另一种替代方法是使用 halaxa/json -机器.

Another alternative is to use halaxa/json-machine.

通过JSON进行迭代的用法与json_decode的用法相同，但是无论文件有多大，它都不会达到内存限制.无需执行任何操作，只需执行foreach.

Usage in case of iteration over JSON is the same as in case of json_decode, but it will not hit memory limit no matter how big your file is. No need to implement anything, just your foreach.

示例:

$users = \JsonMachine\JsonMachine::fromFile('500MB-users.json');

foreach ($users as $id => $user) {
    // process $user as usual
}

有关更多详细信息，请参见github自述文件.

See github readme for more details.