There also is the possibility of it not being an infinite loop at all. 10 iterations is not a sufficiently large number to conclude that with any amount of certainty. So before going on a wild-goose chase it may be advisable to rule out that possibility first.
The easiest method to do so is increasing the maximum digest loop count to a much larger number, which can be done in the module.config
method, using the $rootScopeProvider.digestTtl(limit)
method. If the infdig
error does no longer appear you simply have some sufficiently complex update logic.
If you build data or views relying on recursive watches you may want to search for iterative solutions (i.e. not relying on new digest loops to be started) using while
, for
or Array.forEach
. Sometimes the structure is just highly nested and not even recursive, there probably is not much to be done in those cases except raising the limit.
Another method of debugging the error is looking at the digest data. If you pretty print the JSON you get an array of arrays. Each top level entry represents an iteration, each iteration consists of a list of watch entries.
If you for example have a property which is modified in a $watch
on itself it is easy to see that the value is changing infinitely:
$scope.vm.value1 = true;
$scope.$watch("vm.value1", function(newValue)
{
$scope.vm.value1 = !newValue;
});
[
[
{
"msg":"vm.value1",
"newVal":true,
"oldVal":false
}
],
[
{
"msg":"vm.value1",
"newVal":false,
"oldVal":true
}
],
[
{
"msg":"vm.value1",
"newVal":true,
"oldVal":false
}
],
[
{
"msg":"vm.value1",
"newVal":false,
"oldVal":true
}
],
[
{
"msg":"vm.value1",
"newVal":true,
"oldVal":false
}
]
]
Of course in larger project this may not be as simple, especially since the msg
field often has the value "fn: regularInterceptedExpression"
if the watch is a {{ }}
interpolation.
Other than that the already mentioned methods, like cutting down the HTML to find the source of the problem, are of course helpful.