Bugzilla4Python

This project aims to be a bugzilla-interface for python 3. The python-libraries that exist for bugzilla either use the old XMLRPC-interface or have a limited set of features or both. The best library I found was https://github.com/gdestuynder/simple_bugzilla.

Despite the features of that library (which you should use if you don't like mine) I find several things missing. Some of these are that every bugzilla-object is of the same class and that only a subset of API-calls is supported. My answer to the big question "work with the existing code or rewrite from scratch" was "rewrite", so this project was created.

How to use

A connection to bugzilla is provided through the Bugzilla-object.

import bugzilla
b = bugzilla.Bugzilla(bugzilla_url, api_key)

The url is the "urlbase" of your bugzilla-installation, e.g. "https://bugzilla.mozilla.org/". The "/rest"-part is added automatically. If you want to use an api_key you can specify it as second argument.
Other methods of login will not be supported as they are deprecated.

a = bugzilla.Attachment()
a["data"] = b"Hello World"
a.content_type = "text/plain"
a.update({
    "summary": "A summary",
    "file_name": "hello.txt"
})
b.add_attachment(a, bug_id, comment = "no comment")

Here an attachment is created and all three ways to set fields are shown. All bugzilla-objects derive from dict, so they can be treated as such. The second way is simply setting the attribute. Internally a.x = value will be converted to a["x"] = value. To update several fields at once, the dict.update-method can be used.
In the end the attachment is added to a bug. Some API-calls accept more parameters than what belong to the object. Therefore the comment-field is passed to the method.

Are there bugs?

Yes. I just don't see them because they are out of my line of sight. I tested the methods (roughly) in python 3.4 with an bugzilla-5.0-installation, I expect multiple bugs once this code will be used. If you find bugs send them to me for extermination or kill them yourselves.

TODO's and done's

I plan to add more features to this code, when this will happen is unknown. These features include:

• A good documentation. While the link to the documentation is given for every method every method should get a complete documentation of all possible parameters.
• Lazy-fetching objects. Some API-calls return a object with only a subsets of it attributes (e.g. only an users email). An example would be an user-object which only knows its email. If any other attribute is accessed the object loads its attributes.
• Find out why TODO is written in aggressive uppercase while done is written in frail lowercase

Done's:

• Supporting all (not deprecated) API-methods