Make sure these boxes are checked before submitting/approving the PR
- The code works
- The code is easy to understand
- Follows coding conventions
- Names are simple and if possible short
- Names are spelt correctly
- Names contain units where applicable
- There are no usages of magic numbers
- No hard coded constants that could possibly change in the future
- All variables are in the smallest scope possible
- There is no commented out code
- There is no dead code (inaccessible at Runtime)
- No code that can be replaced with library functions
- Variables are not accidentally used with null values
- Variables are immutable where possible
- Code is not repeated or duplicated
- No complex/long boolean expressions
- No negatively named boolean variables
- No empty blocks of code
- Ideal data structures are used
- Constructors do not accept null/none values
- Catch clauses are fine grained and catch specific exceptions
- Exceptions are not eaten if caught, unless explicitly documented otherwise
- Files/Sockets and other resources are properly closed even when an exception occurs in using them
-
null
is not returned from any method - == operator and === (and its inverse !==) are not mixed up
- Floating point numbers are not compared for equality
- Loops have a set length and correct termination conditions
- Blocks of code inside loops are as small as possible
- No methods with boolean parameters
- No object exists longer than necessary
- No memory leaks
- Code is unit testable
- Test cases are written wherever possible
- Methods return early without compromising code readability
- Performance is considered
- Loop iteration and off by one are taken care of
- Design patterns if used are correctly applied
- Law of Demeter is not violated
- A class should have only a single responsibility (i.e. only one potential change in the software's specification should be able to affect the specification of the class)
- Classes, modules, functions, etc. should be open for extension, but closed for modification
- Objects in a program should be replaceable with instances of their subtypes without altering the correctness of that program
- Many client-specific interfaces are better than one general-purpose interface.
- Depend upon Abstractions. Do not depend upon concretions.
- APIs and other public contracts check input values and fail fast
- API checks for correct oauth scope / user permissions
- Any API change should be reflected in the API documentation
- APIs return correct status codes in responses
- Logging should be easily discoverable
- Required logs are present
- Frivolous logs are absent
- Debugging code is absent
- No
print_r
,var_dump
or similar calls exist - No stack traces are printed
- Comments should indicate WHY rather that WHAT the code is doing
- All methods are commented in clear language.
- Comments exist and describe rationale or reasons for decisions in code
- All public methods/interfaces/contracts are commented describing usage
- All edge cases are described in comments
- All unusual behaviour or edge case handling is commented
- Data structures and units of measurement are explained
- All data inputs are checked (for the correct type, length/size, format, and range)
- Invalid parameter values handled such that exceptions are not thrown
- No sensitive information is logged or visible in a stacktrace
yep