code-contributor-README.rst 8.36 KB
Newer Older
1 2
.. code-contributor-README:

3
###########################
4 5 6
Core Contributor Guidelines
###########################

7 8 9 10 11 12 13 14
License
=======

All contributed code must be compatible with the `Apache 2`_ license,
preferably by being contributed under the Apache 2 license. Code
contributed with another license will need the license reviewed by
Intel before it can be accepted.

15 16 17
Code formatting
================

18 19 20
All C/C++ source code in the repository, including the test code, must
adhere to the source-code formatting and style guidelines described
here.  The coding style described here applies to the nGraph
L.S. Cook's avatar
L.S. Cook committed
21
repository. Related repositories may make adjustments to better match
22 23
the coding styles of libraries they are using.

24

25 26
Adding ops to nGraph Core
-------------------------
27

28 29 30 31
Our design philosophy is that the graph is not a script for running
optimized kernels; rather, the graph is a specification for a
computation composed of basic building blocks which we call
``ops``. Compilation should match groups of ``ops`` to appropriate
L.S. Cook's avatar
L.S. Cook committed
32
optimal semantically equivalent groups of kernels for the backend(s)
33 34
in use. Thus, we expect that adding of new Core ops should be
infrequent and that most functionality instead gets added with new
35
functions that build sub-graphs from existing core ops. 
36 37


38
Coding style
39 40
-------------

41 42 43 44 45
We have a coding standard to help us to get development done. If part of the
standard is impeding progress, we either adjust that part or remove it. To this
end, we employ coding standards that facilitate understanding of *what nGraph
components are doing*. Programs are easiest to understand when they can be
understood locally; if most local changes have local impact, you do not need to
46 47
dig through multiple files to understand what something does and if it
is safe to modify.
48 49 50 51

Names
~~~~~

52 53
Names should *briefly* describe the thing being named and follow these casing
standards:
54 55 56

- Define C++ class or type names with ``CamelCase``.
- Assign template parameters with ``UPPER_SNAKE_CASE``.
57
- Case variable and function names with ``lower_snake_case``.
58

59
Method names for basic accessors are prefixed by ``get_``, ``is_``, or ``set_`` and
60
should have simple :math:`\mathcal{O}(1)` implementations:
61

62
- A ``get_`` method should be externally idempotent. It may perform some simple
63 64 65
  initialization and cache the result for later use.  Trivial ``get_``
  methods can be defined in a header file. If a method is
  non-trivial, that is often a sign that it is not a basic accessor.
66

67
- An ``is_`` may be used instead of ``get_`` for boolean accessors.
68

69
- A ``set_`` method should change the value returned by the corresponding ``get_``
70
  method.
71

72 73 74 75
  * Use ``set_is_`` if using ``is_`` to get a value.
  * Trivial ``set_`` methods may be defined in a header file.

- Names of variables should indicate the use of the variable.
76

77 78 79 80 81
  * Member variables should be prefixed with ``m_``.
  * Static member variables should be rare and be prefixed with ``s_``.

- Do not use ``using`` to define a type alias at top-level in header file.
  If the abstraction is useful, give it a class.
82

83 84 85 86 87 88 89 90 91 92
  * C++ does not enforce the abstraction. For example if ``X`` and ``Y`` are
    aliases for the same type, you can pass an ``X`` to something expecting a ``Y``.
  * If one of the aliases were later changed, or turned into a real type, many
    callers could require changes.


Namespaces
~~~~~~~~~~

- ``ngraph`` is for the public API, although this is not currently enforced.
93

94
  * Use a nested namespace for implementation classes.
95 96
  * Use an unnamed namespace or ``static`` for file-local names. This helps
    prevent unintended name collisions during linking and when using shared
97 98
    and dynamically-loaded libraries.
  * Never use ``using`` at top-level in a header file.
99

100 101
    - Doing so leaks the alias into users of the header, including headers that
      follow.
102
    - It is okay to use ``using`` with local scope, such as inside a class
103
      definiton.
104 105 106
  * Be careful of C++'s implicit namespace inclusions. For example, if a
    parameter's type is from another namespace, that namespace can be visible
    in the body.
107 108 109 110 111 112 113
  * Only use ``using std`` and/or ``using ngraph`` in ``.cpp`` files. ``using`` a
    nested namespace has can result in unexpected behavior.


File Names
~~~~~~~~~~

114
- Do not use the same file name in multiple directories. At least one
115 116 117 118 119 120 121
  IDE/debugger ignores the directory name when setting breakpoints.

- Use ``.hpp`` for headers and ``.cpp`` for implementation.

- Reflect the namespace nesting in the directory hierarchy.

- Unit test files are in the ``tests`` directory.
122

123
  * Transformer-dependent tests are tests running on the default transformer or
124 125 126 127
    specifying a transformer. For these, use the form

    .. code-block:: cpp

L.S. Cook's avatar
L.S. Cook committed
128
       TEST(file_name, test_name)
129 130

  * Transformer-independent tests:
131

132
    - File name is ``file_name.in.cpp``
133 134
    - Add ``#include "test_control.hpp"`` to the file's includes
    - Add the line ``static std::string s_manifest = "${MANIFEST}";`` to the top of the file.
135 136 137 138
    - Use

      .. code-block:: sh

139
         NGRAPH_TEST(${BACKEND_NAME}, test_name)
140

141 142
      for each test. Files are
      generated for each transformer and the ``${BACKEND_NAME}`` is replaced
143 144
      with the transformer name.

145 146 147 148
      Individual unit tests may be disabled by adding the name of the test to the
      ``unit_test.manifest`` file found in
      the transformer's source file directory.

149 150 151 152

Formatting
~~~~~~~~~~

153 154
Things that look different should look different because they are different. We
use **clang format** to enforce certain formatting. Although not always ideal,
155 156
it is automatically enforced and reduces merge conflicts.

157
- The :file:`.clang-format` file located in the root of the project specifies
158
  our format.
159

160 161
  * The script :file:`maint/apply-code-format.sh` enforces that formatting
    at the C/C++ syntactic level.
162 163 164
  * The script at :file:`maint/check-code-format.sh` verifies that the formatting
    rules are met by all C/C++ code (again, at the syntax level). The script has
    an exit  code of ``0`` when code meets the standard and non-zero otherwise.
165 166 167
    This script does *not* modify the source code.

- Formatting with ``#include`` files:
168

169 170 171
  * Put headers in groups separated by a blank line. Logically order the groups
    downward from system-level to 3rd-party to ``ngraph``.
  * Formatting will keep the files in each group in alphabetic order.
172
  * Use this syntax for files that **do not change during nGraph development**; they
173
    will not be checked for changes during builds. Normally this will be
174 175 176 177 178
    everything but the ngraph files:

    .. code-block:: cpp

       #include <file>
179

180
  * Use this syntax for files that **are changing during nGraph development**; they will
L.S. Cook's avatar
L.S. Cook committed
181
    be checked for changes during builds. Normally this will be ngraph headers:
182 183 184 185 186 187 188 189 190 191 192

    .. code-block:: cpp

       #include "file"

  * Use this syntax for system C headers with C++ wrappers:

    .. code-block:: cpp

       #include <c...>

193
- To guard against multiple inclusion, use:
194 195 196 197 198

  .. code-block:: cpp

     #pragma once

199 200 201
  * The syntax is a compiler extension that has been adopted by all
    supported compilers.

202 203 204 205 206 207
- The initialization

  .. code-block:: cpp

     Foo x{4, 5};

L.S. Cook's avatar
L.S. Cook committed
208
  is preferred over
209 210 211 212 213

  .. code-block:: cpp

     Foo x(4, 5);

214
- Indentation should be accompanied by braces; this includes single-line bodies
215 216 217
  for conditionals and loops.

- Exception checking:
218

219
  * Throw an exception to report a problem.
220
  * Nothing that calls ``abort``, ``exit`` or ``terminate`` should be used. Remember
221 222
    that ngraph is a guest of the framework.
  * Do not use exclamation points in messages!
223 224
  * Be as specific as practical. Keep in mind that the person who sees the error
    is likely to be on the other side of the framework and the message might be
225 226
    the only information they see about the problem.

227 228
- If you use ``auto``, know what you are doing. ``auto`` uses the same
  type-stripping rules as template parameters. If something returns a reference,
229
  ``auto`` will strip the reference unless you use ``auto&``:
230

231 232 233 234 235 236
  * Don't do things like

    .. code-block:: cpp

       auto s = Shape{2,3};

L.S. Cook's avatar
L.S. Cook committed
237
    Instead, use
238

L.S. Cook's avatar
L.S. Cook committed
239
    .. code-block:: cpp
240

L.S. Cook's avatar
L.S. Cook committed
241
       Shape s{2, 3};
242

243
  * Indicate the type in the variable name.
244

245
- One variable declaration/definition per line
246

247
  - Don't use the C-style
248

249
    .. code-block:: cpp
250

251
       int x, y, *z;
252

253
    Instead, use:
254

255
    .. code-block:: cpp
256

257 258 259
       int x;
       int y;
       int* z;
260

261

262 263
.. _Apache 2: https://www.apache.org/licenses/LICENSE-2.0
.. _repo wiki: