首页 > itarticle > beautiful_idiomatic_python

beautiful_idiomatic_python

admin 11月 13, 2020 0

https://github.com/JeffPaine/beautiful_idiomatic_python

Notes from Raymond Hettinger’s talk at pycon US 2013 video, slides.

The code examples and direct quotes are all from Raymond’s talk. I’ve reproduced them here for my own edification and the hopes that others will find them as handy as I have!

Looping over a range of numbers

1
2
3
4
5
 
for i in [0, 1, 2, 3, 4, 5]:
    print i**2

for i in range(6):
    print i**2

Better

1
2
 
for i in xrange(6):
    print i**2

xrange creates an iterator over the range producing the values one at a time. This approach is much more memory efficient than range. xrange was renamed to range in python 3.

Looping over a collection

1
2
3
4
 
colors = ['red', 'green', 'blue', 'yellow']

for i in range(len(colors)):
    print colors[i]

Better

1
2
 
for color in colors:
    print color

Looping backwards

1
2
3
4
 
colors = ['red', 'green', 'blue', 'yellow']

for i in range(len(colors)-1, -1, -1):
    print colors[i]

Better

1
2
 
for color in reversed(colors):
    print color

Looping over a collection and indices

1
2
3
4
 
colors = ['red', 'green', 'blue', 'yellow']

for i in range(len(colors)):
    print i, '--->', colors[i]

Better

1
2
 
for i, color in enumerate(colors):
    print i, '--->', color

It’s fast and beautiful and saves you from tracking the individual indices and incrementing them.

Whenever you find yourself manipulating indices [in a collection], you’re probably doing it wrong.

Looping over two collections

1
2
3
4
5
6
7
8
9
 
names = ['raymond', 'rachel', 'matthew']
colors = ['red', 'green', 'blue', 'yellow']

n = min(len(names), len(colors))
for i in range(n):
    print names[i], '--->', colors[i]

for name, color in zip(names, colors):
    print name, '--->', color

Better

1
2
 
for name, color in izip(names, colors):
    print name, '--->', color

zip creates a new list in memory and takes more memory. izip is more efficient than zip.
Note: in python 3 izip was renamed to zip and promoted to a builtin replacing the old zip.

Looping in sorted order

1
2
3
4
5
6
7
8
9
 
colors = ['red', 'green', 'blue', 'yellow']


for color in sorted(colors):
    print colors

# Backwards sorted order
for color in sorted(colors, reverse=True):
    print colors

Custom Sort Order

1
2
3
4
5
6
7
8
 
colors = ['red', 'green', 'blue', 'yellow']

def (c1, c2):
    if len(c1) < len(c2): return -1
    if len(c1) > len(c2): return 1
    return 0

print sorted(colors, cmp=compare_length)

Better

1
 
print sorted(colors, key=len)

The original is slow and unpleasant to write. Also, comparison functions are no longer available in python 3.

Call a function until a sentinel value

1
2
3
4
5
6
 
blocks = []
while True:
    block = f.read(32)
    if block == '':
        break
    blocks.append(block)

Better

1
2
3
 
blocks = []
for block in iter(partial(f.read, 32), ''):
    blocks.append(block)

iter takes two arguments. The first you call over and over again and the second is a sentinel value.

Distinguishing multiple exit points in loops

1
2
3
4
5
6
7
8
9
 
def find(seq, target):
    found = False
    for i, value in enumerate(seq):
        if value == target:
            found = True
            break
    if not found:
        return -1
    return i

Better

1
2
3
4
5
6
7
 
def find(seq, target):
    for i, value in enumerate(seq):
        if value == target:
            break
    else:
        return -1
    return i

Inside of every for loop is an else.

Looping over dictionary keys

1
2
3
4
5
6
7
8
 
d = {'matthew': 'blue', 'rachel': 'green', 'raymond': 'red'}

for k in d:
    print k

for k in d.keys():
    if k.startswith('r'):
        del d[k]

When should you use the second and not the first? When you’re mutating the dictionary.

If you mutate something while you’re iterating over it, you’re living in a state of sin and deserve what ever happens to you.

d.keys() makes a copy of all the keys and stores them in a list. Then you can modify the dictionary.
Note: in python 3 to iterate through a dictionary you have to explicitly write: list(d.keys()) because d.keys() returns a “dictionary view” (an iterable that provide a dynamic view on the dictionary’s keys). See documentation.

Looping over dictionary keys and values

1
2
3
4
5
6
7
 
# Not very fast, has to re-hash every key and do a lookup
for k in d:
    print k, '--->', d[k]

# Makes a big huge list
for k, v in d.items():
    print k, '--->', v

Better

1
2
 
for k, v in d.iteritems():
    print k, '--->', v

iteritems() is better as it returns an iterator.
Note: in python 3 there is no iteritems() and items() behaviour is close to what iteritems() had. See documentation.

Construct a dictionary from pairs

1
2
3
4
5
 
names = ['raymond', 'rachel', 'matthew']
colors = ['red', 'green', 'blue']

d = dict(izip(names, colors))
# {'matthew': 'blue', 'rachel': 'green', 'raymond': 'red'}

For python 3: d = dict(zip(names, colors))

Counting with dictionaries

1
2
3
4
5
6
7
8
9
10
 
colors = ['red', 'green', 'red', 'blue', 'green', 'red']

# Simple, basic way to count. A good start for beginners.
d = {}
for color in colors:
    if color not in d:
        d[color] = 0
    d[color] += 1

# {'blue': 1, 'green': 2, 'red': 3}

Better

1
2
3
4
5
6
7
8
9
 
d = {}
for color in colors:
    d[color] = d.get(color, 0) + 1

# Slightly more modern but has several caveats, better for advanced users
# who understand the intricacies
d = collections.defaultdict(int)
for color in colors:
    d[color] += 1

Grouping with dictionaries – Part I and II

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
 
names = ['raymond', 'rachel', 'matthew', 'roger',
         'betty', 'melissa', 'judith', 'charlie']

# In this example, we're grouping by name length
d = {}
for name in names:
    key = len(name)
    if key not in d:
        d[key] = []
    d[key].append(name)

# {5: ['roger', 'betty'], 6: ['rachel', 'judith'], 7: ['raymond', 'matthew', 'melissa', 'charlie']}

d = {}
for name in names:
    key = len(name)
    d.setdefault(key, []).append(name)

Better

1
2
3
4
 
d = collections.defaultdict(list)
for name in names:
    key = len(name)
    d[key].append(name)

Is a dictionary popitem() atomic?

1
2
3
4
5
 
d = {'matthew': 'blue', 'rachel': 'green', 'raymond': 'red'}

while d:
    key, value = d.popitem()
    print key, '-->', value

popitem is atomic so you don’t have to put locks around it to use it in threads.

Linking dictionaries

1
2
3
4
5
6
7
8
9
10
11
12
13
 
defaults = {'color': 'red', 'user': 'guest'}
parser = argparse.ArgumentParser()
parser.add_argument('-u', '--user')
parser.add_argument('-c', '--color')
namespace = parser.parse_args([])
command_line_args = {k:v for k, v in vars(namespace).items() if v}

# The common approach below allows you to use defaults at first, then override them
# with environment variables and then finally override them with command line arguments.
# It copies data like crazy, unfortunately.
d = defaults.copy()
d.update(os.environ)
d.update(command_line_args)

Better

1
 
d = ChainMap(command_line_args, os.environ, defaults)

ChainMap has been introduced into python 3. Fast and beautiful.

Improving Clarity

  • Positional arguments and indicies are nice
  • Keywords and names are better
  • The first way is convenient for the computer
  • The second corresponds to how human’s think

Clarify function calls with keyword arguments

1
 
twitter_search('@obama', False, 20, True)

Better

1
 
twitter_search('@obama', retweets=False, numtweets=20, popular=True)

Is slightly (microseconds) slower but is worth it for the code clarity and developer time savings.

Clarify multiple return values with named tuples

1
2
3
4
 
# Old testmod return value
doctest.testmod()
# (0, 4)
# Is this good or bad? You don't know because it's not clear.

Better

1
2
3
 
# New testmod return value, a namedTuple
doctest.testmod()
# TestResults(failed=0, attempted=4)

A namedTuple is a subclass of tuple so they still work like a regular tuple, but are more friendly.

To make a namedTuple:

1
 
TestResults = namedTuple('TestResults', ['failed', 'attempted'])

Unpacking sequences

1
2
3
4
5
6
7
 
p = 'Raymond', 'Hettinger', 0x30, '[email protected]'

# A common approach / habit from other languages
fname = p[0]
lname = p[1]
age = p[2]
email = p[3]

Better

1
 
fname, lname, age, email = p

The second approach uses tuple unpacking and is faster and more readable.

Updating multiple state variables

1
2
3
4
5
6
7
8
 
def fibonacci(n):
    x = 0
    y = 1
    for i in range(n):
        print x
        t = y
        y = x + y
        x = t

Better

1
2
3
4
5
 
def fibonacci(n):
    x, y = 0, 1
    for i in range(n):
        print x
        x, y = y, x + y

Problems with first approach

  • x and y are state, and state should be updated all at once or in between lines that state is mis-matched and a common source of issues
  • ordering matters
  • it’s too low level

The second approach is more high-level, doesn’t risk getting the order wrong and is fast.

Simultaneous state updates

1
2
3
4
5
6
7
8
9
10
11
 
tmp_x = x + dx * t
tmp_y = y + dy * t
# NOTE: The "influence" function here is just an example function, what it does
# is not important. The important part is how to manage updating multiple
# variables at once.
tmp_dx = influence(m, x, y, dx, dy, partial='x')
tmp_dy = influence(m, x, y, dx, dy, partial='y')
x = tmp_x
y = tmp_y
dx = tmp_dx
dy = tmp_dy

Better

1
2
3
4
5
6
7
 
# NOTE: The "influence" function here is just an example function, what it does
# is not important. The important part is how to manage updating multiple
# variables at once.
x, y, dx, dy = (x + dx * t,
                y + dy * t,
                influence(m, x, y, dx, dy, partial='x'),
                influence(m, x, y, dx, dy, partial='y'))

Efficiency

  • An optimization fundamental rule
  • Don’t cause data to move around unnecessarily
  • It takes only a little care to avoid O(n**2) behavior instead of linear behavior

Basically, just don’t move data around unecessarily.

Concatenating strings

1
2
3
4
5
6
7
 
names = ['raymond', 'rachel', 'matthew', 'roger',
         'betty', 'melissa', 'judith', 'charlie']

s = names[0]
for name in names[1:]:
    s += ', ' + name
print s

Better

1
 
print ', '.join(names)

Updating sequences

1
2
3
4
5
6
7
 
names = ['raymond', 'rachel', 'matthew', 'roger',
         'betty', 'melissa', 'judith', 'charlie']

del names[0]
# The below are signs you're using the wrong data structure
names.pop(0)
names.insert(0, 'mark')

Better

1
2
3
4
5
6
7
 
names = collections.deque(['raymond', 'rachel', 'matthew', 'roger',
               'betty', 'melissa', 'judith', 'charlie'])

# More efficient with collections.deque
del names[0]
names.popleft()
names.appendleft('mark')

Decorators and Context Managers

  • Helps separate business logic from administrative logic
  • Clean, beautiful tools for factoring code and improving code reuse
  • Good naming is essential.
  • Remember the Spiderman rule: With great power, comes great responsibility!

Using decorators to factor-out administrative logic

1
2
3
4
5
6
7
 
# Mixes business / administrative logic and is not reusable
def web_lookup(url, saved={}):
    if url in saved:
        return saved[url]
    page = urllib.urlopen(url).read()
    saved[url] = page
    return page

Better

1
2
3
 

def web_lookup(url):
    return urllib.urlopen(url).read()

Note: since python 3.2 there is a decorator for this in the standard library: functools.lru_cache.

Factor-out temporary contexts

1
2
3
4
5
 
# Saving the old, restoring the new
old_context = getcontext().copy()
getcontext().prec = 50
print Decimal(355) / Decimal(113)
setcontext(old_context)

Better

1
2
 
with localcontext(Context(prec=50)):
    print Decimal(355) / Decimal(113)

How to open and close files

1
2
3
4
5
 
f = open('data.txt')
try:
    data = f.read()
finally:
    f.close()

Better

1
2
 
with open('data.txt') as f:
    data = f.read()

How to use locks

1
2
3
4
5
6
7
8
9
10
 
# Make a lock
lock = threading.Lock()

# Old-way to use a lock
lock.acquire()
try:
    print 'Critical section 1'
    print 'Critical section 2'
finally:
    lock.release()

Better

1
2
3
4
 
# New-way to use a lock
with lock:
    print 'Critical section 1'
    print 'Critical section 2'

Factor-out temporary contexts

1
2
3
4
 
try:
    os.remove('somefile.tmp')
except OSError:
    pass

Better

1
2
 
with ignored(OSError):
    os.remove('somefile.tmp')

ignored is is new in python 3.4, documentation.
Note: ignored is actually called suppress in the standard library.

To make your own ignored context manager in the meantime:

1
2
3
4
5
6
 
@contextmanager
def ignored(*exceptions):
    try:
        yield
    except exceptions:
        pass

Stick that in your utils directory and you too can ignore exceptions

Factor-out temporary contexts

1
2
3
4
5
6
7
8
 
# Temporarily redirect standard out to a file and then return it to normal
with open('help.txt', 'w') as f:
    oldstdout = sys.stdout
    sys.stdout = f
    try:
        help(pow)
    finally:
        sys.stdout = oldstdout

Better

1
2
3
 
with open('help.txt', 'w') as f:
    with redirect_stdout(f):
        help(pow)

redirect_stdout is proposed for python 3.4, bug report.

To roll your own redirect_stdout context manager

1
2
3
4
5
6
7
8
 
@contextmanager
def redirect_stdout(fileobj):
    oldstdout = sys.stdout
    sys.stdout = fileobj
    try:
        yield fileobj
    finally:
        sys.stdout = oldstdout

Concise Expressive One-Liners

Two conflicting rules:

  • Don’t put too much on one line
  • Don’t break atoms of thought into subatomic particles

Raymond’s rule:

  • One logical line of code equals one sentence in English

List Comprehensions and Generator Expressions

1
2
3
4
5
 
result = []
for i in range(10):
    s = i ** 2
    result.append(s)
print sum(result)

Better

1
 
print sum(i**2 for i in xrange(10))

First way tells you what to do, second way tells you what you want.

微海报分享

近期文章

近期评论

标签

热门

文章归档

分类目录

功能