module HTTParty::ClassMethods
Common Request Options¶ ↑
Request methods (get, post, patch, put, delete, head, options) all take a common set of options. These are:
- :
body
: -
Body of the request. If passed a Hash, will try to normalize it first, by default passing it to ActiveSupport::to_params. Any other kind of object will get used as-is.
- :
http_proxyaddr
: -
Address of proxy server to use.
- :
http_proxyport
: -
Port of proxy server to use.
- :
http_proxyuser
: -
User for proxy server authentication.
- :
http_proxypass
: -
Password for proxy server authentication.
- :
limit
: -
Maximum number of redirects to follow. Takes precedences over :
no_follow
. - :
query
: -
Query string, or a Hash representing it. Normalized according to the same rules as :
body
. If you specify this on a POST, you must use a Hash. See also #default_params. - :
timeout
: -
Timeout for opening connection and reading data.
- :+local_host:
-
Local address to bind to before connecting.
- :+local_port:
-
Local port to bind to before connecting.
There are also another set of options with names corresponding to various class methods. The methods in question are those that let you set a class-wide default, and the options override the defaults on a request-by-request basis. Those options are:
-
:
base_uri
: see #base_uri. -
:
basic_auth
: see #basic_auth. Only one of :basic_auth
and :digest_auth
can be used at a time; if you try using both, you'll get an ArgumentError. -
:
debug_output
: see #debug_output. -
:
digest_auth
: see #digest_auth. Only one of :basic_auth
and :digest_auth
can be used at a time; if you try using both, you'll get an ArgumentError. -
:
format
: see #format. -
:
headers
: see #headers. Must be a Hash. -
:
maintain_method_across_redirects
: see #maintain_method_across_redirects. -
:
no_follow
: see #no_follow. -
:
parser
: see #parser. -
:
connection_adapter
: see #connection_adapter. -
:
pem
: see #pem. -
:
query_string_normalizer
: see #query_string_normalizer -
:
ssl_ca_file
: see #ssl_ca_file. -
:
ssl_ca_path
: see #ssl_ca_path.
Public Instance Methods
Allows setting a base uri to be used for each request. Will normalize uri to include http, etc.
class Foo include HTTParty base_uri 'twitter.com' end
# File lib/httparty.rb, line 107 def base_uri(uri=nil) return default_options[:base_uri] unless uri default_options[:base_uri] = HTTParty.normalize_base_uri(uri) end
Allows setting basic authentication username and password.
class Foo include HTTParty basic_auth 'username', 'password' end
# File lib/httparty.rb, line 118 def basic_auth(u, p) default_options[:basic_auth] = {:username => u, :password => p} end
Allows setting of SSL ciphers to use. This only works in Ruby 1.9+. You can get a list of valid specific ciphers from OpenSSL::Cipher.ciphers. You also can specify a cipher suite here, listed here at openssl.org: www.openssl.org/docs/apps/ciphers.html#CIPHER_SUITE_NAMES
class Foo include HTTParty ciphers "RC4-SHA" end
# File lib/httparty.rb, line 371 def ciphers(cipher_names) default_options[:ciphers] = cipher_names end
Allows setting a custom #connection_adapter for the http connections
@example
class Foo include HTTParty connection_adapter Proc.new {|uri, options| ... } end
@example provide optional configuration for your #connection_adapter
class Foo include HTTParty connection_adapter Proc.new {|uri, options| ... }, {:foo => :bar} end
@see HTTParty::ConnectionAdapter
# File lib/httparty.rb, line 435 def connection_adapter(custom_adapter = nil, options = nil) if custom_adapter.nil? default_options[:connection_adapter] else default_options[:connection_adapter] = custom_adapter default_options[:connection_adapter_options] = options end end
Perform a COPY request to a path
# File lib/httparty.rb, line 497 def copy(path, options={}, &block) perform_request Net::HTTP::Copy, path, options, &block end
Set an output stream for debugging, defaults to $stderr. The output stream is passed on to Net::HTTP#set_debug_output.
class Foo include HTTParty debug_output $stderr end
# File lib/httparty.rb, line 208 def debug_output(stream = $stderr) default_options[:debug_output] = stream end
Allows setting default parameters to be appended to each request. Great for api keys and such.
class Foo include HTTParty default_params :api_key => 'secret', :another => 'foo' end
# File lib/httparty.rb, line 161 def default_params(h={}) raise ArgumentError, 'Default params must be a hash' unless h.is_a?(Hash) default_options[:default_params] ||= {} default_options[:default_params].merge!(h) end
Allows setting a default timeout for all HTTP calls Timeout is specified in seconds.
class Foo include HTTParty default_timeout 10 end
# File lib/httparty.rb, line 174 def default_timeout(t) raise ArgumentError, 'Timeout must be an integer or float' unless t && (t.is_a?(Integer) || t.is_a?(Float)) default_options[:timeout] = t end
Perform a DELETE request to a path
# File lib/httparty.rb, line 487 def delete(path, options={}, &block) perform_request Net::HTTP::Delete, path, options, &block end
Allows setting digest authentication username and password.
class Foo include HTTParty digest_auth 'username', 'password' end
# File lib/httparty.rb, line 128 def digest_auth(u, p) default_options[:digest_auth] = {:username => u, :password => p} end
Do not send rails style query strings. Specically, don't use bracket notation when sending an array
For a query:
get '/', :query => {:selected_ids => [1,2,3]}
The default query string looks like this:
/?selected_ids[]=1&selected_ids[]=2&selected_ids[]=3
Call `disable_rails_query_string_format` to transform the query string into:
/?selected_ids=1&selected_ids=2&selected_ids=3
@example
class Foo include HTTParty disable_rails_query_string_format end
# File lib/httparty.rb, line 150 def disable_rails_query_string_format query_string_normalizer Request::NON_RAILS_QUERY_STRING_NORMALIZER end
Proceed to the location header when an HTTP response dictates a redirect. Redirects are always followed by default.
@example
class Foo include HTTParty base_uri 'http://google.com' follow_redirects true end
# File lib/httparty.rb, line 238 def follow_redirects(value = true) default_options[:follow_redirects] = value end
Allows setting the format with which to parse. Must be one of the allowed formats ie: json, xml
class Foo include HTTParty format :json end
# File lib/httparty.rb, line 249 def format(f = nil) if f.nil? default_options[:format] else parser(Parser) if parser.nil? default_options[:format] = f validate_format end end
Allows making a get request to a url.
class Foo include HTTParty end # Simple get with full url Foo.get('http://foo.com/resource.json') # Simple get with full url and query parameters # ie: http://foo.com/resource.json?limit=10 Foo.get('http://foo.com/resource.json', :query => {:limit => 10})
# File lib/httparty.rb, line 456 def get(path, options={}, &block) perform_request Net::HTTP::Get, path, options, &block end
Perform a HEAD request to a path
# File lib/httparty.rb, line 502 def head(path, options={}, &block) perform_request Net::HTTP::Head, path, options, &block end
Allows setting HTTP headers to be used for each request.
class Foo include HTTParty headers 'Accept' => 'text/html' end
# File lib/httparty.rb, line 218 def headers(h={}) raise ArgumentError, 'Headers must be a hash' unless h.is_a?(Hash) default_options[:headers] ||= {} default_options[:headers].merge!(h) end
Allows setting http proxy information to be used
class Foo include HTTParty http_proxy 'http://foo.com', 80, 'user', 'pass' end
# File lib/httparty.rb, line 93 def http_proxy(addr=nil, port=nil, user=nil, pass=nil) default_options[:http_proxyaddr] = addr default_options[:http_proxyport] = port default_options[:http_proxyuser] = user default_options[:http_proxypass] = pass end
Turns on logging
class Foo include HTTParty logger Logger.new('http_logger'), :info, :apache end
# File lib/httparty.rb, line 81 def logger(logger, level=:info, format=:apache) default_options[:logger] = logger default_options[:log_level] = level default_options[:log_format] = format end
Declare that you wish to maintain the chosen HTTP method across redirects. The default behavior is to follow redirects via the GET method. If you wish to maintain the original method, you can set this option to true.
@example
class Foo include HTTParty base_uri 'http://google.com' maintain_method_across_redirects true end
# File lib/httparty.rb, line 293 def maintain_method_across_redirects(value = true) default_options[:maintain_method_across_redirects] = value end
Perform a MOVE request to a path
# File lib/httparty.rb, line 492 def move(path, options={}, &block) perform_request Net::HTTP::Move, path, options, &block end
Declare whether or not to follow redirects. When true, an {HTTParty::RedirectionTooDeep} error will raise upon encountering a redirect. You can then gain access to the response object via HTTParty::ResponseError#response.
@see HTTParty::ResponseError#response
@example
class Foo include HTTParty base_uri 'http://google.com' no_follow true end begin Foo.get('/') rescue HTTParty::RedirectionTooDeep => e puts e.response.body end
# File lib/httparty.rb, line 278 def no_follow(value = false) default_options[:no_follow] = value end
Allows setting a default #open_timeout for all HTTP calls in seconds
class Foo include HTTParty open_timeout 10 end
# File lib/httparty.rb, line 185 def open_timeout(t) raise ArgumentError, 'open_timeout must be an integer or float' unless t && (t.is_a?(Integer) || t.is_a?(Float)) default_options[:open_timeout] = t end
Perform an OPTIONS request to a path
# File lib/httparty.rb, line 507 def options(path, options={}, &block) perform_request Net::HTTP::Options, path, options, &block end
Allows setting a custom parser for the response.
class Foo include HTTParty parser Proc.new {|data| ...} end
# File lib/httparty.rb, line 411 def parser(custom_parser = nil) if custom_parser.nil? default_options[:parser] else default_options[:parser] = custom_parser validate_format end end
Perform a PATCH request to a path
# File lib/httparty.rb, line 477 def patch(path, options={}, &block) perform_request Net::HTTP::Patch, path, options, &block end
Allows setting a PEM file to be used
class Foo include HTTParty pem File.read('/home/user/my.pem'), "optional password" end
# File lib/httparty.rb, line 303 def pem(pem_contents, password=nil) default_options[:pem] = pem_contents default_options[:pem_password] = password end
Allows setting a PKCS12 file to be used
class Foo include HTTParty pkcs12 File.read('/home/user/my.p12'), "password" end
# File lib/httparty.rb, line 314 def pkcs12(p12_contents, password) default_options[:p12] = p12_contents default_options[:p12_password] = password end
Allows making a post request to a url.
class Foo include HTTParty end # Simple post with full url and setting the body Foo.post('http://foo.com/resources', :body => {:bar => 'baz'}) # Simple post with full url using :query option, # which gets set as form data on the request. Foo.post('http://foo.com/resources', :query => {:bar => 'baz'})
# File lib/httparty.rb, line 472 def post(path, options={}, &block) perform_request Net::HTTP::Post, path, options, &block end
Perform a PUT request to a path
# File lib/httparty.rb, line 482 def put(path, options={}, &block) perform_request Net::HTTP::Put, path, options, &block end
Override the way query strings are normalized. Helpful for overriding the default rails normalization of Array queries.
For a query:
get '/', :query => {:selected_ids => [1,2,3]}
The default query string normalizer returns:
/?selected_ids[]=1&selected_ids[]=2&selected_ids[]=3
Let's change it to this:
/?selected_ids=1&selected_ids=2&selected_ids=3
Pass a Proc to the query normalizer which accepts the yielded query.
@example Modifying Array query strings
class ServiceWrapper include HTTParty query_string_normalizer proc { |query| query.map do |key, value| value.map {|v| "#{key}=#{v}"} end.join('&') } end
@param [Proc] normalizer custom query string normalizer. @yield [Hash, String] query string @yieldreturn [Array] an array that will later be joined with '&'
# File lib/httparty.rb, line 347 def query_string_normalizer(normalizer) default_options[:query_string_normalizer] = normalizer end
Allows setting a default #read_timeout for all HTTP calls in seconds
class Foo include HTTParty read_timeout 10 end
# File lib/httparty.rb, line 196 def read_timeout(t) raise ArgumentError, 'read_timeout must be an integer or float' unless t && (t.is_a?(Integer) || t.is_a?(Float)) default_options[:read_timeout] = t end
Allows setting an OpenSSL certificate authority file. The file should contain one or more certificates in PEM format.
Setting this option enables certificate verification. All certificates along a chain must be available in #ssl_ca_file or #ssl_ca_path for verification to succeed.
class Foo include HTTParty ssl_ca_file '/etc/ssl/certs/ca-certificates.crt' end
# File lib/httparty.rb, line 387 def ssl_ca_file(path) default_options[:ssl_ca_file] = path end
Allows setting an OpenSSL certificate authority path (directory).
Setting this option enables certificate verification. All certificates along a chain must be available in #ssl_ca_file or #ssl_ca_path for verification to succeed.
class Foo include HTTParty ssl_ca_path '/etc/ssl/certs/' end
# File lib/httparty.rb, line 401 def ssl_ca_path(path) default_options[:ssl_ca_path] = path end
Allows setting of SSL version to use. This only works in Ruby 1.9+. You can get a list of valid versions from OpenSSL::SSL::SSLContext::METHODS.
class Foo include HTTParty ssl_version :SSLv3 end
# File lib/httparty.rb, line 358 def ssl_version(version) default_options[:ssl_version] = version end
Private Instance Methods
# File lib/httparty.rb, line 524 def process_headers(options) if options[:headers] && headers.any? options[:headers] = headers.merge(options[:headers]) end end
# File lib/httparty.rb, line 536 def validate_format if format && parser.respond_to?(:supports_format?) && !parser.supports_format?(format) raise UnsupportedFormat, "'#{format.inspect}' Must be one of: #{parser.supported_formats.map{|f| f.to_s}.sort.join(', ')}" end end