What's new in PHP 7? - Coolblue Behind The Scenes 2016

PHP 7 talk I presented at Coolblue Behind The Scenes.

Go to https://talks.feryn.eu/talks/161/what-s-new-in-php-7-behind-the-scenes-rotterdam to see more details.

Transcript

1. By Thijs Feryn
   What's new in
   

   View Slide

2. Hi, I’m Thijs
   

   View Slide

3. I’m
   @ThijsFeryn
   on Twitter
   

   View Slide

4. I’m an
   Evangelist
   At
   

   View Slide

5. I’m a
   at
   board member
   

   View Slide

6. To know where PHP is heading, you have
   to know what happened in the past
   

   View Slide

7. PHP 7.0: December 3rd 2015
   PHP 5.6: 2014
   PHP 5.5: 2013
   PHP 5.4: 2012
   PHP 5.3: 2009
   PHP 5.2: 2006
   PHP 5.1: 2005
   PHP 5.0: 2004
   

   View Slide

8. We've been
   using PHP 5 for
   11 years
   

   View Slide

9. Why is there no
   PHP 6?
   

   View Slide

10. Windows 10?
    

    View Slide

11. PHP 6 did once exist
    

    View Slide

12. But was never
    released
    

    View Slide

13. A history
    lesson
    

    View Slide

14. First ideas in 2005
    by Andrei Zmievski
    at Yahoo
    

    View Slide

15. PHP 6 was all about
    unicode support
    

    View Slide

16. And some extra
    language features
    

    View Slide

17. ✓Namespaces
    ✓Late static binding
    ✓Goto labels
    ✓Object casting to primitives
    ✓Type hinting
    ✓APC
    ✓Non-fatal errors should become exceptions
    PHP 6 extra language features
    

    View Slide

18. Delays, delays, delays …
    

    View Slide

19. Most
    non-Unicode features
    were released in PHP 5.3
    

    View Slide

20. The rest was eventually
    shipped off to PHP 5.4
    

    View Slide

21. View Slide

22. ✓UTF-16 or UTF-8? Or even UTF-32?
    ✓Few people know intricacies of Unicode & ICU
    ✓Major performance penalty (300%)
    ✓Too much effort to port all code
    PHP 6 unicode difficulties
    

    View Slide

23. Controversy
    

    View Slide

24. In 2010 the core team
    concluded that
    Unicode support for
    PHP was too hard to
    implement
    

    View Slide

25. Ah, Jani went a little crazy today in his typical style to force a
    decision. The real decision is not whether to have a version 5.4 or
    not, it is all about solving the Unicode problem. The current effort
    has obviously stalled. We need to figure out how to get development
    back on track in a way that people can get on board. We knew the
    Unicode effort was hugely ambitious the way we approached it. There are
    other ways.
    So I think Lukas and others are right, let's move the PHP 6 trunk to a
    branch since we are still going to need a bunch of code from it and move
    development to trunk and start exploring lighter and more approachable
    ways to attack Unicode. We have a few already. Enhanced mbstring and
    ext/intl. Let's see some good ideas around that and work on those in
    trunk. Other features necessarily need to play along with these in the
    same branch. I refuse to go down the path of a 5.4 branch and a
    separate Unicode branch again.
    The main focus here needs to be to get everyone working in the same branch.
    -Rasmus
    March 11 2010
    

    View Slide

26. March 11 2010
    PHP6
    

    View Slide

27. PHP 6 trunk
    PHP 6 branch PHP 5.4 trunk
    

    View Slide

28. The PHP 5.4 era sparked
    the PHP Renaissance
    

    View Slide

29. How did we end up with PHP 7?
    

    View Slide

30. Controversy
    

    View Slide

31. PHP Next Generation
    

    View Slide

32. To JIT or not to JIT?
    

    View Slide

33. PHPNG vs HHVM
    

    View Slide

34. Tackling
    misinformation
    

    View Slide

35. When we aren't looking for pictures of kittens on the internet, internals developers are
    nearly always looking for ways to improve PHP, a few developers have a focus on performance.
    Over the last year, some research into the possibility of introducing JIT compilation
    capabilities to PHP has been conducted.
    During this research, the realization was made that in order to achieve optimal performance
    from PHP, some internal API's should be changed.
    This necessitated the birth of the phpng branch, initially authored by Dmitry Stogov, Xinchen
    Hui, and Nikita Popov. This branch does not include JIT capabilities, but rather seeks to
    solve those problems that prohibit the current, and any future implementation of a JIT
    capable executor achieving optimal performance by improving memory usage and cleaning up some
    core API's.
    By making these improvements, the phpng branch gives us a considerable performance gain in
    real world applications, for example a 20% increase in throughput for Wordpress. The door may
    well now be open for a JIT capable compiler that can perform as we expect, but it's necessary
    to say that these changes stand strong on their own, without requiring a JIT capable compiler
    in the future to validate them.
    The name "Next Generation" was optimistically presumptuous; in reality phpng is an internal
    project that we are working on, it is not a production ready branch that anyone should
    deploy, or judge as they would a release of PHP.
    The work on phpng, the doors it opens, the conversations it has started, the collaboration it
    is inspiring, are all worth getting excited about. But, we need to stay grounded, honest, and
    open; and say that there is much work to do in order to make the "Next Generation" a reality,
    this is only the start.
    May 27th 2014 by Joe Watkins
    

    View Slide

36. Backlash
    

    View Slide

37. Hi,
    Joe made a post about the introduction on phpng, what it is, and what it
    isn't.
    Some people (myself included) didn't liked that post for various reasons
    (some says it is opiniated, some doesn't like the tone and the wording,
    others feel that it is too early to made official announcement about phpng)..
    There were a couple of iteration on improving the text, but it is still not
    up to our standards imo:
    http://git.php.net/?p=web/php.git;a=history;f=archive/entries/2014-05-27-1.xml
    ;
    It is already on hackernews and reddit, so while some people suggested, I
    think it would be a bad move to just remove it.
    Would like to know what do you guys think about the best step, I see the
    following possible options:
    - keep it as is
    - remove it
    - rewrite it to be more formal and factual(only talk about what it is
    atm. not what it could be in the future).
    - create a post explaining that this post is controversional among the
    core-devs, so it is reflects more of the authors opinion than the projects
    official view on the topic.
    I'm mostly interested on the opinion of the core devs, but others also
    welcome to reply.
    --
    Ferenc Kovács
    @Tyr43l - http://tyrael.hu
    May 28th 2014
    

    View Slide

38. Hackernews
    Reddit
    Blogs
    Social media
    

    View Slide

39. Sparked progress on the
    next major version of PHP
    

    View Slide

40. Controversy
    

    View Slide

41. PHP 6 or PHP 7?
    Or even PHP 5.7?
    

    View Slide

42. Decision was made: PHP 7
    

    View Slide

43. PHP 7 is all about
    performance
    

    View Slide

44. Healthy competition
    with HHVM
    

    View Slide

45. All this
    controversy
    

    View Slide

46. Old boys club
    vs
    Young guns
    

    View Slide

47. Balance between PHP as
    a easy-to-use scripting
    language and a mature
    future-proof
    programming language
    

    View Slide

48. New language
    features
    

    View Slide

49. ✓Scalar type hints
    ✓Return type declarations
    ✓Anonymous classes
    ✓Closure::call()
    ✓Generator delegation
    ✓Generator return expressions
    ✓Null coalesce operator
    ✓Space ship operator
    ✓Throwables
    ✓Uniform variable syntax
    ✓…
    PHP 7 new language features
    

    View Slide

50. Controversy
    

    View Slide

51. Loosely typed
    vs
    Strict types
    

    View Slide

52. Compromise
    

    View Slide

53. declare(strict_types=1);
    

    View Slide

54. /**
    * Scalar type declarations
    */
    //declare(strict_types=1);
    function add(int $a, int $b) {
    return $a + $b;
    }
    var_dump(add(1,2));
    var_dump(add("1","2"));
    

    View Slide

55. Fatal error: Uncaught TypeError: Argument 1 passed to
    add() must be of the type integer, string given
    TypeError: Argument 1 passed to add() must be of the type
    integer, string given
    

    View Slide

56. /**
    * Return type declarations
    */
    //declare(strict_types=1);
    function add(int $a, int $b): int{
    return (string)($a + $b);
    }
    var_dump(add(1,2));
    

    View Slide

57. Fatal error: Uncaught TypeError: Return value of add() must
    be of the type integer, string returned
    TypeError: Return value of add() must be of the type
    integer, string returned
    

    View Slide

58. Type Errors
    

    View Slide

59. Throwables
    Errors Exceptions
    Type error
    

    View Slide

60. //Type error
    try {
    function add(int $a, int $b):int {
    return $a + $b;
    }
    echo add(array(), array());
    } catch (TypeError $t) {
    echo "Type error: ".$t->getMessage().PHP_EOL;
    }
    

    View Slide

61. Throwables
    

    View Slide

62. /**
    * Throwable interface
    */
    //Error as Throwable
    try {
    sqdf();
    } catch (Throwable $t) {
    echo "Throwable: ".$t->getMessage().PHP_EOL;
    }
    //Exception as Throwable
    try {
    throw new Exception("Bla");
    } catch (Throwable $t) {
    echo "Throwable: ".$t->getMessage().PHP_EOL;
    }
    

    View Slide

63. /**
    * Throwable interface
    */
    //Exception
    try {
    throw new Exception("Bla");
    } catch (Error $e) {
    echo "Error: ".$e->getMessage().PHP_EOL;
    } catch (Exception $e) {
    echo "Exception: ".$e->getMessage().PHP_EOL;
    }
    //Type error
    try {
    function add(int $a, int $b):int {
    return $a + $b;
    }
    echo add(array(), array());
    } catch (TypeError $t) {
    echo "Type error: ".$t->getMessage().PHP_EOL;
    }
    

    View Slide

64. Anonymous classes
    

    View Slide

65. /**
    * Anonymous classes
    */
    $foo = new class {
    public function foo() {
    return "bar";
    }
    };
    var_dump($foo,$foo->foo());
    

    View Slide

66. Generators
    Since
    PHP 5.5
    Iterator
    shorthand
    Less
    overhead
    

    View Slide

67. Yield
    

    View Slide

68. function gen_one_to_three() {
    for ($i = 1; $i <= 3; $i++) {
    yield $i;
    }
    }
    $generator = gen_one_to_three();
    foreach ($generator as $value) {
    echo "$value\n";
    }
    

    View Slide

69. /**
    * Generator delegation
    */
    function gen()
    {
    yield 1;
    yield 2;
    yield from gen2();
    }
    function gen2()
    {
    yield 3;
    yield 4;
    }
    foreach (gen() as $val)
    {
    echo $val, PHP_EOL;
    }
    

    View Slide

70. /**
    * Generator return expressions
    */
    $gen = (function() {
    yield 1;
    yield 2;
    return 3;
    })();
    foreach ($gen as $val) {
    echo $val, PHP_EOL;
    }
    echo $gen->getReturn(), PHP_EOL;
    

    View Slide

71. Space ship operator
    

    View Slide

72. /**
    * Space ship operator
    */
    $array = [
    "1 <=> 1" => 1 <=> 1,
    "1 <=> 2" =>1 <=> 2,
    "2 <=> 1" => 2 <=> 1
    ];
    var_dump($array);
    array(3) {
    '1 <=> 1' =>
    int(0)
    '1 <=> 2' =>
    int(-1)
    '2 <=> 1' =>
    int(1)
    }
    Output
    

    View Slide

73. Null coalesce
    operator
    

    View Slide

74. /**
    * Null coalesce operator
    */
    $array = ['foo'=>'bar'];
    //PHP5 style
    $message = isset($array['foo']) ? $array['foo'] : 'not set';
    echo $message.PHP_EOL;
    //PHP7 style
    $message = $array['foo'] ?? 'not set';
    echo $message.PHP_EOL;
    

    View Slide

75. Closure::call()
    

    View Slide

76. /**
    * Closure::call()
    */
    class Foo
    {
    private $foo = 'bar';
    }
    $getFooCallback = function() {
    return $this->foo;
    };
    //PHP5 style
    $binding = $getFooCallback->bindTo(new Foo,'Foo');
    echo $binding().PHP_EOL;
    //PHP7 style
    echo $getFooCallback->call(new Foo).PHP_EOL;
    

    View Slide

77. Uniform variable
    syntax
    

    View Slide

78. Uniform variable syntax
    

    View Slide

79. PHP 7 is fast
    

    View Slide

80. View Slide

81. Speed depends on
    the your code and
    your case
    

    View Slide

82. Up to 300% speed
    increase with lower
    memory usage
    

    View Slide

83. Don't believe the
    banchmarks though
    

    View Slide

84. Why is PHP 7 so fast,
    even without the JIT?
    

    View Slide

85. ZVAL
    Internal
    variable
    representation
    Loosely
    typed variables in
    a strictly typed
    language
    

    View Slide

86. Typedef struct _zval_struct {
    union {
    long lval;
    double dval;
    struct {
    char *val;
    int len;
    } str;
    HashTable *ht;
    struct {
    zend_object_handle handle;
    zend_object_handlers *handlers;
    } obj;
    } value;
    zend_uint refcount;
    zend_uchar type;
    zend_uchar is_ref;
    } zval;
    Typedef struct _zval_struct {
    union {
    long lval;
    double dval;
    zend_refcounted *counted;
    zend_string *str;
    zend_array *arr;
    zend_object *obj;
    zend_resource *res;
    zend_reference *ref;
    void *ptr;
    } value;
    union {
    struct {
    zend_uchar type;
    zend_uchar flags;
    };
    zend_uint type_info;
    };
    zend_uint reserved;
    } zval;
    PHP 5 vs PHP 7
    

    View Slide

87. Remove pointers
    (to pointers)
    to ZVALs
    

    View Slide

88. Direct use of ZVAL
    in most cases
    

    View Slide

89. From heap to stack
    

    View Slide

90. Less refcounting
    Less garbage collection
    

    View Slide

91. New memory
    manager
    Less CPU
    time
    Less
    memory
    overhead
    

    View Slide

92. A lot of small
    optimizations
    

    View Slide

93. View Slide

94. https://blog.feryn.eu
    https://talks.feryn.eu
    https://youtube.com/thijsferyn
    https://soundcloud.com/thijsferyn
    https://twitter.com/thijsferyn
    http://itunes.feryn.eu
    

    View Slide