This document aims to be a reference for the desired style for patches to mgmt,
and the associated mcl language. In particular it describes conventions which
are not officially enforced by tools and in test cases, or that aren't clearly
defined elsewhere. We try to turn as many of these into automated tests as we
can. If something here is not defined in a test, or you think it should be,
please write one! Even better, you can write a tool to automatically fix it,
since this is more useful and can easily be turned into a test!
Most style issues are enforced by the gofmt tool. Other style aspects are
often common sense to seasoned programmers, and we hope this will be a useful
reference for new programmers.
There are a lot of useful code review comments described here. We don't necessarily follow everything strictly, but it is in general a very good guide.
- All of our golang code is formatted with
gofmt.
All of our code is commented with the minimums required for godoc to function,
and so that our comments pass golint. Code comments should either be full
sentences (which end with a period, use proper punctuation, and capitalize the
first word when it is not a lower cased identifier), or are short one-line
comments in the source which are not full sentences and don't end with a period.
They should explain algorithms, describe non-obvious behaviour, or situations which would otherwise need explanation or additional research during a code review. Notes about use of unfamiliar API's is a good idea for a code comment.
Here you can see a function with the correct godoc string. The first word must
match the name of the function. It is not capitalized because the function is
private.
// square multiplies the input integer by itself and returns this product.
func square(x int) int {
return x * x // we don't care about overflow errors
}In general we try to stick to 80 character lines when it is appropriate. It is
almost always appropriate for function godoc comments and most longer
paragraphs. Exceptions are always allowed based on the will of the maintainer.
It is usually better to exceed 80 characters than to break code unnecessarily. If your code often exceeds 80 characters, it might be an indication that it needs refactoring.
Occasionally inline, two line source code comments are used within a function. These should usually be balanced so that you don't have one line with 78 characters and the second with only four. Split the comment between the two.
Whenever a constant or function parameter is defined, try and have the safer or
default value be the zero value. For example, instead of const NoDanger, use
const AllowDanger so that the false value is the safe scenario.
You almost always want any method receivers to be declared on the pointer to the
struct. There are only a few rare situations where this is not the case. This
makes it easier to merge future changes that mutate the state without wondering
why you now have two different copies of a struct. When you do need to copy a
a struct, you can add a Copy() method to it. It's true that in many situations
adding the pointer adds a small performance penalty, but we haven't found them
to be significant in practice. If you do have a performance sensitive patch
which benefits from skipping the pointer, please demonstrate this need with
data first.
type Foo struct {
Whatever string
// ...
}
// Bar is implemented correctly as a pointer on Foo.
func (obj *Foo) Bar(baz string) int {
// ...
}
// Bar is implemented *incorrectly* without a pointer to Foo.
func (obj Foo) Bar(baz string) int {
// ...
}Contrary
to the specialized naming of the method receiver variable, we usually name all
of these obj for ease of code copying throughout the project, and for faster
identification when reviewing code. Some anecdotal studies have shown that it
makes the code easier to read since you don't need to remember the name of the
method receiver variable in each different method. This is very similar to what
is done in python.
// Bar does a thing, and returns the number of baz results found in our
database.
func (obj *Foo) Bar(baz string) int {
if len(obj.s) > 0 {
return strings.Count(obj.s, baz)
}
return -1
}We prefer shorter, scoped variables rather than unnecessarilyLongIdentifiers.
Remember the scoping rules and feel free to use new variables where appropriate.
For example, in a short string snippet you can use s instead of myString, as
well as other common choices. i is a common int counter, f for files, fn
for functions, x for something else and so on.
Feel free to create and use new variables instead of attempting to re-use the
same string. For example, if a function input arg is named s, you can use a
new variable to receive the first computation result on s instead of storing
it back into the original s. This avoids confusion if a different part of the
code wants to read the original input, and it avoids any chance of edit by
reference of the original callers copy of the variable.
MyNotIdealFunc(s string, b bool) string {
if !b {
return s + "hey"
}
s = strings.Replace(s, "blah", "", -1) // not ideal (re-use of `s` var)
return s
}
MyOkayFunc(s string, b bool) string {
if !b {
return s + "hey"
}
s2 := strings.Replace(s, "blah", "", -1) // doesn't re-use `s` variable
return s2
}
MyGreatFunc(s string, b bool) string {
if !b {
return s + "hey"
}
return strings.Replace(s, "blah", "", -1) // even cleaner
}If a function takes a specifier (often a bool) it's sometimes better to name
that variable (often with a const) rather than leaving a naked bool in the
code. For example, x := MyFoo("blah", false) is less clear than
const useMagic = false; x := MyFoo("blah", useMagic).
When declaring a new empty slice, there are three different mechanisms:
-
a := []string{} -
var a []string -
a := make([]string, 0)
In general, we prefer the first method because we find that it is succinct, and very readable. The third method is the least recommended because you're adding extra data that a smart compiler could probably figure out on its own. There are performance implications between these three methods, so unless your code is in a fast path or memory constrained environment where this matters (and that you ideally have proof of this) please use the methods as ordered as much as possible.
In general we try to preserve a logical ordering in source files which usually matches the common order of execution that a lazy evaluator would follow.
This is also the order which is recommended when creating interface types. When implementing an interface, arrange your methods in the same order that they are declared in the interface.
When implementing code for the various types in the language, please follow this
order: bool, str, int, float, list, map, struct, func.
For other aspects where you have a set of items, try to be internally consistent
as well. For example, if you have two switch statements with A, B, and C,
please use the same ordering for these elements elsewhere that they appear in
the code and in the commentary if it is not illogical to do so.
Try to avoid references in the code to mgmt or a specific program name string
if possible. This makes it easier to rename code if we ever pick a better name
or support libmgmt better if we embed it. You can use the Program variable
which is available in numerous places if you want a string to put in the logs.
It is also recommended to avoid the go (programming language name) string if
possible. Try to use golang if required, since the word go is already
overloaded, and in particular it was even already used by the
go!.
The mcl language is quite new, so this guide will probably change over time as
we find what's best, and hopefully we'll be able to add an mclfmt tool in the
future so that less of this needs to be documented. (Patches welcome!)
Code indentation is done with tabs. The tab-width is a private preference, which
is the beauty of using tabs: you can have your own personal preference. The
inventor of mgmt uses and recommends a width of eight, and that is what should
be used if your tool requires a modeline to be publicly committed.
We recommend you stick to 80 char line width. If you find yourself with deeper nesting, it might be a hint that your code could be refactored in a more pleasant way.
At the moment, variables, function names, and classes are all lowercase and do not contain underscores. We will probably figure out what style to recommend when the language is a bit further along. For example, we haven't decided if we should have a notion of public and private variables, and if we'd like to reserve capitalization for this situation.
We recommend you name your modules with an mgmt- prefix. For example, a module
about bananas might be named mgmt-banana. This is helpful for the useful magic
built-in to the module import code, which will by default take a remote import
like: import "https://github.com/purpleidea/mgmt-banana/" and namespace it as
banana. Of course you can always pick the namespace yourself on import with:
import "https://github.com/purpleidea/mgmt-banana/" as tomato or something
similar.
We believe that sharing code helps reduce unnecessary re-invention, so that we can stand on the shoulders of giants and hopefully make faster progress in science, medicine, exploration, etc... As a result, we recommend releasing your modules under the LGPLv3+ license for the maximum balance of freedom and re-usability. We strongly oppose any CLA requirements and believe that the "inbound==outbound" rule applies. Lastly, we do not support software patents and we hope you don't either!
If you have any ideas for suggestions or other improvements to this guide, please let us know!