From Newsgroup: comp.lang.python

Did something change in python buggering up my use of a "\ " sequence in a triple quoted string?

I have yet to go through my archive on the program, but I tried to run it
today and it crashed quite spectacularly when it hit a """ .... """ line
being used as a comment at the top of a function. I changed the "\" to a
"/" and all is well now.
--

**** Listen to my FREE CD at http://www.mellowood.ca/music/cedars ****
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bob@mellowood.ca
WWW: http://www.mellowood.ca
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python

I think it could be this:
A backslash-character pair that is not a valid escape sequence now
generates a SyntaxWarning, instead of DeprecationWarning. For example, re.compile("\d+\.\d+") now emits a SyntaxWarning ("\d" is an invalid
escape sequence, use raw strings for regular expression: re.compile(r"\d+\.\d+")). In a future Python version, SyntaxError will eventually be raised, instead of SyntaxWarning. (Contributed by Victor
Stinner in gh-98401.)
Found in:
https://docs.python.org/3/whatsnew/3.12.html#other-language-changes
It's not supposed to crash your program though. If the program crashes
because of it, it's a bug in Python.
On Thu, May 8, 2025 at 7:00 AM Bob van der Poel via Python-list <python-list@python.org> wrote:

Did something change in python buggering up my use of a "\ " sequence in a triple quoted string?

I have yet to go through my archive on the program, but I tried to run it today and it crashed quite spectacularly when it hit a """ .... """ line being used as a comment at the top of a function. I changed the "\" to a
"/" and all is well now.

--

**** Listen to my FREE CD at http://www.mellowood.ca/music/cedars ****
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bob@mellowood.ca
WWW: http://www.mellowood.ca
--
https://mail.python.org/mailman/listinfo/python-list

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python

Also, it appears that the change linked above is a lie: https://docs.python.org/3/reference/lexical_analysis.html#grammar-token-python-grammar-longstringitem
According to the grammar, any character can follow backslash in a
valid Python program. The warning / error raised by this code should
not be a syntax error / warning because the syntax is correct.
On Thu, May 8, 2025 at 7:54 AM Left Right <olegsivokon@gmail.com> wrote:

I think it could be this:

A backslash-character pair that is not a valid escape sequence now
generates a SyntaxWarning, instead of DeprecationWarning. For example, re.compile("\d+\.\d+") now emits a SyntaxWarning ("\d" is an invalid
escape sequence, use raw strings for regular expression: re.compile(r"\d+\.\d+")). In a future Python version, SyntaxError will eventually be raised, instead of SyntaxWarning. (Contributed by Victor Stinner in gh-98401.)

Found in:
https://docs.python.org/3/whatsnew/3.12.html#other-language-changes

It's not supposed to crash your program though. If the program crashes because of it, it's a bug in Python.

On Thu, May 8, 2025 at 7:00 AM Bob van der Poel via Python-list <python-list@python.org> wrote:

Did something change in python buggering up my use of a "\ " sequence in a triple quoted string?

I have yet to go through my archive on the program, but I tried to run it today and it crashed quite spectacularly when it hit a """ .... """ line being used as a comment at the top of a function. I changed the "\" to a "/" and all is well now.

--

**** Listen to my FREE CD at http://www.mellowood.ca/music/cedars ****
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bob@mellowood.ca
WWW: http://www.mellowood.ca
--
https://mail.python.org/mailman/listinfo/python-list

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python

Left Right <olegsivokon@gmail.com> wrote or quoted:

Also, it appears that the change linked above is a lie:

The Python documentation is a bit, "informal".

A "more formal specification" can be found in, maybe (if I found
the right place), the function definition "_PyBytes_DecodeEscape"
in the source file "Objects\bytesobject.c":

|switch (*s++) {
|/* XXX This assumes ASCII! */
|case '\n': break;
|case '\\': *p++ = '\\'; break;
|case '\'': *p++ = '\''; break;
|case '\"': *p++ = '\"'; break;
. . .

. In 2015, Jakub Przywoski started to write a new specification:

|Python is such a well-designed, clean and enjoyable to code in
|language so it sure deserves to have a decent syntax reference.

. It seems, however, that he was not yet able to finish his project
or to update it to newer versions of Python. It would be a full-time
project I think!

Effectively, Python is what CPython does.


--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python

ram@zedat.fu-berlin.de (Stefan Ram) wrote or quoted:

A "more formal specification" can be found in, maybe (if I found
the right place), the function definition "_PyBytes_DecodeEscape"
in the source file "Objects\bytesobject.c":

See also:

Grammar/python.gram
Grammar/Tokens

in the CPython sources, which does not seem to speak about
escapes in string literals, though.


--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python


--l6dx3oqrzzts6slg
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On 2025-05-08 08:05:54 +0200, Left Right via Python-list wrote:

Also, it appears that the change linked above is a lie:

Such strong words ...

https://docs.python.org/3/reference/lexical_analysis.html#grammar-token-p=

ython-grammar-longstringitem

=20
According to the grammar, any character can follow backslash in a
valid Python program. The warning / error raised by this code should
not be a syntax error / warning because the syntax is correct.

Warnings are about technically correct but probably unintended usage.

The documentation you linked to describes (a bit further down) which
escape sequences are recognized and what happens if you use an
unrecognized escape sequence. It also mentions that using an
unrecognized escape sequence *will* be an error in future versions of
Python.

A warning is appropriate here. It gives the programmer a chance to fix
the program now before it breaks.

One could argue that it should say 'unrecognized escape sequence'
instead of 'invalid escape sequence', since it isn't invalid yet, but
that's nitpicking.

hjp

--=20
_ | Peter J. Holzer | Story must make more sense than reality.
|_|_) | |
| | | hjp@hjp.at | -- Charles Stross, "Creative writing
__/ | http://www.hjp.at/ | challenge!"

--l6dx3oqrzzts6slg
Content-Type: application/pgp-signature; name="signature.asc"

-----BEGIN PGP SIGNATURE-----

iQIzBAABCgAdFiEETtJbRjyPwVTYGJ5k8g5IURL+KF0FAmggdNsACgkQ8g5IURL+ KF3CiQ/+M1YsR/nTAzRyauQ2RiUTR2hw3J3DFSpXSQhsvKvS7S1uPL1+kkBsoUCC 623bVF+5yrPsjAOJVm0S2ieT0465zvnBzDIqcTvKoNmHbAxc3HNEEUyORZ3PGLzN yG7TyUqTLgoHaMGbxEpPfZLQ3DLrslXVKX9JPpMoeDu5jaGah2fv1Y4opCU62Akj QlQP/9rj9k67CGPSr2ndnDu9fqL6v79p57IUC9/SVKOCCaYTG9Wz+LFIyvybP8qp nEfwU+qZw2tKKEn8r78JlP7WkpqBFV2vtmuvFCFsTYLPD7JHihyq9J8fiG1OKy5V t2EPCcFiXshcIUxgiWoEwhBD8NsuqLsvg9Z3z96BzJ8MDB6S6dbp0sLFOdpQiSPT xL7QRB3LvG8oyW63JNCkhuDhKhAf2yPCfezVAiZfcF6Ef3s+o+dv+L0alEXFTpJ1 YFVMlRv5+n/NLAZP7zdpEiNoRlZceTC955MznS40eK7Zc5QK99tzX1BYEBcTNeNR VmNmYma/a4FRo2mFiPFeAyTlDOOCw7G7a3NvRmchHrqWXTZ+WoPmDq/e1DeuDXgi 69Piv0YvwJ2BrN7Dj30R9DmC1Y8K5swg9aUJJNQw3Hamvzybwk94rhBFn1piViQv YAKAEH7NdGYssMgxEbMcqmCpavTVVen7ZxH4WNQ2BIRhs4iKvdY=
=k/a9
-----END PGP SIGNATURE-----

--l6dx3oqrzzts6slg--
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python

Then it just means that the grammar lies. The two claims are mutually exclusive, so either one is a lie or the other or both.
My comment was more of an irony really. It's plenty obvious that the
grammar is a lie. The reason is that it's tedious to put the actual
intender rules into the grammar, and so whoever wrote the grammar
decided to cut corners. But, the grammar is supposed to be the
authoritative source for how the language is parsed, that's why even
though it's clear that the grammar is a lie, blaming whoever doesn't
follow it makes it ironic.
In other words, the grammar author didn't put enough effort into
making grammar actually work, but seeing how many other things are
done in Python, this is not an exception. It would've been strange to
have it done properly when "properly" means doing copious amounts of
tedious work.
On Sun, May 11, 2025 at 12:11 PM Peter J. Holzer via Python-list <python-list@python.org> wrote:

On 2025-05-08 08:05:54 +0200, Left Right via Python-list wrote:

Also, it appears that the change linked above is a lie:

Such strong words ...

https://docs.python.org/3/reference/lexical_analysis.html#grammar-token-python-grammar-longstringitem

According to the grammar, any character can follow backslash in a
valid Python program. The warning / error raised by this code should
not be a syntax error / warning because the syntax is correct.

Warnings are about technically correct but probably unintended usage.

The documentation you linked to describes (a bit further down) which
escape sequences are recognized and what happens if you use an
unrecognized escape sequence. It also mentions that using an
unrecognized escape sequence *will* be an error in future versions of
Python.

A warning is appropriate here. It gives the programmer a chance to fix
the program now before it breaks.

One could argue that it should say 'unrecognized escape sequence'
instead of 'invalid escape sequence', since it isn't invalid yet, but
that's nitpicking.

hjp

--
_ | Peter J. Holzer | Story must make more sense than reality.
|_|_) | |
| | | hjp@hjp.at | -- Charles Stross, "Creative writing
__/ | http://www.hjp.at/ | challenge!"
--
https://mail.python.org/mailman/listinfo/python-list

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python

On Sun, 11 May 2025 at 20:38, Left Right via Python-list <python-list@python.org> wrote:

My comment was more of an irony really. It's plenty obvious that the
grammar is a lie. The reason is that it's tedious to put the actual
intender rules into the grammar, and so whoever wrote the grammar
decided to cut corners. But, the grammar is supposed to be the
authoritative source for how the language is parsed, that's why even
though it's clear that the grammar is a lie, blaming whoever doesn't
follow it makes it ironic.

Have you ever built a language parser? I'm going to guess you haven't.

ChrisA
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python



On 05/11/2025 6:36 AM EDT Left Right via Python-list
<[1]python-list@python.org> wrote:


Then it just means that the grammar lies. The two claims are mutually
exclusive, so either one is a lie or the other or both.



No, it more points out that not all errors are grammatical. The grammar
does not (and can not) fully define what is a legal program. Some forms of
error are semantic, like undefined symbols.

It appears that rather than try to make the grammar complicated enough to
describe what is a valid string, that operation was moved into the
semantics of a string, which simplifies the rules quite a bit.

References

Visible links
1. mailto:python-list@python.org
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python


--7esxoziucknnix62
Content-Type: text/plain; charset=iso-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On 2025-05-11 12:36:31 +0200, Left Right via Python-list wrote:

Then it just means that the grammar lies.

No, because the parser accepts the sequence. And it produces exactly
what the description says.

The program

#!/usr/bin/python3

print("start")
for i in range(3):
print("\copy")
print("end")

is valid and prints

start
\copy
\copy
\copy
end

to stdout, as you would expect from reading the docs.

But in addition to that it also prints the *warning* (not error, that's
a difference)

/home/hjp/tmp/./foo:5: SyntaxWarning: invalid escape sequence '\c'
print("\copy")

to stderr (this is also documented).

One can quibble over the exact wording of the warning ("unrecognized"
would be more in line with the documentation than "invalid"), but there
are reasons for the warning.

The two claims are mutually exclusive, so either one is a lie or the
other or both.

No they are not. The grammar says that the program will compile and run,
which it does. The warning tells you that this probably wasn't a good
idea.

Warnings about stuff that is technically correct aren't unusual.
For example, in C something like=20

#v+
int a =3D 5, b =3D 3;
if (a =3D b) {
printf("a =3D %d\n", a);
}
#v-

is totally legal (and will print "a =3D 3"). However, most compilers will
warn about this because it is very likely that the programmer wanted to
write =ABif (a =3D=3D b) ...=BB.

My comment was more of an irony really. It's plenty obvious that the
grammar is a lie. The reason is that it's tedious to put the actual
intender rules into the grammar, and so whoever wrote the grammar
decided to cut corners. But, the grammar is supposed to be the
authoritative source for how the language is parsed,

Which it is.

that's why even though it's clear that the grammar is a lie, blaming
whoever doesn't follow it makes it ironic.
=20
In other words, the grammar author didn't put enough effort into
making grammar actually work,

But the grammar *does* describe how it actually works. You *can* write
any character after a backslash, and your programm will compile and run.

It is likely that in some future version of Python that will not be the
case any more. *Then* the grammar must be changed. Until then the
documentation should match the current implementation and not something
which may or may not be implemented at some point in the future.

but seeing how many other things are done in Python, this is not an exception. It would've been strange to have it done properly when
"properly" means doing copious amounts of tedious work.

As usual you're making up your lack of knowledge with an abundance of
opinion.

hjp

--=20
_ | Peter J. Holzer | Story must make more sense than reality.
|_|_) | |
| | | hjp@hjp.at | -- Charles Stross, "Creative writing
__/ | http://www.hjp.at/ | challenge!"

--7esxoziucknnix62
Content-Type: application/pgp-signature; name="signature.asc"

-----BEGIN PGP SIGNATURE-----

iQIzBAABCgAdFiEETtJbRjyPwVTYGJ5k8g5IURL+KF0FAmggvpUACgkQ8g5IURL+ KF3XhA//WCebtFiblmx70chiX8tPHbM6bGbkHGK53Gejjn/QCmObeyorApVjMZRN bKuRQrbCnffEXFKC9A6mKV2LqFwqB7aEUQVSZZC9/hyTSUgsYWwowAgPrDzkvNHm vgqLQHFKwHTs+6ysvbf7PiiVTT6w9//3YWhDh30KT/gbppLzXnXWPsLLM111CheF j5Q+vlMZJSKgxdZQjPxBUN7079j2ajuDrwoCgspnGF9MlcnG+37sQ/jDFjlP1gMU Omr9HncQZ3gjjSytU2nvAcn406aufnawy3UamC67DYlUXJotu3UgLYJ/8DwFKDCZ 5XJzd86KxVxKvZXjlNkpOdF2lpnj40IB49ALcfUkiI/cnmf7DBvLbGT346awGfYo vCJPFCWbQgA/D1P8i0fHe65/sXpq5SDtEXADksmGpIh28MBgGCPHjjDobqMwiZ13 xsVYIY+Lr4hKTtzsv3R4OxbKpQFg1CRPr9hZd4rWlQMcXuw1h3To0uaGaxP6NnM6 JcOhU7odDs314E1taG189hgBYHwPjP+DTrEfIiMky59IQcKQfSvoE60vc2JK/ncg eHD+hED4ADJlNDdH9EQCqAw4xeTMYr760MoOmZICXsDQsDmLlhTqxWZRC403I+Wx KzkLYNHvLBMOlIZLjK0LoHShbrnNrvZQyhx1JmlA+DKc4nGac+c=
=Sw2c
-----END PGP SIGNATURE-----

--7esxoziucknnix62--
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python

Have you ever built a language parser?

I've lost count by now. Probably fewer than hundred times though.
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python

Hahah... what a pile of rubbish. The point is that the error is wrong.
It cannot be a syntax error and at the same time the program compiles.
You need to choose one. But, sure, go ahead, foam at the mouth, if it
makes you feel better about it.
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python

On Mon, 12 May 2025 at 01:19, Left Right <olegsivokon@gmail.com> wrote:

Have you ever built a language parser?

I've lost count by now. Probably fewer than hundred times though.

Show me then. Prove it. You're all hot air and opinions and bluster.
Show some actual code, and show that you can do right what you're
complaining that Python has done wrong.

I'm not holding my breath.

ChrisA
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python

On Mon, 12 May 2025 at 01:24, Left Right via Python-list <python-list@python.org> wrote:

But, sure, go ahead, foam at the mouth, if it
makes you feel better about it.

Projecting, much?

ChrisA
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python

Oh, so this is where 4chan relocated after they were hacked?
What a refined discussion!
On Sun, May 11, 2025 at 5:28 PM Chris Angelico <rosuav@gmail.com> wrote:

On Mon, 12 May 2025 at 01:24, Left Right via Python-list <python-list@python.org> wrote:

But, sure, go ahead, foam at the mouth, if it
makes you feel better about it.

Projecting, much?

ChrisA

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python

https://gitlab.com/doodles-archive/protopy/-/blob/master/protopy/lib/protopy.y?ref_type=heads
Here's one. It's not great, not awful. https://gitlab.com/doodles-archive/protopy/-/blob/master/cris-angelico-is-a-moron.org?ref_type=heads
Here's a proof that I have commit rights to this repo
On Sun, May 11, 2025 at 5:24 PM Chris Angelico <rosuav@gmail.com> wrote:

On Mon, 12 May 2025 at 01:19, Left Right <olegsivokon@gmail.com> wrote:

Have you ever built a language parser?

I've lost count by now. Probably fewer than hundred times though.

Show me then. Prove it. You're all hot air and opinions and bluster.
Show some actual code, and show that you can do right what you're
complaining that Python has done wrong.

I'm not holding my breath.

ChrisA

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python

On Mon, 12 May 2025 at 03:35, Left Right <olegsivokon@gmail.com> wrote:

https://gitlab.com/doodles-archive/protopy/-/blob/master/cris-angelico-is-a-moron.org?ref_type=heads
Here's a proof that I have commit rights to this repo

I didn't ask for that, and you don't even have the respect to spell my
name correctly.

You can leave that there, I don't really care. All it really proves is
that you have zero respect for anyone else.

ChrisA
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python

Why of all people would I respect you? What did you ever do other than
insult me? :) Now you are acting surprised?
On Sun, May 11, 2025 at 7:38 PM Chris Angelico <rosuav@gmail.com> wrote:

On Mon, 12 May 2025 at 03:35, Left Right <olegsivokon@gmail.com> wrote:

https://gitlab.com/doodles-archive/protopy/-/blob/master/cris-angelico-is-a-moron.org?ref_type=heads
Here's a proof that I have commit rights to this repo

I didn't ask for that, and you don't even have the respect to spell my
name correctly.

You can leave that there, I don't really care. All it really proves is
that you have zero respect for anyone else.

ChrisA

--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python

Left Right <olegsivokon@gmail.com> writes:

Then it just means that the grammar lies. The two claims are mutually exclusive, so either one is a lie or the other or both.

[...]

A couple of points.

First, the convention in this and most other Usenet newsgroups
is to write new text *below* any quoted text. This is known as "bottom-posting". The alternative, "top-posting" is common in
email in some environments, but tends to cause confusion on Usenet.
It's also a good idea to trim any quoted text that's not relevant to
your followup. See most of the other followups in this newsgroup,
including this one, for examples. Even if you happen to prefer
top-posting, I suggest trying to follow the existing conventions
observed by the vast majority of participants here.

Second, the word "lie" is far more harsh than what I presume
you meant to say. For me, and I think for most people, the word
"lie" implies a deliberate intent to deceive. I don't think you
actually believe that the authors of the documentation you're
complaining about deliberately inserted false information with the
goal of deceiving readers. If you want to say that the grammar
is incorrect, or contains an error, that's something that can be
discussed reasonably. If you say that it "lies", you're making a
claim of malice and making assumptions about someone else's state
of mind with no real basis. Perhaps that's not what the word "lie"
means to you, but I suggest that it explains the harsh reaction to
your initial statement.
--
Keith Thompson (The_Other_Keith) Keith.S.Thompson+u@gmail.com
void Void(void) { Void(); } /* The recursive call of the void */
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python

On 12/05/25 3:21 am, Left Right wrote:

The point is that the error is wrong.
It cannot be a syntax error and at the same time the program compiles.

But the message doesn't say it's an error. It uses the word "warning",
not "error". You're tilting at a straw horse here.
--
Greg
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python

But the message doesn't say it's an error. It uses the word "warning",
not "error". You're tilting at a straw horse here.

Read the associate release note.
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python

Second, the word "lie" is far more harsh than what I presume
you meant to say. For me, and I think for most people, the word
"lie" implies a deliberate intent to deceive.

No, it doesn't. Consider Joseph Conrand's Heart of Darkness, the final
episode where Marlow comes to Kurtz' widow and tells her about how her
husband died. He lies to her, but his intent is not to deceive her,
instead, he intends to make sure that her delusion of her late husband
is unharmed and that she continues to live that delusion because he
judges she will be better off for it.

In fact, you yourself used the word deceit, which is to lie with
intention to benefit from a lie. But people tell lies for all sorts of
reasons. People can lie by omission, through embellishment, by
choosing to focus on less relevant aspects of the event. All of these
are lies. When a painter mixes white paint into the form shadow of a
plaster ball, she lies. When Google Maps puts the destination marker
of the restaurant you have a reservation in in the middle of a sea, it
lies.

English literature lessons aside, even if you believe what you believe
about the meaning of the word, you could at least try to find the
irony, that was the larger goal, than to immediately presume you are
being attacked, and start retaliating instead of looking into the
problem.
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python

Chris and Oleg (sp?), please control your tempers; your latter posts added nothing useful to the conversation.

(Apologies for the late reply, I was out of town.)

--
~Ethan~
Moderator
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python

Left Right <olegsivokon@gmail.com> writes:

Second, the word "lie" is far more harsh than what I presume
you meant to say. For me, and I think for most people, the word
"lie" implies a deliberate intent to deceive.

No, it doesn't. Consider Joseph Conrand's Heart of Darkness,

No, thank you.

I am trying to help you to understand how people are reacting to
your posts. I suggested that saying that the grammar is incorrect
or contains an error would have expressed exactly what you intended
without triggering hostile reactions, which I'm assuming you
don't want.

[...]

English literature lessons aside, even if you believe what you believe
about the meaning of the word, you could at least try to find the
irony, that was the larger goal, than to immediately presume you are
being attacked, and start retaliating instead of looking into the
problem.

I'm not aware that I have been attacked or that I have retaliated.
(No need to explain why you might think I have.)

One more thing: your Usenet or email client probably adds an
attribution line above any quoted text. I urge you to leave it
in place. It makes the discussion easier to follow, especially for
those of us who read the comp.lang.python Usenet newsgroup rather
than the mailing list.

I do not intend to reply further.
--
Keith Thompson (The_Other_Keith) Keith.S.Thompson+u@gmail.com
void Void(void) { Void(); } /* The recursive call of the void */
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python

On 13/05/25 6:28 am, Left Right wrote:

Read the associate release note.

I take it you're referring to this:

In a future Python version, SyntaxError will
eventually be raised, instead of SyntaxWarning. (Contributed by Victor Stinner in gh-98401.)

That doesn't contradict what I said. Currently it's a warning. If and
when it becomes an error, presumably the grammar documentation will be
updated to reflect that. If it isn't, you'll have cause to complain, but
not before.
--
Greg
--- Synchronet 3.21a-Linux NewsLink 1.2

From Newsgroup: comp.lang.python

On 5/8/2025 2:05 AM, Left Right via Python-list wrote:

Also, it appears that the change linked above is a lie:

https://docs.python.org/3/reference/lexical_analysis.html#grammar-token-python-grammar-longstringitem

According to the grammar, any character can follow backslash in a
valid Python program. The warning / error raised by this code should
not be a syntax error / warning because the syntax is correct.

"Changed in version 3.12: Unrecognized escape sequences produce a SyntaxWarning. In a future Python version they will be eventually a SyntaxError."

See https://docs.python.org/3/reference/lexical_analysis.html#strings

Test case for "\" in triple-quoted string:

[code]

def f():

... """\hello"""
<stdin>:2: SyntaxWarning: invalid escape sequence '\h'
... print('hi')
...

f()

hi
[/code]

This example fits the documentation exactly. A SyntaxWarning is emitted
for the unrecognized escape sequence "\h". The program runs as intended.

Please let's drop all the OT back-and-forth.

On Thu, May 8, 2025 at 7:54 AM Left Right <olegsivokon@gmail.com> wrote:

I think it could be this:

A backslash-character pair that is not a valid escape sequence now
generates a SyntaxWarning, instead of DeprecationWarning. For example,
re.compile("\d+\.\d+") now emits a SyntaxWarning ("\d" is an invalid
escape sequence, use raw strings for regular expression:
re.compile(r"\d+\.\d+")). In a future Python version, SyntaxError will
eventually be raised, instead of SyntaxWarning. (Contributed by Victor
Stinner in gh-98401.)

Found in:
https://docs.python.org/3/whatsnew/3.12.html#other-language-changes

It's not supposed to crash your program though. If the program crashes
because of it, it's a bug in Python.

On Thu, May 8, 2025 at 7:00 AM Bob van der Poel via Python-list
<python-list@python.org> wrote:

Did something change in python buggering up my use of a "\ " sequence in a >>> triple quoted string?

I have yet to go through my archive on the program, but I tried to run it >>> today and it crashed quite spectacularly when it hit a """ .... """ line >>> being used as a comment at the top of a function. I changed the "\" to a >>> "/" and all is well now.


--

**** Listen to my FREE CD at http://www.mellowood.ca/music/cedars ****
Bob van der Poel ** Wynndel, British Columbia, CANADA **
EMAIL: bob@mellowood.ca
WWW: http://www.mellowood.ca
--
https://mail.python.org/mailman/listinfo/python-list

--- Synchronet 3.21a-Linux NewsLink 1.2