Python for Ruby Programmers

A presentation comparing the Python and Ruby programming languages. I gave this talk at LA RubyConf 2013.

1. Python for Ruby Programmers MIKE LEONE LA RUBYCONF 2013 FEB 23 2013
2. This presentation: 1. Fast-paced 2. Python in 30 mins 3. It's easy!
3. About Me
4. Web Mobile Embedded
12. That's it.
13. Ruby: 2006 Python: 2010
14. New appreciation Better Rubyist
15. What we'll cover
16. 1. Example program 2. 13 Design similarities 3. 13 Differences 4. 5 Problem domains 5. “Feelings”
17. What we WON'T cover
18. 1. Implementations 2. Performance (Good enough) 3. Parallelism (GIL) 4. Minor syntax (You'll get it)
19. What I'll try to convince you:
20. 1. VERY similar! 2. You're already ready 3. Python = Better Rubyist 4. Hire a Pythonista
21. Zen of Python (abridged)
22. 1. Explicit is better than implicit. 2. Flat is better than nested. 3. Readability counts. 4. There should be [only] one obvious way to do it. 5. Namespaces are a great idea -- let's do more of those!
23. Example: Create a Person class
24. 1. Name and age 2. Can greet you 3. Knows if it's Beiber 4. Greets Google directors
27. 13 Similarities
28. 1. Dynamically Typed
29. def print_object(some_sorta_object): print some_sorta_object.read() print_object(my_file) >>> “This is the file contents”
30. 2. Everything is an Object (even functions!)
31. Python example: 5 + 3 >>> 8 (5).__add__(3) >>> 8 (5).__str__() >>> “5”
32. 3. Arrays
33. things = [1, "hello", 2.5, 43] things[1] >> 'hello' Ruby (array)
34. things = [1, "hello", 2.5, 43] things[1] >> 'hello' Python (list)
35. 4. Hash Tables
36. color_map = { red: "#FF0000", purple: "#800080"} color_map[:red] >> "#FF0000" Ruby (hash)
37. color_map = { "red": "#FF0000", "purple": "#800080"} color_map["red"] >> "#FF0000" Python (list)
38. 5. No special line terminators (no semicolons!)
39. 6. Strong functional programming paradigms
40. def triple(number): return number * 3 list = [1, 2, 3, 4] map(triple, list) >>> [3, 6, 9, 12] Map
41. def is_even(number): return number % 2 == 0 list = [1, 2, 3, 4] filter(is_even, list) >>> [2, 4] Filter
42. list = [1, 2, 3, 4] map(lambda x: x *3, list) >>> [3, 6, 9, 12] Anonymous Functions (lambda)
43. 7. Awesome function parameters
44. Optional / named arguments
45. def greet(name, greeting="Hello"): print greeting + ", " + name greet("Bob") >>> Hello, Bob greet("Bob", greeting="Bonjour") >>> Bonjour, Bob greet("Bob", "Bonjour") >>> Bonjour, Bob
46. Argument Unpacking
47. def greet_everyone(*names): for name in names: print "Hello, " + name names = ["Bob", "Steve", "Jim"] greet_everyone(*names) Hello, Bob Hello, Steve Hello, Jim
48. 8. Raising Exceptions
49. inputted_age = 30 if inputted_age > 25 raise ArgumentError, "must be under 26" end ArgumentError: must be under 26 (stack trace...) Ruby
50. inputted_age = 30 if inputted_age > 25: raise ValueError("must be under 26") (stack trace...) ValueError: must be under 26 Python
51. 9. Handling Exceptions
52. begin 23 / 0 puts "all good" rescue ZeroDivisionError puts "can't do that, bro" end >> can't do that, bro Ruby (rescue)
53. try: 23 / 0 print “all good” except ZeroDivisionError: print "can't do that, bro" >> can't do that, bro Python (except)
54. 10. One-line conditionals
55. “equal” if 3 == 3 >> “equal” 2 == 3 ? “equal” : “not equal” >> “not equal” Ruby
56. if 3 == 3: “hello” >>> “hello” “equal” if 2==3 else “not equal” >> “not equal” Python
57. 11. Nice interactive shells
58. $ python Python 2.7.3 (default) >>> 2 + 2 4 $ irb irb(main):001:0> 2 + 2 => 4
59. 12. Strong object reflection features
60. steve = Person("Steve Ballmer", 54) steve.__dict__ >>> {'age': 54, 'name': 'Steve Ballmer'} steve.__class__ >>> <class 'person.Person'> dir(steve) # all methods/attrbutes ['GOOGLE_DIRECTORS', '__class__', '__delattr__', '__dict__', '__doc__', '__format__', '__getattribute__', '__hash__', '__init__', '__module__', '__new__', '__reduce__', '__reduce_ex__', '__repr__', '__setattr__', '__sizeof__', '__str__', '__subclasshook__', '__weakref__', 'age', 'greet', 'greet_google_directors', 'is_justin_beiber', 'name', 'say_name_and_age']
61. 13. Radically simpler than C++/Java!
62. 13 Differences
63. 1. No blocks
64. Blocks are the killer feature of Ruby
65. 1. Powerful metaprogramming 2. DSLs 3. Great frameworks (Rails)
66. Decorators are a powerful Python feature similar to blocks.
67. They let you inject or modify code in a function, like blocks. But not enough time.
68. 2. First-class Functions
69. Functions as variables
70. def add(a, b): return a + b def subtract(a, b): return a – b def process_numbers(a, b, func): return func(a, b) process_numbers(20, 5, add) >>> 25 process_numbers(20, 5, subtract) >>> 15
71. Similar behavior is possible in Ruby, but you must wrap functions in procs.
72. 3. One-line lambdas, by design
73. Can't do this in Python: [1, 2, 3, 4, 5].map do |num| if num % 3 == 0 0 elsif num % 4 == 0 num * 2 else num end end
74. But do you really want to do that? 1. Harder to test 2. Harder to follow 3. Accidentally duplicate
75. Python pushes you to factor things out into granular methods.
76. def process_num(num): if num % 3 == 0: return 0 elif num %4 == 0: return num * 2 else: return num return map(process_num, [1,2])
77. Python even supports inner methods to this end.
78. def start_program(): def process_num(num): if num % 3 == 0: return 0 elif num %4 == 0: return num * 2 else: return num return map(process_num, [1,2])
79. 4. Python has “tuple,” an immutable list.
80. [1, 2, 3] # can change (1, 2, 3) # can't change (Should be homogeneous, but not enforced)
81. 5. Python has more fine-grained importing
82. require 'open-uri' result = open(“http://google.com”) Ruby You bring in everything from open-uri With lots of requires, which one is responsible?
83. from urllib2 import urlopen response = urlopen(“http://google.com”) Python “Explicit is better than implicit.” You can write from urllib2 import * , but not a best practice.
84. You can import any number of specific functions or variables from a python module(file).
85. 6. Python enforces indentation
86. In practice, this is pretty awesome.
87. Doesn't cause headaches. Makes code style more consistent.
88. 7. Python has more values that evaluate to False.
89. False None (nil) [ ], (), {} ' ' 0
90. 8. Instead of Enumerable, Python has built-in functions
91. filter() map() reduce()
92. Convenient, but not as powerful as Enumerable.
93. 9. Python has simpler conditionals
94. No case statements (use if... elif... else instead) if n == 0: print "You typed zero.\n" elif n== 1 or n == 9 or n == 4: print "n is a perfect square\n" elif n == 2: print "n is an even number\n" elif n== 3 or n == 5 or n == 7: print "n is a prime number\n"
95. No “unless” unless !person.present? && !company.present? puts "do you even know what you're doing?" else puts "and now we're really confused" end
96. 10. No automatic return values
97. Functions return None if they don't have an explicit “return” statement. def add(a, b): a + b add(2, 3) >>> None
98. 11. “self” as an argument to every class instance method
99. Pro: Explicit No clashes with local variables or other imported objects Con: More verbose
100. Class Person(object): def __init__(self, name): self.name = name def say_hi_to(self, name): print self.name + “ says hi to ” + name Person(“Steve”).say_hi_to(“Bob”) >>> “Steve says hi to Bob”
101. 12. No powerful module mixins like Ruby.
102. Multiple inheritence only (yuck)
103. 13. Ruby has stronger metaprogramming features
104. Python doesn't have define_method class_eval method_missing
105. Can't reopen a class and extend it Can't extend built-in types (maybe for the better?)
106. 5 Problem Domains
107. 1. Web Development
108. Ruby: Rails, Sinatra Ruby community is dominated by web development
109. Python: Django, Zope, Flask, Pylons. Web dev is just one piece of the gigantic python community.
110. 2. Mobile Applications
111. Neither language is popular for mobile apps. Interesting Ruby projects: Rhodes, RubyMotion.
112. 3. Desktop Applications
113. Python has PyQT, wxPython, PyGTK. Ruby has shoes, wxRuby, RubyGTK, but not popular or mature.
114. 4. Scientific Programming
115. Python has numpy, scipy. Widely used. Ruby doesn't have any mature scientifc/numeric libraries.
116. 5. Windows Deployment (if you must)
117. Ruby: RailsInstaller is great. Doesn't target deployment. JRuby makes windows deployment much easier.
118. But Python on windows is generally easier. Why? Because entire toolchain is well- supported. Windows neglect is self-perpetuating.
119. The Python community is just bigger.
120. “Feelings”
121. It's harder to write code that pisses off other developers in Python.
122. 1. Smaller grammar 2. Forced indentation 3. Fewer ways to handle conditionals 4. The source of imported functionality is obvious 5. Explicit class and method structure makes code easier to follow.
123. Ruby is more elegant, but in practice, Python is easier to read.
124. Rails still wins for web development.
125. Nothing as good as Rails will ever be implemented in Python. (metaprogramming, reflection)
126. Need to hire a Rubyist? Hire a Pythonista instead.
127. 1. Ruby developer market is out of control. 2. There are more Python developers 3. Pythonistas can ramp up EASILY (hours/days, not weeks/months)
128. Wrap-up
129. 1. VERY similar! 2. You're already ready 3. Python = Better Rubyist 4. Hire a Pythonista
130. MIKE LEONE panopticdev.com leone.panopticdev.com @panopticdev @facebiff THANK YOU!