Replace deprecated functions from string module

Replace deprecated functions from string module

These methods have been deprecated since at least Python 2.4 and their
replacements are available since at least that version.

Individual changes:
 * Replace string.atoi() by int()
 * Replace string.split() by str.split()
 * Replace string.join() by sep.join()
 * Replace string.strip() by str.strip()
 * Replace string.replace() by str.replace()
 * Drop now unused 'import string'

[Malte Meyn, 2019-10-23 21:22:16]

For someone who doesn’t know python two questions come up:

1. Why do you sometimes use " " and sometimes ' '?
2. Is a space (" " or ' ') the default for the old string.join? Or should you
use an empty string ("".join()) instead in some places of the new version?

https://codereview.appspot.com/566920044/diff/549120043/python/musicexp.py
File python/musicexp.py (right):

https://codereview.appspot.com/566920044/diff/549120043/python/musicexp.py#ne...
python/musicexp.py:694: % " ".join (map (lambda x: x.lisp_expression(),
Here you use " " …

https://codereview.appspot.com/566920044/diff/549120043/python/musicexp.py#ne...
python/musicexp.py:1040: printer ('<%s>' % ' '.join (pitches))
… and here ' '.

[Malte Meyn, 2019-10-23 21:29:27]

On 2019/10/23 21:22:16, Malte Meyn wrote:
> 2. Is a space (" " or ' ') the default for the old string.join? Or should you
> use an empty string ("".join()) instead in some places of the new version?

Ok, I found an answer to that question myself. The default is a space indeed.
Weird. https://docs.python.org/2/library/string.html

[hahnjo, 2019-10-24 06:33:50]

On 2019/10/23 21:22:16, Malte Meyn wrote:
> For someone who doesn’t know python two questions come up:
> 
> 1. Why do you sometimes use " " and sometimes ' '?

AFAIK there is no difference between the two ways of writing a constant string
(or not relevant here), so I've tried to stay consistent with the surrounding
code.

> 2. Is a space (" " or ' ') the default for the old string.join? Or should you
> use an empty string ("".join()) instead in some places of the new version?

A space is the old default. I quoted the relevant documentation in the commit
message of the separate patches I posted in the issue, but unfortunately this is
lost when uploading for review here :-(

[dak, 2019-10-24 07:36:50]

jonas.hahnfeld@gmail.com writes:

> Reviewers: Malte Meyn,
>
> Message:
> On 2019/10/23 21:22:16, Malte Meyn wrote:
>> For someone who doesn’t know python two questions come up:
>
>> 1. Why do you sometimes use " " and sometimes ' '?
>
> AFAIK there is no difference between the two ways of writing a constant
> string (or not relevant here), so I've tried to stay consistent with the
> surrounding code.
>
>> 2. Is a space (" " or ' ') the default for the old string.join? Or
> should you
>> use an empty string ("".join()) instead in some places of the new
> version?
>
> A space is the old default. I quoted the relevant documentation in the
> commit message of the separate patches I posted in the issue, but
> unfortunately this is lost when uploading for review here :-(

git cl upload  allows you to edit the message.  You can edit in a
suitably edited version of git log (typically I put something like
Contains commits:
at the end of the principal text and then a lightly edited list of
reverse commit messages)

Tastes differ, but one has the opportunity to give more information.

-- 
David Kastrup

[hahnjo, 2019-10-24 07:40:30]

On 2019/10/24 07:36:50, dak wrote:
> mailto:jonas.hahnfeld@gmail.com writes:
> 
> > Reviewers: Malte Meyn,
> >
> > Message:
> > On 2019/10/23 21:22:16, Malte Meyn wrote:
> >> For someone who doesn’t know python two questions come up:
> >
> >> 1. Why do you sometimes use " " and sometimes ' '?
> >
> > AFAIK there is no difference between the two ways of writing a constant
> > string (or not relevant here), so I've tried to stay consistent with the
> > surrounding code.
> >
> >> 2. Is a space (" " or ' ') the default for the old string.join? Or
> > should you
> >> use an empty string ("".join()) instead in some places of the new
> > version?
> >
> > A space is the old default. I quoted the relevant documentation in the
> > commit message of the separate patches I posted in the issue, but
> > unfortunately this is lost when uploading for review here :-(
> 
> git cl upload  allows you to edit the message.  You can edit in a
> suitably edited version of git log (typically I put something like
> Contains commits:
> at the end of the principal text and then a lightly edited list of
> reverse commit messages)
> 
> Tastes differ, but one has the opportunity to give more information.

Yeah, messed this up I guess. I had expected that I only entered the cover
description and that the separate commits would be uploaded as sub-revisions or
something. Will pay more attention in the future.

[dak, 2019-10-24 11:30:21]

jonas.hahnfeld@gmail.com writes:

> On 2019/10/24 07:36:50, dak wrote:
>
>> git cl upload allows you to edit the message.  You can edit in a
>> suitably edited version of git log (typically I put something like
>> Contains commits: at the end of the principal text and then a lightly
>> edited list of reverse commit messages)
>
>> Tastes differ, but one has the opportunity to give more information.
>
> Yeah, messed this up I guess. I had expected that I only entered the
> cover description and that the separate commits would be uploaded as
> sub-revisions or something. Will pay more attention in the future.

That would be a reasonable expectation given a Git-centric project tool.
But we are working with some Google tools on a code review site created
for Subversion and have migrated the Google code part more or less to
SourceForge as a stopgap measure before managing self-hosting on the
Free version of its software, a step that we never managed due to a lack
of manpower and ongoing resource needs.  All the while having the
repository itself on Savannah.

So it turns out that a few things that would be obvious candidates for
automation just aren't up to scratch.

On the plus side, "messed this up" seems way overblown for "didn't score
a perfect 10 in a quagmire I got into for the first time".  Our
procedures tend to avoid annoying more than one person at a time, so
it's pretty uncommon that tempers flare up.

> https://codereview.appspot.com/566920044/

-- 
David Kastrup

[dak, 2020-01-12 01:13:51]

https://codereview.appspot.com/566920044/diff/549120043/python/convertrules.py
File python/convertrules.py (right):

https://codereview.appspot.com/566920044/diff/549120043/python/convertrules.p...
python/convertrules.py:3898: mods = str.join (re.findall ("\n(" + wordsyntax +
r")\s*=\s*\\with(?:\s|\\|\{)")
  File
"/usr/local/tmp/lilypond/out/lib/lilypond/current/python/convertrules.py", line
3892, in conv
    mods = str.join (re.findall ("\n(" + wordsyntax +
r")\s*=\s*\\with(?:\s|\\|\{)")
TypeError: findall() takes at least 2 arguments (1 given)

I'd call that a case of insufficient testing...  Sometimes our tests don't cover
everything and one has to check out things on one's own.

[dak, 2020-01-12 02:19:20]

On 2020/01/12 01:13:51, dak wrote:
> https://codereview.appspot.com/566920044/diff/549120043/python/convertrules.py
> File python/convertrules.py (right):
> 
>
https://codereview.appspot.com/566920044/diff/549120043/python/convertrules.p...
> python/convertrules.py:3898: mods = str.join (re.findall ("\n(" + wordsyntax +
> r")\s*=\s*\\with(?:\s|\\|\{)")
>   File
> "/usr/local/tmp/lilypond/out/lib/lilypond/current/python/convertrules.py",
line
> 3892, in conv
>     mods = str.join (re.findall ("\n(" + wordsyntax +
> r")\s*=\s*\\with(?:\s|\\|\{)")
> TypeError: findall() takes at least 2 arguments (1 given)
> 
> I'd call that a case of insufficient testing...  Sometimes our tests don't
cover
> everything and one has to check out things on one's own.

Just pushed a fix to staging.

[hahnjo, 2020-01-12 10:24:05]

On 2020/01/12 02:19:20, dak wrote:
> On 2020/01/12 01:13:51, dak wrote:
> >
https://codereview.appspot.com/566920044/diff/549120043/python/convertrules.py
> > File python/convertrules.py (right):
> > 
> >
>
https://codereview.appspot.com/566920044/diff/549120043/python/convertrules.p...
> > python/convertrules.py:3898: mods = str.join (re.findall ("\n(" + wordsyntax
+
> > r")\s*=\s*\\with(?:\s|\\|\{)")
> >   File
> > "/usr/local/tmp/lilypond/out/lib/lilypond/current/python/convertrules.py",
> line
> > 3892, in conv
> >     mods = str.join (re.findall ("\n(" + wordsyntax +
> > r")\s*=\s*\\with(?:\s|\\|\{)")
> > TypeError: findall() takes at least 2 arguments (1 given)
> > 
> > I'd call that a case of insufficient testing...  Sometimes our tests don't
> cover
> > everything and one has to check out things on one's own.
> 
> Just pushed a fix to staging.

Whoops, thanks for handling this!