You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This leads to an error in JSLint and a warning in JSHint, "Expected an assignment or function call and instead saw an expression”. As JSLint explains its expr option:
This option suppresses warnings about the use of expressions where normally you would expect to see assignments or function calls. Most of the time, such code is a typo. However, it is not forbidden by the spec and that's why this warning is optional.
The warning is useful and worth enabling as it does catch typos like JSLint explains. The line you wrote, callback && callback(); is an expression and not a function call. This may be contrived, but take code similar to your example:
JSLint will give the same warning for the code, and this time it caught a typo: callback is not actually being called, it's missing parentheses. This code would raise the same warning for the same reason:
The second example would likely be more quickly noticed by the developer, but again JSLint/JSHint is able to point it out. I find using an explicit if more readable for the next developer too. The intention of the conditional in that case is unmistakable whereas callback && callback(); or something like it is not as straightforward.
There is nothing "wrong" with writing callback && callback(); in the sense that it will run just fine. However, I am a fan of keeping it out of the style guide because endorsing a pattern that looks the same as a typo to tools can lead to frustrating debugging and code that is harder to read for the next developer.
I always prefer to use if statements while developing. When the project is built, uglifyjs will short-circuit the if statements to use the construct like yours.
I don't think either one adequately covers safely attempting to run your callback
if(typeof callback === 'function'){
callback();
}
is what I use to be safe. I don't think a simple null check is a good idea especially when the arguments are variable or when you intend to use arguments[arguments.length - 1] convention:
function myFn(firstName, lastName){
var name = firstName + " " + lastName;
// Is the last argument a callback function?
var callback = arguments[arguments.length - 1];
// If it is, use it, otherwise, technically callback is just a string equivalent to lastName.
if(typeof callback === 'function'){
callback(name);
}
return name;
}
myFn("Keith", "Rosenberg")
// "Keith Rosenberg"
myFn("Keith", "Rosenberg", function(name){ console.log("Callback: " + name); });
// Callback: Keith Rosenberg
Otherwise:
function myFn(firstName, lastName){
var name = firstName + " " + lastName;
// Is the last argument a callback function?
var callback = arguments[arguments.length - 1];
// If it is, use it, otherwise, technically callback is just a string equivalent to lastName.
if(callback){
callback(name);
}
return name;
}
myFn("Keith", "Rosenberg")
// TypeError: string is not a function
@yisibl thats equivalent to if (callback != null) { callback(); }, so that’d be fine (once eslint supports parsing that syntax, and once this guide permits it)
Activity
bishopZ commentedon Feb 21, 2013
+1
ssorallen commentedon Feb 21, 2013
This leads to an error in JSLint and a warning in JSHint, "Expected an assignment or function call and instead saw an expression”. As JSLint explains its
expr
option:The warning is useful and worth enabling as it does catch typos like JSLint explains. The line you wrote,
callback && callback();
is an expression and not a function call. This may be contrived, but take code similar to your example:JSLint will give the same warning for the code, and this time it caught a typo:
callback
is not actually being called, it's missing parentheses. This code would raise the same warning for the same reason:The second example would likely be more quickly noticed by the developer, but again JSLint/JSHint is able to point it out. I find using an explicit
if
more readable for the next developer too. The intention of the conditional in that case is unmistakable whereascallback && callback();
or something like it is not as straightforward.There is nothing "wrong" with writing
callback && callback();
in the sense that it will run just fine. However, I am a fan of keeping it out of the style guide because endorsing a pattern that looks the same as a typo to tools can lead to frustrating debugging and code that is harder to read for the next developer.timofurrer commentedon Feb 21, 2013
I think it is more readable to make an
if
instead of this shorter statement, although it is not wrong!Like @ssorallen pointed out!
medikoo commentedon Feb 22, 2013
I make such if's, a one liners:
Still lint will scream, but I take such approach as much more justified (and readable) than hacky
cb && cb()
.hshoff commentedon Feb 25, 2013
I added a link to this discussion so folks can come back and read through this and add to the discussion.
prashantpalikhe commentedon Mar 11, 2014
I always prefer to use if statements while developing. When the project is built, uglifyjs will short-circuit the if statements to use the construct like yours.
netpoetica commentedon Apr 6, 2014
I don't think either one adequately covers safely attempting to run your callback
is what I use to be safe. I don't think a simple null check is a good idea especially when the arguments are variable or when you intend to use arguments[arguments.length - 1] convention:
Otherwise:
收藏页
yisibl commentedon Nov 1, 2019
What about
callback?.()
?ljharb commentedon Nov 1, 2019
@yisibl thats equivalent to
if (callback != null) { callback(); }
, so that’d be fine (once eslint supports parsing that syntax, and once this guide permits it)[other styles] add link to conditional callback discussion. fixes air…