->lessThan("wins", 50);

// Restricts to wins <= 50
$query->lessThanOrEqualTo("wins", 50);

// Restricts to wins > 50
$query->greaterThan("wins", 50);

// Restricts to wins >= 50
$query->greaterThanOrEqualTo("wins", 50);

If you want to retrieve objects matching several different values, you can use containedIn, providing an array of acceptable values. This is often useful to replace multiple queries with a single query. For example, if you want to retrieve scores made by any player in a particular array:

// Finds scores from any of Jonathan, Dario, or Shawn
$query->containedIn("playerName",
                  ["Jonathan Walsh", "Dario Wunsch", "Shawn Simon"]);

If you want to retrieve objects that do not match any of several values you can use notContainedIn, providing an array of acceptable values. For example if you want to retrieve scores from players besides those in an array:

// Finds scores from anyone who is neither Jonathan, Dario, nor Shawn
$query->notContainedIn("playerName",
                     ["Jonathan Walsh", "Dario Wunsch", "Shawn Simon"]);

If you want to retrieve objects that have a particular key set, you can use exists. Conversely, if you want to retrieve objects without a particular key set, you can use doesNotExist.

// Finds objects that have the score set
$query->exists("score");

// Finds objects that don't have the score set
$query->doesNotExist("score");

You can use the matchesKeyInQuery method to get objects where a key matches the value of a key in a set of objects resulting from another query. For example, if you have a class containing sports teams and you store a user's hometown in the user class, you can issue one query to find the array of users whose hometown teams have winning records. The query would look like:

$teamQuery = new ParseQuery("Team");
$teamQuery->greaterThan("winPct", 0.5);
$userQuery = ParseUser::query();
$userQuery->matchesKeyInQuery("hometown", "city", $teamQuery);
$results = $userQuery->find();
// results has the array of users with a hometown team with a winning record

Conversely, to get objects where a key does not match the value of a key in a set of objects resulting from another query, use doesNotMatchKeyInQuery. For example, to find users whose hometown teams have losing records:

$losingUserQuery = ParseUser::query();
$losingUserQuery->doesNotMatchKeyInQuery("hometown", "city", teamQuery);
$results = $losingUserQuery->find();
// results has the array of users with a hometown team with a losing record

You can restrict the fields returned by calling select with an array of keys. To retrieve documents that contain only the score and playerName fields (and also special built-in fields such as objectId, createdAt, and updatedAt):

$query = new ParseQuery("GameScore");
$query->select(["score", "playerName"]);
$results = $query->find();
// each of results will only have the selected fields available.

The remaining fields can be fetched later by calling fetch on the returned objects:

$result = $query->first();
// only the selected fields of the object will now be available here.
$result->fetch();
// all fields of the object will now be available here.

Queries on Array Values

For keys with an array type, you can find objects where the key's array value contains 2 by:

// Find objects where the array in arrayKey contains 2.
$query->equalTo("arrayKey", 2);

You can also find objects where the key's array value contains each of the elements 2, 3, and 4 with the following:

// Find objects where the array in arrayKey contains all of the elements 2, 3, and 4.
$query->containsAll("arrayKey", [2, 3, 4]);

Queries on String Values

If you're trying to implement a generic search feature, we recommend taking a look at this blog post: Implementing Scalable Search on a NoSQL Backend.

Use startsWith to restrict to string values that start with a particular string. Similar to a MySQL LIKE operator, this is indexed so it is efficient for large datasets:

// Finds barbecue sauces that start with "Big Daddy's".
$query = new ParseQuery("BarbecueSauce");
$query->startsWith("name", "Big Daddy's");

The above example will match any BarbecueSauce objects where the value in the "name" String key starts with "Big Daddy's". For example, both "Big Daddy's" and "Big Daddy's BBQ" will match, but "big daddy's" or "BBQ Sauce: Big Daddy's" will not.

Queries that have regular expression constraints are very expensive, especially for classes with over 100,000 records. Parse restricts how many such operations can be run on a particular app at any given time.

Relational Queries

There are several ways to issue queries for relational data. If you want to retrieve objects where a field matches a particular ParseObject, you can use equalTo just like for other data types. For example, if each Comment has a Post object in its post field, you can fetch comments for a particular Post:

// Assume ParseObject $myPost was previously created.
$query = new ParseQuery("Comment");
$query->equalTo("post", $myPost);
$comments = $query->find();
// comments now contains the comments for myPost

If you want to retrieve objects where a field contains a ParseObject that matches a different query, you can use matchesQuery. Note that the default limit of 100 and maximum limit of 1000 apply to the inner query as well, so with large data sets you may need to construct queries carefully to get the desired behavior. In order to find comments for posts containing images, you can do:

$innerQuery = new ParseQuery("Post");
$innerQuery->exists("image");
$query = new ParseQuery("Comment");
$query->matchesQuery("post", $innerQuery);
$comments = $query->find();
// comments now contains the comments for posts with images.

If you want to retrieve objects where a field contains a ParseObject that does not match a different query, you can use doesNotMatchQuery. In order to find comments for posts without images, you can do:

$innerQuery = new ParseQuery("Post");
$innerQuery->exists("image");
$query = new ParseQuery("Comment");
$query->doesNotMatchQuery("post", $innerQuery);
$query->find();
// comments now contains the comments for posts without images.

You can also do relational queries by objectId:

$post = new ParseObject("Post", "1zEcyElZ80");
$query->equalTo("post", $post);

In some situations, you want to return multiple types of related objects in one query. You can do this with the include method. For example, let's say you are retrieving the last ten comments, and you want to retrieve their related posts at the same time:

$query = new ParseQuery("Comment");

// Retrieve the most recent ones
query->descending("createdAt");

// Only retrieve the last ten
query->limit(10);

// Include the post data with each comment
query->includeKey("post");

$comments = query->find();
// Comments now contains the last ten comments, and the "post" field
// has been populated. For example:
for ($i = 0; $i < count($comments); $i++) {
  // This does not require a network access.
  $post = $comments[$i]->get("post");
}

You can also do multi level includes using dot notation. If you wanted to include the post for a comment and the post's author as well you can do:

$query->includeKey("post.author");

You can issue a query with multiple fields included by calling includeKey multiple times. This functionality also works with ParseQuery helpers like first and get.

Counting Objects

Caveat: Count queries are rate limited to a maximum of 160 requests per minute. They can also return inaccurate results for classes with more than 1,000 objects. Thus, it is preferable to architect your application to avoid this sort of count operation (by using counters, for example.)

If you just need to count how many objects match a query, but you do not need to retrieve all the objects that match, you can use count instead of find. For example, to count how many games have been played by a particular player:

$query = new ParseQuery("GameScore");
$query->equalTo("playerName", "Sean Plott");
$count = $query->count();
// The count request succeeded. Show the count
echo "Sean has played " . $count . " games";

Compound Queries

If you want to find objects that match one of several queries, you can use ParseQuery.or method to construct a query that is an OR of the queries passed in. For instance if you want to find players who either have a lot of wins or a few wins, you can do:

$lotsOfWins = new ParseQuery("Player");
$lotsOfWins->greaterThan("wins", 150);

$fewWins = new ParseQuery("Player");
$fewWins->lessThan("wins", 5);

$mainQuery = ParseQuery::orQueries([$lotsOfWins, $fewWins]);
$results = $mainQuery->find();
// results contains an array of players that either have won a lot of games or won only a few games.

You can add additional constraints to the newly created ParseQuery that act as an 'and' operator.

Note that we do not, however, support GeoPoint or non-filtering constraints (e.g. near, withinGeoBox, limit, skip, ascending/descending, includeKey) in the subqueries of the compound query.

Each ParseObject is an instance of a specific subclass with a class name that you can use to distinguish different sorts of data. For example, we could call the high score object a GameScore. We recommend that you NameYourClassesLikeThis and nameYourKeysLikeThis, just to keep your code looking pretty.

To create a new subclass, create a new class which extends the ParseObject class, add the $parseClassName static property, and call the registerSubclass method before use. Any ParseQuery will return instances of the new class for any ParseObject with the same class name.

class GameScore extends ParseObject
{
  public static $parseClassName = "GameScore";
}
// Do this once, at the start of your app, before ParseClient::initialize(...);
GameScore::registerSubclass();

// Create a new instance of that class.
$gameScore = new GameScore();

You can add additional methods and properties to your subclasses of ParseObject.

// A complex subclass of ParseObject
class Monster extends ParseObject
{
  public static $parseClassName = "Monster";

  public function hasSuperHumanStrength() {
    return this->get("strength") > 18;
  }

  public static function spawn($strength) {
    $monster = new Monster();
    $monster->set("strength", $strength);
    return $monster;
  }
}
$monster = Monster::spawn(200);
echo monster->strength();  // Displays 200.
echo monster->hasSuperHumanStrength();  // Displays true.
Want to contribute to this doc? Edit this section.
Was this section helpful? NO YES Thanks for your feedback, we’re always working to make our docs better.

Users

At the core of many apps, there is a notion of user accounts that lets users access their information in a secure manner. We provide a specialized user class called ParseUser that automatically handles much of the functionality required for user account management.

With this class, you'll be able to add user account functionality in your app.

ParseUser is a subclass of ParseObject, and has all the same features, such as flexible schema, automatic persistence, and a key value interface. All the methods that are on ParseObject also exist in ParseUser. The difference is that ParseUser has some special additions specific to user accounts.

Properties

ParseUser has several values that set it apart from ParseObject:

  • username: The username for the user (required).
  • password: The password for the user (required on signup).
  • email: The email address for the user (optional).

We'll go through each of these in detail as we run through the various use cases for users.

Signing Up

The first thing your app will do is probably ask the user to sign up. The following code illustrates a typical sign up:


$user = new ParseUser();
$user->set("username", "my name");
$user->set("password", "my pass");
$user->set("email", "email@example.com");

// other fields can be set just like with ParseObject
$user->set("phone", "415-392-0202");

try {
  $user->signUp();
  // Hooray! Let them use the app now.
} catch (ParseException $ex) {
  // Show the error message somewhere and let the user try again.
  echo "Error: " . $ex->getCode() . " " . $ex->getMessage();
}

This call will asynchronously create a new user in your Parse App. Before it does this, it also checks to make sure that both the username and email are unique. Also, it securely hashes the password in the cloud using bcrypt. We never store passwords in plaintext, nor will we ever transmit passwords back to the client in plaintext.

Note that we used the signUp method, not the save method. New ParseUsers should always be created using the signUp method. Subsequent updates to a user can be done by calling save.

If a signup isn't successful, you should read the error object that is returned. The most likely case is that the username or email has already been taken by another user. You should clearly communicate this to your users, and ask them try a different username.

You are free to use an email address as the username. Simply ask your users to enter their email, but fill it in the username property — ParseUser will work as normal. We'll go over how this is handled in the reset password section.

Logging In

Of course, after you allow users to sign up, you need to let them log in to their account in the future. To do this, you can use the class method logIn.

try {
  $user = ParseUser::logIn("myname", "mypass");
  // Do stuff after successful login.
} catch (ParseException $error) {
  // The login failed. Check error to see why.
}

Verifying Emails

Enabling email verification in an application's settings allows the application to reserve part of its experience for users with confirmed email addresses. Email verification adds the emailVerified key to the ParseUser object. When a ParseUser's email is set or modified, emailVerified is set to false. Parse then emails the user a link which will set emailVerified to true.

There are three emailVerified states to consider:

  1. true - the user confirmed his or her email address by clicking on the link Parse emailed them. ParseUsers can never have a true value when the user account is first created.
  2. false - at the time the ParseUser object was last refreshed, the user had not confirmed his or her email address. If emailVerified is false, consider calling fetch on the ParseUser.
  3. missing - the ParseUser was created when email verification was off or the ParseUser does not have an email.

Current User

It would be bothersome if the user had to log in every time they open your app. You can avoid this by using the cached current ParseUser object.

By default, whenever you use any signup or login methods, the user will be saved in PHP Session storage (The $_SESSION superglobal.)

$currentUser = ParseUser::getCurrentUser();
if ($currentUser) {
    // do stuff with the user
} else {
    // show the signup or login page
}

You can clear the current user by logging them out:

ParseUser::logOut();

$currentUser = ParseUser::getCurrentUser();  // this will now be null

Setting the Current User

If you’ve created your own authentication routines, or otherwise logged in a user on the server side, you can now pass the session token to the client and use the become method. This method will ensure the session token is valid before setting the current user.

try {
  $user = ParseUser::become("session-token-here");
  // The current user is now set to user.
} catch (ParseException $ex) {
  // The token could not be validated.
}

Session Storage Interface

If you do not want to use the default PHP $_SESSION for storage, you can add your own storage mechanism. We provide the ParseSessionStorageInterface and a default implementation, ParseSessionStorage. Simply create your own storage class that implements ParseSessionStorageInterface and inject it to the SDK when you initialize:

    ParseClient::setStorage(new MyStorageClass());

Security For User Objects

The ParseUser class is secured by default. Data stored in a ParseUser can only be modified by that user. By default, the data can still be read by any client. Thus, some ParseUser objects are authenticated and can be modified, whereas others are read-only.

Specifically, you are not able to invoke any of the save or delete methods unless the ParseUser was obtained using an authenticated method, like logIn or signUp. This ensures that only the user can alter their own data.

The following illustrates this security policy:

$user = ParseUser::logIn("my_username", "my_password");
$user->set("username", "my_new_username");  // attempt to change username
$user->save();
// This succeeds, since the user was authenticated on the device

// Get the user from a non-authenticated method
$query = ParseUser::query();
$userAgain = $query->get($user->getObjectId());
$userAgain->set("username", "another_username");
// This will throw a ParseException, since the ParseUser is not authenticated
$userAgain->save();

The ParseUser obtained from ParseUser::getCurrentUser() will always be authenticated.

If you need to check if a ParseUser is authenticated, you can invoke the authenticated method. You do not need to check authenticated with ParseUser objects that are obtained via an authenticated method.

Security For Other Objects

The same security model that applies to the ParseUser can be applied to other objects. For any object, you can specify which users are allowed to read the object, and which users are allowed to modify an object. To support this type of security, each object has an access control list, implemented by the ParseACL class.

The simplest way to use a ParseACL is to specify that an object may only be read or written by a single user. To create such an object, there must first be a logged in ParseUser. Then, new ParseACL::createACLWithUser($user) generates a ParseACL that limits access to that user. An object's ACL is updated when the object is saved, like any other property. Thus, to create a private note that can only be accessed by the current user:

$privateNote = new ParseObject("Note");
$privateNote->set("content", "This note is private!");
$privateNote->setACL(ParseACL::createACLWithUser(ParseUser::getCurrentUser()));
$privateNote->save();

This note will then only be accessible to the current user, although it will be accessible to any device where that user is signed in. This functionality is useful for applications where you want to enable access to user data across multiple devices, like a personal todo list.

Permissions can also be granted on a per-user basis. You can add permissions individually to a ParseACL using setReadAccess and setWriteAccess. For example, let's say you have a message that will be sent to a group of several users, where each of them have the rights to read and delete that message:

$groupMessage = new ParseObject("Message");
$groupACL = new ParseACL();

// userList is an array with the users we are sending this message to.
for ($i = 0; $i < count($userList); $i++) {
  $groupACL->setReadAccess($userList[$i], true);
  $groupACL->setWriteAccess($userList[$i], true);
}

$groupMessage->setACL($groupACL);
$groupMessage->save();

You can also grant permissions to all users at once using setPublicReadAccess and setPublicWriteAccess. This allows patterns like posting comments on a message board. For example, to create a post that can only be edited by its author, but can be read by anyone:

$publicPost = new ParseObject("Post");
$postACL = ParseACL::createACLWithUser(ParseUser::getCurrentUser());
$postACL->setPublicReadAccess(true);
$publicPost->setACL($postACL);
$publicPost->save();

Operations that are forbidden, such as deleting an object that you do not have write access to, result in a ParseError.OBJECT_NOT_FOUND error code. For security purposes, this prevents clients from distinguishing which object ids exist but are secured, versus which object ids do not exist at all.

Resetting Passwords

It's a fact that as soon as you introduce passwords into a system, users will forget them. In such cases, our library provides a way to let them securely reset their password.

To kick off the password reset flow, ask the user for their email address, and call:

try {
  ParseUser::requestPasswordReset("email@example.com");
    // Password reset request was sent successfully
} catch (ParseException $ex) {
  // Password reset failed, check the exception message
}

This will attempt to match the given email with the user's email or username field, and will send them a password reset email. By doing this, you can opt to have users use their email as their username, or you can collect it separately and store it in the email field.

The flow for password reset is as follows:

  1. User requests that their password be reset by typing in their email.
  2. Parse sends an email to their address, with a special password reset link.
  3. User clicks on the reset link, and is directed to a special Parse page that will allow them type in a new password.
  4. User types in a new password. Their password has now been reset to a value they specify.

Note that the messaging in this flow will reference your app by the name that you specified when you created this app on Parse.

Querying

To query for users, you can get a new ParseQuery for ParseUsers:

$query = ParseUser::query();
$query->equalTo("gender", "female"); 
$results = $query->find();

Associations

Associations involving a ParseUser work right out of the box. For example, let's say you're making a blogging app. To store a new post for a user and retrieve all of their posts:

$user = ParseUser::getCurrentUser()

// Make a new post
$post = new ParseObject("Post");
$post->set("title", "My New Post");
$post->set("body", "This is some great content.");
$post->set("user", $user);
$post->save();

// Find all posts by the current user
$query = new ParseQuery("Post");
$query->equalTo("user", $user);
$userPosts = $query->find();
// $userPosts contains all of the posts by the current user.
Want to contribute to this doc? Edit this section.
Was this section helpful? NO YES Thanks for your feedback, we’re always working to make our docs better.

Sessions

Sessions represent an instance of a user logged into a device. Sessions are automatically created when users log in or sign up. They are automatically deleted when users log out. There is one distinct ParseSession object for each user-installation pair; if a user issues a login request from a device they're already logged into, that user's previous ParseSession object for that Installation is automatically deleted. ParseSession objects are stored on Parse in the Session class, and you can view them on the Parse.com Data Browser. We provide a set of APIs to manage ParseSession objects in your app.

Session APIs are only available in apps with revocable sessions enabled. Parse apps created after March 25, 2015 have this enabled by default ("Require Revocable Sessions" toggle in your Parse.com app settings page). If you have an existing app, you can upgrade to revocable sessions by following the Session Migration Tutorial.

ParseSession is a subclass of ParseObject, so you can query, update, and delete sessions in the same way that you manipulate normal objects on Parse. Because the Parse Cloud automatically creates sessions when you log in or sign up users, you should not manually create ParseSession objects unless you are building a "Parse for IoT" app (e.g. Arduino or Embedded C). Deleting a ParseSession will log the user out of the device that is currently using this session's token.

Unlike other Parse objects, the ParseSession class does not have Cloud Code triggers. So you cannot register a beforeSave or afterSave handler for the Session class.

Properties

The ParseSession object has these special fields:

  • sessionToken (readonly): String token for authentication on Parse API requests. In the response of ParseSession queries, only your current ParseSession object will contain a session token.
  • user: (readonly) Pointer to the ParseUser object that this session is for.
  • createdWith (readonly): Information about how this session was created (e.g. { "action": "login", "authProvider": "password"}).
    • action could have values: login, signup, create, or upgrade. The create action is when the developer manually creates the session by saving a ParseSession object. The upgrade action is when the user is upgraded to revocable session from a legacy session token.
    • authProvider could have values: password, anonymous, facebook, or twitter.
  • restricted (readonly): Boolean for whether this session is restricted.
    • Restricted sessions do not have write permissions on ParseUser, ParseSession, and ParseRole classes on Parse. Restricted sessions also cannot read unrestricted sessions.
    • All sessions that the Parse Cloud automatically creates during user login/signup will be unrestricted. All sessions that the developer manually creates by saving a new ParseSession object from the client (only needed for "Parse for IoT" apps) will be restricted.
  • expiresAt (readonly): Approximate UTC date when this ParseSession object will be automatically deleted. You can configure session expiration settings (either 1-year inactivity expiration or no expiration) in your app's Parse.com dashboard settings page.
  • installationId (can be set only once): String referring to the ParseInstallation where the session is logged in from. For Parse SDKs, this field will be automatically set when users log in or sign up. All special fields except installationId can only be set automatically by the Parse Cloud. You can add custom fields onto ParseSession objects, but please keep in mind that any logged-in device (with session token) can read other sessions that belong to the same user (unless you disable Class-Level Permissions, see below).

Handling Invalid Session Token Error

Apps created before March 25, 2015 use legacy session tokens until you migrate them to use the new revocable sessions. On API requests with legacy tokens, if the token is invalid (e.g. User object was deleted), then the request is executed as a non-logged in user and no error was returned. On API requests with revocable session tokens, an invalid session token will always fail with the "invalid session token" error. This new behavior lets you know when you need to ask the user to log in again.

With revocable sessions, your current session token could become invalid if its corresponding ParseSession object is deleted from the Parse Cloud. This could happen if you implement a Session Manager UI that lets users log out of other devices, or if you manually delete the session via Cloud Code, REST API, or Data Browser. Sessions could also be deleted due to automatic expiration (if configured in app settings). When a device's session token no longer corresponds to a ParseSession object on the Parse Cloud, all API requests from that device will fail with “Error 209: invalid session token”.

To handle this error, we recommend writing a global utility function that is called by all of your Parse request error callbacks. You can then handle the "invalid session token" error in this global function. You should prompt the user to login again so that they can obtain a new session token. This code could look like this:

@interface ParseErrorHandlingController : NSObject

+ (void)handleParseError:(NSError *)error;

@end

@implementation ParseErrorHandlingController

+ (void)handleParseError:(NSError *)error {
  if (![error.domain isEqualToString:PFParseErrorDomain]) {
    return;
  }

  switch (error.code) {
    case kPFErrorInvalidSessionToken: {
      [self _handleInvalidSessionTokenError];
      break;
    }
    ... // Other Parse API Errors that you want to explicitly handle.
  }
}

+ (void)_handleInvalidSessionTokenError {
  //--------------------------------------
  // Option 1: Show a message asking the user to log out and log back in.
  //--------------------------------------
  // If the user needs to finish what they were doing, they have the opportunity to do so.
  //
  // UIAlertView *alertView = [[UIAlertView alloc] initWithTitle:@"Invalid Session"
  //                                                     message:@"Session is no longer valid, please log out and log in again."
  //                                                    delegate:self
  //                                           cancelButtonTitle:@"Not Now"
  //                                           otherButtonTitles:@"OK"];
  // [alertView show];

  //--------------------------------------
  // Option #2: Show login screen so user can re-authenticate.
  //--------------------------------------
  // You may want this if the logout button is inaccessible in the UI.
  //
  // UIViewController *presentingViewController = [UIApplication sharedApplication].keyWindow.rootViewController;
  // PFLogInViewController *logInViewController = [[PFLogInViewController alloc] init];
  // [presentingViewController presentViewController:logInViewController animated:YES completion:nil];
}

@end

// In all API requests, call the global error handler, e.g.
[[PFQuery queryWithClassName:@"Object"] findInBackgroundWithBlock:^(NSArray *objects, NSError *error) {
  if (!error) {
    // Query succeeded - continue your app logic here.
  } else {
    // Query failed - handle an error.
    [ParseErrorHandlingController handleParseError:error];
  }
}];
class ParseErrorHandlingController {
  class func handleParseError(error: NSError) {
    if error.domain != PFParseErrorDomain {
      return
    }

    switch (error.code) {
    case kPFErrorInvalidSessionToken:
      handleInvalidSessionTokenError()

    ... // Other Parse API Errors that you want to explicitly handle.
  }

  private class func handleInvalidSessionTokenError() {
    //--------------------------------------
    // Option 1: Show a message asking the user to log out and log back in.
    //--------------------------------------
    // If the user needs to finish what they were doing, they have the opportunity to do so.
    //
    // let alertView = UIAlertView(
    //   title: "Invalid Session",
    //   message: "Session is no longer valid, please log out and log in again.",
    //   delegate: nil,
    //   cancelButtonTitle: "Not Now",
    //   otherButtonTitles: "OK"
    // )
    // alertView.show()

    //--------------------------------------
    // Option #2: Show login screen so user can re-authenticate.
    //--------------------------------------
    // You may want this if the logout button is inaccessible in the UI.
    //
    // let presentingViewController = UIApplication.sharedApplication().keyWindow?.rootViewController
    // let logInViewController = PFLogInViewController()
    // presentingViewController?.presentViewController(logInViewController, animated: true, completion: nil)
  }
}

// In all API requests, call the global error handler, e.g.
let query = PFQuery(className: "Object")
query.findObjectsInBackgroundWithBlock { (objects: [AnyObject]!, error: NSError!) -> Void in
  if error == nil {
    // Query Succeeded - continue your app logic here.
  } else {
    // Query Failed - handle an error.
    ParseErrorHandlingController.handleParseError(error)
  }
}
public class ParseErrorHandler {
  public static void handleParseError(ParseException e) {
    switch (e.getCode()) {
      case INVALID_SESSION_TOKEN: handleInvalidSessionToken()
        break;

      ... // Other Parse API errors that you want to explicitly handle
    }
  }

  private static void handleInvalidSessionToken() {
    //--------------------------------------
    // Option 1: Show a message asking the user to log out and log back in.
    //--------------------------------------
    // If the user needs to finish what they were doing, they have the opportunity to do so.
    //
    // new AlertDialog.Builder(getActivity())
    //   .setMessage("Session is no longer valid, please log out and log in again.")
    //   .setCancelable(false).setPositiveButton("OK", ...).create().show();

    //--------------------------------------
    // Option #2: Show login screen so user can re-authenticate.
    //--------------------------------------
    // You may want this if the logout button could be inaccessible in the UI.
    //
    // startActivityForResult(new ParseLoginBuilder(getActivity()).build(), 0);
  }
}

// In all API requests, call the global error handler, e.g.
query.findInBackground(new FindCallback<ParseObject>() {
  public void done(List<ParseObject> results, ParseException e) {
    if (e == null) {
      // Query successful, continue other app logic
    } else {
      // Query failed
      ParseErrorHandler.handleParseError(e);
    }
  }
});
function handleParseError(err) {
  switch (err.code) {
    case Parse.Error.INVALID_SESSION_TOKEN:
      Parse.User.logOut();
      ... // If web browser, render a log in screen
      ... // If Express.js, redirect the user to the log in route
      break;

    ... // Other Parse API errors that you want to explicitly handle
  }
}

// For each API request, call the global error handler
query.find().then(function() {
  ...
}, function(err) {
  handleParseError(err);
});
public class ParseErrorHandler {
  public static void HandleParseError(ParseException e) {
    switch (e.Code) {
      case ParseException.ErrorCode.InvalidSessionToken:
        HandleInvalidSessionToken()
        break;

      ... // Other Parse API errors that you want to explicitly handle
    }
  }

  private static void HandleInvalidSessionToken() {
    //--------------------------------------
    // Option 1: Show a message asking the user to log out and log back in.
    //--------------------------------------
    // If the user needs to finish what they were doing, they have the opportunity to do so.

    //--------------------------------------
    // Option #2: Show login screen so user can re-authenticate.
    //--------------------------------------
    // You may want this if the logout button is inaccessible in the UI.
  }
});

// In all API requests, call the global error handler, e.g.
query.FindAsync().ContinueWith(t => {
  if (t.IsFaulted) {
    // Query Failed - handle an error.
    ParseErrorHandler.HandleParseError(t.Exception.InnerException as ParseException);
  } else {
    // Query Succeeded - continue your app logic here.
  }
});
public class ParseErrorHandler {
  public static handleParseError(ParseException $e) {
    $code = $e->getCode();
    switch ($code) {
      case: 209: // INVALID_SESSION_TOKEN
        ParseUser::logOut();
        ... // Redirect the to login page.
        break;

      ... // Other Parse API errors that you want to explicitly handle
    }
  }
});

// For each API request, call the global error handler
try {
  $results = $query->find();
  // ...
} catch (ParseException $e) {
  ParseErrorHandler::handleParseError($e)
}
// No command line example
// No C++ example

Security

ParseSession objects can only be accessed by the user specified in the user field. All ParseSession objects have an ACL that is read and write by that user only. You cannot change this ACL. This means querying for sessions will only return objects that match the current logged-in user.

When you log in a user via ParseUser::logIn(), Parse will automatically create a new unrestricted ParseSession object in the Parse Cloud. Same for signups and Facebook/Twitter logins.

Session objects manually created from client SDKs (by creating an instance of ParseSession, and saving it) are always restricted. You cannot manually create an unrestricted sessions using the object creation API.

Restricted sessions are prohibited from creating, modifying, or deleting any data in the ParseUser, ParseSession, and ParseRole classes. Restricted session also cannot read unrestricted sessions. Restricted Sessions are useful for "Parse for IoT" devices (e.g Arduino or Embedded C) that may run in a less-trusted physical environment than mobile apps. However, please keep in mind that restricted sessions can still read data on ParseUser, ParseSession, and ParseRole classes, and can read/write data in any other class just like a normal session. So it is still important for IoT devices to be in a safe physical environment and ideally use encrypted storage to store the session token.

If you want to prevent restricted Sessions from modifying classes other than ParseUser, ParseSession, or ParseRole, you can write a Cloud Code beforeSave handler for that class:

Parse.Cloud.beforeSave("MyClass", function(request, response) {
  Parse.Session.current().then(function(session) {
    if (session.get('restricted')) {
      response.error('write operation not allowed');
    }
    response.success();
  });
});

You can configure Class-Level Permissions (CLPs) for the Session class just like other classes on Parse. CLPs restrict reading/writing of sessions via the ParseSession API, but do not restrict Parse Cloud's automatic session creation/deletion when users log in, sign up, and log out. We recommend that you disable all CLPs not needed by your app. Here are some common use cases for Session CLPs:

  • Find, Delete — Useful for building a UI screen that allows users to see their active session on all devices, and log out of sessions on other devices. If your app does not have this feature, you should disable these permissions.
  • Create — Useful for "Parse for IoT" apps (e.g. Arduino or Embedded C) that provision restricted user sessions for other devices from the phone app. You should disable this permission when building apps for mobile and web. For "Parse for IoT" apps, you should check whether your IoT device actually needs to access user-specific data. If not, then your IoT device does not need a user session, and you should disable this permission.
  • Get, Update, Add Field — Unless you need these operations, you should disable these permissions.
Want to contribute to this doc? Edit this section.
Was this section helpful? NO YES Thanks for your feedback, we’re always working to make our docs better.

Roles

As your app grows in scope and user-base, you may find yourself needing more coarse-grained control over access to pieces of your data than user-linked ACLs can provide. To address this requirement, Parse supports a form of Role-based Access Control. Roles provide a logical way of grouping users with common access privileges to your Parse data. Roles are named objects that contain users and other roles. Any permission granted to a role is implicitly granted to its users as well as to the users of any roles that it contains.

For example, in your application with curated content, you may have a number of users that are considered "Moderators" and can modify and delete content created by other users. You may also have a set of users that are "Administrators" and are allowed all of the same privileges as Moderators, but can also modify the global settings for the application. By adding users to these roles, you can ensure that new users can be made moderators or administrators, without having to manually grant permission to every resource for each user.

We provide a specialized class called ParseRole that represents these role objects in your client code. ParseRole is a subclass of ParseObject, and has all of the same features, such as a flexibl schema, automatic persistence, and a key value interface. All the methods that are on ParseObject also exist on ParseRole. The difference is that ParseRole has some additions specific to management of roles.

Properties

ParseRole has several properties that set it apart from ParseObject:

  • name: The name for the role. This value is required, and can only be set once as a role is being created. The name must consist of alphanumeric characters, spaces, -, or _. This name will be used to identify the Role without needing its objectId.
  • users: A relation to the set of users that will inherit permissions granted to the containing role.
  • roles: A relation to the set of roles whose users and roles will inherit permissions granted to the containing role.

Security for Role Objects

The ParseRole uses the same security scheme (ACLs) as all other objects on Parse, except that it requires an ACL to be set explicitly. Generally, only users with greatly elevated privileges (e.g. a master user or Administrator) should be able to create or modify a Role, so you should define its ACLs accordingly. Remember, if you give write-access to a ParseRole to a user, that user can add other users to the role, or even delete the role altogether.

To create a new ParseRole, you would write:

// By specifying no write privileges for the ACL, we can ensure the role cannot be altered.
$roleACL = new ParseACL();
$roleACL->setPublicReadAccess(true);
$role = ParseRole::createRole("Administrator", $roleACL);
$role->save();

You can add users and roles that should inherit your new role's permissions through the "users" and "roles" relations on ParseRole:

$role = ParseRole::createRole($roleName, $roleACL);
for ($i = 0; $i < count($usersToAddToRole); $i++) {
  $role->getUsers()->add($usersToAddToRole[$i]);
}
for ($i = 0; $i < count($rolesToAddToRole); $i++) {
  $role->getRoles()->add($rolesToAddToRole[$i]);
}
$role->save();

Take great care when assigning ACLs to your roles so that they can only be modified by those who should have permissions to modify them.

Security for Other Objects

Now that you have created a set of roles for use in your application, you can use them with ACLs to define the privileges that their users will receive. Each ParseObject can specify a ParseACL, which provides an access control list that indicates which users and roles should be granted read or write access to the object.

Giving a role read or write permission to an object is straightforward. You can either use the ParseRole:

$moderators = /* Query for some ParseRole */;
$wallPost = new ParseObject("WallPost");
$postACL = new ParseACL();
$postACL->setRoleWriteAccess($moderators, true);
$wallPost->setACL($postACL);
$wallPost->save();

You can avoid querying for a role by specifying its name for the ACL:

$wallPost = new ParseObject("WallPost");
$postACL = new ParseACL();
$postACL->setRoleWriteAccessWithName("Moderators", true);
$wallPost->setACL($postACL);
$wallPost->save();

Role Hierarchy

As described above, one role can contain another, establishing a parent-child relationship between the two roles. The consequence of this relationship is that any permission granted to the parent role is implicitly granted to all of its child roles.

These types of relationships are commonly found in applications with user-managed content, such as forums. Some small subset of users are "Administrators", with the highest level of access to tweaking the application's settings, creating new forums, setting global messages, and so on. Another set of users are "Moderators", who are responsible for ensuring that the content created by users remains appropriate. Any user with Administrator privileges should also be granted the permissions of any Moderator. To establish this relationship, you would make your "Administrators" role a child role of "Moderators", like this:

$administrators = /* Your "Administrators" role */;
$moderators = /* Your "Moderators" role */;
$moderators->getRoles()->add($administrators);
$moderators->save();
Want to contribute to this doc? Edit this section.
Was this section helpful? NO YES Thanks for your feedback, we’re always working to make our docs better.

Files

Creating a ParseFile

ParseFile lets you store application files in the cloud that would otherwise be too large or cumbersome to fit into a regular ParseObject. The most common use case is storing images, but you can also use it for documents, videos, music, and any other binary data (up to 10 megabytes).

Getting started with ParseFile is easy. There are a couple of ways to create a file. The first is to provide the contents of the file.

$contents = "Hello World.";
$file = ParseFile::createFromData($contents, "myfile.txt");

Alternatively, you can create a file from the contents of a local file:

$localFilePath = "/tmp/myFile.txt";
$file = ParseFile::createFromFile($localFilePath, "myfile.txt");

Notice in this example that we give the file a name of myfile.txt. There's two things to note here:

  • You don't need to worry about filename collisions. Each upload gets a unique identifier so there's no problem with uploading multiple files named photo.jpg.
  • It's important that you give a name to the file that has a file extension. This lets Parse figure out the file type and handle it accordingly. So, if you're storing PNG images, make sure your filename ends with .png.

Next you'll want to save the file up to the cloud. As with ParseObject, the save method is the way to go.

$file->save();
// The file has been saved to Parse and now has a URL.
$url = $file->getURL();

Finally, after the save completes, you can associate a ParseFile with a ParseObject just like any other piece of data:

$jobApplication = new ParseObject("JobApplication");
$jobApplication->set("applicantName", "Joe Smith");
$jobApplication->set("applicantResumeFile", $file);
$jobApplication->save();

Retrieving File Contents

How to best retrieve the file contents back depends on the context of your application. It's best if you can make the visitors browser do the work for you. Typically, that means rendering the file's URL into your output. Here we output an uploaded profile photo in an image tag:

$profilePhoto = $profile->get("photoFile");
echo '<img src="' . $profilePhoto->getURL() . '">';

If you want to fetch the contents of the file, you can retrieve it like this:

$contents = $file->getData();
Want to contribute to this doc? Edit this section.
Was this section helpful? NO YES Thanks for your feedback, we’re always working to make our docs better.

GeoPoints

Parse allows you to associate real-world latitude and longitude coordinates with an object. Adding a ParseGeoPoint to a ParseObject allows queries to take into account the proximity of an object to a reference point. This allows you to easily do things like find out what user is closest to another user or which places are closest to a user.

ParseGeoPoint

To associate a point with an object you first need to create a ParseGeoPoint. For example, to create a point with latitude of 40.0 degrees and -30.0 degrees longitude:

$point = new ParseGeoPoint(40.0, -30.0);

This point is then stored in the object as a regular field.

$placeObject->set("location", $point);

Note: Currently only one key in a class may be a ParseGeoPoint.

Geo Queries

Now that you have a bunch of objects with spatial coordinates, it would be nice to find out which objects are closest to a point. This can be done by adding another restriction to ParseQuery using near. Getting an array of ten places that are closest to a user may look something like:

// User's location
$userGeoPoint = $userObject->get("location");
// Create a query for places
$query = new ParseQuery("PlaceObject");
// Interested in locations near user.
$query->near("location", $userGeoPoint);
// Limit what could be a lot of points.
$query->limit(10);
// Final array of objects
$placesObjects = $query->find();

At this point $placesObjects will be an array of objects ordered by distance (nearest to farthest) from $userGeoPoint. Note that if an additional ascending()/descending() order-by constraint is applied, it will take precedence over the distance ordering.

To limit the results using distance, check out withinMiles, withinKilometers, and withinRadians.

It's also possible to query for the set of objects that are contained within a particular area. To find the objects in a rectangular bounding box, add the withinGeoBox restriction to your ParseQuery.

$southwestOfSF = new ParseGeoPoint(37.708813, -122.526398);
$northeastOfSF = new ParseGeoPoint(37.822802, -122.373962);

$query = new ParseQuery("PizzaPlaceObject");
$query->withinGeoBox("location", $southwestOfSF, $northeastOfSF);
$pizzaPlacesInSF = $query->find();

Caveats

At the moment there are a couple of things to watch out for:

  1. Each ParseObject class may only have one key with a ParseGeoPoint object.
  2. Using the near constraint will also limit results to within 100 miles.
  3. Points should not equal or exceed the extreme ends of the ranges. Latitude should not be -90.0 or 90.0. Longitude should not be -180.0 or 180.0. Attempting to set latitude or longitude out of bounds will cause an error.
Want to contribute to this doc? Edit this section.
Was this section helpful? NO YES Thanks for your feedback, we’re always working to make our docs better.

Push Notifications

Introduction

Push Notifications are a great way to keep your users engaged and informed about your app. You can reach your entire user base quickly and effectively. This guide will help you through the setup process and the general usage of Parse to send push notifications.

The PHP SDK does not currently support receiving pushes. It can only be used to send notifications to iOS and Android applications. A common use case is to send pushes from Cloud Code.

Setting Up Push

There is no setup required to use the PHP SDK for sending push notifications. If you haven't configured your iOS or Android clients to use Push, take a look at their respective setup instruction using the platform toggle at the top.

Installations

Every Parse application installed on a device registered for push notifications has an associated Installation object. The Installation object is where you store all the data needed to target push notifications. For example, in a baseball app, you could store the teams a user is interested in to send updates about their performance.

Note that Installation data can only be modified by the client SDKs, the data browser, or the REST API.

This class has several special fields that help you manage and target devices.

  • badge: The current value of the icon badge for iOS apps. Changes to this value on the server will be used for future badge-increment push notifications.
  • channels: An array of the channels to which a device is currently subscribed.
  • timeZone: The current time zone where the target device is located. This value is synchronized every time an Installation object is saved from the device (readonly).
  • deviceType: The type of device, "ios" or "android" (readonly).
  • installationId: Unique Id for the device used by Parse (readonly).
  • deviceToken: The Apple generated token used for iOS devices (readonly).

Sending Pushes

There are two ways to send push notifications using Parse: channels and advanced targeting. Channels offer a simple and easy to use model for sending pushes, while advanced targeting offers a more powerful and flexible model. Both are fully compatible with each other and will be covered in this section.

You can view your past push notifications on the Parse.com push console for up to 30 days after creating your push. For pushes scheduled in the future, you can delete the push on the web console as long as no sends have happened yet. After you send the push, the web console shows push analytics graphs.

Using Channels

The simplest way to start sending notifications is using channels. This allows you to use a publisher-subscriber model for sending pushes. Devices start by subscribing to one or more channels, and notifications can later be sent to these subscribers. The channels subscribed to by a given Installation are stored in the channels field of the Installation object.

Subscribing to Channels

The PHP SDK does not currently support subscribing iOS and Android devices for pushes. Take a look at the iOS, Android or REST Push guide using the platform toggle at the top.

Sending Pushes to Channels

With the PHP SDK, the following code can be used to alert all subscribers of the "Giants" and "Mets" channels about the results of the game. This will display a notification center alert to iOS users and a system tray notification to Android users.

$data = array("alert" => "Hi!");

ParsePush::send(array(
  "channels" => ["PHPFans"],
  "data" => $data
));

Using Advanced Targeting

While channels are great for many applications, sometimes you need more precision when targeting the recipients of your pushes. Parse allows you to write a query for any subset of your Installation objects using the querying API and to send them a push.

Since Installation objects are just like any other object stored in Parse, you can save any data you want and even create relationships between Installation objects and your other objects. This allows you to send pushes to a very customized and dynamic segment of your user base.

Saving Installation Data

The PHP SDK does not currently support modifying Installation objects. Take a look at the iOS, Android or REST Push guide using the platform toggle at the top.

Sending Pushes to Queries

Once you have your data stored on your Installation objects, you can use a query to target a subset of these devices. Parse.Installation queries work just like any other Parse query.

$query = ParseInstallation::query();
$query->equalTo("design", "rad");
ParsePush::send(array(
  "where" => $query,
  "data" => $data
));

We can even use channels with our query. To send a push to all subscribers of the "Giants" channel but filtered by those who want score update, we can do the following:

$query = ParseInstallation::query();
$query->equalTo("channels", "Giants");
$query->equalTo("scores", true);

ParsePush::send(array(
  "where" => $query,
  "data" => array(
    "alert" => "Giants scored against the A's! It's now 2-2."
  )
));

If we store relationships to other objects in our Installation class, we can also use those in our query. For example, we could send a push notification to all users near a given location like this.

// Find users near a given location
$userQuery = ParseUser::query();
$userQuery->withinMiles("location", $stadiumLocation, 1.0);

// Find devices associated with these users
$pushQuery = ParseInstallation::query();
$pushQuery->matchesQuery('user', $userQuery);

// Send push notification to query
ParsePush::send(array(
  "where" => $pushQuery,
  "data" => array(
    "alert" => "Free hotdogs at the Parse concession stand!"
  )
));

Sending Options

Push notifications can do more than just send a message. In iOS, pushes can also include the sound to be played, the badge number to display as well as any custom data you wish to send. In Android, it is even possible to specify an Intent to be fired upon receipt of a notification. An expiration date can also be set for the notification in case it is time sensitive.

Customizing your Notifications

If you want to send more than just a message, you can set other fields in the data dictionary. There are some reserved fields that have a special meaning.

  • alert: the notification's message.
  • badge: (iOS only) the value indicated in the top right corner of the app icon. This can be set to a value or to Increment in order to increment the current value by 1.
  • sound: (iOS only) the name of a sound file in the application bundle.
  • content-available: (iOS only) If you are a writing a Newsstand app, or an app using the Remote Notification Background Mode introduced in iOS7 (a.k.a. "Background Push"), set this value to 1 to trigger a background download.
  • category: (iOS only) the identifier of the UIUserNotificationCategory for this push notification.
  • uri: (Android only) an optional field that contains a URI. When the notification is opened, an Activity associated with opening the URI is launched.
  • title: (Android only) the value displayed in the Android system tray notification.

For example, to send a notification that increases the current badge number by 1 and plays a custom sound for iOS devices, and displays a particular title for Android users, you can do the following:

ParsePush::send(array(
  "channels" => [ "Mets" ],
  "data" => array(
    "alert" => "The Mets scored! The game is now tied 1-1.",
    "badge" => "Increment",
    "sound" => "cheering.caf",
    "title" => "Mets Score!"
  )
));

It is also possible to specify your own data in this dictionary. As explained in the Receiving Notifications section for iOS and Android, iOS will give you access to this data only when the user opens your app via the notification and Android will provide you this data in the Intent if one is specified.

$query = ParseInstallation::query();
$query->equalTo('channels', 'Indians');
$query->equalTo('injuryReports', true);

ParsePush::send(array(
  "where" => $query,
  "data" => array(
    "action" => "com.example.UPDATE_STATUS"
    "alert" => "Ricky Vaughn was injured in last night's game!",
    "name" => "Vaughn",
    "newsItem" => "Man bites dog"
  )
));

Setting an Expiration Date

When a user's device is turned off or not connected to the internet, push notifications cannot be delivered. If you have a time sensitive notification that is not worth delivering late, you can set an expiration date. This avoids needlessly alerting users of information that may no longer be relevant.

There are two parameters provided by Parse to allow setting an expiration date for your notification. The first is expiration_time which takes a DateTime specifying when Parse should stop trying to send the notification.

Alternatively, you can use the expiration_interval parameter to specify a duration of time before your notification expires. This value is relative to the push_time parameter used to schedule notifications. This means that a push notification scheduled to be sent out in 1 day and an expiration interval of 6 days can be received up to a week from now.

Targeting by Platform

If you build a cross platform app, it is possible you may only want to target iOS or Android devices. There are two methods provided to filter which of these devices are targeted. Note that both platforms are targeted by default.

The following examples would send a different notification to Android and iOS users.

// Notification for Android users
$queryAndroid = ParseInstallation::query();
$queryAndroid->equalTo('deviceType', 'android');

ParsePush::send(array(
  "where" => $queryAndroid,
  "data" => array(
    "alert" => "Your suitcase has been filled with tiny robots!" 
  )
));

// Notification for iOS users
$queryIOS = ParseInstallation::query();
$queryIOS->equalTo('deviceType', 'ios');

ParsePush::send(array(
  "where" => $queryIOS,
  "data" => array(
    "alert" => "Your suitcase has been filled with tiny apples!" 
  )
));

// Notification for Windows 8 users
$queryWindows = ParseInstallation::query();
$queryWindows->equalTo('deviceType', 'winrt');

ParsePush::send(array(
  "where" => $queryWindows,
  "data" => array(
    "alert" => "Your suitcase has been filled with tiny surfaces!" 
  )
));

// Notification for Windows Phone 8 users
$queryWP8 = ParseInstallation::query();
$queryWP8->equalTo('deviceType', 'winphone');

ParsePush::send(array(
  "where" => $queryWP8,
  "data" => array(
    "alert" => "Your suitcase is very hip; very metro." 
  )
));

Scheduling Pushes

You can schedule a push in advance by specifying a push_time parameter of type DateTime.

If you also specify an expiration_interval, it will be calculated from the scheduled push time, not from the time the push is submitted. This means a push scheduled to be sent in a week with an expiration interval of a day will expire 8 days after the request is sent.

The scheduled time cannot be in the past, and can be up to two weeks in the future. It can be an ISO 8601 date with a date, time, and timezone, as in the example above, or it can be a numeric value representing a UNIX epoch time in seconds (UTC).

Receiving Pushes

The PHP SDK does not currently support receiving pushes. To learn more about handling received notifications in iOS or Android, use the platform toggle at the top.

Troubleshooting

For tips on troubleshooting push notifications, check the troubleshooting sections for iOS, Android, and .NET using the platform toggle at the top.

Want to contribute to this doc? Edit this section.
Was this section helpful? NO YES Thanks for your feedback, we’re always working to make our docs better.

Analytics

Parse provides a number of hooks for you to get a glimpse into the ticking heart of your app. We understand that it's important to understand what your app is doing, how frequently, and when.

While this section will cover different ways to instrument your app to best take advantage of Parse's analytics backend, developers using Parse to store and retrieve data can already take advantage of metrics on Parse.

Custom Analytics

ParseAnalytics also allows you to track free-form events, with a handful of string keys and values. These extra dimensions allow segmentation of your custom events via your app's Dashboard.

Say your app offers search functionality for apartment listings, and you want to track how often the feature is used, with some additional metadata.

// Define ranges to bucket data points into meaningful segments
$dimensions = [
  "priceRange" => '1000-1500', 
  "source" => 'craigslist', 
  "dayType": 'weekday'
];
// Send the dimensions to Parse along with the 'search' event
ParseAnalytics::track('search', $dimensions);

ParseAnalytics can even be used as a lightweight error tracker — simply invoke the following and you'll have access to an overview of the rate and frequency of errors, broken down by error code, in your application:

$codeString = '' + $error->getCode();
ParseAnalytics::track('error', ["code" => codeString]);

Note that Parse currently only stores the first eight dimension pairsper call to ParseAnalytics::track().

Want to contribute to this doc? Edit this section.
Was this section helpful? NO YES Thanks for your feedback, we’re always working to make our docs better.

Data

We've designed the Parse SDKs so that you typically don't need to worry about how data is saved while using the client SDKs. Simply add data to the ParseObject, and it'll be saved correctly.

Nevertheless, there are some cases where it's useful to be aware of how data is stored on the Parse platform.

Data Storage

Internally, Parse stores data as JSON, so any datatype that can be converted to JSON can be stored on Parse. Refer to the Data Types in Objects section of this guide to see platform-specific examples.

Keys including the characters $ or ., along with the key __type key, are reserved for the framework to handle additional types, so don't use those yourself. Key names must contain only numbers, letters, and underscore, and must start with a letter. Values can be anything that can be JSON-encoded.

Data Type Lock-in

When a class is initially created, it doesn't have an inherent schema defined. This means that for the first object, it could have any types of fields you want.

However, after a field has been set at least once, that field is locked into the particular type that was saved. For example, if a ParseUser object is saved with field name of type String, that field will be restricted to the String type only (the server will return an error if you try to save anything else).

One special case is that any field can be set to null, no matter what type it is.

The Data Browser

The Data Browser is the web UI where you can update and create objects in each of your apps. Here, you can see the raw JSON values that are saved that represents each object in your class.

When using the interface, keep in mind the following:

  • The objectId, createdAt, updatedAt fields cannot be edited (these are set automatically).
  • The value "(empty)" denotes that the field has not been set for that particular object (this is different than null).
  • You can remove a field's value by hitting your Delete key while the value is selected.

The Data Browser is also a great place to test the Cloud Code validations contained in your Cloud Code functions (such as beforeSave). These are run whenever a value is changed or object is deleted from the Data Browser, just as they would be if the value was changed or deleted from your client code.

Importing Data

You may import data into your Parse app by using CSV or JSON files. To create a new class with data from a CSV or JSON file, go to the Data Browser and click the "Import" button on the left hand column.

The JSON format is an array of objects in our REST format or a JSON object with a results that contains an array of objects. It must adhere to the JSON standard. A file containing regular objects could look like:

{ "results": [
  {
    "score": 1337,
    "playerName": "Sean Plott",
    "cheatMode": false,
    "createdAt": "2012-07-11T20:56:12.347Z",
    "updatedAt": "2012-07-11T20:56:12.347Z",
    "objectId": "fchpZwSuGG"
  }]
}

Objects in either format should contain keys and values that also satisfy the following:

  • Key names must contain only numbers, letters, and underscore, and must start with a letter.
  • No value may contain a hard newline '\n'.

Normally, when objects are saved to Parse, they are automatically assigned a unique identifier through the objectId field, as well as a createdAt field and updatedAt field which represent the time that the object was created and last modified in the Parse Cloud. These fields can be manually set when data is imported from a JSON file. Please keep in mind the following:

  • Use a unique 10 character alphanumeric string as the value of your objectId fields.
  • Use a UTC timestamp in the ISO 8601 format when setting a value for the createdAt field or the updatedAt field.

In addition to the exposed fields, objects in the Parse User class can also have the bcryptPassword field set. The value of this field is a String that is the bcrypt hashed password + salt in the modular crypt format described in this StackOverflow answer. Most OpenSSL based bcrypt implementations should have built-in methods to produce these strings.

A file containing a ParseUser object could look like:

{ "results":
  [{
    "username": "cooldude",
    "createdAt": "1983-09-13T22:42:30.548Z",
    "updatedAt": "2015-09-04T10:12:42.137Z",
    "objectId": "ttttSEpfXm",
    "sessionToken": "dfwfq3dh0zwe5y2sqv514p4ib",
    "bcryptPassword": "$2a$10$ICV5UeEf3lICfnE9W9pN9.O9Ved/ozNo7G83Qbdk5rmyvY8l16MIK"
  }]
}

Note that in CSV the import field types are limited to String, Boolean, and Number.

Exporting your Data

You can request an export of your data at any time from your app's Settings page. The data export runs at a lower priority than production queries, so if your app is still serving queries, production traffic will always be given a higher priority, which may slow down the delivery of your data export.

Export Formats

Each collection will be exported in the same JSON format used by our REST API and delivered in a single zipped file. Since data is stored internally as JSON, this allows us to ensure that the export closely matches how the data is saved to Parse. Other formats such as CSV cannot represent all of the data types supported by Parse without losing information. If you'd like to work with your data in CSV format, you can use any of the JSON-to-CSV converters available widely on the web.

Offline Analysis

For offline analysis of your data, we highly recommend using alternate ways to access your data that do not require extracting the entire collection at once. For example, you can try exporting only the data that has changed since your last export. Here are some ways of achieving this:

  • Use the JavaScript SDK in a node app. Parse.Query.each() will allow you to extract every single object that matches a query. You can use date constraints to make sure the query only matches data that has been updated since you last ran this app. Your node app can write this data to disk for offline analysis.

  • Use the REST API in a script. You can run queries against your class and use skip/limit to page through results, which can then be written to disk for offline analysis. You can again use date constraints to make sure only newly updated data is extracted.

  • If the above two options do not fit your needs, you can try using the Data Browser to export data selectively. Use the Funnel icon to create a filter for the specific data that you need to export, such as newly updated objects. Once the filter has been applied, click on the Export data icon on the upper right of your Data Browser. This type of export will only include the objects that match your criteria.

Want to contribute to this doc? Edit this section.
Was this section helpful? NO YES Thanks for your feedback, we’re always working to make our docs better.

Relations

There are three kinds of relationships. One-to-one relationships enable one object to be associated with another object. One-to-many relationships enable one object to have many related objects. Finally, many-to-many relationships enable complex relationships among many objects.

There are four ways to build relationships in Parse:

One-to-Many

When you’re thinking about one-to-many relationships and whether to implement Pointers or Arrays, there are several factors to consider. First, how many objects are involved in this relationship? If the "many" side of the relationship could contain a very large number (greater than 100 or so) of objects, then you have to use Pointers. If the number of objects is small (fewer than 100 or so), then Arrays may be more convenient, especially if you typically need to get all of the related objects (the "many" in the "one-to-many relationship") at the same time as the parent object.

Using Pointers

Let's say we have a game app. The game keeps track of the player's score and achievements every time she chooses to play. In Parse, we can store this data in a single Game object. If the game becomes incredibly successful, each player will store thousands of Game objects in the system. For circumstances like this, where the number of relationships can be arbitrarily large, Pointers are the best option.

Suppose in this game app, we want to make sure that every Game object is associated with a Parse User. We can implement this like so:

ParseObject game = new ParseObject("Game");
game.put("createdBy", ParseUser.getCurrentUser());
PFObject *game= [PFObject objectWithClassName:@"Game"];
[game setObject:[PFUser currentUser] forKey:@"createdBy"];
let game = PFObject(className:"Game")
game["createdBy"] = PFUser.currentUser()
$game = ParseObject::create("Game");
$game->set("createdBy", ParseUser::getCurrentUser());
var game = new ParseObject("Game");
game["createdBy"] = ParseUser.CurrentUser;
var game = new Parse.Object("Game");
game.set("createdBy", Parse.User.current());
# No REST API example
// No C++ example

We can obtain all of the Game objects created by a Parse User with a query:

ParseQuery<ParseObject> gameQuery = ParseQuery.getQuery("Game");
gameQuery.whereEqualTo("createdBy", ParseUser.getCurrentUser());
PFQuery *gameQuery = [PFQuery queryWithClassName:@"Game"];
[gameQuery whereKey:@"createdBy" equalTo:[PFUser currentUser]];
let gameQuery = PFQuery(className:"Game")
if let user = PFUser.currentUser() {
  gameQuery.whereKey("createdBy", equalTo: user)
}
$gameQuery = new ParseQuery("Game");
$gameQuery->equalTo("createdBy", ParseUser::getCurrentUser());
var query = ParseObject.getQuery("Game").WhereEqualTo("createdBy", ParseUser.CurrentUser);
var query = new Parse.Query("Game");
query.equalTo("createdBy", Parse.User.current());
# No REST API example
// No C++ example

And, if we want to find the Parse User who created a specific Game, that is a lookup on the createdBy key:

// say we have a Game object
ParseObject game = ...

// getting the user who created the Game
ParseUser createdBy = game.getUser("createdBy");
// say we have a Game object
PFObject *game = ...

// getting the user who created the Game
PFUser *createdBy = [game objectForKey@"createdBy"];
// say we have a Game object
let game = ...

// getting the user who created the Game
let createdBy = game["createdBy"]
// say we have a Game object
$game = ...

// getting the user who created the Game
$user = $game["createdBy"];
// say we have a Game object
ParseObject game = ...

// getting the user who created the Game
ParseUser user = game["createdBy"];
// say we have a Game object
var game = ...

// getting the user who created the Game
var user = game.get("createdBy");
# No REST API example
// No C++ example

For most scenarios, Pointers will be your best bet for implementing one-to-many relationships.

Using Arrays

Arrays are ideal when we know that the number of objects involved in our one-to-many relationship are going to be small. Arrays will also provide some productivity benefit via the includeKey parameter. Supplying the parameter will enable you to obtain all of the "many" objects in the "one-to-many" relationship at the same time that you obtain the "one" object. However, the response time will be slower if the number of objects involved in the relationship turns out to be large.

Suppose in our game, we enabled players to keep track of all the weapons their character has accumulated as they play, and there can only be a dozen or so weapons. In this example, we know that the number of weapons is not going to be very large. We also want to enable the player to specify the order in which the weapons will appear on screen. Arrays are ideal here because the size of the array is going to be small and because we also want to preserve the order the user has set each time they play the game:

Let's start by creating a column on our Parse User object called weaponsList.

Now let's store some Weapon objects in the weaponsList:

// let's say we have four weapons
ParseObject scimitar = ...
ParseObject plasmaRifle = ...
ParseObject grenade = ...
ParseObject bunnyRabbit = ...

// stick the objects in an array
ArrayList<ParseObject> weapons = new ArrayList<ParseObject>();
weapons.add(scimitar);
weapons.add(plasmaRifle);
weapons.add(grenade);
weapons.add(bunnyRabbit);

// store the weapons for the user
ParseUser.getCurrentUser().put("weaponsList", weapons);
// let's say we have four weapons
PFObject *scimitar = ...
PFObject *plasmaRifle = ...
PFObject *grenade = ...
PFObject *bunnyRabbit = ...

// stick the objects in an array
NSArray *weapons = @[scimitar, plasmaRifle, grenade, bunnyRabbit];

// store the weapons for the user
[[PFUser currentUser] setObject:weapons forKey:@weaponsList"];
// let's say we have four weapons
let scimitar = ...
let plasmaRifle = ...
let grenade = ...
let bunnyRabbit = ...

// stick the objects in an array
let weapons = [scimitar, plasmaRifle, grenade, bunnyRabbit]

// store the weapons for the user
let user = PFUser.currentUser()
user["weaponsList"] = weapons
// let's say we have four weapons
$scimitar = ...
$plasmaRifle = ...
$grenade = ...
$bunnyRabbit = ...

// stick the objects in an array
$weapons = [$scimitar, $plasmaRifle, $grenade, $bunnyRabbit];

// store the weapons for the user
$user = ParseUser::getCurrentUser();
$user->set("weaponsList", weapons);
// let's say we have four weapons
var scimitar = ...
var plasmaRifle = ...
var grenade = ...
var bunnyRabbit = ...

// stick the objects in an array
var weapons = new List<ParseObject>();
weapons.Add(scimitar);
weapons.Add(plasmaRifle);
weapons.Add(grenade);
weapons.Add(bunnyRabbit);

// store the weapons for the user
var user = ParseUser.CurrentUser;
user.AddRangeToList("weaponsList", weapons);
// let's say we have four weapons
var scimitar = ...
var plasmaRifle = ...
var grenade = ...
var bunnyRabbit = ...

// stick the objects in an array
var weapons = [scimitar, plasmaRifle, grenade, bunnyRabbit];

// store the weapons for the user
var user = Parse.User.current();
user.set("weaponsList", weapons);
# No REST API example
// No C++ example

Later, if we want to retrieve the Weapon objects, it's just one line of code:

ArrayList<ParseObject> weapons = ParseUser.getCurrentUser().get("weaponsList");
NSArray *weapons = [[PFUser currentUser] objectForKey:@"weaponsList"];
let weapons = PFUser.currentUser()?.objectForKey("weaponsList")
$weapons = ParseUser::getCurrentUser()->get("weaponsList");
var weapons = ParseUser.CurrentUser.Get<IList<Object>>("weaponsList");
var weapons = Parse.User.current().get("weaponsList")
# No REST API example
// No C++ example

Sometimes, we will want to fetch the "many" objects in our one-to-many relationship at the same time as we fetch the "one" object. One trick we could employ is to use the includeKey (or include in Android) parameter whenever we use a Parse Query to also fetch the array of Weapon objects (stored in the weaponsList column) along with the Parse User object:

// set up our query for a User object
ParseQuery<ParseUser> userQuery = ParseUser.getQuery();

// configure any constraints on your query...
// for example, you may want users who are also playing with or against you
// tell the query to fetch all of the Weapon objects along with the user
// get the "many" at the same time that you're getting the "one"
userQuery.include("weaponsList");

// execute the query
userQuery.findInBackground(new FindCallback<ParseUser>() {
  public void done(List<ParseUser> userList, ParseException e) {
    // userList contains all of the User objects, and their associated Weapon objects, too
  }
});
// set up our query for a User object
PFQuery *userQuery = [PFUser query];

// configure any constraints on your query...
// for example, you may want users who are also playing with or against you

// tell the query to fetch all of the Weapon objects along with the user
// get the "many" at the same time that you're getting the "one"
[userQuery includeKey:@"weaponsList"];

// execute the query
[userQuery findObjectsInBackgroundWithBlock:^(NSArray *objects, NSError *error) {
    // objects contains all of the User objects, and their associated Weapon objects, too
}];
// set up our query for a User object
let userQuery = PFUser.query();

// configure any constraints on your query...
// for example, you may want users who are also playing with or against you

// tell the query to fetch all of the Weapon objects along with the user
// get the "many" at the same time that you're getting the "one"
userQuery?.includeKey("weaponsList");

// execute the query
userQuery?.findObjectsInBackgroundWithBlock {
    (objects: [AnyObject]?, error: NSError?) -> Void in
    // objects contains all of the User objects, and their associated Weapon objects, too
}
// set up our query for a User object
$userQuery = ParseUser::query();

// configure any constraints on your query...
// for example, you may want users who are also playing with or against you

// tell the query to fetch all of the Weapon objects along with the user
// get the "many" at the same time that you're getting the "one"
$userQuery->includeKey("weaponsList");

// execute the query
$results = $userQuery->find();
// results contains all of the User objects, and their associated Weapon objects, too
// set up our query for a User object
var userQuery = ParseUser.Query;

// configure any constraints on your query...
// for example, you may want users who are also playing with or against you

// tell the query to fetch all of the Weapon objects along with the user
// get the "many" at the same time that you're getting the "one"
userQuery = userQuery.Include("weaponsList");

// execute the query
IEnumerable<ParseUser> results = await userQuery.FindAsync();
// results contains all of the User objects, and their associated Weapon objects, too
// set up our query for a User object
var userQuery = new Parse.Query(Parse.User);

// configure any constraints on your query...
// for example, you may want users who are also playing with or against you

// tell the query to fetch all of the Weapon objects along with the user
// get the "many" at the same time that you're getting the "one"
userQuery.include("weaponsList");

// execute the query
userQuery.find({
  success: function(results){
    // results contains all of the User objects, and their associated Weapon objects, too
  }
});
# No REST API example
// No C++ example

You can also get the "one" side of the one-to-many relationship from the "many" side. For example, if we want to find all Parse User objects who also have a given Weapon, we can write a constraint for our query like this:

// add a constraint to query for whenever a specific Weapon is in an array
userQuery.whereEqualTo("weaponsList", scimitar);

// or query using an array of Weapon objects...
userQuery.whereEqualTo("weaponsList", arrayOfWeapons);
// add a constraint to query for whenever a specific Weapon is in an array
[userQuery whereKey:@"weaponsList" equalTo:scimitar];

// or query using an array of Weapon objects...
[userQuery whereKey:@"weaponsList" containedIn:arrayOfWeapons];
// add a constraint to query for whenever a specific Weapon is in an array
userQuery?.whereKey("weaponsList", equalTo: scimitar);

// or query using an array of Weapon objects...
userQuery?.whereKey("weaponsList", containedIn: arrayOfWeapons)
// add a constraint to query for whenever a specific Weapon is in an array
$userQuery->equalTo("weaponsList", $scimitar);

// or query using an array of Weapon objects...
$userQuery->containedIn("weaponsList", $arrayOfWeapons);
// add a constraint to query for whenever a specific Weapon is in an array
userQuery = userQuery.WhereEqualTo("weaponsList", scimitar);

// or query using an array of Weapon objects...
userQuery = userQuery.WhereContainedIn("weaponsList", arrayOfWeapons);
// add a constraint to query for whenever a specific Weapon is in an array
userQuery.equalTo("weaponsList", scimitar);

// or query using an array of Weapon objects...
userQuery.containedIn("weaponsList", arrayOfWeapons);
# No REST API example
// No C++ example

Many-to-Many

Now let’s tackle many-to-many relationships. Suppose we had a book reading app and we wanted to model Book objects and Author objects. As we know, a given author can write many books, and a given book can have multiple authors. This is a many-to-many relationship scenario where you have to choose between Arrays, Parse Relations, or creating your own Join Table.

The decision point here is whether you want to attach any metadata to the relationship between two entities. If you don’t, Parse Relation or using Arrays are going to be the easiest alternatives. In general, using arrays will lead to higher performance and require fewer queries. If either side of the many-to-many relationship could lead to an array with more than 100 or so objects, then, for the same reason Pointers were better for one-to-many relationships, Parse Relation or Join Tables will be better alternatives.

On the other hand, if you want to attach metadata to the relationship, then create a separate table (the "Join Table") to house both ends of the relationship. Remember, this is information about the relationship, not about the objects on either side of the relationship. Some examples of metadata you may be interested in, which would necessitate a Join Table approach, include:

Using Parse Relations

Using Parse Relations, we can create a relationship between a Book and a few Author objects. In the Data Browser, you can create a column on the Book object of type relation and name it authors.

After that, we can associate a few authors with this book:

// let’s say we have a few objects representing Author objects
ParseObject authorOne =
ParseObject authorTwo =
ParseObject authorThree =

// now we create a book object
ParseObject book = new ParseObject("Book");

// now let’s associate the authors with the book
// remember, we created a "authors" relation on Book
ParseRelation<ParseObject> relation = book.getRelation("authors");
relation.add(authorOne);
relation.add(authorTwo);
relation.add(authorThree);

// now save the book object
book.saveInBackground();
// let’s say we have a few objects representing Author objects
PFObject *authorOne = …
PFObject *authorTwo = …
PFObject *authorThree = …

// now we create a book object
PFObject *book= [PFObject objectWithClassName:@"Book"];

// now let’s associate the authors with the book
// remember, we created a "authors" relation on Book
PFRelation *relation = [book relationForKey:@"authors"];
[relation addObject:authorOne];
[relation addObject:authorTwo];
[relation addObject:authorThree];

// now save the book object
[book saveInBackground];
// let’s say we have a few objects representing Author objects
let authorOne = ...
let authorTwo = ...
let authorThree = ...

// now we create a book object
let book = PFObject(className: "Book")

// now let’s associate the authors with the book
// remember, we created a "authors" relation on Book
let relation = book.relationForKey("authors")
relation.addObject(authorOne)
relation.addObject(authorTwo)
relation.addObject(authorThree)

// now save the book object
book.saveInBackground()
// let’s say we have a few objects representing Author objects
$authorOne = ...
$authorTwo = ...
$authorThree = ...

// now we create a book object
$book = new ParseObject("Book");

// now let’s associate the authors with the book
// remember, we created a "authors" relation on Book
$relation = $book->getRelation("authors");
$relation->add($authorOne);
$relation->add($authorTwo);
$relation->add($authorThree);

// now save the book object
$book->save();
// let’s say we have a few objects representing Author objects
var authorOne = ...
var authorTwo = ...
var authorThree = ...

// now we create a book object
var book = new ParseObject("Book");

// now let’s associate the authors with the book
// remember, we created a "authors" relation on Book
var relation = book.GetRelation<ParseObject>("authors");
relation.Add(authorOne);
relation.Add(authorTwo);
relation.Add(authorThree);

// now save the book object
await book.SaveAsync();
// let’s say we have a few objects representing Author objects
var authorOne = ...
var authorTwo = ...
var authorThree = ...

// now we create a book object
var book = new Parse.Object("Book");

// now let’s associate the authors with the book
// remember, we created a "authors" relation on Book
var relation = book.relation("authors");
relation.add(authorOne);
relation.add(authorTwo);
relation.add(authorThree);

// now save the book object
book.save();
# No REST API example
// No C++ example

To get the list of authors who wrote a book, create a query:

// suppose we have a book object
ParseObject book = ...

// create a relation based on the authors key
ParseRelation relation = book.getRelation("authors");

// generate a query based on that relation
ParseQuery query = relation.getQuery();

// now execute the query
// suppose we have a book object
PFObject *book = ...

// create a relation based on the authors key
PFRelation *relation = [book relationForKey:@"authors"];

// generate a query based on that relation
PFQuery *query = [relation query];

// now execute the query
// suppose we have a book object
let book = ...

// create a relation based on the authors key
let relation = book.relationForKey("authors")

// generate a query based on that relation
let query = relation.query()

// now execute the query
// suppose we have a book object
$book = ...

// create a relation based on the authors key
$relation = $book->getRelation("authors");

// generate a query based on that relation
$query = $relation->getQuery();

// now execute the query
// suppose we have a book object
var book = ...

// create a relation based on the authors key
var relation = book.GetRelation<ParseObject>("authors");

// generate a query based on that relation
var query = relation.Query;

// now execute the query
// suppose we have a book object
var book = ...

// create a relation based on the authors key
var relation = book.relation("authors");

// generate a query based on that relation
var query = relation.query();

// now execute the query
# No REST API example
// No C++ example

Perhaps you even want to get a list of all the books to which an author contributed. You can create a slightly different kind of query to get the inverse of the relationship:

// suppose we have a author object, for which we want to get all books
ParseObject author = ...

// first we will create a query on the Book object
ParseQuery<ParseObject> query = ParseQuery.getQuery("Book");

// now we will query the authors relation to see if the author object we have
// is contained therein
query.whereEqualTo("authors", author);
// suppose we have a author object, for which we want to get all books
PFObject *author = ...

// first we will create a query on the Book object
PFQuery *query = [PFQuery queryWithClassName:@"Book"];

// now we will query the authors relation to see if the author object
// we have is contained therein
[query whereKey:@"authors" equalTo:author];
// suppose we have a author object, for which we want to get all books
let author = ...

// first we will create a query on the Book object
let query = PFQuery(className: "Book")

// now we will query the authors relation to see if the author object
// we have is contained therein
query?.whereKey("authors", equalTo: author)
// suppose we have a author object, for which we want to get all books
$author = ...

// first we will create a query on the Book object
$query = new ParseQuery("Book");

// now we will query the authors relation to see if the author object we have
// is contained therein
$query->equalTo("authors", $author);
// suppose we have a author object, for which we want to get all books
var author = ...

// first we will create a query on the Book object
var query = ParseObject.GetQuery("Book");

// now we will query the authors relation to see if the author object we have
// is contained therein
query = query.WhereEqualTo("authors", author);
// suppose we have a author object, for which we want to get all books
var author = ...

// first we will create a query on the Book object
var query = new Parse.Query("Book");

// now we will query the authors relation to see if the author object we have
// is contained therein
query.equalTo("authors", author);
# No REST API example
// No C++ example

Using Join Tables

There may be certain cases where we want to know more about a relationship. For example, suppose we were modeling a following/follower relationship between users: a given user can follow another user, much as they would in popular social networks. In our app, we not only want to know if User A is following User B, but we also want to know when User A started following User B. This information could not be contained in a Parse Relation. In order to keep track of this data, you must create a separate table in which the relationship is tracked. This table, which we will call Follow, would have a from column and a to column, each with a pointer to a Parse User. Alongside the relationship, you can also add a column with a Date object named date.

Now, when you want to save the following relationship between two users, create a row in the Follow table, filling in the from, to, and date keys appropriately:

// suppose we have a user we want to follow
ParseUser otherUser = ...

// create an entry in the Follow table
ParseObject follow = new ParseObject("Follow");
follow.put("from", ParseUser.getCurrentUser());
follow.put("to", otherUser);
follow.put("date", Date());
follow.saveInBackground();
// suppose we have a user we want to follow
PFUser *otherUser = ...

// create an entry in the Follow table
PFObject *follow = [PFObject objectWithClassName:@"Follow"];
[follow setObject:[PFUser currentUser]  forKey:@"from"];
[follow setObject:otherUser forKey:@"to"];
[follow setObject:[NSDate date] forKey@"date"];
[follow saveInBackground];
// suppose we have a user we want to follow
let otherUser = ...

// create an entry in the Follow table
let follow = PFObject(className: "Follow")
follow.setObject(PFUser.currentUser()!, forKey: "from")
follow.setObject(otherUser, forKey: "to")
follow.setObject(NSDate(), forKey: "date")
follow.saveInBackground()
// suppose we have a user we want to follow
$otherUser = ...

// create an entry in the Follow table
$follow = new ParseObject("Follow");
$follow->set("from", ParseUser::getCurrentUser());
$follow->set("to", $otherUser);
$follow->set("date", new DateTime());
$follow->save();
// suppose we have a user we want to follow
ParseUser otherUser = ...

// create an entry in the Follow table
var follow = new ParseObject("Follow");
follow["from"] = ParseUser.CurrentUser;
follow["to"] = otherUser;
follow["date"] = DateTime.UtcNow;
await follow.SaveAsync();
var otherUser = ...

// create an entry in the Follow table
var follow = new Parse.Object("Follow");
follow.set("from", Parse.User.current());
follow.set("to", otherUser);
follow.set("date", Date());
follow.save();
# No REST API example
// No C++ example

If we want to find all of the people we are following, we can execute a query on the Follow table:

// set up the query on the Follow table
ParseQuery<ParseObject> query = ParseQuery.getQuery("Follow");
query.whereEqualTo("from", ParseUser.getCurrentUser());

// execute the query
query.findInBackground(newFindCallback<ParseObject>() {
    public void done(List<ParseObject> followList, ParseException e) {

    }
});
// set up the query on the Follow table
PFQuery *query = [PFQuery queryWithClassName:@"Follow"];
[query whereKey:@"from" equalTo:[PFUser currentUser]];

// execute the query
[query findObjectsInBackgroundWithBlock:^(NSArray *objects, NSError *error) {
  for(PFObject *o in objects) {
    // o is an entry in the Follow table
    // to get the user, we get the object with the to key
    PFUser *otherUser = [o objectForKey@"to"];

    // to get the time when we followed this user, get the date key
    PFObject *when = [o objectForKey@"date"];
  }
}];
// set up the query on the Follow table
let query = PFQuery(className: "Follow")
query.whereKey("from", equalTo: PFUser.currentUser()!)

// execute the query
query.findObjectsInBackgroundWithBlock{
    (objects: [AnyObject]?, error: NSError?) -> Void in
    if let objects = objects {
        for o in objects {
            // o is an entry in the Follow table
            // to get the user, we get the object with the to key
            let otherUse = o.objectForKey("to") as? PFUser

            // to get the time when we followed this user, get the date key
            let when = o.objectForKey("date") as? PFObject
        }
    }
}
// set up the query on the Follow table
$query = new ParseQuery("Follow");
$query->equalTo("from", ParseUser::getCurrentUser());

// execute the query
$results = $query->find();
// set up the query on the Follow table
ParseQuery<ParseObject> query = ParseQuery.getQuery("Follow");
query = query.WhereEqualTo("from", ParseUser.CurrentUser);

// execute the query
IEnumerable<ParseObject> results = await query.FindAsync();
var query = new Parse.Query("Follow");
query.equalTo("from", Parse.User.current());
query.find({
  success: function(users){
    ...
  }
});
# No REST API example
// No C++ example

It’s also pretty easy to find all the users that are following the current user by querying on the to key:

// set up the query on the Follow table
ParseQuery<ParseObject> query = ParseQuery.getQuery("Follow");
query.whereEqualTo("to", ParseUser.getCurrentUser());

// execute the query
query.findInBackground(newFindCallback<ParseObject>() {
    public void done(List<ParseObject> followList, ParseException e) {

    }
});
// set up the query on the Follow table
PFQuery *query = [PFQuery queryWithClassName:@"Follow"];
[query whereKey:@"to" equalTo:[PFUser currentUser]];

[query findObjectsInBackgroundWithBlock:^(NSArray *objects, NSError *error) {
  for(PFObject *o in objects) {
     // o is an entry in the Follow table
     // to get the user, we get the object with the from key
    PFUser *otherUser = [o objectForKey@"from"];

    // to get the time the user was followed, get the date key
    PFObject *when = [o objectForKey@"date"];
  }
}];
// set up the query on the Follow table
let query = PFQuery(className: "Follow")
query.whereKey("to", equalTo: PFUser.currentUser()!)

query.findObjectsInBackgroundWithBlock{
    (objects: [AnyObject]?, error: NSError?) -> Void in
    if let objects = objects {
        for o in objects {
            // o is an entry in the Follow table
            // to get the user, we get the object with the to key
            let otherUse = o.objectForKey("to") as? PFUser

            // to get the time when we followed this user, get the date key
            let when = o.objectForKey("date") as? PFObject
        }

    }
}
// create an entry in the Follow table
$query = new ParseQuery("Follow");
$query->equalTo("to", ParseUser::getCurrentUser());
$results = $query->find();
// create an entry in the Follow table
var query = ParseObject.GetQuery("Follow")
    .WhereEqualTo("to", ParseUser.CurrentUser);
IEnumerable<ParseObject> results = await query.FindAsync();
// create an entry in the Follow table
var query = new Parse.Query("Follow");
query.equalTo("to", Parse.User.current());
query.find({
  success: function(users){
    ...
  }
});
# No REST API example
// No C++ example

Using an Array

Arrays are used in Many-to-Many relationships in much the same way that they are for One-to-Many relationships. All objects on one side of the relationship will have an Array column containing several objects on the other side of the relationship.

Suppose we have a book reading app with Book and Author objects. The Book object will contain an Array of Author objects (with a key named authors). Arrays are a great fit for this scenario because it's highly unlikely that a book will have more than 100 or so authors. We will put the Array in the Book object for this reason. After all, an author could write more than 100 books.

Here is how we save a relationship between a Book and an Author.

// let's say we have an author
ParseObject author = ...

// and let's also say we have an book
ParseObject book = ...

// add the author to the authors list for the book
book.put("authors", author);
// let's say we have an author
PFObject *author = ...

// and let's also say we have an book
PFObject *book = ...

// add the author to the authors list for the book
[book addObject:author forKey:@"authors"];
// let's say we have an author
let author = ...

// and let's also say we have an book
let book = ...

// add the author to the authors list for the book
book.addObject(author, forKey: "authors")
// let's say we have an author
$author = ...

// and let's also say we have an book
$book = ...

// add the author to the authors list for the book
$book->addUnique("authors", array($author));
// let's say we have an author
var author = ...

// and let's also say we have an book
var book = ...

// add the author to the authors list for the book
book.AddToList("authors", author);
// let's say we have an author
var author = ...

// and let's also say we have an book
var book = ...

// add the author to the authors list for the book
book.add("authors", author);
# No REST API example
// No C++ example

Because the author list is an Array, you should use the includeKey (or include on Android) parameter when fetching a Book so that Parse returns all the authors when it also returns the book:

// set up our query for the Book object
ParseQuery bookQuery = ParseQuery.getQuery("Book");

// configure any constraints on your query...
// tell the query to fetch all of the Author objects along with the Book
bookQuery.include("authors");

// execute the query
bookQuery.findInBackground(newFindCallback<ParseObject>() {
    public void done(List<ParseObject> bookList, ParseException e) {
    }
});
// set up our query for the Book object
PFQuery *bookQuery = [PFQuery queryWithClassName:@"Book"];

// configure any constraints on your query...
// tell the query to fetch all of the Author objects along with the Book
[bookQuery includeKey:@"authors"];

// execute the query
[bookQuery findObjectsInBackgroundWithBlock:^(NSArray *objects, NSError *error) {
    // objects is all of the Book objects, and their associated
    // Author objects, too
}];
// set up our query for the Book object
let bookQuery = PFQuery(className: "Book")

// configure any constraints on your query...
// tell the query to fetch all of the Author objects along with the Book
bookQuery.includeKey("authors")

// execute the query
bookQuery.findObjectsInBackgroundWithBlock{
    (objects: [AnyObject]?, error: NSError?) -> Void in
    // objects is all of the Book objects, and their associated
    // Author objects, too
}
// set up our query for the Book object
$bookQuery = new ParseQuery("Book");

// configure any constraints on your query...
// tell the query to fetch all of the Author objects along with the Book
$bookQuery->includeKey("authors");

// execute the query
$books= $bookQuery->find();
// set up our query for the Book object
var bookQuery = ParseObject.GetQuery("Book");

// configure any constraints on your query...
// tell the query to fetch all of the Author objects along with the Book
bookQuery = bookQuery.Include("authors");

// execute the query
IEnumerable<ParseObject> books= await bookQuery.FindAsync();
// set up our query for the Book object
var bookQuery = new Parse.Query("Book");

// configure any constraints on your query...
// tell the query to fetch all of the Author objects along with the Book
bookQuery.include("authors");

// execute the query
bookQuery.find({
  success: function(books){
    ...
  }
});
# No REST API example
// No C++ example

At that point, getting all the Author objects in a given Book is a pretty straightforward call:

ArrayList<ParseObject> authorList = book.getList("authors");
NSArray *authorList = [book objectForKey@"authors"];
let authorList = book.objectForKey("authors") as? NSArray
$authorList = $book->get("authors");
var authorList = book.Get<List<ParseObject>>("authors");
var authorList = book.get("authors")
# No REST API example
// No C++ example

Finally, suppose you have an Author and you want to find all the Book objects in which she appears. This is also a pretty straightforward query with an associated constraint:

// set up our query for the Book object
ParseQuery bookQuery = ParseQuery.getQuery("Book");

// configure any constraints on your query...
booKQuery.whereEqualTo("authors", author);

// tell the query to fetch all of the Author objects along with the Book
bookQuery.include("authors");

// execute the query
bookQuery.findInBackground(newFindCallback<ParseObject>() {
    public void done(List<ParseObject> bookList, ParseException e) {

    }
});
// suppose we have an Author object
PFObject *author = ...

// set up our query for the Book object
PFQuery *bookQuery = [PFQuery queryWithClassName:@"Book"];

// configure any constraints on your query...
[bookQuery whereKey:@"authors" equalTo:author];

// tell the query to fetch all of the Author objects along with the Book
[bookQuery includeKey:@"authors"];

// execute the query
[bookQuery findObjectsInBackgroundWithBlock:^(NSArray *objects, NSError *error) {
    // objects is all of the Book objects, and their associated Author objects, too
}];
// suppose we have an Author object
let author = ...

// set up our query for the Book object
let bookQuery = PFQuery(className: "Book")

// configure any constraints on your query...
bookQuery.whereKey("authors", equalTo: author)

// tell the query to fetch all of the Author objects along with the Book
bookQuery.includeKey("authors")

// execute the query
bookQuery.findObjectsInBackgroundWithBlock{
    (objects: [AnyObject]?, error: NSError?) -> Void in
    // objects is all of the Book objects, and their associated Author objects, too
}
// set up our query for the Book object
$bookQuery = new ParseQuery("Book");

// configure any constraints on your query...
$bookQuery->equalTo("authors", $author);

// tell the query to fetch all of the Author objects along with the Book
$bookQuery->includeKey("authors");

// execute the query
$books = $bookQuery->find();
// set up our query for the Book object
var bookQuery = ParseObject.GetQuery("Book");

// configure any constraints on your query...
bookQuery = bookQuery.WhereEqualTo("authors", author);

// tell the query to fetch all of the Author objects along with the Book
bookQuery = bookQuery.Include("authors");

// execute the query
IEnumerable<ParseObject> books = await bookQuery.FindAsync();
// set up our query for the Book object
var bookQuery = new Parse.Query("Book");

// configure any constraints on your query...
bookQuery.equalTo("authors", author);

// tell the query to fetch all of the Author objects along with the Book
bookQuery.include("authors");

// execute the query
bookQuery.find({
  success: function(books){
    ...
  }
});
# No REST API example
// No C++ example

One-to-One

In Parse, a one-to-one relationship is great for situations where you need to split one object into two objects. These situations should be rare, but two examples include:

  • Limiting visibility of some user data. In this scenario, you would split the object in two, where one portion of the object contains data that is visible to other users, while the related object contains data that is private to the original user (and protected via ACLs).
  • Splitting up an object for size. In this scenario, your original object is greater than the 128K maximum size permitted for an object, so you decide to create a secondary object to house extra data. It is usually better to design your data model to avoid objects this large, rather than splitting them up. If you can't avoid doing so, you can also consider storing large data in a Parse File.

Thank you for reading this far. We apologize for the complexity. Modeling relationships in data is a hard subject, in general. But look on the bright side: it's still easier than relationships with people.

Want to contribute to this doc? Edit this section.
Was this section helpful? NO YES Thanks for your feedback, we’re always working to make our docs better.

Handling Errors

The Parse PHP SDK throws ParseExceptions when errors are returned from the Parse API. For other errors, the base Exception class will be thrown. It is recommended to wrap your Parse calls in try/catch blocks to handle any errors which occur.

$query = new ParseQuery("Note");
try {
  // This will throw a ParseException, the object is not found.
  $result = $query->get("aBcDeFgH")
} catch (ParseException $error) {
  // $error is an instance of ParseException with details about the error.
  echo $error->getCode();
  echo $error->getMessage();
}

For a list of all possible error codes, scroll down to Error Codes.

Want to contribute to this doc? Edit this section.
Was this section helpful? NO YES Thanks for your feedback, we’re always working to make our docs better.

Security

As your app development progresses, you will want to use Parse's security features in order to safeguard data. This document explains the ways in which you can secure your apps.

If your app is compromised, it's not only you as the developer who suffers, but potentially the users of your app as well. Continue reading for our suggestions for sensible defaults and precautions to take before releasing your app into the wild.

Client vs. Server

When an app first connects to Parse, it identifies itself with an Application ID and a Client key (or REST Key, or .NET Key, or JavaScript Key, depending on which platform you're using). These are not secret and by themselves they do not secure an app. These keys are shipped as a part of your app, and anyone can decompile your app or proxy network traffic from their device to find your client key. This exploit is even easier with JavaScript — one can simply "view source" in the browser and immediately find your client key.

This is why Parse has many other security features to help you secure your data. The client key is given out to your users, so anything that can be done with just the client key is doable by the general public, even malicious hackers.

The master key, on the other hand, is definitely a security mechanism. Using the master key allows you to bypass all of your app's security mechanisms, such as class-level permissions and ACLs. Having the master key is like having root access to your app's servers, and you should guard your master key with the same zeal with which you would guard your production machines' root password.

The overall philosophy is to limit the power of your clients (using client keys), and to perform any sensitive actions requiring the master key in Cloud Code. You'll learn how to best wield this power in the section titled Implementing Business Logic in Cloud Code.

A final note: All connections are made with HTTPS and SSL, and Parse will reject all non-HTTPS connections. As a result, you don't need to worry about man-in-the-middle attacks.

Class-Level Permissions

The second level of security is at the schema and data level. Enforcing security measures at this level will restrict how and when client applications can access and create data on Parse. When you first begin developing your Parse application, all of the defaults are set so that you can be a more productive developer. For example:

  • A client application can create new classes on Parse
  • A client application can add fields to classes
  • A client application can modify or query for objects on Parse

You can configure any of these permissions to apply to everyone, no one, or to specific users or roles in your app. Roles are groups that contain users or other roles, which you can assign to an object to restrict its use. Any permission granted to a role is also granted to any of its children, whether they are users or other roles, enabling you to create an access hierarchy for your apps. Each of the Parse guides includes a detailed description of employing Roles in your apps.

Once you are confident that you have the right classes and relationships between classes in your app, you should begin to lock it down by doing the following:

Almost every class that you create should have these permissions tweaked to some degree. For classes where every object has the same permissions, class-level settings will be most effective. For example, one common use case entails having a class of static data that can be read by anyone but written by no one.

Restricting class creation

As a start, you can configure your application so that clients cannot create new classes on Parse. This is done from the Settings tab on the Data Browser. Scroll down to the App Permissions section and turn off Allow client class creation. Once enabled, classes may only be created from the Data Browser. This will prevent attackers from filling your database with unlimited, arbitrary new classes.

Configuring Class-Level Permissions

Parse lets you specify what operations are allowed per class. This lets you restrict the ways in which clients can access or modify your classes. To change these settings, go to the Data Browser, select a class, and click the "Security" button.

You can configure the client's ability to perform each of the following operations for the selected class:

  • Read:

    • Get: With Get permission, users can fetch objects in this table if they know their objectIds.

    • Find: Anyone with Find permission can query all of the objects in the table, even if they don’t know their objectIds. Any table with public Find permission will be completely readable by the public, unless you put an ACL on each object.

  • Write:

    • Update: Anyone with Update permission can modify the fields of any object in the table that doesn't have an ACL. For publicly readable data, such as game levels or assets, you should disable this permission.

    • Create: Like Update, anyone with Create permission can create new objects of a class. As with the Update permission, you'll probably want to turn this off for publicly readable data.

    • Delete: With this permission, people can delete any object in the table that doesn't have an ACL. All they need is its objectId.

  • Add fields: Parse classes have schemas that are inferred when objects are created. While you're developing your app, this is great, because you can add a new field to your object without having to make any changes on the backend. But once you ship your app, it's very rare to need to add new fields to your classes automatically. You should pretty much always turn off this permission for all of your classes when you submit your app to the public.

For each of the above actions, you can grant permission to all users (which is the default), or lock permissions down to a list of roles and users. For example, a class that should be available to all users would be set to read-only by only enabling get and find. A logging class could be set to write-only by only allowing creates. You could enable moderation of user-generated content by providing update and delete access to a particular set of users or roles.

Object-Level Access Control

Once you've locked down your schema and class-level permissions, it's time to think about how data is accessed by your users. Object-level access control enables one user's data to be kept separate from another's, because sometimes different objects in a class need to be accessible by different people. For example, a user’s private personal data should be accessible only to them.

Parse also supports the notion of anonymous users for those apps that want to store and protect user-specific data without requiring explicit login.

When a user logs into an app, they initiate a session with Parse. Through this session they can add and modify their own data but are prevented from modifying other users' data.

Access Control Lists

The easiest way to control who can access which data is through access control lists, commonly known as ACLs. The idea behind an ACL is that each object has a list of users and roles along with what permissions that user or role has. A user needs read permissions (or must belong to a role that has read permissions) in order to retrieve an object's data, and a user needs write permissions (or must belong to a role that has write permissions) in order to update or delete that object.

Once you have a User, you can start using ACLs. Remember: Users can be created through traditional username/password signup, through a third-party login system like Facebook or Twitter, or even by using Parse's automatic anonymous users functionality. To set an ACL on the current user's data to not be publicly readable, all you have to do is:

PFUser *user = [PFUser currentUser];
user.ACL = [PFACL ACLWithUser:user];
if let user = PFUser.currentUser() {
    user.ACL = PFACL(user: user)
}
ParseUser user = ParseUser.getCurrentUser();
user.setACL(new ParseACL(user));
var user = Parse.User.current();
user.setACL(new Parse.ACL(user));
var user = ParseUser.CurrentUser;
user.ACL = new ParseACL(user);
$user = ParseUser::getCurrentUser();
$user->setACL(new ParseACL($user))
# No command line example
// No C++ example

Most apps should do this. If you store any sensitive user data, such as email addresses or phone numbers, you need to set an ACL like this so that the user's private information isn't visible to other users. If an object doesn't have an ACL, it's readable and writeable by everyone. The only exception is the _User class. We never allow users to write each other's data, but they can read it by default. (If you as the developer need to update other _User objects, remember that your master key can provide the power to do this.)

To make it super easy to create user-private ACLs for every object, we have a way to set a default ACL that will be used for every new object you create:

[PFACL setDefaultACL:[PFACL ACL] withAccessForCurrentUser:YES];
PFACL.setDefaultACL(PFACL(), withAccessForCurrentUser: true)
ParseACL.setDefaultACL(new ParseACL(), true);
// not available in the JavaScript SDK
// not available in the .NET SDK
ParseACL::setDefaultACL(new ParseACL(), true);
# No command line example
// No C++ example

If you want the user to have some data that is public and some that is private, it's best to have two separate objects. You can add a pointer to the private data from the public one.

PFObject *privateData = [PFObject objectWithClassName:@"PrivateUserData"];
privateData.ACL = [PFACL ACLWithUser:[PFUser currentUser]];
[privateData setObject:@"555-5309" forKey:@"phoneNumber"];

[[PFUser currentUser] setObject:privateData forKey:@"privateData"];
if let currentUser = PFUser.currentUser() {
    let privateData = PFObject(className: "PrivateUserData")
    privateData.ACL = PFACL(user: currentUser)
    privateData.setObject("555-5309", forKey: "phoneNumber")
    currentUser.setObject(privateData, forKey: "privateData")
}
ParseObject privateData = new ParseObject("PrivateUserData");
privateData.setACL(new ParseACL(ParseUser.getCurrentUser()));
privateData.put("phoneNumber", "555-5309");

ParseUser.getCurrentUser().put("privateData", privateData);
var privateData = Parse.Object.extend("PrivateUserData");
privateData.setACL(new Parse.ACL(Parse.User.current()));
privateData.set("phoneNumber", "555-5309");

Parse.User.current().set("privateData", privateData);
var privateData = new ParseObject("PrivateUserData");
privateData.ACL = new ParseACL(ParseUser.CurrentUser);
privateData["phoneNumber"] = "555-5309";

ParseUser.CurrentUser["privateData"] =  privateData;
$privateData = ParseObject::create("PrivateUserData");
$privateData->setACL(new ParseACL(ParseUser::getCurrentUser()));
$privateData->set("phoneNumber", "555-5309");

ParseUser::getCurrentUser()->set("privateData", $privateData);
# No command line example
// No C++ example

Of course, you can set different read and write permissions on an object. For example, this is how you would create an ACL for a public post by a user, where anyone can read it:

PFACL *acl = [PFACL ACL];
[acl setPublicReadAccess:true];
[acl setWriteAccess:true forUser:[PFUser currentUser]];
let acl = PFACL()
acl.setPublicReadAccess(true)
if let currentUser = PFUser.currentUser() {
    acl.setWriteAccess(true, forUser: currentUser)
}
ParseACL acl = new ParseACL();
acl.setPublicReadAccess(true);
acl.setWriteAccess(ParseUser.getCurrentUser(), true);
var acl = new Parse.ACL();
acl.setPublicReadAccess(true);
acl.setWriteAccess(Parse.User.current().id, true);
var acl = new ParseACL();
acl.PublicReadAccess = true;
acl.SetRoleWriteAccess(ParseUser.CurrentUser.ObjectId, true);
$acl = new ParseACL();
$acl->setPublicReadAccess(true);
$acl->setWriteAccess(ParseUser::getCurrentUser(), true);
# No command line example
// No C++ example

Sometimes it's inconvenient to manage permissions on a per-user basis, and you want to have groups of users who get treated the same (like a set of admins with special powers). Roles are are a special kind of object that let you create a group of users that can all be assigned to the ACL. The best thing about roles is that you can add and remove users from a role without having to update every single object that is restricted to that role. To create an object that is writeable only by admins:

// Assuming you've already created a role called "admins"...
PFACL *acl = [PFACL ACL];
[acl setPublicReadAccess:true];
[acl setWriteAccess:true forRoleWithName:@"admins"];
let acl = PFACL()
acl.setPublicReadAccess(true)
acl.setWriteAccess(true, forRoleWithName: "admins")
// Assuming you've already created a role called "admins"...
ParseACL acl = new ParseACL();
acl.setPublicReadAccess(true);
acl.setRoleWriteAccess("admins", true);
var acl = new Parse.ACL();
acl.setPublicReadAccess(true);
acl.setRoleWriteAccess("admins", true);
var acl = new ParseACL();
acl.PublicReadAccess = true;
acl.SetRoleWriteAccess("admins", true);
$acl = new ParseACL();
$acl->setPublicReadAccess(true);
$acl->setRoleWriteAccessWithName("admins", true);
# No command line example
// No C++ example

Of course, this snippet assumes you've already created a role named "admins". This is often reasonable when you have a small set of special roles set up while developing your app. Roles can also be created and updated on the fly — for example, adding new friends to a "friendOf_" role after each connection is made.

All this is just the beginning. Applications can enforce all sorts of complex access patterns through ACLs and class-level permissions. For example:

  • For private data, read and write access can be restricted to the owner.
  • For a post on a message board, the author and members of the "Moderators" role can have "write" access, and the general public can have "read" access.
  • For logging data that will only be accessed by the developer through the REST API using the master key, the ACL can deny all permissions.
  • Data created by a privileged group of users or the developer, like a global message of the day, can have public read access but restrict write access to an "Administrators" role.
  • A message sent from one user to another can give "read" and "write" access just to those users.

For the curious, here's the format for an ACL that restricts read and write permissions to the owner (whose objectId is identified by "aSaMpLeUsErId") and enables other users to read the object:

{
    "*": { "read":true },
    "aSaMpLeUsErId": { "read" :true, "write": true }
}

And here's another example of the format of an ACL that uses a Role:

{
    "role:RoleName": { "read": true },
    "aSaMpLeUsErId": { "read": true, "write": true }
}

Pointer Permissions

Pointer permissions are a special type of class-level permission that create a virtual ACL on every object in a class, based on users stored in pointer fields on those objects. For example, given a class with an owner field, setting a read pointer permission on owner will make each object in the class only readable by the user in that object's owner field. For a class with a sender and a reciever field, a read pointer permission on the receiver field and a read and write pointer permission on the sender field will make each object in the class readable by the user in the sender and receiver field, and writable only by the user in the sender field.

Given that objects often already have pointers to the user(s) that should have permissions on the object, pointer permissions provide a simple and fast solution for securing your app using data which is already there, that doesn't require writing any client code or cloud code.

Pointer permissions are like virtual ACLs. They don't appear in the ACL column, buf if you are familiar with how ACLs work, you can think of them like ACLs. In the above example with the sender and receiver, each object will act as if it has an ACL of:

{
    "<SENDER_USER_ID>": {
        "read": true,
        "write": true
    },
    "<RECEIVER_USER_ID>": {
        "read": true
    }
}

Note that this ACL is not actually created on each object. Any existing ACLs will not be modified when you add or remove pointer permissions, and any user attempting to interact with an object can only interact with the object if both the virtual ACL created by the pointer permissions, and the real ACL already on the object allow the interaction. For this reason, it can sometimes be confusing to combine pointer permissions and ACLs, so we recommend using pointer permissions for classes that don't have many ACLs set. Fortunately, it's easy to remove pointer permissions if you later decide to use Cloud Code or ACLs to secure your app.

CLP and ACL interaction

Class-Level Permissions (CLPs) and Access Control Lists (ACLs) are both powerful tools for securing your app, but they don't always interact exactly how you might expect. They actually represent two separate layers of security that each request has to pass through to return the correct information or make the intended change. These layers, one at the class level, and one at the object level, are shown below. A request must pass through BOTH layers of checks in order to be authorized. Note that despite acting similarly to ACLs, Pointer Permissions are a type of class level permission, so a request must pass the pointer permission check in order to pass the CLP check.

As you can see, whether a user is authorized to make a request can become complicated when you use both CLPs and ACLs. Let's look at an example to get a better sense of how CLPs and ACLs can interact. Say we have a Photo class, with an object, photoObject. There are 2 users in our app, user1 and user2. Now lets say we set a Get CLP on the Photo class, disabling public Get, but allowing user1 to perform Get. Now let's also set an ACL on photoObject to allow Read - which includes GET - for only user2.

You may expect this will allow both user1 and user2 to Get photoObject, but because the CLP layer of authentication and the ACL layer are both in effect at all times, it actually makes it so neither user1 nor user2 can Get photoObject. If user1 tries to Get photoObject, it will get through the CLP layer of authentication, but then will be rejected because it does not pass the ACL layer. In the same way, if user2 tries to Get photoObject, it will also be rejected at the CLP layer of authentication.

Now lets look at example that uses Pointer Permissions. Say we have a Post class, with an object, myPost. There are 2 users in our app, poster, and viewer. Lets say we add a pointer permission that gives anyone in the Creator field of the Post class read and write access to the object, and for the myPost object, poster is the user in that field. There is also an ACL on the object that gives read access to viewer. You may expect that this will allow poster to read and edit myPost, and viewer to read it, but viewer will be rejected by the Pointer Permission, and poster will be rejected by the ACL, so again, neither user will be able to access the object.

Because of the complex interaction between CLPs, Pointer Permissions, and ACLs, we recommend being careful when using them together. Often it can be useful to use CLPs only to disable all permissions for a certain request type, and then using Pointer Permissions or ACLs for other request types. For example, you may want to disable Delete for a Photo class, but then put a Pointer Permission on Photo so the user who created it can edit it, just not delete it. Because of the especially complex way that Pointer Permissions and ACLs interact, we usually recommend only using one of those two types of security mechanisms.

Security Edge Cases

There are some special classes in Parse that don't follow all of the same security rules as every other class. Not all classes follow Class-Level Permissions (CLPs) or Access Control Lists (ACLs) exactly how they are defined, and here those exceptions are documented. Here "normal behavior" refers to CLPs and ACLs working normally, while any other special behaviors are described in the footnotes.

_User _Installation
Get normal behaviour [1, 2, 3] ignores CLP, but not ACL
Find normal behavior [3] master key only [6]
Create normal behavior [4] ignores CLP
Update normal behavior [5] ignores CLP, but not ACL [7]
Delete normal behavior [5] master key only [7]
Add Field normal behavior normal behavior
  1. Logging in, or /1/login in the REST API, does not respect the Get CLP on the user class. Login works just based on username and password, and cannot be disabled using CLPs.

  2. Retrieving the current user, or becoming a User based on a session token, which are both /1/users/me in the REST API, do not respect the Get CLP on the user class.

  3. Read ACLs do not apply to the logged in user. For example, if all users have ACLs with Read disabled, then doing a find query over users will still return the logged in user. However, if the Find CLP is disabled, then trying to perform a find on users will still return an error.

  4. Create CLPs also apply to signing up. So disabling Create CLPs on the user class also disables people from signing up without the master key.

  5. Users can only Update and Delete themselves. Public CLPs for Update and Delete may still apply. For example, if you disable public Update for the user class, then users cannot edit themselves. But no matter what the write ACL on a user is, that user can still Update or Delete itself, and no other user can Update or Delete that user. As always, however, using the master key allows users to update other users, independent of CLPs or ACLs.

  6. Get requests on installations follow ACLs normally. Find requests without master key is not allowed unless you supply the installationId as a constraint.

  7. Update requests on installations do adhere to the ACL defined on the installation, but Delete requests are master-key-only. For more information about how installations work, check out the installations section of the REST guide.

Data Integrity in Cloud Code

For most apps, care around keys, class-level permissions, and object-level ACLs are all you need to keep your app and your users' data safe. Sometimes, though, you'll run into an edge case where they aren't quite enough. For everything else, there's Cloud Code.

Cloud Code allows you to upload JavaScript to Parse's servers, where we will run it for you. Unlike client code running on users' devices that may have been tampered with, Cloud Code is guaranteed to be the code that you've written, so it can be trusted with more responsibility.

One particularly common use case for Cloud Code is preventing invalid data from being stored. For this sort of situation, it's particularly important that a malicious client not be able to bypass the validation logic.

To create validation functions, Cloud Code allows you to implement a beforeSave trigger for your class. These triggers are run whenever an object is saved, and allow you to modify the object or completely reject a save. For example, this is how you create a Cloud Code beforeSave trigger to make sure every user has an email address set:

Parse.Cloud.beforeSave(Parse.User, function(request, response) {
  var user = request.object;
  if (!user.get("email")) {
    response.error("Every user must have an email address.");
  } else {
    response.success();
  }
});

Our Cloud Code guide provides instructions on how to upload this trigger to our servers.

Validations can lock down your app so that only certain values are acceptable. You can also use afterSave validations to normalize your data (e.g. formatting all phone numbers or currency identically). You get to retain most of the productivity benefits of accessing Parse data directly from your client applications, but you can also enforce certain invariants for your data on the fly.

Common scenarios that warrant validation include:

  • Making sure phone numbers have the right format
  • Sanitizing data so that its format is normalized
  • Making sure that an email address looks like a real email address
  • Requiring that every user specifies an age within a particular range
  • Not letting users directly change a calculated field
  • Not letting users delete specific objects unless certain conditions are met

Implementing Business Logic in Cloud Code

While validation often makes sense in Cloud Code, there are likely certain actions that are particularly sensitive, and should be as carefully guarded as possible. In these cases, you can remove permissions or the logic from clients entirely and instead funnel all such operations to Cloud Code functions.

When a Cloud Code function is called, it can invoke the useMasterKey function to gain the ability to modify user data. With the master key, your Cloud Code function can override any ACLs and write data. This means that it'll bypass all the security mechanisms you've put in place in the previous sections.

Say you want to allow a user to "like" a Post object without giving them full write permissions on the object. You can do this by having the client call a Cloud Code function instead of modifying the Post itself:

The master key should be used carefully. When invoked, the master key is in effect for the duration of the Cloud Code function in which it is called:

Parse.Cloud.define("like", function(request, response) {
  Parse.Cloud.useMasterKey();
  // Everything after this point will bypass ACLs and other security
  // even if I do things besides just updating a Post object.
});

A more prudent way to use the master key would be to pass it as a parameter on a per-function basis. For example, instead of the above, set useMasterKey to true in each individual API function:

Parse.Cloud.define("like", function(request, response) {
  var post = new Parse.Object("Post");
  post.id = request.params.postId;
  post.increment("likes");
  post.save(null, { useMasterKey: true }).then(function() {
    // If I choose to do something else here, it won't be using
    // the master key and I'll be subject to ordinary security measures.
    response.success();
  }, function(error) {
    response.error(error);
  });
});

One very common use case for Cloud Code is sending push notifications to particular users. In general, clients can't be trusted to send push notifications directly, because they could modify the alert text, or push to people they shouldn't be able to. Your app's settings will allow you to set whether "client push" is enabled or not; we recommend that you make sure it's disabled. Instead, you should write Cloud Code functions that validate the data to be pushed and sent before sending a push.

Parse Security Summary

Parse provides a number of ways for you to secure data in your app. As you build your app and evaluate the kinds of data you will be storing, you can make the decision about which implementation to choose.

It is worth repeating that that the Parse User object is readable by all other users by default. You will want to set the ACL on your User object accordingly if you wish to prevent data contained in the User object (for example, the user's email address) from being visible by other users.

Most classes in your app will fall into one of a couple of easy-to-secure categories. For fully public data, you can use class-level permissions to lock down the table to put publicly readable and writeable by no one. For fully private data, you can use ACLs to make sure that only the user who owns the data can read it. But occasionally, you'll run into situations where you don't want data that’s fully public or fully private. For example, you may have a social app, where you have data for a user that should be readable only to friends whom they’ve approved. For this you'll need to a combination of the techniques discussed in this guide to enable exactly the sharing rules you desire.

We hope that you'll use these tools to do everything you can to keep your app's data and your users' data secure. Together, we can make the web a safer place.

Want to contribute to this doc? Edit this section.
Was this section helpful? NO YES Thanks for your feedback, we’re always working to make our docs better.

Performance

As your app scales, you will want to ensure that it performs well under increased load and usage. There are parts of optimizing performance that Parse takes care of but there are some things you can do. This document provides guidelines on how you can optimize your app's performance. While you can use Parse for quick prototyping and not worry about performance, you will want to keep our performance guidelines in mind when you're initially designing your app. We strongly advise that you make sure you've followed all suggestions before releasing your app.

Parse provides services out of the box to help your app scale auto-magically. On top of our MongoDB datastore, we have built an API layer that seamlessly integrates with our client-side SDKs. Our cloud infrastructure uses online learning algorithms to automatically rewrite inefficient queries and generate database indexes based on your app’s realtime query stream.

In addition to what Parse provides, you can improve your app's performance by looking at the following:

  • Writing efficient queries.
  • Writing restrictive queries.
  • Using client-side caching.
  • Using Cloud Code.
  • Avoiding count queries.
  • Using efficient search techniques.

Keep in mind that not all suggestions may apply to your app. Let's look into each one of these in more detail.

Write Efficient Queries

Parse objects are stored in a database. A Parse query retrieves objects that you are interested in based on conditions you apply to the query. To avoid looking through all the data present in a particular Parse class for every query, the database can use an index. An index is a sorted list of items matching a given criteria. Indexes help because they allow the database to do an efficient search and return matching results without looking at all of the data. Indexes are typically smaller in size and available in memory, resulting in faster lookups.

Smart Indexing

The key to writing efficient queries is understanding our indexing strategy. If your data is not indexed, every query will have to go through the the entire data for a class to return a query result. On the other hand, if your data is indexed appropriately, the number of documents scanned to return a correct query result should be low.

One of the advantages to using Parse if that you don't have to worry about managing your own database and maintaining indexes. We've built an abstraction to manage all that complexity. However, you do have to organize your data model and use performant queries to take advantage of this. To better understand how to go about doing this, you need to understand how our systems are operating behind the abstraction. The key strategy you will want to understand here is our use of smart indexing.

Smart indexing means that we algorithmically generate indexes for the apps that we host. The sheer number of apps hosted on Parse means that we cannot manually generate indexes for each app. This would not scale well as developers can change their schemas or query patterns at any time. This is why we rely on smart indexes.

We perform two types of index creation logic. The first generates simple (single field) indexes for each API request, and the second does offline processing to pick good compound indexes based on real API traffic patterns. In each case the goal is to pick indexes that result in the smallest search space for the query, that is, there will be less data to scan to find results.

The simple indexing strategy looks at every API request and attempts to pick good indexes based on the following:

  • The query constraint's likelihood of providing the smallest search space.
  • The number of possible values a field may have based on its data type. We consider data types with a larger number of possible values to have a higher entropy.

The order of a query constraint's usefulness is:

  • Equal to
  • Contained In
  • Less than, Less than or Equal to, Greater than, Greater than or Equal to
  • Prefix string matches
  • Not equal to
  • Not contained in
  • Everything else

Take a look at the following query to retrieve GameScore objects:

var GameScore = Parse.Object.extend("GameScore");
var query = new Parse.Query(GameScore);
query.equalTo("score", 50);
query.containedIn("playerName",
    ["Jonathan Walsh", "Dario Wunsch", "Shawn Simon"]);
PFQuery *query = [PFQuery queryWithClassName:@"GameScore"];
[query whereKey:@"score" equalTo:@50];
[query whereKey:@"playerName"
    containedIn:@[@"Jonathan Walsh", @"Dario Wunsch", @"Shawn Simon"]];
let query = PFQuery.queryWithClassName("GameScore")
query.whereKey("score", equalTo: 50)
query.whereKey("playerName", containedIn: ["Jonathan Walsh", "Dario Wunsch", "Shawn Simon"])
ParseQuery<ParseObject> query = ParseQuery.getQuery("GameScore");
query.whereEqualTo("score", 50);
query.whereContainedIn("playerName", Arrays.asList("Jonathan Walsh", "Dario Wunsch", "Shawn Simon"));
var names = new[] { "Jonathan Walsh", "Dario Wunsch", "Shawn Simon" };
var query = new ParseObject.GetQuery("GameScore")
    .WhereEqualTo("score", 50)
    .WhereContainedIn("playerName", names);
# No REST API example
// No C++ example

Creating an index query based on the score field would yield a smaller search space in general than creating one on the playerName field.

When examining data types, booleans have a very low entropy and and do not make good indexes. Take the following query constraint:

query.equalTo("cheatMode", false);
[query whereKey:@"cheatMode" equalTo:@NO];
query.whereKey("cheatMode", equalTo: false)
query.whereEqualTo("cheatMode", false);
query.WhereEqualTo("cheatMode", false);
# No REST API example
// No C++ example

The two possible values for "cheatMode" are true and false. If an index was added on this field it would be of little use because it's likely that 50% of the records will have to be looked at to return query results.

We also throw out relations and join tables, since no values are stored for these keys. We heavily promote GeoPoints since MongoDB won’t run a geo query without a geo index. Other data types are ranked by their expected entropy of the value space for the key:

  • Array
  • Pointer
  • Date
  • String
  • Number
  • Other

We score each query according to the above metrics, and make sure we create a unique index on the three top-scoring fields for each query. For a compound query that consists of an OR of subqueries, we compute the top three indexes for each subquery.

Even the best indexing strategy can be defeated by suboptimal queries. You will need to design queries that work hand in hand with smart indexing to deliver performant apps.

Efficient Query Design

Writing efficient queries means taking full advantage of indexes. Let's take a look at some query constraints that negate the use of indexes:

  • Not Equal To
  • Not Contained In

Additionally, the following queries under certain scenarios may result in slow query responses if they can't take advantage of indexes:

  • Regular Expressions
  • Ordered By

Not Equal To

For example, let's say you're tracking high scores for a game in a GameScore class. Now say you want to retrieve the scores for all players except a certain one. You could create this query:

var GameScore = Parse.Object.extend("GameScore");
var query = new Parse.Query(GameScore);
query.notEqualTo("playerName", "Michael Yabuti");
query.find().then(function(results) {
  // Retrieved scores successfully
});
PFQuery *query = [PFQuery queryWithClassName:@"GameScore"];
[query whereKey:@"playerName" notEqualTo:@"Michael Yabuti"];
[query findObjectsInBackgroundWithBlock:^(NSArray *objects, NSError *error) {
  if (!error) {
    // Retrieved scores successfully
  }
}];
let query = PFQuery.queryWithClassName("GameScore")
query.whereKey("playerName", notEqualTo: "Michael Yabuti")
query.findObjectsInBackgroundWithBlock {
  (objects, error) in
  if !error {
    // Retrieved scores successfully
  }
}
ParseQuery<ParseObject> query = ParseQuery.getQuery("GameScore");
query.whereNotEqualTo("playerName", "Michael Yabuti");
query.findInBackground(new FindCallback<ParseObject>() {
  @Override
  public void done(List<ParseObject> list, ParseException e) {
    if ( e == null) {
      // Retrieved scores successfully
    }
  }
});
var results = await ParseObject.GetQuery("GameScore")
    .WhereNotEqualTo("playerName", "Michael Yabuti")
    .FindAsync();
# No REST API example
// No C++ example

This query can't take advantage of indexes. The database has to look at all the objects in the "GameScore" class to satisfy the constraint and retrieve the results. As the number of entries in the class grows, the query takes longer to run.

Luckily, most of the time a “Not Equal To” query condition can be rewritten as a “Contained In” condition. Instead of querying for the absence of values, you ask for values which match the rest of the column values. Doing this allows the database to use an index and your queries will be faster.

For example if the User class has a column called state which has values “SignedUp”, “Verified”, and “Invited”, the slow way to find all users who have used the app at least once would be to run the query:

var query = new Parse.Query(Parse.User);
query.notEqualTo("state", "Invited");
PFQuery *query = [PFUser query];
[query whereKey:@"state" notEqualTo:@"Invited"];
var query = PFUser.query()
query.whereKey("state", notEqualTo: "Invited")
ParseQuery<ParseUser> query = ParseQuery.getQuery(ParseUser.class);
query.whereNotEqualTo("state", "Invited");
var query = ParseUser.Query
    .WhereNotEqualTo("state", "Invited");
# No REST API example
// No C++ example

It would be faster to use the “Contained In” condition when setting up the query:

query.containedIn("state", ["SignedUp", "Verified"]);
[query whereKey:@"state"
    containedIn:@[@"SignedUp", @"Verified"]];
query.whereKey("state", containedIn: ["SignedUp", "Verified"])
query.whereContainedIn("state", Arrays.asList("SignedUp", "Verified"));
query.WhereContainedIn("state", new[] { "SignedUp", "Verified" });
# No REST API example
// No C++ example

Sometimes, you may have to completely rewrite your query. Going back to the "GameScore" example, let's say we were running that query to display players who had scored higher than the given player. We could do this differently, by first getting the given player's high score and then using the following query:

var GameScore = Parse.Object.extend("GameScore");
var query = new Parse.Query(GameScore);
// Previously retrieved highScore for Michael Yabuti
query.greaterThan("score", highScore);
query.find().then(function(results) {
  // Retrieved scores successfully
});
PFQuery *query = [PFQuery queryWithClassName:@"GameScore"];
// Previously retrieved highScore for Michael Yabuti
[query whereKey:@"score" greaterThan:highScore];
[query findObjectsInBackgroundWithBlock:^(NSArray *objects, NSError *error) {
  if (!error) {
    // Retrieved scores successfully
  }
}];
let query = PFQuery.queryWithClassName("GameScore")
// Previously retrieved highScore for Michael Yabuti
query.whereKey("score", greaterThan: highScore)
query.findObjectsInBackgroundWithBlock {
  (objects, error) in
  if !error {
    // Retrieved scores successfully
  }
}
ParseQuery<ParseObject> query = ParseQuery.getQuery("GameScore");
// Previously retrieved highScore for Michael Yabuti
query.whereGreaterThan("score", highScore);
query.findInBackground(new FindCallback<ParseObject>() {
  @Override
  public void done(List<ParseObject> list, ParseException e) {
    if (e == null) {
      // Retrieved scores successfully
    }
  }
});
// Previously retrieved highScore for Michael Yabuti
var results = await ParseObject.GetQuery("GameScore")
    .WhereGreaterThan("score", highScore)
    .FindAsync();
# No REST API example
// No C++ example

The new query you use depends on your use case. This may sometimes mean a redesign of your data model.

Not Contained In

Similar to “Not Equal To”, the “Not Contained In” query constraint can't use an index. You should try and use the complementary “Contained In” constraint. Building on the User example, if the state column had one more value, “Blocked”, to represent blocked users, a slow query to find active users would be:

var query = new Parse.Query(Parse.User);
query.notContainedIn("state", ["Invited", "Blocked"];
PFQuery *query = [PFUser query];
[query whereKey:@"state" notContainedIn:@[@"Invited", @"Blocked"]];
var query = PFUser.query()
query.whereKey("state", notContainedIn: ["Invited", "Blocked"])
ParseQuery<ParseUser> query = ParseQuery.getQuery(ParseUser.class);
query.whereNotContainedIn("state", Arrays.asList("Invited", "Blocked"));
var query = ParseUser.Query
    .WhereNotContainedIn("state", new[] { "Invited", "Blocked" });
# No REST API example
// No C++ example

Using a complimentary “Contained In” query constraint will always be faster:

query.containedIn("state", ["SignedUp", "Verified"]);
[query whereKey:@"state" containedIn:@[@"SignedUp", @"Verified"]];
query.whereKey("state", containedIn: ["SignedUp", "Verified"])
query.whereContainedIn("state", Arrays.asList("SignedUp", "Verified"));
query.WhereContainedIn("state", new[] { "SignedUp", "Verified"});
# No REST API example
// No C++ example

This means rewriting your queries accordingly. Your query rewrites will depend on your schema set up. It may mean redoing that schema.

Regular Expressions

Most regular expression queries in Parse are heavily throttled due to performance considerations. MongoDB is not efficient for doing partial string matching except for the special case where you only want a prefix match. Queries that have regular expression constraints are therefore very expensive, especially for classes with over 100,000 records. Parse restricts how many such operations can be run on a particular app at any given time.

You should avoid using regular expression constraints that don't use indexes. For example, the following query looks for data with a given string in the "playerName" field. The string search is case insensitive and therefore cannot be indexed:

query.matches("playerName", "Michael", “i”);
[query whereKey:@"playerName" matchesRegex:@"Michael" modifiers:@"i"];
query.whereKey("playerName", matchesRegex: "Michael", modifiers: "i")
query.whereMatches("playerName", "Michael", "i");
query.WhereMatches("playerName", "Michael", "i")
# No REST API example
// No C++ example

The following query, while case sensitive, looks for any occurrence of the string in the field and cannot be indexed:

query.contains("playerName", "Michael");
[query whereKey:@"playerName" containsString:@"Michael"];
query.whereKey("playerName", containsString: "Michael")
query.whereContains("playerName", "Michael");
query.WhereContains("playerName", "Michael")
# No REST API example
// No C++ example

These queries are both slow. In fact, the matches and contains query constraints are not covered in our querying guides on purpose and we do not recommend using them. Depending on your use case, you should switch to using the following constraint that uses an index, such as:

query.startsWith("playerName", "Michael");
[query whereKey:@"playerName" hasPrefix:@"Michael"];
query.whereKey("playerName", hasPrefix: "Michael")
query.whereStartsWith("playerName", "Michael");
query.WhereStartsWith("playerName", "Michael")
# No REST API example
// No C++ example

This looks for data that starts with the given string. This query will use the backend index, so it will be faster even for large datasets.

As a best practice, when you use regular expression constraints, you'll want to ensure that other constraints in the query reduce the result set to the order of hundreds of objects to make the query efficient. If you must use the matches or contains constraints for legacy reasons, then use case sensitive, anchored queries where possible, for example:

query.matches("playerName", "^Michael");
[query whereKey:@"playerName" matchesRegex:@"^Michael"];
query.whereKey("playerName", matchesRegex: "^Michael")
query.whereMatches("playerName", "^Michael");
query.WhereMatches("playerName", "^Michael")
# No REST API example
// No C++ example

Most of the use cases around using regular expressions involve implementing search. A more performant way of implementing search is detailed later.

Write Restrictive Queries

Writing restrictive queries allows you to return only the data that the client needs. This is critical in a mobile environment were data usage can be limited and network connectivity unreliable. You also want your mobile app to appear responsive and this is directly affected by the objects you send back to the client. The Querying Guide shows the types of constraints you can add to your existing queries to limit the data returned. When adding constraints, you want to pay attention and design efficient queries.

You can limit the number of query results returned. The limit is 100 by default but anything from 1 to 1000 is a valid limit:

query.limit(10); // limit to at most 10 results
query.limit = 10; // limit to at most 10 results
query.limit = 10 // limit to at most 10 results
query.setLimit(10); // limit to at most 10 results
query.Limit(10); // limit to at most 10 results
# No REST API example
// No C++ example

If you're issuing queries on GeoPoints, make sure you specify a reasonable radius:

var query = new Parse.Query(PlaceObject);
query.withinMiles("location", userGeoPoint, 10.0);
query.find().then(function(placesObjects) {
  // Get a list of objects within 10 miles of a user's location
});
PFQuery *query = [PFQuery queryWithClassName:@"Place"];
[query whereKey:@"location" nearGeoPoint:userGeoPoint withinMiles:10.0];
[query findObjectsInBackgroundWithBlock:^(NSArray *places, NSError *error) {
  if (!error) {
    // List of objects within 10 miles of a user's location
  }
}];
let query = PFQuery.queryWithClassName("Place")
query.whereKey("location", nearGeoPoint: userGeoPoint, withinMiles: 10.0)
query.findObjectsInBackgroundWithBlock {
  (places, error) in
  if !error {
    // List of places within 10 miles of a user's location
  }
}
ParseQuery<ParseObject> query = ParseQuery.getQuery("Place");
query.whereWithinMiles("location", userGeoPoint, 10.0);
query.findInBackground(new FindCallback<ParseObject>() {
  @Override
  public void done(List<ParseObject> list, ParseException e) {
    if (e == null) {
      // List of places within 10 miles of a user's location
    }
  }
});
var results = await ParseObject.GetQuery("GameScore")
    .WhereWithinDistance("location", userGeoPoint, ParseGeoDistance.FromMiles(10.0))
    .FindAsync();
# No REST API example
// No C++ example

You can further limit the fields returned by calling select:

var GameScore = Parse.Object.extend("GameScore");
var query = new Parse.Query(GameScore);
query.select("score", "playerName");
query.find().then(function(results) {
  // each of results will only have the selected fields available.
});
PFQuery *query = [PFQuery queryWithClassName:@"GameScore"];
[query selectKeys:@[@"score", @"playerName"]];
[query findObjectsInBackgroundWithBlock:^(NSArray *objects, NSError *error) {
  if (!error) {
    // each of results will only have the selected fields available.
  }
}];
let query = PFQuery.queryWithClassName("GameScore")
query.selectKeys(["score", "playerName"])
query.findObjectsInBackgroundWithBlock {
  (objects, error) in
  if !error {
    // each of results will only have the selected fields available.
  }
}
ParseQuery<ParseObject> query = ParseQuery.getQuery("GameScore");
query.selectKeys(Arrays.asList("score", "playerName"));
query.findInBackground(new FindCallback<ParseObject>() {
  @Override
  public void done(List<ParseObject> list, ParseException e) {
    if (e == null) {
      // each of results will only have the selected fields available.
    }
  }
});
var results = await ParseObject.GetQuery("GameScore")
     .Select(new[] { "score", "playerName" })
     .FindAsync();
// each of results will only have the selected fields available.
# No REST API example
// No C++ example

Client-side Caching

For queries run from iOS and Android, you can turn on query caching. See the iOS and Android guides for more details. Caching queries will increase your mobile app's performance especially in cases where you want to display cached data while fetching the latest data from Parse.

Use Cloud Code

Cloud Code allows you to run custom JavaScript logic on Parse instead of on the client.

You can use this to off load processing to the Parse servers thus increasing your app's perceived performance. You can create hooks that run whenever an object is saved or deleted. This is useful if you want to validate or sanitize your data. You can also use Cloud Code to modify related objects or kick off other processes such as sending off a push notification. There are time limits to how long Cloud Code can run, for example an afterSave hook only has 3 seconds to run. You can use Background Jobs if you need to run more time consuming processes such as data migrations.

We saw examples of limiting the data returned by writing restrictive queries. You can also use Cloud Functions to help limit the amount of data returned to your app. In the following example, we use a Cloud Function to get a movie's average rating:

Parse.Cloud.define("averageStars", function(request, response) {
  var Review = Parse.Object.extend("Review");
  var query = new Parse.Query(Review);
  query.equalTo("movie", request.params.movie);
  query.find().then(function(results) {
    var sum = 0;
    for (var i = 0; i < results.length; ++i) {
      sum += results[i].get("stars");
    }
    response.success(sum / results.length);
  }, function(error) {
    response.error("movie lookup failed");
  });
});

You could have ran a query on the Review class on the client, returned only the stars field data and computed the result on the client. As the number of reviews for a movie increases you can see that the data being returned to the device using this methodology also increases. Implementing the functionality through a Cloud Function returns the one result if successful.

As you look at optimizing your queries, you'll find that you may have to change the queries - sometimes even after you've shipped your app to the App Store or Google Play. The ability to change your queries without a client update is possible if you use Cloud Functions. Even if you have to redesign your schema, you could make all the changes in your Cloud Functions while keeping the client interface the same to avoid an app update. Take the average stars Cloud Function example from before, calling it from a client SDK would look like this:

Parse.Cloud.run("averageStars", { "movie": "The Matrix" }).then(function(ratings) {
  // ratings is 4.5
});
[PFCloud callFunctionInBackground:@"averageStars"
                  withParameters:@{@"movie": @"The Matrix"}
                           block:^(NSNumber *ratings, NSError *error) {
  if (!error) {
    // ratings is 4.5
  }
}];
PFCloud.callFunctionInBackground("averageStars", withParameters: ["movie": "The Matrix"]) {
  (ratings, error) in
  if !error {
    // ratings is 4.5
  }
}
HashMap<String, String> params = new HashMap();
params.put("movie", "The Matrix");
ParseCloud.callFunctionInBackground("averageStars", params, new FunctionCallback<Float>() {
  @Override
  public void done(Float aFloat, ParseException e) {
    if (e == null) {
      // ratings is 4.5
    }
  }
});
IDictionary<string, object> dictionary = new Dictionary<string, object>
{
    { "movie", "The Matrix" }
};

ParseCloud.CallFunctionAsync<float>("averageStars", dictionary).ContinueWith(t => {
  var result = t.Result;
  // result is 4.5
});
# No REST API example
// No C++ example

If later on, you need to modify the underlying data model, your client call can remain the same, as long as you return back a number that represents the ratings result.

Avoid Count Operations

For classes with over 1,000 objects, count operations are limited by timeouts. They may routinely yield timeout errors or return results that are only approximately correct. Thus, it is preferable to architect your application to avoid this count operation.

Suppose you are displaying movie information in your app and your data model consists of a Movie class and a Review class that contains a pointer to the corresponding movie. You might want to display the review count for each movie on the top-level navigation screen using a query like this:

var Review = Parse.Object.extend("Review");
var query = new Parse.Query("Review");
// movieId corresponds to a given movie's id
query.equalTo(“movie”, movieId);
query.count().then(function(count) {
  // Request succeeded
});
PFQuery *query = [PFQuery queryWithClassName:@"Review"];
// movieId corresponds to a given movie's id
[query whereKey:@"movie" equalTo:movieId];
[query countObjectsInBackgroundWithBlock:^(int number, NSError *error) {
  if (!error) {
    // Request succeeded
  }
}];
let query = PFQuery.queryWithClassName("Review")
// movieId corresponds to a given movie's id
query.whereKey("movie", equalTo: movieId)
query.countObjectsInBackgroundWithBlock {
  (number, error) in
  if !error {
    // Request succeeded
  }
}
ParseQuery<ParseObject> query = ParseQuery.getQuery("Review");
// movieId corresponds to a given movie's id
query.whereEqualTo("movie", movieId);
query.countInBackground(new CountCallback() {
  @Override
  public void done(int i, ParseException e) {
    if ( e == null) {
      // Request succeeded
    }
  }
});
var count = await ParseObject.GetQuery("Review")
// movieId corresponds to a given movie's id
    .WhereEqualTo("movie", movieId)
    .CountAsync();
# No REST API example
// No C++ example

If you run the count query for each of the UI elements, they will not run efficiently on large data sets. One approach to avoid using the count() operator could be to add a field to the Movie class that represents the review count for that movie. When saving an entry to the Review class you could increment the corresponding movie's review count field. This can be done in an afterSave handler:

Parse.Cloud.afterSave("Review", function(request) {
  // Get the movie id for the Review
  var movieId = request.object.get("movie").id;
  // Query the Movie represented by this review
  var Movie = Parse.Object.extend("Movie");
  var query = new Parse.Query(Movie);
  query.get(movieId).then(function(movie) {
    // Increment the reviews field on the Movie object
    movie.increment("reviews");
    movie.save();
  }, function(error) {
    throw "Got an error " + error.code + " : " + error.message;
  });
});

Your new optimized query would not need to look at the Review class to get the review count:

var Movie = Parse.Object.extend("Movie");
var query = new Parse.Query(Movie);
query.find().then(function(results) {
  // Results include the reviews count field
}, function(error) {
  // Request failed
});
PFQuery *query = [PFQuery queryWithClassName:@"Movie"];
[query findObjectsInBackgroundWithBlock:^(NSArray *objects, NSError *error) {
  if (!error) {
    // Results include the reviews count field
  }
}];
let query = PFQuery.queryWithClassName("Movie")
query.findObjectsInBackgroundWithBlock {
  (objects, error) in
  if !error {
    // Results include the reviews count field
  }
}
ParseQuery<ParseObject> query = ParseQuery.getQuery("Movie");
query.findInBackground(new FindCallback<ParseObject>() {
  @Override
  public void done(List<ParseObject> list, ParseException e) {
    if (e == null) {
      // Results include the reviews count field
    }
  }
});
var results = await ParseObject.GetQuery("Movie")
    .FindAsync();
// Results include the reviews count field
# No REST API example
// No C++ example

You could also use a separate Parse Object to keep track of counts for each review. Whenever a review gets added or deleted, you can increment or decrement the counts in an afterSave or afterDelete Cloud Code handler. The approach you choose depends on your use case.

Implement Efficient Searches

As mentioned previously, MongoDB is not efficient for doing partial string matching. However, this is an important use case when implementing search functionality that scales well in production.

Simplistic search algorithms simply scan through all the class data and executes the query on each entry. The key to making searches run efficiently is to minimize the number of data that has to be examined when executing each query by using an index as we've outlined earlier. You’ll need to build your data model in a way that it’s easy for us to build an index for the data you want to be searchable. For example, string matching queries that don’t match an exact prefix of the string won’t be able to use an index leading to timeout errors as the data set grows.

Let's walk through an example of how you could build an efficient search. You can apply the concepts you learn in this example to your use case. Say your app has users making posts, and you want to be able to search those posts for hashtags or particular keywords. You’ll want to pre-process your posts and save the list of hashtags and words into array fields. You can do this processing either in your app before saving the posts, or you can use a Cloud Code beforeSave hook to do this on the fly:

var _ = require("underscore");
Parse.Cloud.beforeSave("Post", function(request, response) {
  var post = request.object;
  var toLowerCase = function(w) { return w.toLowerCase(); };
  var words = post.get("text").split(/\b/);
  words = _.map(words, toLowerCase);
  var stopWords = ["the", "in", "and"]
  words = _.filter(words, function(w) {
    return w.match(/^\w+$/) && !   _.contains(stopWords, w);
  });
  var hashtags = post.get("text").match(/#.+?\b/g);
  hashtags = _.map(hashtags, toLowerCase);
  post.set("words", words);
  post.set("hashtags", hashtags);
  response.success();
});

This saves your words and hashtags in array fields, which MongoDB will store with a multi-key index. There are some important things to notice about this. First of all it’s converting all words to lower case so that we can look them up with lower case queries, and get case insensitive matching. Secondly, it’s filtering out common words like ‘the’, ‘in’, and ‘and’ which will occur in a lot of posts, to additionally reduce useless scanning of the index when executing the queries.

Once you've got the keywords set up, you can efficiently look them up using “All” constraint on your query:

var Post = Parse.Object.extend("Post");
var query = new Parse.Query(Post);
query.containsAll("hashtags", [“#parse”, “#ftw”]);
query.find().then(function(results) {
  // Request succeeded
}, function(error) {
  // Request failed
});
PFQuery *query = [PFQuery queryWithClassName:@"Post"];
[query whereKey:@"hashtags" containsAllObjectsInArray:@[@"#parse", @"#ftw"]];
[query findObjectsInBackgroundWithBlock:^(NSArray *objects, NSError *error) {
  if (!error) {
    // Request succeeded
  }
}];
let query = PFQuery.queryWithClassName("Post")
query.whereKey("hashtags", containsAllObjectsInArray: ["#parse", "#ftw"])
query.findObjectsInBackgroundWithBlock {
  (objects, error) in
  if !error {
    // Request succeeded
  }
}
ParseQuery<ParseObject> query = ParseQuery.getQuery("Post");
query.whereContainsAll("hashtags", Arrays.asList("#parse", "#ftw"));
query.findInBackground(new FindCallback<ParseObject>() {
  @Override
  public void done(List<ParseObject> list, ParseException e) {
    if (e == null) {
      // Request succeeded
    }
  }
});
var results = await ParseObject.GetQuery("Post")
    .WhereContainsAll("hashtags", new[] { "#parse", "#ftw" })
    .FindAsync();
# No REST API example
// No C++ example

Limits and Other Considerations

There are some limits in place to ensure the API can provide the data you need in a performant manner. We may adjust these in the future. Please take a moment to read through the following list:

Objects

  • Parse Objects are limited in size to 128 KB.
  • We recommend against creating more than 64 fields on a single Parse Object to ensure that we can build effective indexes for your queries.
  • We recommend against using field names that are longer than 1,024 characters, otherwise an index for the field will not be created.
  • An app may only create up to 100 Parse Config keys. Use Parse Objects if you need more.
  • An app may only create up to 200 classes.
  • Only the first 100 classes will be displayed in the Data Browser.

Queries

  • Queries return 100 objects by default. Use the limit parameter to change this, up to a value of 1,000.
  • Queries can only return up to 1,000 objects in a single result set. This includes any resolved pointers. You can use skip and limit to page through results.
  • The maximum value accepted by skip is 10,000. If you need to get more objects, we recommend sorting the results and then using a constraint on the sort column to filter out the first 10,000 results. You will then be able to continue paging through results starting from a skip value of 0. For example, you can sort your results by createdAt ASC and then filter out any objects older than the createdAt value of the 10,000th object when starting again from 0.
  • Alternatively, you may use the each() method in the JavaScript SDK to page through all objects that match the query.
  • Skips and limits can only be used on the outer query.
  • You may increase the limit of a inner query to 1,000, but skip cannot be used to get more results past the first 1,000.
  • Constraints that collide with each other will result in only one of the constraint being applied. An example of this would be two equalTo constraints over the same key with two different values, which contradicts itself (perhaps you're looking for 'contains').
  • Regular expression queries are deprecated and an app can make no more than 80 regex queries / minute.
  • Count operations are limited to 160 count queries / minute period for each application. The limit applies to all requests made by all clients of the application. If you need to query counts more frequently, you may want to design your system so the results can be cached in your client application.
  • No geo-queries inside compound OR queries.
  • Using $exists: false is not advised.
  • The each query method in the JavaScript SDK cannot be used in conjunction with queries using geo-point constraints.
  • A maximum of 500,000 objects will be scanned per query. If your constraints do not successfully limit the scope of the search, this can result in queries with incomplete results.
  • A containsAll query constraint can only take up to 9 items in the comparison array.

Files

  • Parse Files are limited to 10 MB each and the limit cannot be increased. There is no limit on how many Parse Files your app can create.
  • Parse Hosting, on the other hand, supports files up to 500 MB, ideal for Unity projects.
  • Parse Files cannot be accessed via HTTPS.

Push Notifications

Cloud Code

  • JavaScript scripts in Cloud Code are limited to 128 KB.
  • Cloud functions must return within 15 seconds. Use webhooks if you need more time.
  • Cloud save/delete hooks must return within 3 seconds. Use webhooks if you need more time.
  • Background jobs will be terminated after 15 minutes.
  • Cloud Code invocations (such as functions, save/delete hooks, background jobs, and custom endpoint handlers) must use a finite amount of memory (currently 384 MB, but subject to change in the future). This memory is garbage collected by the JavaScript engine, so it is generally not an issue. To allow the garbage collector to do its work, make sure that you aren't saving references to large objects, HTTP response buffers, and other objects to global variables in your script.
  • Webhooks must return within 30 seconds.
  • Apps can run one job concurrently by default, but this can be increased by increasing your requests/second in your Account Overview page.
  • Additional background jobs over quota will fail to run. They will not be queued.
  • The params payload that is passed to a Cloud Function is limited to 50 MB.
  • When using console to log information and errors, the message string will be limited to 1 KB.
  • Only the first 100 messages logged by a Cloud function will be persisted in the Cloud Code logs.
  • Parse.Cloud.httpRequest has a 2 second connection timeout. If the connection cannot be established within 2 seconds, you may see a 124 request timed out error.
  • Parse.Cloud.httpRequest has a maximum total timeout of 1 minute. This timeout can be further restricted by the maximum timeout of the particular Cloud Code invocation. For instance, before/after save triggers have a maximum timeout of 3 seconds, so the maximum timeout for an HTTP request made within a before/after save trigger is 3 seconds.
  • There is a limit of 8 concurrent outbound Parse.Cloud.httpRequest requests per Cloud Code request.
Want to contribute to this doc? Edit this section.
Was this section helpful? NO YES Thanks for your feedback, we’re always working to make our docs better.

Error Codes

The following is a list of all the error codes that can be returned by the Parse API. You may also refer to RFC2616 for a list of http error codes. Make sure to check the error message for more details.

API Issues

Name Code Description
UserInvalidLoginParams 101 Invalid login parameters. Check error message for more details.
ObjectNotFound 101 The specified object or session doesn't exist or could not be found. Can also indicate that you do not have the necessary permissions to read or write this object. Check error message for more details.
InvalidQuery 102 There is a problem with the parameters used to construct this query. This could be an invalid field name or an invalid field type for a specific constraint. Check error message for more details.
InvalidClassName 103 Missing or invalid classname. Classnames are case-sensitive. They must start with a letter, and a-zA-Z0-9_ are the only valid characters.
MissingObjectId 104 An unspecified object id.
InvalidFieldName 105 An invalid field name. Keys are case-sensitive. They must start with a letter, and a-zA-Z0-9_ are the only valid characters. Some field names may be reserved. Check error message for more details.
InvalidPointer 106 A malformed pointer was used. You would typically only see this if you have modified a client SDK.
InvalidJSON 107 Badly formed JSON was received upstream. This either indicates you have done something unusual with modifying how things encode to JSON, or the network is failing badly. Can also indicate an invalid utf-8 string or use of multiple form encoded values. Check error message for more details.
CommandUnavailable 108 The feature you tried to access is only available internally for testing purposes.
NotInitialized 109 You must call Parse.initialize before using the Parse library. Check the Quick Start guide for your platform.
ObjectTooLarge 116 The object is too large. ParseObjects have a max size of 128 kilobytes.
ExceededConfigParamsError 116 You have reached the limit of 100 config parameters.
InvalidLimitError 117 An invalid value was set for the limit. Check error message for more details.
InvalidSkipError 118 An invalid value was set for skip. Check error message for more details.
OperationForbidden 119 The operation isn't allowed for clients due to class-level permissions. Check error message for more details.
CacheMiss 120 The result was not found in the cache.
InvalidNestedKey 121 An invalid key was used in a nested JSONObject. Check error message for more details.
InvalidACL 123 An invalid ACL was provided.
InvalidEmailAddress 125 The email address was invalid.
DuplicateValue 137 Unique field was given a value that is already taken.
InvalidRoleName 139 Role's name is invalid.
ReservedValue 139 Field value is reserved.
ExceededCollectionQuota 140 You have reached the quota on the number of classes in your app. Please delete some classes if you need to add a new class.
ScriptFailed 141 Cloud Code script failed. Usually points to a JavaScript error. Check error message for more details.
FunctionNotFound 141 Cloud function not found. Check that the specified Cloud function is present in your Cloud Code script and has been deployed.
JobNotFound 141 Background job not found. Check that the specified job is present in your Cloud Code script and has been deployed.
SuccessErrorNotCalled 141 success/error was not called. A cloud function will return once response.success() or response.error() is called. A background job will similarly finish execution once status.success() or status.error() is called. If a function or job never reaches either of the success/error methods, this error will be returned. This may happen when a function does not handle an error response correctly, preventing code execution from reaching the success() method call.
MultupleSuccessErrorCalls 141 Can't call success/error multiple times. A cloud function will return once response.success() or response.error() is called. A background job will similarly finish execution once status.success() or status.error() is called. If a function or job calls success() and/or error() more than once in a single execution path, this error will be returned.
ValidationFailed 142 Cloud Code validation failed.
WebhookError 143 Webhook error.
InvalidImageData 150 Invalid image data.
UnsavedFileError 151 An unsaved file.
InvalidPushTimeError 152 An invalid push time was specified.
HostingError 158 Hosting error.
InvalidEventName 160 The provided analytics event name is invalid.
ClassNotEmpty 255 Class is not empty and cannot be dropped.
AppNameInvalid 256 App name is invalid.
MissingAPIKeyError 902 The request is missing an API key.
InvalidAPIKeyError 903 The request is using an invalid API key.

Push related errors

Name Code Description
IncorrectType 111 A field was set to an inconsistent type. Check error message for more details.
InvalidChannelName 112 Invalid channel name. A channel name is either an empty string (the broadcast channel) or contains only a-zA-Z0-9_ characters and starts with a letter.
InvalidSubscriptionType 113 Bad subscription type. Check error message for more details.
InvalidDeviceToken 114 The provided device token is invalid.
PushMisconfigured 115 Push is misconfigured in your app. Check error message for more details.
PushWhereAndChannels 115 Can't set channels for a query-targeted push. You can fix this by moving the channels into your push query constraints.
PushWhereAndType 115 Can't set device type for a query-targeted push. You can fix this by incorporating the device type constraints into your push query.
PushMissingData 115 Push is missing a 'data' field.
PushMissingChannels 115 Non-query push is missing a 'channels' field. Fix by passing a 'channels' or 'query' field.
ClientPushDisabled 115 Client-initiated push is not enabled. Check your Parse app's push notification settings.
RestPushDisabled 115 REST-initiated push is not enabled. Check your Parse app's push notification settings.
ClientPushWithURI 115 Client-initiated push cannot use the "uri" option.
PushQueryOrPayloadTooLarge 115 Your push query or data payload is too large. Check error message for more details.
InvalidExpirationError 138 Invalid expiration value.
MissingPushIdError 156 A push id is missing. Deprecated.
MissingDeviceTypeError 157 The device type field is missing. Deprecated.

File related errors

Name Code Description
InvalidFileName 122 An invalid filename was used for ParseFile. A valid file name contains only a-zA-Z0-9_. characters and is between 1 and 128 characters.
MissingContentType 126 Missing content type.
MissingContentLength 127 Missing content length.
InvalidContentLength 128 Invalid content length.
FileTooLarge 129 File size exceeds maximum allowed.
FileSaveError 130 Error saving a file.
FileDeleteError 131 File could not be deleted.

Installation related errors

Name Code Description
InvalidInstallationIdError 132 Invalid installation id.
InvalidDeviceTypeError 133 Invalid device type.
InvalidChannelsArrayError 134 Invalid channels array value.
MissingRequiredFieldError 135 Required field is missing.
ChangedImmutableFieldError 136 An immutable field was changed.

Purchase related errors

Name Code Description
ReceiptMissing 143 Product purchase receipt is missing.
InvalidPurchaseReceipt 144 Product purchase receipt is invalid.
PaymentDisabled 145 Payment is disabled on this device.
InvalidProductIdentifier 146 The product identifier is invalid.
ProductNotFoundInAppStore 147 The product is not found in the App Store.
InvalidServerResponse 148 The Apple server response is not valid.
ProductDownloadFilesystemError 149 The product fails to download due to file system error.

User related errors

Name Code Description
UsernameMissing 200 The username is missing or empty.
PasswordMissing 201 The password is missing or empty.
UsernameTaken 202 The username has already been taken.
UserEmailTaken 203 Email has already been used.
UserEmailMissing 204 The email is missing, and must be specified.
UserWithEmailNotFound 205 A user with the specified email was not found.
SessionMissing 206 A user object without a valid session could not be altered.
MustCreateUserThroughSignup 207 A user can only be created through signup.
AccountAlreadyLinked 208 An account being linked is already linked to another user.
InvalidSessionToken 209 The device's session token is no longer valid. The application should ask the user to log in again.

Linked services errors

Name Code Description
LinkedIdMissing 250 A user cannot be linked to an account because that account's id could not be found.
InvalidLinkedSession 251 A user with a linked (e.g. Facebook or Twitter) account has an invalid session. Check error message for more details.
InvalidGeneralAuthData 251 Invalid auth data value used.
BadAnonymousID 251 Anonymous id is not a valid lowercase UUID.
FacebookBadToken 251 The supplied Facebook session token is expired or invalid.
FacebookBadID 251 A user with a linked Facebook account has an invalid session.
FacebookWrongAppID 251 Unacceptable Facebook application id.
TwitterVerificationFailed 251 Twitter credential verification failed.
TwitterWrongID 251 Submitted Twitter id does not match the id associated with the submitted access token.
TwitterWrongScreenName 251 Submitted Twitter handle does not match the handle associated with the submitted access token.
TwitterConnectFailure 251 Twitter credentials could not be verified due to problems accessing the Twitter API.
UnsupportedService 252 A service being linked (e.g. Facebook or Twitter) is unsupported. Check error message for more details.
UsernameSigninDisabled 252 Authentication by username and password is not supported for this application. Check your Parse app's authentication settings.
AnonymousSigninDisabled 252 Anonymous users are not supported for this application. Check your Parse app's authentication settings.
FacebookSigninDisabled 252 Authentication by Facebook is not supported for this application. Check your Parse app's authentication settings.
TwitterSigninDisabled 252 Authentication by Twitter is not supported for this application. Check your Parse app's authentication settings.
InvalidAuthDataError 253 An invalid authData value was passed. Check error message for more details.
LinkingNotSupportedError 999 Linking to an external account not supported yet with signup_or_login. Use update instead.

Client-only errors

Name Code Description
ConnectionFailed 100 The connection to the Parse servers failed.
AggregateError 600 There were multiple errors. Aggregate errors have an "errors" property, which is an array of error objects with more detail about each error that occurred.
FileReadError 601 Unable to read input for a ParseFile on the client.
XDomainRequest 602 A real error code is unavailable because we had to use an XDomainRequest object to allow CORS requests in Internet Explorer, which strips the body from HTTP responses that have a non-2XX status code.

Operational issues

Name Code Description
RequestTimeout 124 The request was slow and timed out. Typically this indicates that the request is too expensive to run. You may see this when a Cloud function did not finish before timing out, or when a Parse.Cloud.httpRequest connection times out.
InefficientQueryError 154 An inefficient query was rejected by the server. Refer to the Performance Guide and slow query log.
RequestLimitExceeded 155 This application has exceeded its request limit. Please retry in one minute or raise your request limit at https://parse.com/account. Check error message for more details.
TemporaryRejectionError 159 An application's requests are temporary rejected by the server.

If your requests are rejected with a 155 error code, your application may be exceeding its request limit. You can learn more about how the request limit is calculated in the Pricing FAQ. If waiting one minute or increasing the request limit does not resolve the issue, please file a bug report. If your requests are being rejected with a 159 error code, please file a bug report for further instructions on how to resolve the issue.

Other issues

Name Code Description
OtherCause -1 An unknown error or an error unrelated to Parse occurred.
InternalServerError 1 Internal server error. No information available.
ServiceUnavailable 2 The service is currently unavailable.
ClientDisconnected 4 Connection failure.
Want to contribute to this doc? Edit this section.
Was this section helpful? NO YES Thanks for your feedback, we’re always working to make our docs better.