I've always thought that the best way to introduce new programmers to Python is to show them small code examples.
When you go to the tutorial, though, you have to wade through quite a bit of English before seeing any Python examples.
Below is my attempt at generating ten fairly simple, representative Python programs that expose new users to most basic concepts, as well as the overall syntax.
It was an interesting exercise. I constrained myself to ten lines or less, and it was pretty easy to incorporate loops, conditionals, print, open(), lists, tuples, dictionaries, and imported modules.
It was harder to show classes, and my ShoppingCart class is nothing more than an encapsulation of a list, which has dubious value (although it's the start of something more useful).
Anyway, here goes:
------ print 'hello world'
------ for name in ('peter', 'paul', 'mary'): print name
------ # This is a Python comment. \n is a newline name = raw_input('What is your name?\n') print 'Hi', name
------ import re for test_string in [ '555-1212', 'ILL-EGAL']: if re.match('\d\d\d-\d\d\d\d$', test_string): print test_string, 'is a valid US local phone number' else: print test_string, 'rejected'
------ prices = {'apple': 0.40, 'banana': 0.50} myPurchase = { 'apple': 1, 'banana': 6} groceryBill = sum([prices[fruit] * myPurchase[fruit] for fruit in myPurchase]) print 'I owe the grocer $%.2f' % groceryBill
------ # indent your Python code to put into an email import glob pythonFiles = glob.glob('*.py') pythonFiles.sort() for fn in pythonFiles: print ' ------' for line in open(fn): print ' ' + line.rstrip() print
> I've always thought that the best way to introduce new > programmers to Python is to show them small code > examples.
> When you go to the tutorial, though, you have to wade > through quite a bit of English before seeing any > Python examples.
> Below is my attempt at generating ten fairly simple, > representative Python programs that expose new users > to most basic concepts, as well as the overall syntax.
> It was an interesting exercise. I constrained myself > to ten lines or less, and it was pretty easy to > incorporate loops, conditionals, print, open(), lists, > tuples, dictionaries, and imported modules.
> It was harder to show classes, and my ShoppingCart > class is nothing more than an encapsulation of a list, > which has dubious value (although it's the start of > something more useful).
> Anyway, here goes:
Many thanks for that. This is exactly what is missing in most introductory books. Simple, relevant and concise examples.
Regards, John -- War is God's way of teaching Americans geography Ambrose Bierce (1842 - 1914)
Steve Howell wrote: > I've always thought that the best way to introduce new > programmers to Python is to show them small code > examples.
> When you go to the tutorial, though, you have to wade > through quite a bit of English before seeing any > Python examples.
> Below is my attempt at generating ten fairly simple, > representative Python programs that expose new users > to most basic concepts, as well as the overall syntax.
Very cool! Do you mind putting this up on the Wiki somewhere so that we can link to it more easily? Maybe something like:
I think boughtItems() is probably not a good example of Python code since in this case, you should probably just write ``my_cart.items``. Maybe it should define ``__len__`` instead? Or maybe something like::
def select_items(self, prefix): return [item for item in self.items if item.startswith(prefix)]
> I think boughtItems() is probably not a good example > of Python code > since in this case, you should probably just write > ``my_cart.items``. > Maybe it should define ``__len__`` instead? Or maybe > something like::
> def select_items(self, prefix): > return [item for item in self.items if > item.startswith(prefix)]
I think the problem here is that it's hard to write a useful class in less than 10 lines of code. Can somebody else give it a try?
Although I didn't call it out in the email, I tried to make each program progressively one line longer, so if somebody wants to write, say, an 11-line class example, then I will try fill in the gap with another 8-liner.
Did I miss any basic concepts in the first 10 programs? Maybe an 8-liner could demonstrate command line arguments?
Really? Underscore-separated words preferred over camel case? What is the rationale for this? This style is so retro/80's/C-ish. It seems more like a Java backlash to me than anything else. If we (or Guido) don't like changing case to indicate word breaks, why are class names to be UpperCamelCase, and not Capitalized_with_underscores? If there is a casing convention nit to pick, I'd focus on UpperCamelCase for class names, lower case (either with underscores or mixed case) for attributes and method names, and UNDERSCORE_SEPARATED_ALL_CAPS for constants.
If we want to just say "well, PEP-8 says such and such," I think this is an area where the thinking has possibly evolved since 2001. Also, I think the PEP would benefit from explicitly discouraging some practices, such as Hungarian notation.
I guess pyparsing with its mixedCase functions and attributes is doomed for the Dunce Corner. Too bad for BeautifulSoup, cElementTree, and wxPython that are also at variance with this canon of Python coding style. ("Modules should have short, all-lowercase names. ... Python packages should also have short, all-lowercase names, although the use of underscores is discouraged.")
> I guess pyparsing with its mixedCase functions and attributes is > doomed for the Dunce Corner. Too bad for BeautifulSoup, cElementTree, > and wxPython that are also at variance with this canon of Python > coding style. ("Modules should have short, all-lowercase names. ... > Python packages should also have short, all-lowercase names, although > the use of underscores is discouraged.")
Although the names in wxPython are indeed contrary to PEP 8 (because they are the same as the names used in wxWidgets) I should point out that nowadays the name you import is "wx".
regards Steve -- Steve Holden +1 571 484 6266 +1 800 494 3119 Holden Web LLC/Ltd http://www.holdenweb.com Skype: holdenweb http://del.icio.us/steve.holden ------------------ Asciimercial --------------------- Get on the web: Blog, lens and tag your way to fame!! holdenweb.blogspot.com squidoo.com/pythonology tagged items: del.icio.us/steve.holden/python All these services currently offer free registration! -------------- Thank You for Reading ----------------
> I ***love*** this "10 Little Programs" idea! As > soon as I get a > breathing space, I'm going to add a "10 Little > Parsers" page to the > pyparsing wiki!
Thanks. :)
I'm thinking you could actually have a progression from a 1 line program up to a 50-line program. The number 50 is kind of arbitrary, but my gut says that by a 50-line program, you will have demonstrated almost every useful concept.
Yep, that sounds more in the ballpark of what I'd want to show, versus a weakish class that just encapulates a list.
Here's my challenge to whoever wants to take it--write (or find) a program with 20 or fewer lines that sufficiently motivates the need for classes, has decent Python style, and is newbie friendly.
The tutorial has this example, which is useful for demonstrating the syntax of classes, but it doesn't actually do anything interesting:
class MyClass: "A simple example class" i = 12345 def f(self): return 'hello world'
It also has a ComplexNumber class, but I don't want to scare away mathphobes.
It does have this idiom, which I think is worth putting somewhere into the progression.
class Employee: pass
john = Employee() # Create an empty employee record
# Fill the fields of the record john.name = 'John Doe' john.dept = 'computer lab' john.salary = 1000
___________________________________________________________________________ _________Take the Internet to Go: Yahoo!Go puts the Internet in your pocket: mail, news, photos & more. http://mobile.yahoo.com/go?refer=1GNXIC
On May 26, 8:48 pm, Steve Howell <showel...@yahoo.com> wrote:
> I'm thinking you could actually have a progression > from a 1 line program up to a 50-line program. The > number 50 is kind of arbitrary, but my gut says that > by a 50-line program, you will have demonstrated > almost every useful concept.
If there is anything arbitrary here, I'd say it is your "increment each example by one source line" constraint. This can force you to use some bad coding practices to meet your target line count for a given example.
Maybe try this approach: pick your top 10/20/50 language features and develop concise examples. Then order the examples by length as a first cut (longer examples probably *are* more complex), and then reorder a bit to handle pre-requisites (introduce a minimum of new features, preferably 1, per example). Overall, I'd have a tough time picking just 10 language features to illustrate, but there are probably 10-20 basic features that will get new people onto fairly productive ground. Pick 20 as your example count (50 sounds a bit long), and stick to it, and then later add "20 More Little Programs" for the next wave of examples in increasing complexity.
One other nit to pick: have your example classes inherit from object, to get new people using new-style classes from the get-go.
Paul McGuire wrote: > I ***love*** this "10 Little Programs" idea! As soon as I get a > breathing space, I'm going to add a "10 Little Parsers" page to the > pyparsing wiki!
> On May 26, 2:38 pm, Steven Bethard <steven.beth...@gmail.com> wrote: >> <nitpick> >> Though the code should probably follow PEP 8 guidelines, e.g. >> under_scores instead of camelCase for object and method names:
> Really? Underscore-separated words preferred over camel case? What > is the rationale for this?
Rationale? It's a style guide. There is no rationale. ;-)
> If we want to just say "well, PEP-8 says such and such," I think this > is an area where the thinking has possibly evolved since 2001.
I really don't think so. If anything, it's gotten more strict. PEP 8 used to allow either camelCase or under_scores. Now it only allows the latter.
> I guess pyparsing with its mixedCase functions and attributes is > doomed for the Dunce Corner. Too bad for BeautifulSoup, cElementTree, > and wxPython that are also at variance with this canon of Python > coding style.
Many (if not all) of these modules were written before the most recent incarnation of PEP 8. Thus, they fall under the second good reason "to break a particular rule":
(2) To be consistent with surrounding code that also breaks it
Of course, for new code, such as that in this thread, there's no reason to break from the PEP 8 guidelines.
> On May 26, 8:48 pm, Steve Howell > <showel...@yahoo.com> wrote:
> > I'm thinking you could actually have a progression > > from a 1 line program up to a 50-line program. > The > > number 50 is kind of arbitrary, but my gut says > that > > by a 50-line program, you will have demonstrated > > almost every useful concept.
> If there is anything arbitrary here, I'd say it is > your "increment > each example by one source line" constraint. This > can force you to > use some bad coding practices to meet your target > line count for a > given example.
I understand your point, but I'm sticking to the concept for now. My intent with the progression isn't so much for each example to thoroughly teach a concept (although I could certainly hyperlink to a more in-depth treatment), but really more to give a bird's eye view of the language very quickly.
I recently helped teach a Java programmer to program in Python, and he learned a lot just by seeing simple examples. So I guess my target audience isn't so much people learning how to program; it's more for programmers getting their first exposure to Python.
On the other side of the fence, I recently tried to relearn a bit of Ruby, and I remember being frustrated by their tutorials, as really, I just wanted to see a bunch of simple programs, and I can figure out mostly what they're doing. Instead, I had to wade through verbose descriptions of what a variable is, rules for how you construct identifiers, etc.
> Maybe try this approach: pick your top 10/20/50 > language features and > develop concise examples. Then order the examples by > length as a first > cut (longer examples probably *are* more complex), > and then reorder a > bit to handle pre-requisites (introduce a minimum of > new features, > preferably 1, per example). Overall, I'd have a > tough time picking > just 10 language features to illustrate, but there > are probably 10-20 > basic features that will get new people onto fairly > productive > ground. Pick 20 as your example count (50 sounds a > bit long), and > stick to it, and then later add "20 More Little > Programs" for the next > wave of examples in increasing complexity.
My only reluctance with that approach is that it sounds like a little more work than I'm ready to take on right away. But it's on the Wiki now, so maybe other folks can help me grow it.
> One other nit to pick: have your example classes > inherit from object, > to get new people using new-style classes from the > get-go.
My only fear here is that when old classes go away (Py3K? I don't know), that practice may become obsolete.
But on a more general note, I welcome folks to just clean up my examples on the Wiki if I accidentally introduce some bad practices...but preserving line counts. :)
___________________________________________________________________________ _________ Never miss an email again! Yahoo! Toolbar alerts you the instant new Mail arrives. http://tools.search.yahoo.com/toolbar/features/mail/
Out of curiosity, how does this style jibe with the latest embracing of Unicode identifiers? Ever tried to type an underscore on a non-US keyboard? I have a heck of a time finding/typing the '_' character when I visit our clients in Germany, but this may just be my own personal Amerocentric issue (I also get messed up by the transposition of Y and Z on German keyboards, but my German colleagues understandably are not bothered by it). For someone already familiar with that keyboard layout, is typing an underscore any more difficult than my pressing Shift-_ on my US keyboard?
On May 26, 9:58 pm, Paul McGuire <p...@austin.rr.com> wrote:
> Out of curiosity, how does this style jibe with the latest embracing > of Unicode identifiers? Ever tried to type an underscore on a non-US > keyboard? I have a heck of a time finding/typing the '_' character > when I visit our clients in Germany, but this may just be my own > personal Amerocentric issue (I also get messed up by the transposition > of Y and Z on German keyboards, but my German colleagues > understandably are not bothered by it). For someone already familiar > with that keyboard layout, is typing an underscore any more difficult > than my pressing Shift-_ on my US keyboard?
> -- Paul
Steve, sorry for going so far off-topic. I've started a new thread on my questions about this aspect of PEP-8, and if there's more to say about this, people should post it there.
Obviously, it's not surprising that a useful class in a tutorial would have a corresponding implementation in the standard library, but I'm not sure newbies would learn much about classes from this statement:
ComplexNumber = complex
___________________________________________________________________________ _________ Need Mail bonding? Go to the Yahoo! Mail Q&A for great tips from Yahoo! Answers users. http://answers.yahoo.com/dir/?link=list&sid=396546091
--- Steven D'Aprano wrote: > On Sat, 26 May 2007 18:48:45 -0700, Steve Howell > wrote:
> > It also has a ComplexNumber class, but I don't > want to > > scare away mathphobes.
> Is it as short as this one-liner?
> ComplexNumber = complex
Along the idea of not reinventing a class from the standard library in the list of ten small Python programs (which has since grown), I went with the classic BankAccount example for the first program to introduce the "class" statement.
___________________________________________________________________________ _________Be a better Globetrotter. Get better travel answers from someone who knows. Yahoo! Answers - Check it out. http://answers.yahoo.com/dir/?link=list&sid=396545469
For the person new to programming (doesn't come from C or other languages), I think you need to add a separate explanation of string formatting and how it works, or at least add a comment that tells them you are using string formatting so that they can search and find out how it works. If your aim is to teach simple programming concepts, why confuse them so early on with fancy interpolation?
Somebody also fixed a few style things in my other examples, changing a tuple to a list, catching a more specific exception. Whoever you are, thanks, I agree.
___________________________________________________________________________ _________Sick sense of humor? Visit Yahoo! TV's Comedy with an Edge to see what's on, when. http://tv.yahoo.com/collections/222
> For the person new to programming (doesn't come from > C or other > languages), I think you need to add a separate > explanation of string > formatting and how it works, or at least add a > comment that tells them > you are using string formatting so that they can > search and find out > how it works. If your aim is to teach simple > programming concepts, why > confuse them so early on with fancy interpolation?
It's a thought provoking question, and I think my aim here is not exactly to teach simple programming concepts, but more to expose people to what Python looks like. I'm not really intending this page to be a tutorial, as several good tutorials already exist.
I'm really targeting a particular niche of people. There are folks that know how to program, but don't know anything about Python, and they really just want to see a bunch of small examples all in one place, without a lot of explanation cluttering their presentation.
That may sound like I'm narrowing my audience too much, but I do think it's a niche group that's not adequately addressed.
I do hope, though, that folks more in a teaching role can reuse the examples, add better explanation, etc., as needed.
Also, I wouldn't mind at all to add a little link called "Read more..." after each example.
___________________________________________________________________________ _________Ready for the edge of your seat? Check out tonight's top picks on Yahoo! TV. http://tv.yahoo.com/
Steve Howell wrote: > # def defines a method in Python > def say_hello(name): > print 'hello', name > say_hello('Jack') > say_hello('Jill')
Doesn't def define methods *xor* functions, depending on the context? And in this example, say_hello (*yuck*, underscores ...) is certainly a function. Or is it that functions are considered "module-methods"?
> Steve Howell wrote: > > # def defines a method in Python > > def say_hello(name): > > print 'hello', name > > say_hello('Jack') > > say_hello('Jill')
> Doesn't def define methods *xor* functions, > depending on the context? > And in this example, say_hello (*yuck*, underscores > ...) is certainly a > function. Or is it that functions are considered > "module-methods"?
Goodness, I didn't expect such a simple example to be so controversial. But please see the new version here:
Steve Howell wrote: > --- Steven Bethard <steven.beth...@gmail.com> wrote: >> Very cool! Do you mind putting this up on the Wiki >> somewhere so that we >> can link to it more easily? Maybe something like:
I think I would rewrite the current unit-testing example to use the standard library unittest module::
# Let's write reusable code, and unit test it. def add_money(amounts): # do arithmetic in pennies so as not to accumulate float errors pennies = sum([round(int(amount * 100)) for amount in amounts]) return float(pennies / 100.0) import unittest class TestAddMoney(unittest.TestCase): def test_float_errors(self): self.failUnlessEqual(add_money([0.13, 0.02]), 0.15) self.failUnlessEqual(add_money([100.01, 99.99]), 200) self.failUnlessEqual(add_money([0, -13.00, 13.00]), 0) if __name__ == '__main__': unittest.main()
I believe I've still kept it to 13 lines.
STeVe
P.S. The "right" way to add money is using the decimal module, but I couldn't think of a better example.
> I think I would rewrite the current unit-testing > example to use the > standard library unittest module::
> # Let's write reusable code, and unit test it. > def add_money(amounts): > # do arithmetic in pennies so as not to > accumulate float errors > pennies = sum([round(int(amount * 100)) for > amount in amounts]) > return float(pennies / 100.0) > import unittest > class TestAddMoney(unittest.TestCase): > def test_float_errors(self): > self.failUnlessEqual(add_money([0.13, > 0.02]), 0.15) > self.failUnlessEqual(add_money([100.01, > 99.99]), 200) > self.failUnlessEqual(add_money([0, > -13.00, 13.00]), 0) > if __name__ == '__main__': > unittest.main()
> I believe I've still kept it to 13 lines.
I approve this change, although in a sense, it's harder for a Python newbie, because it introduces inheritance a little earlier than I would have liked.
FWIW I'm in the minority (I think) of people that prefer roll-your-own testing, but I don't want to argue that, because I think it mostly comes down to personal preference.
I'll only defend my position by posting this link, which suggests that roll-your-own even has validity in an educational setting:
> P.S. The "right" way to add money is using the > decimal module, but I > couldn't think of a better example.
Agreed. Maybe somebody else will come up with something more creative, but I'm happy enough with our current version.
___________________________________________________________________________ _________Take the Internet to Go: Yahoo!Go puts the Internet in your pocket: mail, news, photos & more. http://mobile.yahoo.com/go?refer=1GNXIC
Steve Howell wrote: > --- Steven Bethard <steven.beth...@gmail.com> wrote: >> I think I would rewrite the current unit-testing >> example to use the >> standard library unittest module::
>> # Let's write reusable code, and unit test it. >> def add_money(amounts): >> # do arithmetic in pennies so as not to >> accumulate float errors >> pennies = sum([round(int(amount * 100)) for >> amount in amounts]) >> return float(pennies / 100.0) >> import unittest >> class TestAddMoney(unittest.TestCase): >> def test_float_errors(self): >> self.failUnlessEqual(add_money([0.13, >> 0.02]), 0.15) >> self.failUnlessEqual(add_money([100.01, >> 99.99]), 200) >> self.failUnlessEqual(add_money([0, >> -13.00, 13.00]), 0) >> if __name__ == '__main__': >> unittest.main()
> Just a minor quibble, but wouldn't you want the import > and test class to only get executed in the ___main__ > context?
That would be fine too. In the real world, I'd put the tests in a different module.
Steve Howell wrote: > --- Steven Bethard <steven.beth...@gmail.com> wrote:
>> I think I would rewrite the current unit-testing >> example to use the >> standard library unittest module::
>> # Let's write reusable code, and unit test it. >> def add_money(amounts): >> # do arithmetic in pennies so as not to >> accumulate float errors >> pennies = sum([round(int(amount * 100)) for >> amount in amounts]) >> return float(pennies / 100.0) >> import unittest >> class TestAddMoney(unittest.TestCase): >> def test_float_errors(self): >> self.failUnlessEqual(add_money([0.13, >> 0.02]), 0.15) >> self.failUnlessEqual(add_money([100.01, >> 99.99]), 200) >> self.failUnlessEqual(add_money([0, >> -13.00, 13.00]), 0) >> if __name__ == '__main__': >> unittest.main()
>> I believe I've still kept it to 13 lines.
> I approve this change, although in a sense, it's > harder for a Python newbie, because it introduces > inheritance a little earlier than I would have liked.
> FWIW I'm in the minority (I think) of people that > prefer roll-your-own testing, but I don't want to > argue that, because I think it mostly comes down to > personal preference.