Python-list Digest, Vol 232, Issue 1
Christman, Roger Graydon
dvl at psu.edu
Sun Jan 1 19:30:51 EST 2023
Re: Nonetype List
In my introductory programming course, I have drawn some attention to this behavior regarding mutating lists. Indeed, Python is very consistent with its behavior:
Any function that mutates a list parameter does not return that list as a return value.
For one thing, there is no need to return that value, because the caller still owns the list parameter that has been modified.
But secondly, (and what I find especially important), is that returning the modified list would lead too many program bugs or misunderstandings.
For example, if append did return the list, you might see this:
x = [1,2,3]
y = x.append(4)
z = y.append(5)
The principal of 'least surprise' would cause a casual reader to believe that x retains the value of [1,2,3], y would have the value of [1,2,3,4], and z would contain [1,2,3,4,5].
So it would be very surprising indeed to discover that x contains [1,2,3,4,5], especially after that statement that makes no reference to x. Since append modifies the list in place, returning that list would make x, y, and z all aliases of each other, and aliasing is a source of many bugs that are very hard to find.
So a recommendation that I make to my class (and which coincides with Python behavior), is to NEVER return a modified list as a return value, but only to return lists that were newly created within the function. So to support this principal of 'least surprise', the append method above would necessarily create new lists for y and z that are not aliases to x. Why Python does not do that is a very obvious cases of run-time efficiency (constant time to append vs. linear to recreate a new list).
And as another observation, I have my students review all of the methods defined for the list object, and they are all very consistent. Most of them either define a return value, or modify the list parameter, but almost none do both. The sole exception is the pop() function that modified a list and returns a value, but that returned value still is not the modified list, so the aliasing problem will never arise.
So, I am very happy with this Python language decision -- it allows for the most efficient means to modify a list in place and also very much reduce the danger of aliasing bugs.
Roger Christman
Pennsylvania State University
________________________________
From: Python-list <python-list-bounces+dvl=psu.edu at python.org> on behalf of python-list-request at python.org <python-list-request at python.org>
Sent: Sunday, January 1, 2023 12:00 PM
To: python-list at python.org <python-list at python.org>
Subject: Python-list Digest, Vol 232, Issue 1
Send Python-list mailing list submissions to
python-list at python.org
To subscribe or unsubscribe via the World Wide Web, visit
https://nam10.safelinks.protection.outlook.com/?url=https%3A%2F%2Fmail.python.org%2Fmailman%2Flistinfo%2Fpython-list&data=05%7C01%7Cdvl%40psu.edu%7C744c83fc485a4b1c79db08daec19a436%7C7cf48d453ddb4389a9c1c115526eb52e%7C0%7C0%7C638081892123929669%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=etYqO01OszhEpqgjLeKQTMC9b3wT0sc%2FcN8oJo9eEhk%3D&reserved=0
or, via email, send a message with subject or body 'help' to
python-list-request at python.org
You can reach the person managing the list at
python-list-owner at python.org
When replying, please edit your Subject line so it is more specific
than "Re: Contents of Python-list digest..."
Today's Topics:
1. Re: NoneType List (Thomas Passin)
2. Re: NoneType List (MRAB)
3. Re: NoneType List (dn)
4. RE: NoneType List (avi.e.gross at gmail.com)
5. Re: NoneType List (Thomas Passin)
6. Re: NoneType List (Greg Ewing)
7. RE: NoneType List (avi.e.gross at gmail.com)
8. Re: NoneType List (Chris Angelico)
9. RE: NoneType List (avi.e.gross at gmail.com)
10. Re: NoneType List (Chris Angelico)
11. Re: NoneType List (Thomas Passin)
----------------------------------------------------------------------
Message: 1
Date: Sat, 31 Dec 2022 12:07:25 -0500
From: Thomas Passin <list1 at tompassin.net>
To: python-list at python.org
Subject: Re: NoneType List
Message-ID: <3eb7480c-88f7-72cf-af66-c0072928bf3f at tompassin.net>
Content-Type: text/plain; charset=UTF-8; format=flowed
Oops, my reply got lost somehow. Here it is:
Everyone's answer to date has been too complicated. What is going on is
that list.append() changes the list in place. It returns nothing. If
you want to append an item and then assign the result to a new list, you
have to do just that:
l1.append(item)
# If we want a *copy* of the appended list:
l2 = l1[:] # Changes to l2 will not change l1
# If we want another name for the appended list:
l2 = l1 # Changes to l2 will change l1 since they are the same object
list.sort() also operates in place. There is a function sorted() that
returns the sorted list (without changing the original list).
The same thing is true of set.add(). The set is changed in place, and
nothing is returned.
On 12/31/2022 10:50 AM, Thomas Passin wrote:
> Happy New Year, everybody!
> I'm new in the Python List, new in Python world, and new in coding.
> A few days (weeks?) ago, I faced a problem trying to write a program for an
> exercise. I asked for help and nobody answered.
> In the meantime, I found a part of the solution, but a part still remains a
> mystery for me. Please run this small snippet, and help.
> Thanks
>
> a = [1, 2]
> print()
>
> print("a ==", a)
> print("type(a) is", type(a))
>
> b = a.append(3)
> print("\nb =", "a.append(3)")
> print("b ==", b)
> print("type(b) is", type(b))
>
> c = ['1', '2']
> print("\nc ==", c)
> print("type(c) is", type(c))
>
> d = c.append('3')
> print("\nd =", "c.append('3')")
> print("d ==", d)
> print("type(d) is", type(d))
>
> """
> I mean: why b = a.append(something) is the None type, and how to make a new
> list
> that contains all the items from a and some new items?
> I wasn't able to solve it in a few hours :(
> Now I know:
> """
>
> crta = '='
> print('\n', 4 * ' ', crta * (len('The solution:')), sep='')
> print(3 * ' ', 'The solution:')
> print(4 * ' ', crta * (len('The solution:')), sep='')
> print('\nThe solution is the slice, an element of Python syntax that
> allows')
> print('to make a brand new copy of a list, or parts of a list.')
> print('The slice actually copies the list\'s contents, not the list\'s
> name:')
>
> print()
> a = [1, 2]
> print("a ==", a)
> print("type(a) is", type(a))
>
> b = a[:]
> print("\nb = a[:]")
> print("b ==", b)
> b.append(3)
> print("\nb =", "b.append(3)")
> print("b ==", b)
> print("type(b) is", type(b))
> print("\na ==", a)
>
> print('But I still don't know why "b = a.append(something)" is the None
> type.')
> print('Is there anybody out there?!')
------------------------------
Message: 2
Date: Sat, 31 Dec 2022 17:37:39 +0000
From: MRAB <python at mrabarnett.plus.com>
To: python-list at python.org
Subject: Re: NoneType List
Message-ID: <2321d9bc-5eb3-c1b9-0425-020b54dd39e8 at mrabarnett.plus.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
On 2022-12-31 05:45, Goran Ikac wrote:
> Happy New Year, everybody!
> I'm new in the Python List, new in Python world, and new in coding.
> A few days (weeks?) ago, I faced a problem trying to write a program for an
> exercise. I asked for help and nobody answered.
> In the meantime, I found a part of the solution, but a part still remains a
> mystery for me. Please run this small snippet, and help.
> Thanks
>
> a = [1, 2]
> print()
>
> print("a ==", a)
> print("type(a) is", type(a))
>
> b = a.append(3)
> print("\nb =", "a.append(3)")
> print("b ==", b)
> print("type(b) is", type(b))
>
> c = ['1', '2']
> print("\nc ==", c)
> print("type(c) is", type(c))
>
> d = c.append('3')
> print("\nd =", "c.append('3')")
> print("d ==", d)
> print("type(d) is", type(d))
>
> """
> I mean: why b = a.append(something) is the None type, and how to make a new
> list
> that contains all the items from a and some new items?
> I wasn't able to solve it in a few hours :(
> Now I know:
> """
>
> crta = '='
> print('\n', 4 * ' ', crta * (len('The solution:')), sep='')
> print(3 * ' ', 'The solution:')
> print(4 * ' ', crta * (len('The solution:')), sep='')
> print('\nThe solution is the slice, an element of Python syntax that
> allows')
> print('to make a brand new copy of a list, or parts of a list.')
> print('The slice actually copies the list\'s contents, not the list\'s
> name:')
>
> print()
> a = [1, 2]
> print("a ==", a)
> print("type(a) is", type(a))
>
> b = a[:]
> print("\nb = a[:]")
> print("b ==", b)
> b.append(3)
> print("\nb =", "b.append(3)")
> print("b ==", b)
> print("type(b) is", type(b))
> print("\na ==", a)
>
> print('But I still don't know why "b = a.append(something)" is the None
> type.')
> print('Is there anybody out there?!')
Methods that modify in-place usually return None. "a.append(something)"
modifies (appends to) the list 'a' and returns None.
If you want to a new line with something at the end try "b = a +
[something]".
------------------------------
Message: 3
Date: Sun, 1 Jan 2023 09:58:35 +1300
From: dn <PythonList at DancesWithMice.info>
To: python-list at python.org
Subject: Re: NoneType List
Message-ID: <2b238dce-3e9f-9bee-27a0-d163cb6c7cef at DancesWithMice.info>
Content-Type: text/plain; charset=UTF-8; format=flowed
On 31/12/2022 18.45, Goran Ikac wrote:
...
> A few days (weeks?) ago, I faced a problem trying to write a program for an
> exercise. I asked for help and nobody answered.
Looking back over the last six months of List-Archives, your name does
not appear against a single post. This may explain why "nobody answered".
However, ten hours after the above/first message, you posted again. This
time as "Thomas Passin". That was followed an hour-or-so later, with a
reply-to-self saying: "Everyone's answer to date has been too
complicated", and then basically repeating the information
previously-provided by a number of contributors.
It then goes on to talk about copying, sorting, adding, and even sets.
Much of which you had earlier said: "I know". Was there a further
question in there?
Which part of which answer did you find "too complicated"?
Did you try the experiments suggested, and read the references-provided?
Please ask a further question detailing what you have understood, and
what is still mystifying. Folk here endeavor to be helpful (see also,
earlier reference to the Python-Tutor list).
When asking a question here, please try to reduce the problem to its
simplest form, so that you're not asking volunteers to read
multiple-screens of irrelevant code. It will help to ask one question at
a time, or to carefully separate multiple questions.
Yes, this can be difficult when one is learning. That said, we all
started somewhere and are happy to help you to become a valued colleague!
--
Regards,
=dn
------------------------------
Message: 4
Date: Sat, 31 Dec 2022 17:36:59 -0500
From: <avi.e.gross at gmail.com>
To: <python-list at python.org>
Subject: RE: NoneType List
Message-ID: <006501d91d68$6538dd10$2faa9730$@gmail.com>
Content-Type: text/plain; charset="us-ascii"
It depends on what people consider too complicated.
I find it a tad complicated when someone posts using two different ID, and
then wonders ...
The question related to taking a list and extending it and using the result
in an assignment statement.
There were several inter-related questions people responded to.
One was asking why something did not work as expected. Several answers
pointed out it was because it was not designed the way expected and reading
the manual page or other reference material would help explain that you need
to use code the way it was intended and not as you wish it. Other aspects of
similar code that do what you expect were shown and you could pick any you
liked.
A second question some were answering is various ways to get the
functionality wanted. Some were simple enough like using "+" and others did
seem complex as they showed many variations on copying an object or how to
make a subclass that effectively substitutes a method that returns the
internally changed object.
And, of course, we had the philosophical question of why the feature was
designed to not return anything (well, NULL) rather than return the changed
object. Returning nothing is arguably slightly more efficient in the many
cases where the return value is simply ignored and thus tossed. But as
mentioned, it would be nice for some purposes, including chaining, to be
able to write something like
result = lst.add_ret("item").sort_ret()
As mentioned, this can still be easily done using something like:
result = sorted(lst + "item")
-----Original Message-----
From: Python-list <python-list-bounces+avi.e.gross=gmail.com at python.org> On
Behalf Of dn
Sent: Saturday, December 31, 2022 3:59 PM
To: python-list at python.org
Subject: Re: NoneType List
On 31/12/2022 18.45, Goran Ikac wrote:
...
> A few days (weeks?) ago, I faced a problem trying to write a program
> for an exercise. I asked for help and nobody answered.
Looking back over the last six months of List-Archives, your name does not
appear against a single post. This may explain why "nobody answered".
However, ten hours after the above/first message, you posted again. This
time as "Thomas Passin". That was followed an hour-or-so later, with a
reply-to-self saying: "Everyone's answer to date has been too complicated",
and then basically repeating the information previously-provided by a number
of contributors.
It then goes on to talk about copying, sorting, adding, and even sets.
Much of which you had earlier said: "I know". Was there a further question
in there?
Which part of which answer did you find "too complicated"?
Did you try the experiments suggested, and read the references-provided?
Please ask a further question detailing what you have understood, and what
is still mystifying. Folk here endeavor to be helpful (see also, earlier
reference to the Python-Tutor list).
When asking a question here, please try to reduce the problem to its
simplest form, so that you're not asking volunteers to read multiple-screens
of irrelevant code. It will help to ask one question at a time, or to
carefully separate multiple questions.
Yes, this can be difficult when one is learning. That said, we all started
somewhere and are happy to help you to become a valued colleague!
--
Regards,
=dn
--
https://nam10.safelinks.protection.outlook.com/?url=https%3A%2F%2Fmail.python.org%2Fmailman%2Flistinfo%2Fpython-list&data=05%7C01%7Cdvl%40psu.edu%7C744c83fc485a4b1c79db08daec19a436%7C7cf48d453ddb4389a9c1c115526eb52e%7C0%7C0%7C638081892123929669%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=etYqO01OszhEpqgjLeKQTMC9b3wT0sc%2FcN8oJo9eEhk%3D&reserved=0
------------------------------
Message: 5
Date: Sat, 31 Dec 2022 18:32:04 -0500
From: Thomas Passin <list1 at tompassin.net>
To: python-list at python.org
Subject: Re: NoneType List
Message-ID: <11625e0f-8975-b929-a2ca-e14ed74a19b0 at tompassin.net>
Content-Type: text/plain; charset=UTF-8; format=flowed
On 12/31/2022 3:58 PM, dn wrote:
> On 31/12/2022 18.45, Goran Ikac wrote:
> ...
>> A few days (weeks?) ago, I faced a problem trying to write a program
>> for an
>> exercise. I asked for help and nobody answered.
>
> Looking back over the last six months of List-Archives, your name does
> not appear against a single post. This may explain why "nobody answered".
>
> However, ten hours after the above/first message, you posted again. This
> time as "Thomas Passin".
That message was probably a mistaken one from me. I had composed a
reply but through some mental glitch had to re-do it. I managed to send
it with the only quoted thread but not the reply I had wanted to
include. So I added my reply and sent it again. It was probably
confusing, and I'm sorry about that.
> That was followed an hour-or-so later, with a
> reply-to-self saying: "Everyone's answer to date has been too
> complicated", and then basically repeating the information
> previously-provided by a number of contributors.
>
> It then goes on to talk about copying, sorting, adding, and even sets.
> Much of which you had earlier said: "I know". Was there a further
> question in there?
>
> Which part of which answer did you find "too complicated"?
>
> Did you try the experiments suggested, and read the references-provided?
>
> Please ask a further question detailing what you have understood, and
> what is still mystifying. Folk here endeavor to be helpful (see also,
> earlier reference to the Python-Tutor list).
>
> When asking a question here, please try to reduce the problem to its
> simplest form, so that you're not asking volunteers to read
> multiple-screens of irrelevant code. It will help to ask one question at
> a time, or to carefully separate multiple questions.
>
> Yes, this can be difficult when one is learning. That said, we all
> started somewhere and are happy to help you to become a valued colleague!
>
------------------------------
Message: 6
Date: Sun, 1 Jan 2023 13:21:05 +1300
From: Greg Ewing <greg.ewing at canterbury.ac.nz>
To: python-list at python.org
Subject: Re: NoneType List
Message-ID: <k1c1vkF15cmU1 at mid.individual.net>
Content-Type: text/plain; charset=UTF-8; format=flowed
On 1/01/23 11:36 am, avi.e.gross at gmail.com wrote:
> And, of course, we had the philosophical question of why the feature was
> designed to not return anything ... rather than return the changed
> object.
My understanding is that Guido designed it that way to keep a
clear separation between mutating and non-mutating methods, and
to help catch mistakes resulting from mixing them up.
--
Greg
------------------------------
Message: 7
Date: Sat, 31 Dec 2022 22:17:56 -0500
From: <avi.e.gross at gmail.com>
To: "'Greg Ewing'" <greg.ewing at canterbury.ac.nz>,
<python-list at python.org>
Subject: RE: NoneType List
Message-ID: <00e601d91d8f$a4e59b20$eeb0d160$@gmail.com>
Content-Type: text/plain; charset="us-ascii"
Agreed, there are lots of pro/con arguments and the feature is what it is
historically and not trivial to change. Inline changes to an object make
sense to just be done "silently" and if there are errors, they propagate the
usual way.
As Guido was a major influence at that time, one view was seen as more
important and prevailed.
Had a language like that been created today, I wonder if some designs might
have looked a bit different so that some functions could be called with
optional arguments that specified what the user wanted returned.
In particular, besides a function where you add a value returning
nothing/None, there may be room for other choices and any choice hard-wired
in would eleicit complaints from others.
lst.add("Value")
could also return one of several other things such as a pointer to the
upgraded object but also of a pointer to the object as it looked before the
change (not a serious proposal) or a True/False value specifying if the
change was able to be completed (such as running out of memory, or the
addition not being something you can put in the object) or even return what
was added or how many objects are now in the object at the top level, or how
many times the method was called so far!
I suspect, at the expense of some overhead, you could just add an argument
to many kinds of methods or even functions such as returning='option' that
guides what you want returned, with the default often being what Guido and
others currently have set.
Python already allows functions to return anything they feel like, so this
probably would not break many things.
Of course there are other paths in that direction, such as setting an
attribute of the list/object first that affects how things get returned but
that seems more cumbersome and makes all kinds of errors more likely. Still,
that is a path often used by some Python modules where objects are created
and then tweaked to behave in various ways when later methods are invoked.
But is any of it needed? The current system generally works fine and we have
seen many ways to get other results without tampering with the current
implementation.
This may be yet another example of people who come to python with
pre-existing bias, such as insisting it work like another language they have
used, or wanting the computer to do what they MEANT rather than what they
explicitly or implicitly programmed!
-----Original Message-----
From: Python-list <python-list-bounces+avi.e.gross=gmail.com at python.org> On
Behalf Of Greg Ewing
Sent: Saturday, December 31, 2022 7:21 PM
To: python-list at python.org
Subject: Re: NoneType List
On 1/01/23 11:36 am, avi.e.gross at gmail.com wrote:
> And, of course, we had the philosophical question of why the feature
> was designed to not return anything ... rather than return the changed
> object.
My understanding is that Guido designed it that way to keep a clear
separation between mutating and non-mutating methods, and to help catch
mistakes resulting from mixing them up.
--
Greg
--
https://nam10.safelinks.protection.outlook.com/?url=https%3A%2F%2Fmail.python.org%2Fmailman%2Flistinfo%2Fpython-list&data=05%7C01%7Cdvl%40psu.edu%7C744c83fc485a4b1c79db08daec19a436%7C7cf48d453ddb4389a9c1c115526eb52e%7C0%7C0%7C638081892123929669%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=etYqO01OszhEpqgjLeKQTMC9b3wT0sc%2FcN8oJo9eEhk%3D&reserved=0
------------------------------
Message: 8
Date: Sun, 1 Jan 2023 14:23:13 +1100
From: Chris Angelico <rosuav at gmail.com>
To: python-list at python.org
Subject: Re: NoneType List
Message-ID:
<CAPTjJmq=n0=B7fHvtydTwe0eZ+GVxTzGH-3TJDX6E9PyVHsEnQ at mail.gmail.com>
Content-Type: text/plain; charset="UTF-8"
On Sun, 1 Jan 2023 at 14:19, <avi.e.gross at gmail.com> wrote:
> Had a language like that been created today, I wonder if some designs might
> have looked a bit different so that some functions could be called with
> optional arguments that specified what the user wanted returned.
Frankly, I doubt it. While you can argue "returning self is more
useful" vs "returning None is more clear if you get it wrong", having
options does NOT improve things, and just makes everything more
complicated, slower, harder to comprehend, and generally worse to work
with.
A language should have some sort of consistent behaviour and stick to
it. If that's not possible, an object type should at least have that.
ChrisA
------------------------------
Message: 9
Date: Sat, 31 Dec 2022 23:16:10 -0500
From: <avi.e.gross at gmail.com>
To: "'Chris Angelico'" <rosuav at gmail.com>, <python-list at python.org>
Subject: RE: NoneType List
Message-ID: <00fd01d91d97$c78994d0$569cbe70$@gmail.com>
Content-Type: text/plain; charset="us-ascii"
Chris,
There is much to say about consistent behavior as compared to flexibility
and convenience.
I have seen other languages provide functions for example, where the result
can vary and often cause confusion. R had a function that would sometimes
notice the result could be simplified and return that. Guess what? It caused
lots of problems and an option was added that said it should NOT simplify
and always return the same kind of thing.
Consider what happens if a calculation that returned a matrix would decide
that is there was only one columns, it would return a vector/array/whatever
as a one-dimensional result and if the calculation produced a 1 by 1 matrix,
it would simply return a scalar value!
But it may be a feature you want in some cases, albeit rarely when the
results of the function are fed into something that is not expecting it.
Various forms of polymorphism can be helpful but can also be very confusing.
Some solutions are what I have described in earlier messages and some
languages blur distinctions. Again, in R, there may not be a scalar as all
you have is a vector of length one. But matrices even of 1-D are not vectors
and I have had to interconvert between them to make some code run.
I am not making comparisons in the sense that nothing other languages choose
to do is binding on what Python should do. Still, I think it is wrong to
suggest that it does not often do partially ambiguous things from some
perspective. Many functions will take a variety of arguments and return
something reasonable but different each time.
As a dumb example, what does a simple function like max() return if fed just
integers, just floating point or a combination of both? It seems to return
whatever type the maximum indicates. But it also accepts characters and
sorts them appropriately returning type 'str' and probably accepts and
returns all kinds of objects if you specify a key function.
Is that so much different than we are discussing in that there isn't any
absolute consistency and things can go various ways in terms of return
value? Heck, I can even make max() return None!
-----Original Message-----
From: Python-list <python-list-bounces+avi.e.gross=gmail.com at python.org> On
Behalf Of Chris Angelico
Sent: Saturday, December 31, 2022 10:23 PM
To: python-list at python.org
Subject: Re: NoneType List
On Sun, 1 Jan 2023 at 14:19, <avi.e.gross at gmail.com> wrote:
> Had a language like that been created today, I wonder if some designs
> might have looked a bit different so that some functions could be
> called with optional arguments that specified what the user wanted
returned.
Frankly, I doubt it. While you can argue "returning self is more useful" vs
"returning None is more clear if you get it wrong", having options does NOT
improve things, and just makes everything more complicated, slower, harder
to comprehend, and generally worse to work with.
A language should have some sort of consistent behaviour and stick to it. If
that's not possible, an object type should at least have that.
ChrisA
--
https://nam10.safelinks.protection.outlook.com/?url=https%3A%2F%2Fmail.python.org%2Fmailman%2Flistinfo%2Fpython-list&data=05%7C01%7Cdvl%40psu.edu%7C744c83fc485a4b1c79db08daec19a436%7C7cf48d453ddb4389a9c1c115526eb52e%7C0%7C0%7C638081892123929669%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=etYqO01OszhEpqgjLeKQTMC9b3wT0sc%2FcN8oJo9eEhk%3D&reserved=0
------------------------------
Message: 10
Date: Sun, 1 Jan 2023 15:23:59 +1100
From: Chris Angelico <rosuav at gmail.com>
To: python-list at python.org
Subject: Re: NoneType List
Message-ID:
<CAPTjJmqqGLqE=FepRMM1P95E3eGv5Qnq_=BfsGJNXuucLH6wMg at mail.gmail.com>
Content-Type: text/plain; charset="UTF-8"
On Sun, 1 Jan 2023 at 15:16, <avi.e.gross at gmail.com> wrote:
>
> Chris,
>
> There is much to say about consistent behavior as compared to flexibility
> and convenience.
>
> I have seen other languages provide functions for example, where the result
> can vary and often cause confusion. R had a function that would sometimes
> notice the result could be simplified and return that. Guess what? It caused
> lots of problems and an option was added that said it should NOT simplify
> and always return the same kind of thing.
The option was presumably added because backward compatibility is
important, but it would have been better to not need the option in the
first place.
> As a dumb example, what does a simple function like max() return if fed just
> integers, just floating point or a combination of both? It seems to return
> whatever type the maximum indicates. But it also accepts characters and
> sorts them appropriately returning type 'str' and probably accepts and
> returns all kinds of objects if you specify a key function.
>
> Is that so much different than we are discussing in that there isn't any
> absolute consistency and things can go various ways in terms of return
> value? Heck, I can even make max() return None!
>
... yes? So? It still always returns the largest item in the list, for
some definition of "largest". You'll never have that value returned
wrapped up in a single-element list, though.
ChrisA
------------------------------
Message: 11
Date: Sun, 1 Jan 2023 08:11:02 -0500
From: Thomas Passin <list1 at tompassin.net>
To: python-list at python.org
Subject: Re: NoneType List
Message-ID: <71caabcd-65a0-8b3f-a971-79cd3f4415e1 at tompassin.net>
Content-Type: text/plain; charset=UTF-8; format=flowed
On 12/31/2022 10:17 PM, avi.e.gross at gmail.com wrote:
> Agreed, there are lots of pro/con arguments and the feature is what it is
> historically and not trivial to change. Inline changes to an object make
> sense to just be done "silently" and if there are errors, they propagate the
> usual way.
>
> As Guido was a major influence at that time, one view was seen as more
> important and prevailed.
>
> Had a language like that been created today, I wonder if some designs might
> have looked a bit different so that some functions could be called with
> optional arguments that specified what the user wanted returned.
Guido had been working on the ABC language for some years before he
developed Python. ABC was intended mainly as a teaching and prototyping
language. Guido probably had a good sense of what things worked well
and what didn't in that usage. IIRC, Python did not originally have
classes or instances, and a "fluent" style of programming probably
wasn't in use yet. Having an object return itself after an operation is
useful (mostly, perhaps) for fluent programming (that is, the style
where one can write X.method1().method2().method3() ...).
> In particular, besides a function where you add a value returning
> nothing/None, there may be room for other choices and any choice hard-wired
> in would eleicit complaints from others.
>
> lst.add("Value")
>
> could also return one of several other things such as a pointer to the
> upgraded object but also of a pointer to the object as it looked before the
> change (not a serious proposal) or a True/False value specifying if the
> change was able to be completed (such as running out of memory, or the
> addition not being something you can put in the object) or even return what
> was added or how many objects are now in the object at the top level, or how
> many times the method was called so far!
>
> I suspect, at the expense of some overhead, you could just add an argument
> to many kinds of methods or even functions such as returning='option' that
> guides what you want returned, with the default often being what Guido and
> others currently have set.
>
> Python already allows functions to return anything they feel like, so this
> probably would not break many things.
>
> Of course there are other paths in that direction, such as setting an
> attribute of the list/object first that affects how things get returned but
> that seems more cumbersome and makes all kinds of errors more likely. Still,
> that is a path often used by some Python modules where objects are created
> and then tweaked to behave in various ways when later methods are invoked.
>
> But is any of it needed? The current system generally works fine and we have
> seen many ways to get other results without tampering with the current
> implementation.
>
> This may be yet another example of people who come to python with
> pre-existing bias, such as insisting it work like another language they have
> used, or wanting the computer to do what they MEANT rather than what they
> explicitly or implicitly programmed!
>
>
> -----Original Message-----
> From: Python-list <python-list-bounces+avi.e.gross=gmail.com at python.org> On
> Behalf Of Greg Ewing
> Sent: Saturday, December 31, 2022 7:21 PM
> To: python-list at python.org
> Subject: Re: NoneType List
>
> On 1/01/23 11:36 am, avi.e.gross at gmail.com wrote:
>> And, of course, we had the philosophical question of why the feature
>> was designed to not return anything ... rather than return the changed
>> object.
>
> My understanding is that Guido designed it that way to keep a clear
> separation between mutating and non-mutating methods, and to help catch
> mistakes resulting from mixing them up.
>
> --
> Greg
> --
> https://nam10.safelinks.protection.outlook.com/?url=https%3A%2F%2Fmail.python.org%2Fmailman%2Flistinfo%2Fpython-list&data=05%7C01%7Cdvl%40psu.edu%7C744c83fc485a4b1c79db08daec19a436%7C7cf48d453ddb4389a9c1c115526eb52e%7C0%7C0%7C638081892123929669%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=etYqO01OszhEpqgjLeKQTMC9b3wT0sc%2FcN8oJo9eEhk%3D&reserved=0
>
------------------------------
Subject: Digest Footer
--
https://nam10.safelinks.protection.outlook.com/?url=https%3A%2F%2Fmail.python.org%2Fmailman%2Flistinfo%2Fpython-list&data=05%7C01%7Cdvl%40psu.edu%7C744c83fc485a4b1c79db08daec19a436%7C7cf48d453ddb4389a9c1c115526eb52e%7C0%7C0%7C638081892123929669%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=etYqO01OszhEpqgjLeKQTMC9b3wT0sc%2FcN8oJo9eEhk%3D&reserved=0
------------------------------
End of Python-list Digest, Vol 232, Issue 1
*******************************************
More information about the Python-list
mailing list