Sections:
- Licensing
- Copyright
- Documentation
- Naming
- Abbreviations
- Multiple names - which to use
- Formatting
- Reserved words
- Language use
- References
- Structure
- New variables
- New features/functions
Details:
- All new files should be based on the gpl.template.* files and have documentation about their intended use
- All files should have the GPL header inside them
- When you work on a file that exists, add your name to the copyright (unless the edit is extremely minor http://www.gnu.org/prep/maintain/html_node/Legally-Significant.html)
- If the the copyright year is not the current year, extend the range
- eg Copyright 2008-2012 Fred Cooke
- Dates in the code should be DD/MM/YY for easy reading
- Dates in file names should be YYYY-MM-DD for easy sorting
- Use Doxygen style comments
- Document all parameters
- Document return object details
- use @ instead of \
- to start a function comment block use ** instead of *!
- when declaring a pointer make it (type* name) not (type *name)
- Never use int, always char, short or long
- Always use full types, eg unsigned char not char and signed short instead of short
- use // style comments for temporary work/documentation and block comments for permanent stuff.
- order for doc tags : author, param(s), return
- tabs should be used for indentation
- spaces should be used to space out from variable width text before more tabs
- generally, two blank lines between functions, one blank line between logical blocks
- default tab size is 4 spaces, currently best viewed with that, but working on making it clean for any.
- Include order should have the most general thing first and most specific last :
#include <string.h>
#include "inc/thisFilesHeader.h"
- braces should be positioned as follows with the opening brace on the same line as the function name and no leading space:
- case statements should always use braces :
switch(var){
case 666:
{
}
}
- parameter lists should look like this with a space after each comma and no leading or trailing spaces
void function(unsigned char one, signed short another){
}
function(255, -32768);
Discussion:
We need to figure out a consistent way of naming variables and functions and files etc. The main thing that needs deciding is which scheme from the following:
- UPPER_CASE
- lower_case
- lowerCamelCase
- UpperCamelCase
- Other ways?
The categories of names required to be discussed are as follows:
- hash defines
- register aliases
- register masks
- custom masks
- error codes
- payload IDs
- constant values (would be literals)
- what else?
- structs (and unions)
- struct typedefs
- struct members
- struct instances
- local variables
- global variables
- files
- header see source :
- source - RPM ISRs and data initialisers UpperCamelCase - the rest lowerCamelCase
- doxygen
- standalone documentation
- JSON files and schemas
- linker scripts
- makefiles
- readme/readme.txt/README/README.txt
Please see the following Mantis issue for the original mandate to produce this document: