summaryrefslogtreecommitdiff
path: root/vendor/guzzle/guzzle/docs/webservice-client
diff options
context:
space:
mode:
Diffstat (limited to 'vendor/guzzle/guzzle/docs/webservice-client')
-rw-r--r--vendor/guzzle/guzzle/docs/webservice-client/guzzle-service-descriptions.rst619
-rw-r--r--vendor/guzzle/guzzle/docs/webservice-client/using-the-service-builder.rst316
-rw-r--r--vendor/guzzle/guzzle/docs/webservice-client/webservice-client.rst659
3 files changed, 1594 insertions, 0 deletions
diff --git a/vendor/guzzle/guzzle/docs/webservice-client/guzzle-service-descriptions.rst b/vendor/guzzle/guzzle/docs/webservice-client/guzzle-service-descriptions.rst
new file mode 100644
index 0000000..ad6070b
--- /dev/null
+++ b/vendor/guzzle/guzzle/docs/webservice-client/guzzle-service-descriptions.rst
@@ -0,0 +1,619 @@
+===========================
+Guzzle service descriptions
+===========================
+
+Guzzle allows you to serialize HTTP requests and parse HTTP responses using a DSL called a service descriptions.
+Service descriptions define web service APIs by documenting each operation, the operation's parameters, validation
+options for each parameter, an operation's response, how the response is parsed, and any errors that can be raised for
+an operation. Writing a service description for a web service allows you to more quickly consume a web service than
+writing concrete commands for each web service operation.
+
+Guzzle service descriptions can be representing using a PHP array or JSON document. Guzzle's service descriptions are
+heavily inspired by `Swagger <http://swagger.wordnik.com/>`_.
+
+Service description schema
+==========================
+
+A Guzzle Service description must match the following JSON schema document. This document can also serve as a guide when
+implementing a Guzzle service description.
+
+Download the schema here: :download:`Guzzle JSON schema document </_downloads/guzzle-schema-1.0.json>`
+
+.. class:: overflow-height-500px
+
+ .. literalinclude:: ../_downloads/guzzle-schema-1.0.json
+ :language: json
+
+Top-level attributes
+--------------------
+
+Service descriptions are comprised of the following top-level attributes:
+
+.. code-block:: json
+
+ {
+ "name": "string",
+ "apiVersion": "string|number",
+ "baseUrl": "string",
+ "description": "string",
+ "operations": {},
+ "models": {},
+ "includes": ["string.php", "string.json"]
+ }
+
++-----------------------------------------+-------------------------+-----------------------------------------------------------------------------------------------------------------------+
+| Property Name | Value | Description |
++=========================================+=========================+=======================================================================================================================+
+| name | string | Name of the web service |
++-----------------------------------------+-------------------------+-----------------------------------------------------------------------------------------------------------------------+
+| apiVersion | string|number | Version identifier that the service description is compatible with |
++-----------------------------------------+-------------------------+-----------------------------------------------------------------------------------------------------------------------+
+| baseUrl or basePath | string | Base URL of the web service. Any relative URI specified in an operation will be merged with the baseUrl using the |
+| | | process defined in RFC 2396. Some clients require custom logic to determine the baseUrl. In those cases, it is best |
+| | | to not include a baseUrl in the service description, but rather allow the factory method of the client to configure |
+| | | the client’s baseUrl. |
++-----------------------------------------+-------------------------+-----------------------------------------------------------------------------------------------------------------------+
+| description | string | Short summary of the web service |
++-----------------------------------------+-------------------------+-----------------------------------------------------------------------------------------------------------------------+
+| operations | object containing | Operations of the service. The key is the name of the operation and value is the attributes of the operation. |
+| | :ref:`operation-schema` | |
+| | | |
++-----------------------------------------+-------------------------+-----------------------------------------------------------------------------------------------------------------------+
+| models | object containing | Schema models that can be referenced throughout the service description. Models can be used to define how an HTTP |
+| | :ref:`model-schema` | response is parsed into a ``Guzzle\Service\Resource\Model`` object when an operation uses a ``model`` ``responseType``|
++-----------------------------------------+-------------------------+-----------------------------------------------------------------------------------------------------------------------+
+| includes | array of .js, | Service description files to include and extend from (can be a .json, .js, or .php file) |
+| | .json, or .php | |
+| | files. | |
++-----------------------------------------+-------------------------+-----------------------------------------------------------------------------------------------------------------------+
+| (any additional properties) | mixed | Any additional properties specified as top-level attributes are allowed and will be treated as arbitrary data |
++-----------------------------------------+-------------------------+-----------------------------------------------------------------------------------------------------------------------+
+
+.. _operation-schema:
+
+Operations
+----------
+
+Operations are the actions that can be taken on a service. Each operation is given a unique name and has a distinct
+endpoint and HTTP method. If an API has a ``DELETE /users/:id`` operation, a satisfactory operation name might be
+``DeleteUser`` with a parameter of ``id`` that is inserted into the URI.
+
+.. class:: overflow-height-250px
+
+ .. code-block:: json
+
+ {
+ "operations": {
+ "operationName": {
+ "extends": "string",
+ "httpMethod": "GET|POST|PUT|DELETE|PATCH|string",
+ "uri": "string",
+ "summary": "string",
+ "class": "string",
+ "responseClass": "string",
+ "responseNotes": "string",
+ "type": "string",
+ "description": "string",
+ "responseType": "primitive|class|(model by name)|documentation|(string)",
+ "deprecated": false,
+ "errorResponses": [
+ {
+ "code": 500,
+ "reason": "Unexpected Error",
+ "class": "string"
+ }
+ ],
+ "data": {
+ "foo": "bar",
+ "baz": "bam"
+ },
+ "parameters": {}
+ }
+ }
+ }
+
+.. csv-table::
+ :header: "Property Name", "Value", "Description"
+ :widths: 20, 15, 65
+
+ "extends", "string", "Extend from another operation by name. The parent operation must be defined before the child."
+ "httpMethod", "string", "HTTP method used with the operation (e.g. GET, POST, PUT, DELETE, PATCH, etc)"
+ "uri", "string", "URI of the operation. The uri attribute can contain URI templates. The variables of the URI template are parameters of the operation with a location value of uri"
+ "summary", "string", "Short summary of what the operation does"
+ "class", "string", "Custom class to instantiate instead of the default Guzzle\\Service\\Command\\OperationCommand. Using this attribute allows you to define an operation using a service description, but allows more customized logic to be implemented in user-land code."
+ "responseClass", "string", "Defined what is returned from the method. Can be a primitive, class name, or model name. You can specify the name of a class to return a more customized result from the operation (for example, a domain model object). When using the name of a PHP class, the class must implement ``Guzzle\Service\Command\ResponseClassInterface``."
+ "responseNotes", "string", "A description of the response returned by the operation"
+ "responseType", "string", "The type of response that the operation creates: one of primitive, class, model, or documentation. If not specified, this value will be automatically inferred based on whether or not there is a model matching the name, if a matching class name is found, or set to 'primitive' by default."
+ "deprecated", "boolean", "Whether or not the operation is deprecated"
+ "errorResponses", "array", "Errors that could occur while executing the operation. Each item of the array is an object that can contain a 'code' (HTTP response status code of the error), 'reason' (reason phrase or description of the error), and 'class' (an exception class that will be raised when this error is encountered)"
+ "data", "object", "Any arbitrary data to associate with the operation"
+ "parameters", "object containing :ref:`parameter-schema` objects", "Parameters of the operation. Parameters are used to define how input data is serialized into a HTTP request."
+ "additionalParameters", "A single :ref:`parameter-schema` object", "Validation and serialization rules for any parameter supplied to the operation that was not explicitly defined."
+
+additionalParameters
+~~~~~~~~~~~~~~~~~~~~
+
+When a webservice offers a large number of parameters that all are set in the same location (for example the query
+string or a JSON document), defining each parameter individually can require a lot of time and repetition. Furthermore,
+some web services allow for completely arbitrary parameters to be supplied for an operation. The
+``additionalParameters`` attribute can be used to solve both of these issues.
+
+As an example, we can define a Twitter API operation quite easily using ``additionalParameters``. The
+GetMentions operation accepts a large number of query string parameters. Defining each of these parameters
+is ideal because it provide much more introspection for the client and opens the possibility to use the description with
+other tools (e.g. a documentation generator). However, you can very quickly provide a "catch-all" serialization rule
+that will place any custom parameters supplied to an operation the generated request's query string parameters.
+
+.. class:: overflow-height-250px
+
+ .. code-block:: json
+
+ {
+ "name": "Twitter",
+ "apiVersion": "1.1",
+ "baseUrl": "https://api.twitter.com/1.1",
+ "operations": {
+ "GetMentions": {
+ "httpMethod": "GET",
+ "uri": "statuses/mentions_timeline.json",
+ "responseClass": "GetMentionsOutput",
+ "additionalParameters": {
+ "location": "query"
+ }
+ }
+ },
+ "models": {
+ "GetMentionsOutput": {
+ "type": "object",
+ "additionalProperties": {
+ "location": "json"
+ }
+ }
+ }
+ }
+
+responseClass
+~~~~~~~~~~~~~
+
+The ``responseClass`` attribute is used to define the return value of an operation (what is returned by calling the
+``getResult()`` method of a command object). The value set in the responseClass attribute can be one of "primitive"
+(meaning the result with be primitive type like a string), a class name meaning the result will be an instance of a
+specific user-land class, or a model name meaning the result will be a ``Guzzle\Service\Resource\Model`` object that
+uses a :ref:`model schema <model-schema>` to define how the HTTP response is parsed.
+
+.. note::
+
+ Using a class name with a ``responseClass`` will only work if it is supported by the ``class`` that is instantiated
+ for the operation. Keep this in mind when specifying a custom ``class`` attribute that points to a custom
+ ``Guzzle\Service\Command\CommandInterface`` class. The default ``class``,
+ ``Guzzle\Service\Command\OperationCommand``, does support setting custom ``class`` attributes.
+
+You can specify the name of a class to return a more customized result from the operation (for example, a domain model
+object). When using the name of a PHP class, the class must implement ``Guzzle\Service\Command\ResponseClassInterface``.
+Here's a very simple example of implementing a custom responseClass object.
+
+.. code-block:: json
+
+ {
+ "operations": {
+ "test": {
+ "responseClass": "MyApplication\\User"
+ }
+ }
+ }
+
+.. code-block:: php
+
+ namespace MyApplication;
+
+ use Guzzle\Service\Command\ResponseClassInterface;
+ use Guzzle\Service\Command\OperationCommand;
+
+ class User implements ResponseClassInterface
+ {
+ protected $name;
+
+ public static function fromCommand(OperationCommand $command)
+ {
+ $response = $command->getResponse();
+ $xml = $response->xml();
+
+ return new self((string) $xml->name);
+ }
+
+ public function __construct($name)
+ {
+ $this->name = $name;
+ }
+ }
+
+errorResponses
+~~~~~~~~~~~~~~
+
+``errorResponses`` is an array containing objects that define the errors that could occur while executing the
+operation. Each item of the array is an object that can contain a 'code' (HTTP response status code of the error),
+'reason' (reason phrase or description of the error), and 'class' (an exception class that will be raised when this
+error is encountered).
+
+ErrorResponsePlugin
+^^^^^^^^^^^^^^^^^^^
+
+Error responses are by default only used for documentation. If you don't need very complex exception logic for your web
+service errors, then you can use the ``Guzzle\Plugin\ErrorResponse\ErrorResponsePlugin`` to automatically throw defined
+exceptions when one of the ``errorResponse`` rules are matched. The error response plugin will listen for the
+``request.complete`` event of a request created by a command object. Every response (including a successful response) is
+checked against the list of error responses for an exact match using the following order of checks:
+
+1. Does the errorResponse have a defined ``class``?
+2. Is the errorResponse ``code`` equal to the status code of the response?
+3. Is the errorResponse ``reason`` equal to the reason phrase of the response?
+4. Throw the exception stored in the ``class`` attribute of the errorResponse.
+
+The ``class`` attribute must point to a class that implements
+``Guzzle\Plugin\ErrorResponse\ErrorResponseExceptionInterface``. This interface requires that an error response class
+implements ``public static function fromCommand(CommandInterface $command, Response $response)``. This method must
+return an object that extends from ``\Exception``. After an exception is returned, it is thrown by the plugin.
+
+.. _parameter-schema:
+
+Parameter schema
+----------------
+
+Parameters in both operations and models are represented using the
+`JSON schema <http://tools.ietf.org/id/draft-zyp-json-schema-04.html>`_ syntax.
+
+.. csv-table::
+ :header: "Property Name", "Value", "Description"
+ :widths: 20, 15, 65
+
+ "name", "string", "Unique name of the parameter"
+ "type", "string|array", "Type of variable (string, number, integer, boolean, object, array, numeric, null, any). Types are using for validation and determining the structure of a parameter. You can use a union type by providing an array of simple types. If one of the union types matches the provided value, then the value is valid."
+ "instanceOf", "string", "When the type is an object, you can specify the class that the object must implement"
+ "required", "boolean", "Whether or not the parameter is required"
+ "default", "mixed", "Default value to use if no value is supplied"
+ "static", "boolean", "Set to true to specify that the parameter value cannot be changed from the default setting"
+ "description", "string", "Documentation of the parameter"
+ "location", "string", "The location of a request used to apply a parameter. Custom locations can be registered with a command, but the defaults are uri, query, statusCode, reasonPhrase, header, body, json, xml, postField, postFile, responseBody"
+ "sentAs", "string", "Specifies how the data being modeled is sent over the wire. For example, you may wish to include certain headers in a response model that have a normalized casing of FooBar, but the actual header is x-foo-bar. In this case, sentAs would be set to x-foo-bar."
+ "filters", "array", "Array of functions to to run a parameter value through."
+
+filters
+~~~~~~~
+
+Each value in the array must be a string containing the full class path to a static method or an array of complex
+filter information. You can specify static methods of classes using the full namespace class name followed by
+"::" (e.g. ``FooBar::baz()``). Some filters require arguments in order to properly filter a value. For complex filters,
+use an object containing a ``method`` attribute pointing to a function, and an ``args`` attribute containing an
+array of positional arguments to pass to the function. Arguments can contain keywords that are replaced when filtering
+a value: ``@value`` is replaced with the value being filtered, and ``@api`` is replaced with the actual Parameter
+object.
+
+.. code-block:: json
+
+ {
+ "filters": [
+ "strtolower",
+ {
+ "method": "MyClass::convertString",
+ "args": [ "test", "@value", "@api" ]
+ }
+ ]
+ }
+
+The above example will filter a parameter using ``strtolower``. It will then call the ``convertString`` static method
+of ``MyClass``, passing in "test", the actual value of the parameter, and a ``Guzzle\Service\Description\Parameter``
+object.
+
+Operation parameter location attributes
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+The location field of top-level parameters control how a parameter is serialized when generating a request.
+
+uri location
+^^^^^^^^^^^^
+
+Parameters are injected into the ``uri`` attribute of the operation using
+`URI-template expansion <http://tools.ietf.org/html/rfc6570>`_.
+
+.. code-block:: json
+
+ {
+ "operations": {
+ "uriTest": {
+ "uri": "/test/{testValue}",
+ "parameters": {
+ "testValue": {
+ "location": "uri"
+ }
+ }
+ }
+ }
+ }
+
+query location
+^^^^^^^^^^^^^^
+
+Parameters are injected into the query string of a request. Query values can be nested, which would result in a PHP
+style nested query string. The name of a parameter is the default name of the query string parameter added to the
+request. You can override this behavior by specifying the ``sentAs`` attribute on the parameter.
+
+.. code-block:: json
+
+ {
+ "operations": {
+ "queryTest": {
+ "parameters": {
+ "testValue": {
+ "location": "query",
+ "sentAs": "test_value"
+ }
+ }
+ }
+ }
+ }
+
+header location
+^^^^^^^^^^^^^^^
+
+Parameters are injected as headers on an HTTP request. The name of the parameter is used as the name of the header by
+default. You can change the name of the header created by the parameter using the ``sentAs`` attribute.
+
+Headers that are of type ``object`` will be added as multiple headers to a request using the key of the input array as
+the header key. Setting a ``sentAs`` attribute along with a type ``object`` will use the value of ``sentAs`` as a
+prefix for each header key.
+
+body location
+^^^^^^^^^^^^^
+
+Parameters are injected as the body of a request. The input of these parameters may be anything that can be cast to a
+string or a ``Guzzle\Http\EntityBodyInterface`` object.
+
+postField location
+^^^^^^^^^^^^^^^^^^
+
+Parameters are inserted as POST fields in a request. Nested values may be supplied and will be represented using
+PHP style nested query strings. The POST field name is the same as the parameter name by default. You can use the
+``sentAs`` parameter to override the POST field name.
+
+postFile location
+^^^^^^^^^^^^^^^^^
+
+Parameters are added as POST files. A postFile value may be a string pointing to a local filename or a
+``Guzzle\Http\Message\PostFileInterface`` object. The name of the POST file will be the name of the parameter by
+default. You can use a custom POST file name by using the ``sentAs`` attribute.
+
+Supports "string" and "array" types.
+
+json location
+^^^^^^^^^^^^^
+
+Parameters are added to the body of a request as top level keys of a JSON document. Nested values may be specified,
+with any number of nested ``Guzzle\Common\ToArrayInterface`` objects. When JSON parameters are specified, the
+``Content-Type`` of the request will change to ``application/json`` if a ``Content-Type`` has not already been specified
+on the request.
+
+xml location
+^^^^^^^^^^^^
+
+Parameters are added to the body of a request as top level nodes of an XML document. Nested values may be specified,
+with any number of nested ``Guzzle\Common\ToArrayInterface`` objects. When XML parameters are specified, the
+``Content-Type`` of the request will change to ``application/xml`` if a ``Content-Type`` has not already been specified
+on the request.
+
+responseBody location
+^^^^^^^^^^^^^^^^^^^^^
+
+Specifies the EntityBody of a response. This can be used to download the response body to a file or a custom Guzzle
+EntityBody object.
+
+No location
+^^^^^^^^^^^
+
+If a parameter has no location attribute, then the parameter is simply used as a data value.
+
+Other locations
+^^^^^^^^^^^^^^^
+
+Custom locations can be registered as new locations or override default locations if needed.
+
+.. _model-schema:
+
+Model Schema
+------------
+
+Models are used in service descriptions to provide generic JSON schema definitions that can be extended from or used in
+``$ref`` attributes. Models can also be referenced in a ``responseClass`` attribute to provide valuable output to an
+operation. Models are JSON schema documents and use the exact syntax and attributes used in parameters.
+
+Response Models
+~~~~~~~~~~~~~~~
+
+Response models describe how a response is parsed into a ``Guzzle\Service\Resource\Model`` object. Response models are
+always modeled as JSON schema objects. When an HTTP response is parsed using a response model, the rules specified on
+each property of a response model will translate 1:1 as keys in a PHP associative array. When a ``sentAs`` attribute is
+found in response model parameters, the value retrieved from the HTTP response is retrieved using the ``sentAs``
+parameter but stored in the response model using the name of the parameter.
+
+The location field of top-level parameters in a response model tell response parsers how data is retrieved from a
+response.
+
+statusCode location
+^^^^^^^^^^^^^^^^^^^
+
+Retrieves the status code of the response.
+
+reasonPhrase location
+^^^^^^^^^^^^^^^^^^^^^
+
+Retrieves the reason phrase of the response.
+
+header location
+^^^^^^^^^^^^^^^
+
+Retrieves a header from the HTTP response.
+
+body location
+^^^^^^^^^^^^^
+
+Retrieves the body of an HTTP response.
+
+json location
+^^^^^^^^^^^^^
+
+Retrieves a top-level parameter from a JSON document contained in an HTTP response.
+
+You can use ``additionalProperties`` if the JSON document is wrapped in an outer array. This allows you to parse the
+contents of each item in the array using the parsing rules defined in the ``additionalProperties`` schema.
+
+xml location
+^^^^^^^^^^^^
+
+Retrieves a top-level node value from an XML document contained in an HTTP response.
+
+Other locations
+^^^^^^^^^^^^^^^
+
+Custom locations can be registered as new locations or override default locations if needed.
+
+Example service description
+---------------------------
+
+Let's say you're interacting with a web service called 'Foo' that allows for the following routes and methods::
+
+ GET/POST /users
+ GET/DELETE /users/:id
+
+The following JSON service description implements this simple web service:
+
+.. class:: overflow-height-500px
+
+ .. code-block:: json
+
+ {
+ "name": "Foo",
+ "apiVersion": "2012-10-14",
+ "baseUrl": "http://api.foo.com",
+ "description": "Foo is an API that allows you to Baz Bar",
+ "operations": {
+ "GetUsers": {
+ "httpMethod": "GET",
+ "uri": "/users",
+ "summary": "Gets a list of users",
+ "responseClass": "GetUsersOutput"
+ },
+ "CreateUser": {
+ "httpMethod": "POST",
+ "uri": "/users",
+ "summary": "Creates a new user",
+ "responseClass": "CreateUserOutput",
+ "parameters": {
+ "name": {
+ "location": "json",
+ "type": "string"
+ },
+ "age": {
+ "location": "json",
+ "type": "integer"
+ }
+ }
+ },
+ "GetUser": {
+ "httpMethod": "GET",
+ "uri": "/users/{id}",
+ "summary": "Retrieves a single user",
+ "responseClass": "GetUserOutput",
+ "parameters": {
+ "id": {
+ "location": "uri",
+ "description": "User to retrieve by ID",
+ "required": true
+ }
+ }
+ },
+ "DeleteUser": {
+ "httpMethod": "DELETE",
+ "uri": "/users/{id}",
+ "summary": "Deletes a user",
+ "responseClass": "DeleteUserOutput",
+ "parameters": {
+ "id": {
+ "location": "uri",
+ "description": "User to delete by ID",
+ "required": true
+ }
+ }
+ }
+ },
+ "models": {
+ "GetUsersOutput": {
+ "type": "array",
+ "items": {
+ "type": "object",
+ "properties": {
+ "name": {
+ "location": "json",
+ "type": "string"
+ },
+ "age": {
+ "location": "json",
+ "type": "integer"
+ }
+ }
+ }
+ },
+ "CreateUserOutput": {
+ "type": "object",
+ "properties": {
+ "id": {
+ "location": "json",
+ "type": "string"
+ },
+ "location": {
+ "location": "header",
+ "sentAs": "Location",
+ "type": "string"
+ }
+ }
+ },
+ "GetUserOutput": {
+ "type": "object",
+ "properties": {
+ "name": {
+ "location": "json",
+ "type": "string"
+ },
+ "age": {
+ "location": "json",
+ "type": "integer"
+ }
+ }
+ },
+ "DeleteUserOutput": {
+ "type": "object",
+ "properties": {
+ "status": {
+ "location": "statusCode",
+ "type": "integer"
+ }
+ }
+ }
+ }
+ }
+
+If you attach this service description to a client, you would completely configure the client to interact with the
+Foo web service and provide valuable response models for each operation.
+
+.. code-block:: php
+
+ use Guzzle\Service\Description\ServiceDescription;
+
+ $description = ServiceDescription::factory('/path/to/client.json');
+ $client->setDescription($description);
+
+ $command = $client->getCommand('DeleteUser', array('id' => 123));
+ $responseModel = $client->execute($command);
+ echo $responseModel['status'];
+
+.. note::
+
+ You can add the service description to your client's factory method or constructor.
diff --git a/vendor/guzzle/guzzle/docs/webservice-client/using-the-service-builder.rst b/vendor/guzzle/guzzle/docs/webservice-client/using-the-service-builder.rst
new file mode 100644
index 0000000..b7113d6
--- /dev/null
+++ b/vendor/guzzle/guzzle/docs/webservice-client/using-the-service-builder.rst
@@ -0,0 +1,316 @@
+=======================
+Using a service builder
+=======================
+
+The best way to instantiate Guzzle web service clients is to let Guzzle handle building the clients for you using a
+ServiceBuilder. A ServiceBuilder is responsible for creating concrete client objects based on configuration settings
+and helps to manage credentials for different environments.
+
+You don't have to use a service builder, but they help to decouple your application from concrete classes and help to
+share configuration data across multiple clients. Consider the following example. Here we are creating two clients that
+require the same API public key and secret key. The clients are created using their ``factory()`` methods.
+
+.. code-block:: php
+
+ use MyService\FooClient;
+ use MyService\BarClient;
+
+ $foo = FooClient::factory(array(
+ 'key' => 'abc',
+ 'secret' => '123',
+ 'custom' => 'and above all'
+ ));
+
+ $bar = BarClient::factory(array(
+ 'key' => 'abc',
+ 'secret' => '123',
+ 'custom' => 'listen to me'
+ ));
+
+The redundant specification of the API keys can be removed using a service builder.
+
+.. code-block:: php
+
+ use Guzzle\Service\Builder\ServiceBuilder;
+
+ $builder = ServiceBuilder::factory(array(
+ 'services' => array(
+ 'abstract_client' => array(
+ 'params' => array(
+ 'key' => 'abc',
+ 'secret' => '123'
+ )
+ ),
+ 'foo' => array(
+ 'extends' => 'abstract_client',
+ 'class' => 'MyService\FooClient',
+ 'params' => array(
+ 'custom' => 'and above all'
+ )
+ ),
+ 'bar' => array(
+ 'extends' => 'abstract_client',
+ 'class' => 'MyService\FooClient',
+ 'params' => array(
+ 'custom' => 'listen to me'
+ )
+ )
+ )
+ ));
+
+ $foo = $builder->get('foo');
+ $bar = $builder->get('bar');
+
+You can make managing your API keys even easier by saving the service builder configuration in a JSON format in a
+.json file.
+
+Creating a service builder
+--------------------------
+
+A ServiceBuilder can source information from an array, an PHP include file that returns an array, or a JSON file.
+
+.. code-block:: php
+
+ use Guzzle\Service\Builder\ServiceBuilder;
+
+ // Source service definitions from a JSON file
+ $builder = ServiceBuilder::factory('services.json');
+
+Sourcing data from an array
+~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+Data can be source from a PHP array. The array must contain an associative ``services`` array that maps the name of a
+client to the configuration information used by the service builder to create the client. Clients are given names
+which are used to identify how a client is retrieved from a service builder. This can be useful for using multiple
+accounts for the same service or creating development clients vs. production clients.
+
+.. code-block:: php
+
+ $services = array(
+ 'includes' => array(
+ '/path/to/other/services.json',
+ '/path/to/other/php_services.php'
+ ),
+ 'services' => array(
+ 'abstract.foo' => array(
+ 'params' => array(
+ 'username' => 'foo',
+ 'password' => 'bar'
+ )
+ ),
+ 'bar' => array(
+ 'extends' => 'abstract.foo',
+ 'class' => 'MyClientClass',
+ 'params' => array(
+ 'other' => 'abc'
+ )
+ )
+ )
+ );
+
+A service builder configuration array contains two top-level array keys:
+
++------------+---------------------------------------------------------------------------------------------------------+
+| Key | Description |
++============+=========================================================================================================+
+| includes | Array of paths to JSON or PHP include files to include in the configuration. |
++------------+---------------------------------------------------------------------------------------------------------+
+| services | Associative array of defined services that can be created by the service builder. Each service can |
+| | contain the following keys: |
+| | |
+| | +------------+----------------------------------------------------------------------------------------+ |
+| | | Key | Description | |
+| | +============+========================================================================================+ |
+| | | class | The concrete class to instantiate that implements the | |
+| | | | ``Guzzle\Common\FromConfigInterface``. | |
+| | +------------+----------------------------------------------------------------------------------------+ |
+| | | extends | The name of a previously defined service to extend from | |
+| | +------------+----------------------------------------------------------------------------------------+ |
+| | | params | Associative array of parameters to pass to the factory method of the service it is | |
+| | | | instantiated | |
+| | +------------+----------------------------------------------------------------------------------------+ |
+| | | alias | An alias that can be used in addition to the array key for retrieving a client from | |
+| | | | the service builder. | |
+| | +------------+----------------------------------------------------------------------------------------+ |
++------------+---------------------------------------------------------------------------------------------------------+
+
+The first client defined, ``abstract.foo``, is used as a placeholder of shared configuration values. Any service
+extending abstract.foo will inherit its params. As an example, this can be useful when clients share the same username
+and password.
+
+The next client, ``bar``, extends from ``abstract.foo`` using the ``extends`` attribute referencing the client from
+which to extend. Additional parameters can be merged into the original service definition when extending a parent
+service.
+
+.. important::
+
+ Each client that you intend to instantiate must specify a ``class`` attribute that references the full class name
+ of the client being created. The class referenced in the ``class`` parameter must implement a static ``factory()``
+ method that accepts an array or ``Guzzle\Common\Collection`` object and returns an instantiated object.
+
+Sourcing from a PHP include
+~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+You can create service builder configurations using a PHP include file. This can be useful if you wish to take
+advantage of an opcode cache like APC to speed up the process of loading and processing the configuration. The PHP
+include file is the same format as an array, but you simply create a PHP script that returns an array and save the
+file with the .php file extension.
+
+.. code-block:: php
+
+ <?php return array('services' => '...');
+ // Saved as config.php
+
+This configuration file can then be used with a service builder.
+
+.. code-block:: php
+
+ $builder = ServiceBuilder::factory('/path/to/config.php');
+
+Sourcing from a JSON document
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+You can use JSON documents to serialize your service descriptions. The JSON format uses the exact same structure as
+the PHP array syntax, but it's just serialized using JSON.
+
+.. code-block:: javascript
+
+ {
+ "includes": ["/path/to/other/services.json", "/path/to/other/php_services.php"],
+ "services": {
+ "abstract.foo": {
+ "params": {
+ "username": "foo",
+ "password": "bar"
+ }
+ },
+ "bar": {
+ "extends": "abstract.foo",
+ "class": "MyClientClass",
+ "params": {
+ "other": "abc"
+ }
+ }
+ }
+ }
+
+Referencing other clients in parameters
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+If one of your clients depends on another client as one of its parameters, you can reference that client by name by
+enclosing the client's reference key in ``{}``.
+
+.. code-block:: javascript
+
+ {
+ "services": {
+ "token": {
+ "class": "My\Token\TokenFactory",
+ "params": {
+ "access_key": "xyz"
+ }
+ },
+ "client": {
+ "class": "My\Client",
+ "params": {
+ "token_client": "{token}",
+ "version": "1.0"
+ }
+ }
+ }
+ }
+
+When ``client`` is constructed by the service builder, the service builder will first create the ``token`` service
+and then inject the token service into ``client``'s factory method in the ``token_client`` parameter.
+
+Retrieving clients from a service builder
+-----------------------------------------
+
+Clients are referenced using a customizable name you provide in your service definition. The ServiceBuilder is a sort
+of multiton object-- it will only instantiate a client once and return that client for subsequent retrievals. Clients
+are retrieved by name (the array key used in the configuration) or by the ``alias`` setting of a service.
+
+Here's an example of retrieving a client from your ServiceBuilder:
+
+.. code-block:: php
+
+ $client = $builder->get('foo');
+
+ // You can also use the ServiceBuilder object as an array
+ $client = $builder['foo'];
+
+Creating throwaway clients
+~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+You can get a "throwaway" client (a client that is not persisted by the ServiceBuilder) by passing ``true`` in the
+second argument of ``ServiceBuilder::get()``. This allows you to create a client that will not be returned by other
+parts of your code that use the service builder. Instead of passing ``true``, you can pass an array of configuration
+settings that will override the configuration settings specified in the service builder.
+
+.. code-block:: php
+
+ // Get a throwaway client and overwrite the "custom" setting of the client
+ $foo = $builder->get('foo', array(
+ 'custom' => 'in this world there are rules'
+ ));
+
+Getting raw configuration settings
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+You can get the raw configuration settings provided to the service builder for a specific service using the
+``getData($name)`` method of a service builder. This method will null if the service was not found in the service
+builder or an array of configuration settings if the service was found.
+
+.. code-block:: php
+
+ $data = $builder->getData('foo');
+ echo $data['key'] . "\n";
+ echo $data['secret'] . "\n";
+ echo $data['custom'] . "\n";
+
+Adding a plugin to all clients
+------------------------------
+
+You can add a plugin to all clients created by a service builder using the ``addGlobalPlugin($plugin)`` method of a
+service builder and passing a ``Symfony\Component\EventDispatcher\EventSubscriberInterface`` object. The service builder
+will then attach each global plugin to every client as it is created. This allows you to, for example, add a LogPlugin
+to every request created by a service builder for easy debugging.
+
+.. code-block:: php
+
+ use Guzzle\Plugin\Log\LogPlugin;
+
+ // Add a debug log plugin to every client as it is created
+ $builder->addGlobalPlugin(LogPlugin::getDebugPlugin());
+
+ $foo = $builder->get('foo');
+ $foo->get('/')->send();
+ // Should output all of the data sent over the wire
+
+.. _service-builder-events:
+
+Events emitted from a service builder
+-------------------------------------
+
+A ``Guzzle\Service\Builder\ServiceBuilder`` object emits the following events:
+
++-------------------------------+--------------------------------------------+-----------------------------------------+
+| Event name | Description | Event data |
++===============================+============================================+=========================================+
+| service_builder.create_client | Called when a client is created | * client: The created client object |
++-------------------------------+--------------------------------------------+-----------------------------------------+
+
+.. code-block:: php
+
+ use Guzzle\Common\Event;
+ use Guzzle\Service\Builder\ServiceBuilder;
+
+ $builder = ServiceBuilder::factory('/path/to/config.json');
+
+ // Add an event listener to print out each client client as it is created
+ $builder->getEventDispatcher()->addListener('service_builder.create_client', function (Event $e) {
+ echo 'Client created: ' . get_class($e['client']) . "\n";
+ });
+
+ $foo = $builder->get('foo');
+ // Should output the class used for the "foo" client
diff --git a/vendor/guzzle/guzzle/docs/webservice-client/webservice-client.rst b/vendor/guzzle/guzzle/docs/webservice-client/webservice-client.rst
new file mode 100644
index 0000000..7ec771e
--- /dev/null
+++ b/vendor/guzzle/guzzle/docs/webservice-client/webservice-client.rst
@@ -0,0 +1,659 @@
+======================
+The web service client
+======================
+
+The ``Guzzle\Service`` namespace contains various abstractions that help to make it easier to interact with a web
+service API, including commands, service descriptions, and resource iterators.
+
+In this chapter, we'll build a simple `Twitter API client <https://dev.twitter.com/docs/api/1.1>`_.
+
+Creating a client
+=================
+
+A class that extends from ``Guzzle\Service\Client`` or implements ``Guzzle\Service\ClientInterface`` must implement a
+``factory()`` method in order to be used with a :doc:`service builder <using-the-service-builder>`.
+
+Factory method
+--------------
+
+You can use the ``factory()`` method of a client directly if you do not need a service builder.
+
+.. code-block:: php
+
+ use mtdowling\TwitterClient;
+
+ // Create a client and pass an array of configuration data
+ $twitter = TwitterClient::factory(array(
+ 'consumer_key' => '****',
+ 'consumer_secret' => '****',
+ 'token' => '****',
+ 'token_secret' => '****'
+ ));
+
+.. note::
+
+ If you'd like to follow along, here's how to get your Twitter API credentials:
+
+ 1. Visit https://dev.twitter.com/apps
+ 2. Click on an application that you've created
+ 3. Click on the "OAuth tool" tab
+ 4. Copy all of the settings under "OAuth Settings"
+
+Implementing a factory method
+-----------------------------
+
+Creating a client and its factory method is pretty simple. You just need to implement ``Guzzle\Service\ClientInterface``
+or extend from ``Guzzle\Service\Client``.
+
+.. code-block:: php
+
+ namespace mtdowling;
+
+ use Guzzle\Common\Collection;
+ use Guzzle\Plugin\Oauth\OauthPlugin;
+ use Guzzle\Service\Client;
+ use Guzzle\Service\Description\ServiceDescription;
+
+ /**
+ * A simple Twitter API client
+ */
+ class TwitterClient extends Client
+ {
+ public static function factory($config = array())
+ {
+ // Provide a hash of default client configuration options
+ $default = array('base_url' => 'https://api.twitter.com/1.1');
+
+ // The following values are required when creating the client
+ $required = array(
+ 'base_url',
+ 'consumer_key',
+ 'consumer_secret',
+ 'token',
+ 'token_secret'
+ );
+
+ // Merge in default settings and validate the config
+ $config = Collection::fromConfig($config, $default, $required);
+
+ // Create a new Twitter client
+ $client = new self($config->get('base_url'), $config);
+
+ // Ensure that the OauthPlugin is attached to the client
+ $client->addSubscriber(new OauthPlugin($config->toArray()));
+
+ return $client;
+ }
+ }
+
+Service Builder
+---------------
+
+A service builder is used to easily create web service clients, provides a simple configuration driven approach to
+creating clients, and allows you to share configuration settings across multiple clients. You can find out more about
+Guzzle's service builder in :doc:`using-the-service-builder`.
+
+.. code-block:: php
+
+ use Guzzle\Service\Builder\ServiceBuilder;
+
+ // Create a service builder and provide client configuration data
+ $builder = ServiceBuilder::factory('/path/to/client_config.json');
+
+ // Get the client from the service builder by name
+ $twitter = $builder->get('twitter');
+
+The above example assumes you have JSON data similar to the following stored in "/path/to/client_config.json":
+
+.. code-block:: json
+
+ {
+ "services": {
+ "twitter": {
+ "class": "mtdowling\\TwitterClient",
+ "params": {
+ "consumer_key": "****",
+ "consumer_secret": "****",
+ "token": "****",
+ "token_secret": "****"
+ }
+ }
+ }
+ }
+
+.. note::
+
+ A service builder becomes much more valuable when using multiple web service clients in a single application or
+ if you need to utilize the same client with varying configuration settings (e.g. multiple accounts).
+
+Commands
+========
+
+Commands are a concept in Guzzle that helps to hide the underlying implementation of an API by providing an easy to use
+parameter driven object for each action of an API. A command is responsible for accepting an array of configuration
+parameters, serializing an HTTP request, and parsing an HTTP response. Following the
+`command pattern <http://en.wikipedia.org/wiki/Command_pattern>`_, commands in Guzzle offer a greater level of
+flexibility when implementing and utilizing a web service client.
+
+Executing commands
+------------------
+
+You must explicitly execute a command after creating a command using the ``getCommand()`` method. A command has an
+``execute()`` method that may be called, or you can use the ``execute()`` method of a client object and pass in the
+command object. Calling either of these execute methods will return the result value of the command. The result value is
+the result of parsing the HTTP response with the ``process()`` method.
+
+.. code-block:: php
+
+ // Get a command from the client and pass an array of parameters
+ $command = $twitter->getCommand('getMentions', array(
+ 'count' => 5
+ ));
+
+ // Other parameters can be set on the command after it is created
+ $command['trim_user'] = false;
+
+ // Execute the command using the command object.
+ // The result value contains an array of JSON data from the response
+ $result = $command->execute();
+
+ // You can retrieve the result of the command later too
+ $result = $command->getResult().
+
+Command object also contains methods that allow you to inspect the HTTP request and response that was utilized with
+the command.
+
+.. code-block:: php
+
+ $request = $command->getRequest();
+ $response = $command->getResponse();
+
+.. note::
+
+ The format and notation used to retrieve commands from a client can be customized by injecting a custom command
+ factory, ``Guzzle\Service\Command\Factory\FactoryInterface``, on the client using ``$client->setCommandFactory()``.
+
+Executing with magic methods
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+When using method missing magic methods with a command, the command will be executed right away and the result of the
+command is returned.
+
+.. code-block:: php
+
+ $jsonData = $twitter->getMentions(array(
+ 'count' => 5,
+ 'trim_user' => true
+ ));
+
+Creating commands
+-----------------
+
+Commands are created using either the ``getCommand()`` method of a client or a magic missing method of a client. Using
+the ``getCommand()`` method allows you to create a command without executing it, allowing for customization of the
+command or the request serialized by the command.
+
+When a client attempts to create a command, it uses the client's ``Guzzle\Service\Command\Factory\FactoryInterface``.
+By default, Guzzle will utilize a command factory that first looks for a concrete class for a particular command
+(concrete commands) followed by a command defined by a service description (operation commands). We'll learn more about
+concrete commands and operation commands later in this chapter.
+
+.. code-block:: php
+
+ // Get a command from the twitter client.
+ $command = $twitter->getCommand('getMentions');
+ $result = $command->execute();
+
+Unless you've skipped ahead, running the above code will throw an exception.
+
+ PHP Fatal error: Uncaught exception 'Guzzle\Common\Exception\InvalidArgumentException' with message
+ 'Command was not found matching getMentions'
+
+This exception was thrown because the "getMentions" command has not yet been implemented. Let's implement one now.
+
+Concrete commands
+~~~~~~~~~~~~~~~~~
+
+Commands can be created in one of two ways: create a concrete command class that extends
+``Guzzle\Service\Command\AbstractCommand`` or
+:doc:`create an OperationCommand based on a service description <guzzle-service-descriptions>`. The recommended
+approach is to use a service description to define your web service, but you can use concrete commands when custom
+logic must be implemented for marshaling or unmarshaling a HTTP message.
+
+Commands are the method in which you abstract away the underlying format of the requests that need to be sent to take
+action on a web service. Commands in Guzzle are meant to be built by executing a series of setter methods on a command
+object. Commands are only validated right before they are executed. A ``Guzzle\Service\Client`` object is responsible
+for executing commands. Commands created for your web service must implement
+``Guzzle\Service\Command\CommandInterface``, but it's easier to extend the ``Guzzle\Service\Command\AbstractCommand``
+class, implement the ``build()`` method, and optionally implement the ``process()`` method.
+
+Serializing requests
+^^^^^^^^^^^^^^^^^^^^
+
+The ``build()`` method of a command is responsible for using the arguments of the command to build and serialize a
+HTTP request and set the request on the ``$request`` property of the command object. This step is usually taken care of
+for you when using a service description driven command that uses the default
+``Guzzle\Service\Command\OperationCommand``. You may wish to implement the process method yourself when you aren't
+using a service description or need to implement more complex request serialization.
+
+.. important::::
+
+ When implementing a custom ``build()`` method, be sure to set the class property of ``$this->request`` to an
+ instantiated and ready to send request.
+
+The following example shows how to implement the ``getMentions``
+`Twitter API <https://dev.twitter.com/docs/api/1.1/get/statuses/mentions_timeline>`_ method using a concrete command.
+
+.. code-block:: php
+
+ namespace mtdowling\Twitter\Command;
+
+ use Guzzle\Service\Command\AbstractCommand;
+
+ class GetMentions extends AbstractCommand
+ {
+ protected function build()
+ {
+ // Create the request property of the command
+ $this->request = $this->client->get('statuses/mentions_timeline.json');
+
+ // Grab the query object of the request because we will use it for
+ // serializing command parameters on the request
+ $query = $this->request->getQuery();
+
+ if ($this['count']) {
+ $query->set('count', $this['count']);
+ }
+
+ if ($this['since_id']) {
+ $query->set('since_id', $this['since_id']);
+ }
+
+ if ($this['max_id']) {
+ $query->set('max_id', $this['max_id']);
+ }
+
+ if ($this['trim_user'] !== null) {
+ $query->set('trim_user', $this['trim_user'] ? 'true' : 'false');
+ }
+
+ if ($this['contributor_details'] !== null) {
+ $query->set('contributor_details', $this['contributor_details'] ? 'true' : 'false');
+ }
+
+ if ($this['include_entities'] !== null) {
+ $query->set('include_entities', $this['include_entities'] ? 'true' : 'false');
+ }
+ }
+ }
+
+By default, a client will attempt to find concrete command classes under the ``Command`` namespace of a client. First
+the client will attempt to find an exact match for the name of the command to the name of the command class. If an
+exact match is not found, the client will calculate a class name using inflection. This is calculated based on the
+folder hierarchy of a command and converting the CamelCased named commands into snake_case. Here are some examples on
+how the command names are calculated:
+
+#. ``Foo\Command\JarJar`` **->** jar_jar
+#. ``Foo\Command\Test`` **->** test
+#. ``Foo\Command\People\GetCurrentPerson`` **->** people.get_current_person
+
+Notice how any sub-namespace beneath ``Command`` is converted from ``\`` to ``.`` (a period). CamelCasing is converted
+to lowercased snake_casing (e.g. JarJar == jar_jar).
+
+Parsing responses
+^^^^^^^^^^^^^^^^^
+
+The ``process()`` method of a command is responsible for converting an HTTP response into something more useful. For
+example, a service description operation that has specified a model object in the ``responseClass`` attribute of the
+operation will set a ``Guzzle\Service\Resource\Model`` object as the result of the command. This behavior can be
+completely modified as needed-- even if you are using operations and responseClass models. Simply implement a custom
+``process()`` method that sets the ``$this->result`` class property to whatever you choose. You can reuse parts of the
+default Guzzle response parsing functionality or get inspiration from existing code by using
+``Guzzle\Service\Command\OperationResponseParser`` and ``Guzzle\Service\Command\DefaultResponseParser`` classes.
+
+If you do not implement a custom ``process()`` method and are not using a service description, then Guzzle will attempt
+to guess how a response should be processed based on the Content-Type header of the response. Because the Twitter API
+sets a ``Content-Type: application/json`` header on this response, we do not need to implement any custom response
+parsing.
+
+Operation commands
+~~~~~~~~~~~~~~~~~~
+
+Operation commands are commands in which the serialization of an HTTP request and the parsing of an HTTP response are
+driven by a Guzzle service description. Because request serialization, validation, and response parsing are
+described using a DSL, creating operation commands is a much faster process than writing concrete commands.
+
+Creating operation commands for our Twitter client can remove a great deal of redundancy from the previous concrete
+command, and allows for a deeper runtime introspection of the API. Here's an example service description we can use to
+create the Twitter API client:
+
+.. code-block:: json
+
+ {
+ "name": "Twitter",
+ "apiVersion": "1.1",
+ "baseUrl": "https://api.twitter.com/1.1",
+ "description": "Twitter REST API client",
+ "operations": {
+ "GetMentions": {
+ "httpMethod": "GET",
+ "uri": "statuses/mentions_timeline.json",
+ "summary": "Returns the 20 most recent mentions for the authenticating user.",
+ "responseClass": "GetMentionsOutput",
+ "parameters": {
+ "count": {
+ "description": "Specifies the number of tweets to try and retrieve",
+ "type": "integer",
+ "location": "query"
+ },
+ "since_id": {
+ "description": "Returns results with an ID greater than the specified ID",
+ "type": "integer",
+ "location": "query"
+ },
+ "max_id": {
+ "description": "Returns results with an ID less than or equal to the specified ID.",
+ "type": "integer",
+ "location": "query"
+ },
+ "trim_user": {
+ "description": "Limits the amount of data returned for each user",
+ "type": "boolean",
+ "location": "query"
+ },
+ "contributor_details": {
+ "description": "Adds more data to contributor elements",
+ "type": "boolean",
+ "location": "query"
+ },
+ "include_entities": {
+ "description": "The entities node will be disincluded when set to false.",
+ "type": "boolean",
+ "location": "query"
+ }
+ }
+ }
+ },
+ "models": {
+ "GetMentionsOutput": {
+ "type": "object",
+ "additionalProperties": {
+ "location": "json"
+ }
+ }
+ }
+ }
+
+If you're lazy, you can define the API in a less descriptive manner using ``additionalParameters``.
+``additionalParameters`` define the serialization and validation rules of parameters that are not explicitly defined
+in a service description.
+
+.. code-block:: json
+
+ {
+ "name": "Twitter",
+ "apiVersion": "1.1",
+ "baseUrl": "https://api.twitter.com/1.1",
+ "description": "Twitter REST API client",
+ "operations": {
+ "GetMentions": {
+ "httpMethod": "GET",
+ "uri": "statuses/mentions_timeline.json",
+ "summary": "Returns the 20 most recent mentions for the authenticating user.",
+ "responseClass": "GetMentionsOutput",
+ "additionalParameters": {
+ "location": "query"
+ }
+ }
+ },
+ "models": {
+ "GetMentionsOutput": {
+ "type": "object",
+ "additionalProperties": {
+ "location": "json"
+ }
+ }
+ }
+ }
+
+You should attach the service description to the client at the end of the client's factory method:
+
+.. code-block:: php
+
+ // ...
+ class TwitterClient extends Client
+ {
+ public static function factory($config = array())
+ {
+ // ... same code as before ...
+
+ // Set the service description
+ $client->setDescription(ServiceDescription::factory('path/to/twitter.json'));
+
+ return $client;
+ }
+ }
+
+The client can now use operations defined in the service description instead of requiring you to create concrete
+command classes. Feel free to delete the concrete command class we created earlier.
+
+.. code-block:: php
+
+ $jsonData = $twitter->getMentions(array(
+ 'count' => 5,
+ 'trim_user' => true
+ ));
+
+Executing commands in parallel
+------------------------------
+
+Much like HTTP requests, Guzzle allows you to send multiple commands in parallel. You can send commands in parallel by
+passing an array of command objects to a client's ``execute()`` method. The client will serialize each request and
+send them all in parallel. If an error is encountered during the transfer, then a
+``Guzzle\Service\Exception\CommandTransferException`` is thrown, which allows you to retrieve a list of commands that
+succeeded and a list of commands that failed.
+
+.. code-block:: php
+
+ use Guzzle\Service\Exception\CommandTransferException;
+
+ $commands = array();
+ $commands[] = $twitter->getCommand('getMentions');
+ $commands[] = $twitter->getCommand('otherCommandName');
+ // etc...
+
+ try {
+ $result = $client->execute($commands);
+ foreach ($result as $command) {
+ echo $command->getName() . ': ' . $command->getResponse()->getStatusCode() . "\n";
+ }
+ } catch (CommandTransferException $e) {
+ // Get an array of the commands that succeeded
+ foreach ($e->getSuccessfulCommands() as $command) {
+ echo $command->getName() . " succeeded\n";
+ }
+ // Get an array of the commands that failed
+ foreach ($e->getFailedCommands() as $command) {
+ echo $command->getName() . " failed\n";
+ }
+ }
+
+.. note::
+
+ All commands executed from a client using an array must originate from the same client.
+
+Special command options
+-----------------------
+
+Guzzle exposes several options that help to control how commands are validated, serialized, and parsed.
+Command options can be specified when creating a command or in the ``command.params`` parameter in the
+``Guzzle\Service\Client``.
+
+=========================== ============================================================================================
+command.request_options Option used to add :ref:`Request options <request-options>` to the request created by a
+ command
+command.hidden_params An array of the names of parameters ignored by the ``additionalParameters`` parameter schema
+command.disable_validation Set to true to disable JSON schema validation of the command's input parameters
+command.response_processing Determines how the default response parser will parse the command. One of "raw" no parsing,
+ "model" (the default method used to parse commands using response models defined in service
+ descriptions)
+command.headers (deprecated) Option used to specify custom headers. Use ``command.request_options`` instead
+command.on_complete (deprecated) Option used to add an onComplete method to a command. Use
+ ``command.after_send`` event instead
+command.response_body (deprecated) Option used to change the entity body used to store a response.
+ Use ``command.request_options`` instead
+=========================== ============================================================================================
+
+Advanced client configuration
+=============================
+
+Default command parameters
+--------------------------
+
+When creating a client object, you can specify default command parameters to pass into all commands. Any key value pair
+present in the ``command.params`` settings of a client will be added as default parameters to any command created
+by the client.
+
+.. code-block:: php
+
+ $client = new Guzzle\Service\Client(array(
+ 'command.params' => array(
+ 'default_1' => 'foo',
+ 'another' => 'bar'
+ )
+ ));
+
+Magic methods
+-------------
+
+Client objects will, by default, attempt to create and execute commands when a missing method is invoked on a client.
+This powerful concept applies to both concrete commands and operation commands powered by a service description. This
+makes it appear to the end user that you have defined actual methods on a client object, when in fact, the methods are
+invoked using PHP's magic ``__call`` method.
+
+The ``__call`` method uses the ``getCommand()`` method of a client, which uses the client's internal
+``Guzzle\Service\Command\Factory\FactoryInterface`` object. The default command factory allows you to instantiate
+operations defined in a client's service description. The method in which a client determines which command to
+execute is defined as follows:
+
+1. The client will first try to find a literal match for an operation in the service description.
+2. If the literal match is not found, the client will try to uppercase the first character of the operation and find
+ the match again.
+3. If a match is still not found, the command factory will inflect the method name from CamelCase to snake_case and
+ attempt to find a matching command.
+4. If a command still does not match, an exception is thrown.
+
+.. code-block:: php
+
+ // Use the magic method
+ $result = $twitter->getMentions();
+
+ // This is exactly the same as:
+ $result = $twitter->getCommand('getMentions')->execute();
+
+You can disable magic methods on a client by passing ``false`` to the ``enableMagicMethod()`` method.
+
+Custom command factory
+----------------------
+
+A client by default uses the ``Guzzle\Service\Command\Factory\CompositeFactory`` which allows multiple command
+factories to attempt to create a command by a certain name. The default CompositeFactory uses a ``ConcreteClassFactory``
+and a ``ServiceDescriptionFactory`` if a service description is specified on a client. You can specify a custom
+command factory if your client requires custom command creation logic using the ``setCommandFactory()`` method of
+a client.
+
+Custom resource Iterator factory
+--------------------------------
+
+Resource iterators can be retrieved from a client using the ``getIterator($name)`` method of a client. This method uses
+a client's internal ``Guzzle\Service\Resource\ResourceIteratorFactoryInterface`` object. A client by default uses a
+``Guzzle\Service\Resource\ResourceIteratorClassFactory`` to attempt to find concrete classes that implement resource
+iterators. The default factory will first look for matching iterators in the ``Iterator`` subdirectory of the client
+followed by the ``Model`` subdirectory of a client. Use the ``setResourceIteratorFactory()`` method of a client to
+specify a custom resource iterator factory.
+
+Plugins and events
+==================
+
+``Guzzle\Service\Client`` exposes various events that allow you to hook in custom logic. A client object owns a
+``Symfony\Component\EventDispatcher\EventDispatcher`` object that can be accessed by calling
+``$client->getEventDispatcher()``. You can use the event dispatcher to add listeners (a simple callback function) or
+event subscribers (classes that listen to specific events of a dispatcher).
+
+.. _service-client-events:
+
+Events emitted from a Service Client
+------------------------------------
+
+A ``Guzzle\Service\Client`` object emits the following events:
+
++------------------------------+--------------------------------------------+------------------------------------------+
+| Event name | Description | Event data |
++==============================+============================================+==========================================+
+| client.command.create | The client created a command object | * client: Client object |
+| | | * command: Command object |
++------------------------------+--------------------------------------------+------------------------------------------+
+| command.before_prepare | Before a command is validated and built. | * command: Command being prepared |
+| | This is also before a request is created. | |
++------------------------------+--------------------------------------------+------------------------------------------+
+| command.after_prepare | After a command instantiates and | * command: Command that was prepared |
+| | configures its request object. | |
++------------------------------+--------------------------------------------+------------------------------------------+
+| command.before_send | The client is about to execute a prepared | * command: Command to execute |
+| | command | |
++------------------------------+--------------------------------------------+------------------------------------------+
+| command.after_send | The client successfully completed | * command: The command that was executed |
+| | executing a command | |
++------------------------------+--------------------------------------------+------------------------------------------+
+| command.parse_response | Called when ``responseType`` is ``class`` | * command: The command with a response |
+| | and the response is about to be parsed. | about to be parsed. |
++------------------------------+--------------------------------------------+------------------------------------------+
+
+.. code-block:: php
+
+ use Guzzle\Common\Event;
+ use Guzzle\Service\Client;
+
+ $client = new Client();
+
+ // create an event listener that operates on request objects
+ $client->getEventDispatcher()->addListener('command.after_prepare', function (Event $event) {
+ $command = $event['command'];
+ $request = $command->getRequest();
+
+ // do something with request
+ });
+
+.. code-block:: php
+
+ use Guzzle\Common\Event;
+ use Guzzle\Common\Client;
+ use Symfony\Component\EventDispatcher\EventSubscriberInterface;
+
+ class EventSubscriber implements EventSubscriberInterface
+ {
+ public static function getSubscribedEvents()
+ {
+ return array(
+ 'client.command.create' => 'onCommandCreate',
+ 'command.parse_response' => 'onParseResponse'
+ );
+ }
+
+ public function onCommandCreate(Event $event)
+ {
+ $client = $event['client'];
+ $command = $event['command'];
+ // operate on client and command
+ }
+
+ public function onParseResponse(Event $event)
+ {
+ $command = $event['command'];
+ // operate on the command
+ }
+ }
+
+ $client = new Client();
+
+ $client->addSubscriber(new EventSubscriber());