software tool seminar ws1516 taming the snake
play

Software Tool Seminar WS1516 - Taming the Snake November 4, 2015 1 - PDF document

Software Tool Seminar WS1516 - Taming the Snake November 4, 2015 1 Taming the Snake 1.1 Understanding how Python works in N simple steps (with N still growing) 1.2 Step 0. What this talk is about (and what it isnt) Four basic things you


  1. Software Tool Seminar WS1516 - Taming the Snake November 4, 2015 1 Taming the Snake 1.1 Understanding how Python works in N simple steps (with N still growing) 1.2 Step 0. What this talk is about (and what it isn’t) Four basic things you need to learn to become a good programmer: 1. Learn how to express algorithms in a computer programming language 2. Learn how to design programs 3. Learn how your tool (the programming language) works and what it can offer you 4. Learn what others already have built for you I will only teach you how Python works. (And this is only the first part of it.) 1.3 Step 1. Everything is an object In Python, everything is an object. Examples: 42, 4.3, ’Hello world’, True, False, None, [0, 1, 2, 3], { ’key’: ’value’, ’other key’: ’other value’ } (’This’, ’is’, ’a’, ’tuple’), np.eye(10) Really everything! math.sin, lambda x: x, class C, math, func.__code__ To understand Python, we should first understand what objects are! 1.4 Step 2. The three properties of an object Objects have: 1. Identity 2. State 3. Methods 1

  2. 1.5 Step 3. The Identity of an object In [1]: a = [0, 1, 2] b = [0, 1, 2] a == b Out[1]: True In [2]: a is b Out[2]: False In [3]: id(a) Out[3]: 140008350782240 In [4]: id(b) Out[4]: 140008350781880 In [5]: (a is b) == (id(a) == id(b)) Out[5]: True In [6]: x = 6 y = 6 x is y Out[6]: True In [7]: x = 666 y = 666 x is y Out[7]: False In [8]: id(a) is id(a) Out[8]: False Do not use is unless you have a good reason! Reasonable exceptions: x is True x is False x is None This works because True , False , None are signletons in Python, i.e. there is only one object True , etc., in the whole Python universe. 2

  3. 1.6 Step 4. Understand assignment In [9]: import numpy as np a = np.ones(10) b = np.zeros(10) print(a) print(b) [ 1. 1. 1. 1. 1. 1. 1. 1. 1. 1.] [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.] In [10]: a is b Out[10]: False In [11]: a = b print(a) print(b) [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.] [ 0. 0. 0. 0. 0. 0. 0. 0. 0. 0.] In [12]: a[0] = 1 print(a) print(b) [ 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.] [ 1. 0. 0. 0. 0. 0. 0. 0. 0. 0.] In [13]: a is b Out[13]: True 1.6.1 Definition of assignment a = b means assign the name a to the object with name b 1.6.2 Let’s repeat! 1.6.3 Definition of assignment a = b means assign the name a to the object with name b 3

  4. 1.7 Step 5. The state of an object Objects have state (data) associated to them, which can change over the lifetime of an object. The state is stored in the objects attributes . In [14]: from email.mime.text import MIMEText m = MIMEText(’Hi, this is an email!’) m Out[14]: <email.mime.text.MIMEText instance at 0x7f563c04cf38> In [15]: print(m.as_string()) Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Hi, this is an email! In [16]: m._payload m._headers Out[16]: [(’Content-Type’, ’text/plain; charset="us-ascii"’), (’MIME-Version’, ’1.0’), (’Content-Transfer-Encoding’, ’7bit’)] The underscore means, payload and headers are private attributes. You, the user, should not mess around with them. Methods can change attributes: In [17]: m.add_header(’From’, ’stephan.rave@uni-muenster.de’) m.add_header(’To’, ’evil.overlord@the.hell.com’) m.add_header(’Subject’, ’World domination’) print(m.as_string()) Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit From: stephan.rave@uni-muenster.de To: evil.overlord@the.hell.com Subject: World domination Hi, this is an email! In [18]: m._headers Out[18]: [(’Content-Type’, ’text/plain; charset="us-ascii"’), (’MIME-Version’, ’1.0’), (’Content-Transfer-Encoding’, ’7bit’), (’From’, ’stephan.rave@uni-muenster.de’), (’To’, ’evil.overlord@the.hell.com’), (’Subject’, ’World domination’)] We can also change attributes, even private ones: (Do not try this at home!) In [19]: m._payload = ’We need to talk!’ print(m.as_string()) 4

  5. Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit From: stephan.rave@uni-muenster.de To: evil.overlord@the.hell.com Subject: World domination We need to talk! In [20]: m.my_attribute = 666 m.my_attribute Out[20]: 666 1.7.1 Step 6. Understand the basics of attribute lookup So where do all these attributes come from? In [21]: m.__dict__ Out[21]: { ’ charset’: us-ascii, ’ default type’: ’text/plain’, ’ headers’: [(’Content-Type’, ’text/plain; charset="us-ascii"’), (’MIME-Version’, ’1.0’), (’Content-Transfer-Encoding’, ’7bit’), (’From’, ’stephan.rave@uni-muenster.de’), (’To’, ’evil.overlord@the.hell.com’), (’Subject’, ’World domination’)], ’ payload’: ’We need to talk!’, ’ unixfrom’: None, ’defects’: [], ’epilogue’: None, ’my attribute’: 666, ’preamble’: None } In [22]: m.__dict__[’favourite_song’] = ’Hotel california’ m.__dict__[’_payload’] = ’WE NEED TO TALK!!!’ In [23]: m.favourite_song Out[23]: ’Hotel california’ In [24]: print(m.as_string()) Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit From: stephan.rave@uni-muenster.de To: evil.overlord@the.hell.com Subject: World domination WE NEED TO TALK!!! 5

  6. 1.7.2 Attribute lookup a.b means a.__dict__[’b’] In [25]: a = np.eye(10) a.secret_answer = 42 Writing out traceback for vim ... --------------------------------------------------------------------------- AttributeError Traceback (most recent call last) <ipython-input-25-9d909bc47556> in <module>() 1 a = np.eye(10) ----> 2 a.secret answer = 42 AttributeError: ’numpy.ndarray’ object has no attribute ’secret answer’ In [26]: a.__dict__ Writing out traceback for vim ... --------------------------------------------------------------------------- AttributeError Traceback (most recent call last) <ipython-input-26-87da7e691200> in <module>() ----> 1 a. dict AttributeError: ’numpy.ndarray’ object has no attribute ’ dict ’ In [27]: (42).__dict__ Writing out traceback for vim ... --------------------------------------------------------------------------- AttributeError Traceback (most recent call last) <ipython-input-27-f22e1a648ac9> in <module>() ----> 1 (42). dict AttributeError: ’int’ object has no attribute ’ dict ’ 6

  7. In [28]: [0, 1, 2].__dict__ Writing out traceback for vim ... --------------------------------------------------------------------------- AttributeError Traceback (most recent call last) <ipython-input-28-cb92ad523bf5> in <module>() ----> 1 [0, 1, 2]. dict AttributeError: ’list’ object has no attribute ’ dict ’ 1.7.3 Attribute lookup If a has a dict, a.b means a.__dict__[’b’] This is not the case for builtin types or C extension types. In [29]: class C(object): pass c = C() c.__dict__ Out[29]: {} In [30]: c.secret_answer = 42 c.__dict__ Out[30]: { ’secret answer’: 42 } In [31]: class C(object): secret_answer = 42 c = C() c.secret_answer Out[31]: 42 In [32]: c.__dict__ Out[32]: {} In [33]: c.__class__ Out[33]: main .C In [34]: c.__class__.__dict__ Out[34]: <dictproxy { ’ dict ’: <attribute ’ dict ’ of ’C’ objects>, ’ doc ’: None, ’ module ’: ’ main ’, ’ weakref ’: <attribute ’ weakref ’ of ’C’ objects>, ’secret answer’: 42 } > 7

  8. 1.7.4 Attribute lookup (simplified) If a has a dict, a.b means if ’b’ in a.__dict__: return a.__dict__[’b’] elif ’b’ in a.__class__.__dict__: return a.__class__.__dict__[’b’] elif ’b’ in ’base class dicts’: return base_class.__dict__[’b’] else: raise AttributeError This is not the case for builtin types or C extension types. 1.7.5 Step 7. Be careful with class attributes In [35]: class C(object): great_list_of_awesomeness = [] c1 = C() c2 = C() In [36]: c1.great_list_of_awesomeness.append(42) c1.great_list_of_awesomeness Out[36]: [42] In [37]: c2.great_list_of_awesomeness Out[37]: [42] This might not be what you want! 1.7.6 Step 8. Understand methods Methods are functions defined in class definitions as follows: In [38]: class C(object): def __init__(self, x): self.x = x def f(self, y): return self.x + y c = C(11) c.f(31) Out[38]: 42 8

  9. c.f(31) translates to C.f(c, 31) i.e. calling a method on an object c magically inserts c as first argument of the method. This also explains this strange error: In [39]: c.f() Writing out traceback for vim ... --------------------------------------------------------------------------- TypeError Traceback (most recent call last) <ipython-input-39-69c69864e152> in <module>() ----> 1 c.f() TypeError: f() takes exactly 2 arguments (1 given) is one of Python’s many special methods. init It is called, after a new object has been created. Thus c = C(11) translates to c = newly_created_C_instance c.__init__(c, 11) We can add new methods to the class at any time: In [40]: def g(ego, y): return ego.x * y C.mult = g c.mult(2) Out[40]: 22 Note that self as first argument name is pure convention. You should really stick to it! Adding functions directly to objects does not work: In [41]: def h(self, y): return self.x - y c.h = h c.h(-31) Writing out traceback for vim ... 9

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend


More recommend