This gem offers read/write access to YieldManager's API tools and ad-hoc reporting through the Reportware tool.
Currently it generates a fresh wsdl from the api.yieldmanager.com site the first time you use a service (in the future it will use locally-cached copies) and re-uses that wsdl for the life of the Yieldmanager::Client object.
This version implements API version 1.33.
Yieldmanager is available as a gem for easy installation.
sudo gem install yieldmanager
or if you're using RVM (and why on earth wouldn't you?)
gem install yieldmanager
The project is available for review/forking on github.com
git clone git://github.com/billgathen/yieldmanager.git
To use in a Rails project, add this to config/environment.rb:
config.gem 'yieldmanager'
require 'yieldmanager'
@ym = Yieldmanager::Client.new(
:user => "bob",
:pass => "secret"
)
The default environment is production. To access the test environment, use this:
@ym = Yieldmanager::Client.new(
:user => "bob",
:pass => "secret"
:env => "test"
)
The keys can also be passed as strings: 'user', 'pass' and 'env'.
NOTE Changing the environment after creation has no effect!
To check (or log) the current API version, execute the following:
Yieldmanager::Client.api_version
@ym.available_services
@ym.session do |token|
@currencies = @ym.dictionary.getCurrencies(token)
end
GOTCHA In projects with ActiveRecord enabled (i.e., Rails projects) SOAP will identify returned data as AR objects if there's a naming collision. For example, if you're running
@ym.creative.get(token,123)
and you have an AR objects for a creatives table in the db, the SOAP parser will interpret the returned SOAP object as an AR Creative object, resulting in bizarre errors. Uniquely re-name your AR object to eliminate the conflict.
Some calls return datasets too large to retrieve all at once. Pagination allows you to pull them back incrementally, handling the partial dataset on-the-fly or accumulating it for later use.
BLOCK_SIZE = 50
id = 1
@ym.session do |token|
@ym.paginate(BLOCK_SIZE) do |block|
(lines,tot) = @ym.line_item.getByBuyer(token,id,BLOCK_SIZE,block)
# ...do something with lines...
tot # remember to return total!
end
end
Accessing reportware assumes you've used the "Get request XML" functionality in the UI to get your request XML, or have crafted one from scratch. Assuming it's in a variable called request_xml, you'd access the data this way:
@ym.session do |token|
report = @ym.pull_report(token, request_xml)
puts report.headers.join("\t")
report.data.each do |row|
puts row.join("\t")
end
end
Column data can be accessed either by index or column name:
report.headers # => ['advertiser_name','seller_imps']
report.data[0][0] # => "Bob's Ads"
report.data[0].by_name('advertiser_name') # => "Bob's Ads"
report.data[0].by_name(:advertiser_name) # => "Bob's Ads"
If you call by_name with a non-existent column, it will throw an ArgumentError telling you so.
Or you can extract the report to an array of named hashes, removing dependencies on the gem for consumers of the data (say, across an API):
hashes = report.to_hashes
hashes[0]['advertiser_name'] # => "Bob's Ads"
NOTE Any totals columns your xml requests will be interpreted as ordinary data.
When simulating report calls without actually hitting Yieldmanager, you can create your own reports.
rpt = Yieldmanager::Report.new
rpt.headers = ["first","second"]
rpt.add_row([1,2])
rpt.data.first.by_name("first").should == 1
rpt.data.first.by_name("second").should == 2
To see the nitty-gritty of what's going over the wire (Yahoo tech support often asks for this), you can activate a "wiredump" on a per-service basis. Typically you just echo it to standard out. For instance:
client.entity.wiredump_dev = $stdout # on
adv = client.entity.get(token,12345)
client.entity.wiredump_dev = nil # off
For Rails in a passenger environment, standard out doesn't end up in the logfiles. Instead, redirect to a file:
wiredump = File.new("#{Rails.root}/log/wiredump_entity_#{Time.new.strftime('%H%M%S')}.log",'w')
client.entity.wiredump_dev = wiredump # on
adv = client.entity.get(token,12345)
wiredump.flush # make sure everything gets in there before it closes
client.entity.wiredump_dev = nil # off
The last 2 lines belong in an ensure block, so the file is created even when there's an error (which is probably why you're doing this).
The session method opens a session, gives you a token to use in your service calls, then closes the session when the block ends, even if an exception is raised during processing. It's the recommended method to ensure you don't hang connections when things go wrong. If you use start/end, make sure you wrap your logic in a begin/ensure clause and call end_session from the ensure.
In an effort to be a good ruby citizen, KarateCode and I have made the gem 1.9 compatible, but it is based on soap4r 1.5.8, which requires a pile of monkey-patches to get working and not every combination of calls and args have been tested in 1.9. If you're interested in what's been done, check out lib/soap4r_19_patch and Tomor Doron's blog post.
- Fork the project.
- Make your feature addition or bug fix.
- Add specs for it. This is important so I don't break it in a future version unintentionally.
- Commit, do not mess with rakefile, version, or history. (if you want to have your own version, that is fine but bump version in a commit by itself I can ignore when I pull)
- Send me a pull request. Bonus points for topic branches.
- manlycode[https://github.com/manlycode] (Chris Rittersdorf)
- KarateCode[https://github.com/KarateCode] (Michael Schneider)
Copyright (c) 2009-2011 Bill Gathen. See LICENSE for details.