Matthew Treinish | 8b37289 | 2012-12-07 17:13:16 -0500 | [diff] [blame] | 1 | Test Data/Configuration |
| 2 | ----------------------- |
| 3 | - Assume nothing about existing test data |
| 4 | - Tests should be self contained (provide their own data) |
| 5 | - Clean up test data at the completion of each test |
| 6 | - Use configuration files for values that will vary by environment |
| 7 | |
| 8 | |
| 9 | General |
| 10 | ------- |
| 11 | - Put two newlines between top-level code (funcs, classes, etc) |
| 12 | - Put one newline between methods in classes and anywhere else |
| 13 | - Long lines should be wrapped in parentheses |
| 14 | in preference to using a backslash for line continuation. |
| 15 | - Do not write "except:", use "except Exception:" at the very least |
| 16 | - Include your name with TODOs as in "#TODO(termie)" |
| 17 | - Do not name anything the same name as a built-in or reserved word Example:: |
| 18 | |
| 19 | def list(): |
| 20 | return [1, 2, 3] |
| 21 | |
| 22 | mylist = list() # BAD, shadows `list` built-in |
| 23 | |
| 24 | class Foo(object): |
| 25 | def list(self): |
| 26 | return [1, 2, 3] |
| 27 | |
| 28 | mylist = Foo().list() # OKAY, does not shadow built-in |
| 29 | |
| 30 | Imports |
| 31 | ------- |
| 32 | - Do not import objects, only modules (*) |
| 33 | - Do not import more than one module per line (*) |
| 34 | - Do not make relative imports |
| 35 | - Order your imports by the full module path |
| 36 | - Organize your imports according to the following template |
| 37 | |
| 38 | Example:: |
| 39 | |
| 40 | # vim: tabstop=4 shiftwidth=4 softtabstop=4 |
| 41 | {{stdlib imports in human alphabetical order}} |
| 42 | \n |
| 43 | {{third-party lib imports in human alphabetical order}} |
| 44 | \n |
| 45 | {{tempest imports in human alphabetical order}} |
| 46 | \n |
| 47 | \n |
| 48 | {{begin your code}} |
| 49 | |
| 50 | |
| 51 | Human Alphabetical Order Examples |
| 52 | --------------------------------- |
| 53 | Example:: |
| 54 | |
| 55 | import httplib |
| 56 | import logging |
| 57 | import random |
| 58 | import StringIO |
| 59 | import time |
| 60 | import unittest |
| 61 | |
| 62 | import eventlet |
| 63 | import webob.exc |
| 64 | |
| 65 | import tempest.config |
| 66 | from tempest.services.compute.json.limits_client import LimitsClientJSON |
| 67 | from tempest.services.compute.xml.limits_client import LimitsClientXML |
| 68 | from tempest.services.volume.volumes_client import VolumesClientJSON |
| 69 | import tempest.test |
| 70 | |
| 71 | |
| 72 | Docstrings |
| 73 | ---------- |
| 74 | Example:: |
| 75 | |
| 76 | """A one line docstring looks like this and ends in a period.""" |
| 77 | |
| 78 | |
| 79 | """A multi line docstring has a one-line summary, less than 80 characters. |
| 80 | |
| 81 | Then a new paragraph after a newline that explains in more detail any |
| 82 | general information about the function, class or method. Example usages |
| 83 | are also great to have here if it is a complex class for function. |
| 84 | |
| 85 | When writing the docstring for a class, an extra line should be placed |
| 86 | after the closing quotations. For more in-depth explanations for these |
| 87 | decisions see http://www.python.org/dev/peps/pep-0257/ |
| 88 | |
| 89 | If you are going to describe parameters and return values, use Sphinx, the |
| 90 | appropriate syntax is as follows. |
| 91 | |
| 92 | :param foo: the foo parameter |
| 93 | :param bar: the bar parameter |
| 94 | :returns: return_type -- description of the return value |
| 95 | :returns: description of the return value |
| 96 | :raises: AttributeError, KeyError |
| 97 | """ |
| 98 | |
| 99 | |
| 100 | Dictionaries/Lists |
| 101 | ------------------ |
| 102 | If a dictionary (dict) or list object is longer than 80 characters, its items |
| 103 | should be split with newlines. Embedded iterables should have their items |
| 104 | indented. Additionally, the last item in the dictionary should have a trailing |
| 105 | comma. This increases readability and simplifies future diffs. |
| 106 | |
| 107 | Example:: |
| 108 | |
| 109 | my_dictionary = { |
| 110 | "image": { |
| 111 | "name": "Just a Snapshot", |
| 112 | "size": 2749573, |
| 113 | "properties": { |
| 114 | "user_id": 12, |
| 115 | "arch": "x86_64", |
| 116 | }, |
| 117 | "things": [ |
| 118 | "thing_one", |
| 119 | "thing_two", |
| 120 | ], |
| 121 | "status": "ACTIVE", |
| 122 | }, |
| 123 | } |
| 124 | |
| 125 | |
| 126 | Calling Methods |
| 127 | --------------- |
| 128 | Calls to methods 80 characters or longer should format each argument with |
| 129 | newlines. This is not a requirement, but a guideline:: |
| 130 | |
| 131 | unnecessarily_long_function_name('string one', |
| 132 | 'string two', |
| 133 | kwarg1=constants.ACTIVE, |
| 134 | kwarg2=['a', 'b', 'c']) |
| 135 | |
| 136 | |
| 137 | Rather than constructing parameters inline, it is better to break things up:: |
| 138 | |
| 139 | list_of_strings = [ |
| 140 | 'what_a_long_string', |
| 141 | 'not as long', |
| 142 | ] |
| 143 | |
| 144 | dict_of_numbers = { |
| 145 | 'one': 1, |
| 146 | 'two': 2, |
| 147 | 'twenty four': 24, |
| 148 | } |
| 149 | |
| 150 | object_one.call_a_method('string three', |
| 151 | 'string four', |
| 152 | kwarg1=list_of_strings, |
| 153 | kwarg2=dict_of_numbers) |
| 154 | |
| 155 | |
| 156 | OpenStack Trademark |
| 157 | ------------------- |
| 158 | |
| 159 | OpenStack is a registered trademark of OpenStack, LLC, and uses the |
| 160 | following capitalization: |
| 161 | |
| 162 | OpenStack |
| 163 | |
| 164 | |
| 165 | Commit Messages |
| 166 | --------------- |
| 167 | Using a common format for commit messages will help keep our git history |
| 168 | readable. Follow these guidelines: |
| 169 | |
| 170 | First, provide a brief summary (it is recommended to keep the commit title |
| 171 | under 50 chars). |
| 172 | |
| 173 | The first line of the commit message should provide an accurate |
| 174 | description of the change, not just a reference to a bug or |
| 175 | blueprint. It must be followed by a single blank line. |
| 176 | |
| 177 | If the change relates to a specific driver (libvirt, xenapi, qpid, etc...), |
| 178 | begin the first line of the commit message with the driver name, lowercased, |
| 179 | followed by a colon. |
| 180 | |
| 181 | Following your brief summary, provide a more detailed description of |
| 182 | the patch, manually wrapping the text at 72 characters. This |
| 183 | description should provide enough detail that one does not have to |
| 184 | refer to external resources to determine its high-level functionality. |
| 185 | |
| 186 | Once you use 'git review', two lines will be appended to the commit |
| 187 | message: a blank line followed by a 'Change-Id'. This is important |
| 188 | to correlate this commit with a specific review in Gerrit, and it |
| 189 | should not be modified. |
| 190 | |
| 191 | For further information on constructing high quality commit messages, |
| 192 | and how to split up commits into a series of changes, consult the |
| 193 | project wiki: |
| 194 | |
| 195 | http://wiki.openstack.org/GitCommitMessages |