forked from OpenLightingProject/ola
-
Notifications
You must be signed in to change notification settings - Fork 1
/
README.developer
263 lines (185 loc) · 8.27 KB
/
README.developer
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
Developer Information
###############################################################################
This file provides some information for developers looking to submit patches to
OLA. The first section contains some general notes, and the latter
sections address particular languages.
Also take a look at README which includes some relevant information for
developers too.
General
===============================================================================
Code Reviews
------------
To have patches reviewed please push your changes to GitHub and create a Pull
Request from within your fork.
If you're using Google Code please enable comments on the repo. Under
Administer -> Source check:
Enable code reviews
Allow non-members to review code
Licensing
---------
Think carefully about the license for each file. Code to be used by clients
(./ola , ./python) should be under the LGPL, so that it may be used in
commercial projects.
Plugins and code used solely in olad should be under the GPL.
scripts/enforce_licence.py will look for a LICENCE file in each directory, and
ensure that all source files in the directory ( & subdirectories) start with
the LICENCE. You can pass --fix to automatically add the licence.
Unittests
---------
Unittests are *highly* encouraged. At the very least please make sure any
changes don't break the tests.
The tests are performed with `make check` in the root ola directory. The same
command can be run within a sub directory to only run a particular set of
tests (although you may experience issues with this method, running from the
root ola directory is guaranteed to work).
Branches, Versioning & Releases
-------------------------------
Version numbers take the form MAJOR.MINOR.PATCH based on http://semver.org/.
The release lifecycle is:
- New feature work occurs on the master branch.
- Once the new features are considered stable or enough time has passed, a new
minor release branch will be created, e.g. 0.10.
- The minor release branch will be stablized with bugfixes, these bug fixes
will also be merged back into master.
- Once declared stable, a new patch branch 0 will be created e.g. 0.10.0
- Release specific changes like the version number, debian files etc. will be
pushed to the patch branch.
- The release will occur.
- Changes on the release branch will be merged back into the minor version
branch.
- Bug fixes will continue on the minor version branch, leading to another patch
release.
What this means for you as a developer is that depending on the scope of your
change, we may request that you merge it into the minor release branch rather
than master. This allows us to get your change out to end uses faster than if
it was merged into master directly.
C++
===============================================================================
The bulk of OLA code is written in C++. This includes the daemon olad, and
plugins.
https://wiki.openlighting.org/index.php/OLA_developer_info describes many of
the underlying classes used.
C++ 11
------
As much as we'd like to start using C++11, Ubuntu 12.04 still comes with g++
4.6 which has partial support. Once 12.04 is end-of-lifed, we'll consider
using C++ 11 constructs.
Endian Issues
-------------
Be aware that OLA runs on big endian systems. When dealing with wire formats
always use the functions in include/ola/network/NetworkUtils.h to convert or
use the BigEndianOutputStream to manage this automatically for you.
Structure Packing
-----------------
Structure packing syntax differs between platforms and compilers. If you declare
a struct or class that needs to be packed, use the PACK() macro from
ola/base/Macro.h . See below for an example of PACK() usage.
Not that this macro doesn't work for enums. See plugins/espnet/EspNetPackets.h
for an example of how to pack an enum.
Non x86 platforms
-----------------
OLA also runs on more than just x86. Some platforms like ARM can't de-reference
pointers which aren't aligned.
For example:
PACK(struct f {
uint8_t value1;
uint16_t value2;
});
struct f foo = {1, 2};
uint16_t *ptr = &foo.value2;
// Bug! Will not be true on all platforms.
if (*ptr == 2)
http://www.aleph1.co.uk/chapter-10-arm-structured-alignment-faq has a good
explanation.
Style / Coding Standards
------------------------
We use the Google C++ Style Guide:
https://google.github.io/styleguide/cppguide.html
We provide formatting configuration files for some IDEs at
https://github.com/OpenLightingProject/editor-configs . If you use an IDE
which isn't listed there please consider submitting the formatting
configuration files.
Please run the cpplint.py script on all files before requesting a review,
alternatively it will be run automatically by Travis CI on pull requests.
cpplint.py can be downloaded here:
https://github.com/google/styleguide/tree/gh-pages/cpplint
Run it with:
--filter=-legal/copyright,-readability/streams,-runtime/arrays
Doxygen
-------
The code is documented using Doxygen. There is an automatically generated
version available from:
https://docs.openlighting.org/ola/doc/latest/
If you want to build it yourself, install Doxygen, run ./configure in the ola
root directory, then run make doxygen-doc. The files will be generated into
html/
Race Conditions
---------------
If possible, code should be tested with slower machines (embedded platforms,
virtual machines etc.). This has discovered race conditions in the past.
Valgrind
--------
All code must be tested with valgrind to ensure it doesn't leak memory.
Coverage (gcov)
---------------
To enable the coverage tools, you need lcov and gcov installed. To enable run
./configure --enable-gcov and then build as normal.
To generate the HTML pages run:
mkdir /tmp/lcov
tmpdir=`mktemp -d /tmp/lcov.XXXXXX`
coverage="${tmpdir}/coverage.info"
lcov --capture --directory ./ --output-file $coverage
genhtml $coverage --output-directory /tmp/lcov
Java
===============================================================================
An experimental Java API is provided.
Style / Coding Standards
------------------------
Please follow the Sun Java style guide.
Javascript
===============================================================================
Javascript is used for the olad web UI. Instructions for building the
javascript can be found in javascript/README.
Closure Compiler
----------------
The closure compiler catches many errors. The javascript code should build
cleanly.
Style / Coding Standards
------------------------
We use the Google Javascript style guide:
https://google.github.io/styleguide/javascriptguide.xml
A javascript linter can be downloaded from:
https://developers.google.com/closure/utilities/docs/linter_howto?csw=1
Please make sure all Javascript code is lint clean.
Python
===============================================================================
Python is used for tools like the RDM responder tests and the device model
collector. To enable these a OLA Python API is provided.
Style / Coding Standards
------------------------
We use the Google Python style guide:
https://google.github.io/styleguide/pyguide.html
However we deviate from the standard and use 2 space indents, so it's
consistent with the C++ code.
It's linted and static analysis is performed using flake8, you can either run
it yourself or await the Travis CI results of your pull request.
flake8 can be installed following the instructions here:
https://gitlab.com/pycqa/flake8/
Run it with:
--max-line-length 80 --exclude *_pb2.py,.git,__pycache \
--ignore E111,E121,E127,E129 data/rdm include/ola python scripts \
tools/ola_mon tools/rdm
Build System
===============================================================================
Autotools is a complex beast. Even after reading two books and a using it for a
decade I still feel like I hardly understand it.
Useful tips
-----------
* Run `make distcheck` to ensure you haven't broken anything.
* Use $(srcdir), $(top_srcdir) to reference files where required. This is to
support VPATH builds:
http://www.gnu.org/software/automake/manual/html_node/VPATH-Builds.html
* Prefer manualy dependencies over BUILT_SOURCES where possible, see
http://www.gnu.org/software/automake/manual/html_node/Built-Sources-Example.html
However this doesn't work in some cases (libraries?) because it overrides the
automake generated rules.