Amazon SNS Subscription Filter Policies - Amazon Simple Notification Service

A subscription filter policy allows you to specify attribute names and assign a list of values to each attribute name. For more information, see Message Filtering.

When Amazon SNS evaluates message attributes against the subscription filter policy, it ignores message attributes that aren't specified in the policy.

A subscription accepts a message under the following conditions:

Each attribute name in a filter policy matches an attribute name assigned to the message.
For each matching attribute name, at least one match exists between the following:
- the values of the attribute name in the filter policy
- the message attributes

Topics

Example Message with Attributes
Example Filter Policies
Filter Policy Constraints
Attribute String Value Matching
Attribute Numeric Value Matching
Attribute Key Matching
AND/OR Logic

Example Message with Attributes

The following example shows a message payload sent by an Amazon SNS topic that publishes customer transactions. The MessageAttributes field includes attributes that describe the transaction:

Customer's interests
Store name
Event state
Purchase price in USD

Because this message includes the MessageAttributes field, any topic subscription that includes a filter policy can selectively accept or reject the message.


{
   "Type": "Notification",
   "MessageId": "a1b2c34d-567e-8f90-g1h2-i345j67klmn8",
   "TopicArn": "arn:aws:sns:us-east-2:123456789012:MyTopic",
   "Message": "message-body-with-transaction-details",
   "Timestamp": "2019-11-03T23:28:01.631Z",
   "SignatureVersion": "4",
   "Signature": "signature",
   "UnsubscribeURL": "unsubscribe-url",
   "MessageAttributes": {
      "customer_interests": {
         "Type": "String.Array",
         "Value": "[\"soccer\", \"rugby\", \"hockey\"]"
      },
      "store": {
         "Type": "String",
         "Value":"example_corp"
      },
      "event": {
         "Type": "String",
         "Value": "order_placed"
      },
      "price_usd": {
         "Type": "Number", 
         "Value":210.75
      }
   }
}

For information about applying attributes to a message, see Amazon SNS Message Attributes.

Example Filter Policies

The following filter policies accept or reject messages based on their attribute names and values.

A Policy that Accepts Messages

The attributes in the following subscription filter policy match the attributes assigned to the example message.

If any single attribute in this policy doesn't match an attribute assigned to the message, the policy rejects the message.


{
   "store": ["example_corp"],
   "event": [{"anything-but": "order_cancelled"}],
   "customer_interests": [
      "rugby",
      "football",
      "baseball"
   ],
   "price_usd": [{"numeric": [">=", 100]}]
}

A Policy that Rejects Messages

The following subscription filter policy has multiple mismatches between its attributes and the attributes assigned to the example message. Because the encrypted attribute name isn't present in the message attributes, this policy attribute causes the message to be rejected regardless of the value assigned to it.

If any mismatches occur, the policy rejects the message.


{
   "store": ["example_corp"],
   "event": ["order_cancelled"],
   "encrypted": [false],
   "customer_interests": [
      "basketball",
      "baseball"
   ]
}

Filter Policy Constraints

When you create a filter policy, keep the following constraints in mind:

For the String data type, the attribute comparison between policy and message is case-sensitive.
A numeric policy attribute can have a value from -10⁹ to 10⁹, with 5 digits of accuracy after the decimal point.
The total combination of values must not exceed 150. Calculate the total combination by multiplying the number of values in each array.

Consider the following policy:
```
{
   "key_a": ["value_one", "value_two", "value_three"],
   "key_b": ["value_one"],
   "key_c": ["value_one", "value_two"]
}
```
The first array has three values, the second has one value, and the third has two values. The total combination is calculated as follows:
```
3 x 1 x 2 = 6
```
Amazon SNS compares policy attributes only to message attributes that have the following data types:
- String
- String.Array
- Number
Amazon SNS ignores message attributes with the Binary data type.
The JSON of the filter policy can contain the following:
- strings enclosed in quotation marks
- numbers
- the keywords true, false, and null, without quotation marks
When you use the Amazon SNS API, you must pass the JSON of the filter policy as a valid UTF-8 string.
A filter policy can have a maximum of 5 attribute names.
The maximum size of a policy is 256 KB.
By default, you can have up to 200 filter policies per AWS account, per region. To increase this limit, submit a limit increase request.

Attribute String Value Matching

You can use string values to match message attributes and filter messages. String values are enclosed in double quotation marks in the JSON policy.

You can use the following string operations to match message attributes.

Exact Matching (Whitelisting)

Exact matching occurs when a policy attribute value matches one or more message attribute values.

Consider the following policy attribute:


"customer_interests": ["rugby", "tennis"]

It matches the following message attributes:


"customer_interests": {"Type": "String", "Value": "rugby"}


"customer_interests": {"Type": "String", "Value": "tennis"}

However, it doesn't match the following message attribute:


"customer_interests": {"Type": "String", "Value": "baseball"}

Anything-But Matching (Blacklisting)

When a policy attribute value includes the keyword anything-but, it matches any message attribute that doesn't include any of the policy attribute values.

Consider the following policy attribute:


"customer_interests": [{"anything-but": ["rugby", "tennis"]}]

It matches either of the following message attributes:


"customer_interests": {"Type": "String", "Value": "baseball"}


"customer_interests": {"Type": "String", "Value": "football"}

It also matches the following message attribute (because it contains a value that isn't rugby or tennis):


"customer_interests": {"Type": "String.Array", "Value": "[\"rugby\", \"baseball\"]"}

However, it doesn't match the following message attribute:


"customer_interests": {"Type": "String", "Value": "rugby"}

Prefix Matching

When a policy attribute includes the keyword prefix, it matches any message attribute value that begins with the specified characters.

Consider the following policy attribute:


"customer_interests": [{"prefix": "bas"}]

It matches either of the following message attributes:


"customer_interests": {"Type": "String", "Value": "baseball"}


"customer_interests": {"Type": "String", "Value": "basketball"}

However, it doesn't match the following message attribute:


"customer_interests": {"Type": "String", "Value": "rugby"}

Attribute Numeric Value Matching

You can use numeric values to match message attributes and filter messages. Numeric values aren't enclosed in double quotation marks in the JSON policy. You can use the following numeric operations to match message attributes.

Exact Matching (Whitelisting)

When a policy attribute value includes the keyword numeric and the operator =, it matches any message attribute that has the same name and an equal numeric value.

Consider the following policy attribute:


"price_usd": [{"numeric": ["=",301.5]}]

It matches either of the following message attributes:


"price_usd": {"Type": "Number", "Value": 301.5}


"price_usd": {"Type": "Number", "Value": 3.015e2}

Anything-But Matching (Blacklisting)

When a policy attribute value includes the keyword anything-but, it matches any message attribute that doesn't include any of the policy attribute values.

Consider the following policy attribute:


"price": [{"anything-but": [100, 500]}]

It matches either of the following message attributes:


"price": {"Type": "Number", "Value": 101}


"price": {"Type": "Number", "Value": 100.1}

It also matches the following message attribute (because it contains a value that isn't 100 or 500):


"price": {"Type": "Number.Array", "Value": "[100, 50]"}

However, it doesn't match the following message attribute:


"price": {"Type": "Number", "Value": 100}

Value Range Matching

In addition to the operator =, a numeric policy attribute can include the following operators: <, <=, >, and >=.

Consider the following policy attribute:


"price_usd": [{"numeric": ["<", 0]}]

It matches any message attributes with negative numeric values.

Consider another message attribute:


"price_usd": [{"numeric": [">", 0, "<=", 150]}]

It matches any message attributes with positive numbers up to and including 150.

Attribute Key Matching

You can use the exists operator to check whether an incoming message has an attribute whose key is listed in the filter policy.

Consider the following policy attribute:


"store": [{"exists": true}]

It matches any message with at least the store attribute key, such as the following:


"store": "fans"
"customer_interests": ["baseball", "basketball"]

However, it doesn't match any message without the store attribute key, such as the following:


"customer_interests": ["baseball", "basketball"]

AND/OR Logic

You can use operations that include AND/OR logic to match message attributes.

AND Logic

You can apply AND logic using multiple attribute names.

Consider the following policy:


{
  "customer_interests": ["rugby"],
  "price_usd": [{"numeric": [">", 100]}]
}

It matches any message attributes with the value of customer_interests set to rugby and the value of price_usd set to a number larger than 100.

OR Logic

You can apply OR logic by assigning multiple values to an attribute name.

Consider the following policy attribute:


"customer_interests": ["rugby", "football", "baseball"]

It matches any message attributes with the value of customer_interests set to rugby, football, or baseball.

Warning Javascript is disabled or is unavailable in your browser.

To use the AWS Documentation, Javascript must be enabled. Please refer to your browser's Help pages for instructions.

Document Conventions

« Previous Next »