gh-94808: Cover `%p` in `PyUnicode_FromFormat` #96677

sobolevn · 2022-09-08T08:29:36Z

Source:

Lines 2531 to 2552 in 4f523a7

    
           case 'p': 
        
           { 
        
               char number[MAX_LONG_LONG_CHARS]; 
        
               len = sprintf(number, "%p", va_arg(*vargs, void*)); 
        
               assert(len >= 0); 
        
               /* %p is ill-defined:  ensure leading 0x. */ 
        
               if (number[1] == 'X') 
        
                   number[1] = 'x'; 
        
               else if (number[1] != 'x') { 
        
                   memmove(number + 2, number, 
        
                           strlen(number) + 1); 
        
                   number[0] = '0'; 
        
                   number[1] = 'x'; 
        
                   len += 2; 
        
               } 
        
               if (_PyUnicodeWriter_WriteASCIIString(writer, number, len) < 0) 
        
                   return NULL; 
        
               break; 
        
           }

Internet says that some systems can return representation with 0x, some with 0X, and some with none. It is reflected in the implementation, but cannot be reflected in tests.

Issue: Metabug: Improving C-level coverage #94808

sobolevn · 2022-09-08T09:08:06Z

And we got a failure: MacOS passes locally and on CI, but other platforms do not.

Windows:

 ======================================================================
FAIL: test_from_format (test.test_unicode.CAPITest.test_from_format)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "D:\a\cpython\cpython\Lib\test\test_unicode.py", line 2817, in test_from_format
    self.assertEqual(len(p_format1), 11)
AssertionError: 10 != 11

Ubuntu:

======================================================================
FAIL: test_from_format (test.test_unicode.CAPITest.test_from_format)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/runner/work/cpython/cpython-ro-srcdir/Lib/test/test_unicode.py", line 2817, in test_from_format
    self.assertEqual(len(p_format1), 11)
AssertionError: 14 != 11

I think it is better to remove len assumptions.

kumaraditya303

I think a regex test would be better here.

sobolevn · 2022-09-08T11:24:51Z

Good idea, thank you.

JelleZijlstra · 2022-10-03T22:30:53Z

Lib/test/test_unicode.py

+        self.assertIsInstance(p_format1, str)
+        self.assertRegex(p_format1, p_format_regex)
+
+        p_format2 = PyUnicode_FromFormat(b'repr=%p', '123456', None, b'xyz')


What does this case test? I don't see what it adds over the previous case.

It has:

Multiple arguments (while p_format1 only has one arg)

All arguments have different types

New types are tested: bytes and None

Aren't the additional arguments ignored though, because there is only one %p in the format string?

I agree, tests can be better :)
Can you please take a look at the updated version?

It now has both:

Extra types

Ignored arguments

sobolevn · 2022-10-06T21:48:08Z

From https://github.com/python/cpython/actions/runs/3200613350/jobs/5227726848

AssertionError: Regex didn't match: '0x[a-zA-Z0-9]{8,} 0x[a-zA-Z0-9]{1,} 0x[a-zA-Z0-9]{8,}' not found in '0x60400191c290 0x(nil) 0x60400191c230'

This is quite interesting. Is 0x(nil) valid?

JelleZijlstra · 2022-10-06T21:57:33Z

Interesting! I think that comes from libc since I can't find code in CPython that would do this.

sobolevn · 2022-10-06T22:03:25Z

From https://pubs.opengroup.org/onlinepubs/009695399/functions/fprintf.html

p
The argument shall be a pointer to void. The value of the pointer is converted to a sequence of printable characters, in an implementation-defined manner.

I think it is better to just leave None out of this test case.
And possibly create a new discussion about this: aren't NULL and None confused here?

Related discussions:

JelleZijlstra · 2022-10-06T22:05:27Z

I think ctypes translates None into NULL, which is reasonable.

Lib/test/test_unicode.py

miss-islington · 2022-10-07T16:53:45Z

Thanks @sobolevn for the PR, and @JelleZijlstra for merging it 🌮🎉.. I'm working now to backport this PR to: 3.10, 3.11.
🐍🍒⛏🤖

miss-islington · 2022-10-07T16:53:47Z

Sorry @sobolevn and @JelleZijlstra, I had trouble checking out the 3.11 backport branch.
Please backport using cherry_picker on command line.
cherry_picker 72c166add89a0cd992d66f75ce94eee5eb675a99 3.11

bedevere-bot · 2022-10-07T16:53:54Z

GH-98032 is a backport of this pull request to the 3.10 branch.

Co-authored-by: Jelle Zijlstra <[email protected]> (cherry picked from commit 72c166a) Co-authored-by: Nikita Sobolev <[email protected]>

miss-islington · 2022-10-07T16:54:16Z

Thanks @sobolevn for the PR, and @JelleZijlstra for merging it 🌮🎉.. I'm working now to backport this PR to: 3.11.
🐍🍒⛏🤖

bedevere-bot · 2022-10-07T16:54:23Z

GH-98033 is a backport of this pull request to the 3.11 branch.

Co-authored-by: Jelle Zijlstra <[email protected]> (cherry picked from commit 72c166a) Co-authored-by: Nikita Sobolev <[email protected]>

bedevere-bot · 2022-10-07T17:02:55Z

⚠️⚠️⚠️ Buildbot failure ⚠️⚠️⚠️

Hi! The buildbot s390x SLES 3.x has failed when building commit 72c166a.

What do you need to do:

Don't panic.
Check the buildbot page in the devguide if you don't know what the buildbots are or how they work.
Go to the page of the buildbot that failed (https://buildbot.python.org/all/#builders/540/builds/3747) and take a look at the build logs.
Check if the failure is related to this commit (72c166a) or if it is a false positive.
If the failure is related to this commit, please, reflect that on the issue and make a new Pull Request with a fix.

You can take a look at the buildbot page here:

https://buildbot.python.org/all/#builders/540/builds/3747

Failed tests:

test_unicode

Failed subtests:

test_from_format - test.test_unicode.CAPITest.test_from_format

Summary of the results of the build (if available):

== Tests result: FAILURE then FAILURE ==

420 tests OK.

10 slowest tests:

test_concurrent_futures: 2 min 51 sec
test_tools: 2 min 34 sec
test_multiprocessing_spawn: 1 min 51 sec
test_asyncio: 1 min 17 sec
test_multiprocessing_forkserver: 1 min 14 sec
test_signal: 1 min 10 sec
test_multiprocessing_fork: 1 min 8 sec
test_nntplib: 1 min 5 sec
test_capi: 52.3 sec
test_tokenize: 48.6 sec

1 test failed:
test_unicode

16 tests skipped:
test_devpoll test_ioctl test_kqueue test_launcher test_msilib
test_nis test_perf_profiler test_startfile test_tix test_tkinter
test_ttk test_winconsoleio test_winreg test_winsound test_wmi
test_zipfile64

1 re-run test:
test_unicode

Total duration: 5 min 53 sec

Click to see traceback logs

Traceback (most recent call last):
  File "/home/dje/cpython-buildarea/3.x.edelsohn-sles-z/build/Lib/test/test_unicode.py", line 2817, in test_from_format
    self.assertRegex(p_format1, p_format_regex)
AssertionError: Regex didn't match: '^0x[a-zA-Z0-9]{8,}$' not found in '0x153b8c0'

JelleZijlstra · 2022-10-07T17:10:40Z

I guess the pointers are shorter on s390. I'll adjust the regex.

JelleZijlstra · 2022-10-07T17:14:37Z

#98036

bedevere-bot · 2022-10-07T17:26:56Z

⚠️⚠️⚠️ Buildbot failure ⚠️⚠️⚠️

Hi! The buildbot AMD64 FreeBSD Non-Debug 3.x has failed when building commit 72c166a.

What do you need to do:

Don't panic.
Check the buildbot page in the devguide if you don't know what the buildbots are or how they work.
Go to the page of the buildbot that failed (https://buildbot.python.org/all/#builders/172/builds/2913) and take a look at the build logs.
Check if the failure is related to this commit (72c166a) or if it is a false positive.
If the failure is related to this commit, please, reflect that on the issue and make a new Pull Request with a fix.

You can take a look at the buildbot page here:

https://buildbot.python.org/all/#builders/172/builds/2913

Failed tests:

test_unicode

Failed subtests:

test_from_format - test.test_unicode.CAPITest.test_from_format

Summary of the results of the build (if available):

== Tests result: FAILURE then FAILURE ==

415 tests OK.

10 slowest tests:

test_multiprocessing_spawn: 4 min 3 sec
test_tokenize: 3 min 40 sec
test_asyncio: 3 min 18 sec
test_concurrent_futures: 3 min 9 sec
test_multiprocessing_forkserver: 2 min 19 sec
test_unparse: 2 min 10 sec
test_lib2to3: 2 min 3 sec
test_unicodedata: 1 min 48 sec
test_subprocess: 1 min 37 sec
test_multiprocessing_fork: 1 min 36 sec

1 test failed:
test_unicode

21 tests skipped:
test_dbm_gnu test_devpoll test_epoll test_idle test_ioctl
test_launcher test_msilib test_perf_profiler test_spwd
test_startfile test_tcl test_tix test_tkinter test_ttk
test_ttk_textonly test_turtle test_winconsoleio test_winreg
test_winsound test_wmi test_zipfile64

1 re-run test:
test_unicode

Total duration: 17 min

Click to see traceback logs

Traceback (most recent call last):
  File "/usr/home/buildbot/python/3.x.koobs-freebsd-9e36.nondebug/build/Lib/test/test_unicode.py", line 2817, in test_from_format
    self.assertRegex(p_format1, p_format_regex)
AssertionError: Regex didn't match: '^0x[a-zA-Z0-9]{8,}$' not found in '0x612bd0'

bedevere-bot · 2022-10-07T17:59:09Z

⚠️⚠️⚠️ Buildbot failure ⚠️⚠️⚠️

Hi! The buildbot x86 Gentoo Installed with X 3.x has failed when building commit 72c166a.

What do you need to do:

Don't panic.
Check the buildbot page in the devguide if you don't know what the buildbots are or how they work.
Go to the page of the buildbot that failed (https://buildbot.python.org/all/#builders/464/builds/3172) and take a look at the build logs.
Check if the failure is related to this commit (72c166a) or if it is a false positive.
If the failure is related to this commit, please, reflect that on the issue and make a new Pull Request with a fix.

You can take a look at the buildbot page here:

https://buildbot.python.org/all/#builders/464/builds/3172

Failed tests:

test_unicode

Failed subtests:

test_from_format - test.test_unicode.CAPITest.test_from_format

Summary of the results of the build (if available):

== Tests result: FAILURE then FAILURE ==

417 tests OK.

1 test failed:
test_unicode

17 tests skipped:
test_asdl_parser test_check_c_globals test_clinic test_dbm_ndbm
test_devpoll test_gdb test_ioctl test_kqueue test_launcher
test_msilib test_perf_profiler test_startfile test_winconsoleio
test_winreg test_winsound test_wmi test_zipfile64

1 re-run test:
test_unicode

Total duration: 53 min 22 sec

Click to see traceback logs

Traceback (most recent call last):
  File "/buildbot/buildarea/cpython/3.x.ware-gentoo-x86.installed/build/target/lib/python3.12/test/test_unicode.py", line 2817, in test_from_format
    self.assertRegex(p_format1, p_format_regex)
AssertionError: Regex didn't match: '^0x[a-zA-Z0-9]{8,}$' not found in '0x8bd9f0'


Traceback (most recent call last):
  File "/buildbot/buildarea/cpython/3.x.ware-gentoo-x86.installed/build/target/lib/python3.12/test/test_unicode.py", line 2817, in test_from_format
    self.assertRegex(p_format1, p_format_regex)
AssertionError: Regex didn't match: '^0x[a-zA-Z0-9]{8,}$' not found in '0x9809f0'

) Co-authored-by: Nikita Sobolev <[email protected]> Co-authored-by: Jelle Zijlstra <[email protected]> (cherry picked from commit 72c166a)

Co-authored-by: Jelle Zijlstra <[email protected]>

pythongh-94808: Cover %p in PyUnicode_FromFormat

45f0bcc

bedevere-bot added awaiting review tests Tests in the Lib/test dir labels Sep 8, 2022

sobolevn added the skip news label Sep 8, 2022

Remove len assumptions from %p

9faf155

kumaraditya303 reviewed Sep 8, 2022

View reviewed changes

Use regex

ad127bd

mdboom mentioned this pull request Sep 8, 2022

Metabug: Improving C-level coverage #94808

Open

JelleZijlstra reviewed Oct 3, 2022

View reviewed changes

JelleZijlstra self-assigned this Oct 3, 2022

sobolevn added 2 commits October 7, 2022 00:30

Address review

0213fab

Merge branch 'main' into issue-94808

111a173

JelleZijlstra approved these changes Oct 6, 2022

View reviewed changes

bedevere-bot added awaiting merge and removed awaiting review labels Oct 6, 2022

JelleZijlstra added needs backport to 3.10 only security fixes needs backport to 3.11 only security fixes labels Oct 6, 2022

Update test_unicode.py

af4ba72

JelleZijlstra reviewed Oct 7, 2022

View reviewed changes

Lib/test/test_unicode.py Outdated Show resolved Hide resolved

Update Lib/test/test_unicode.py

cb76e61

JelleZijlstra merged commit 72c166a into python:main Oct 7, 2022

bedevere-bot removed the awaiting merge label Oct 7, 2022

bedevere-bot removed the needs backport to 3.10 only security fixes label Oct 7, 2022

JelleZijlstra removed the needs backport to 3.11 only security fixes label Oct 7, 2022

JelleZijlstra added the needs backport to 3.11 only security fixes label Oct 7, 2022

bedevere-bot removed the needs backport to 3.11 only security fixes label Oct 7, 2022

mpage pushed a commit to mpage/cpython that referenced this pull request Oct 11, 2022

pythongh-94808: Cover %p in PyUnicode_FromFormat (python#96677)

f100fc3

Co-authored-by: Jelle Zijlstra <[email protected]>

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

gh-94808: Cover `%p` in `PyUnicode_FromFormat` #96677

gh-94808: Cover `%p` in `PyUnicode_FromFormat` #96677

sobolevn commented Sep 8, 2022 •

edited by bedevere-bot

Loading

sobolevn commented Sep 8, 2022 •

edited

Loading

kumaraditya303 left a comment

sobolevn commented Sep 8, 2022

JelleZijlstra Oct 3, 2022

sobolevn Oct 4, 2022

JelleZijlstra Oct 6, 2022

sobolevn Oct 6, 2022

JelleZijlstra Oct 6, 2022

sobolevn commented Oct 6, 2022

JelleZijlstra commented Oct 6, 2022

sobolevn commented Oct 6, 2022

JelleZijlstra commented Oct 6, 2022

miss-islington commented Oct 7, 2022

miss-islington commented Oct 7, 2022

bedevere-bot commented Oct 7, 2022

miss-islington commented Oct 7, 2022

bedevere-bot commented Oct 7, 2022

bedevere-bot commented Oct 7, 2022

JelleZijlstra commented Oct 7, 2022

JelleZijlstra commented Oct 7, 2022

bedevere-bot commented Oct 7, 2022

bedevere-bot commented Oct 7, 2022

	case 'p':
	{
	char number[MAX_LONG_LONG_CHARS];

	len = sprintf(number, "%p", va_arg(vargs, void));
	assert(len >= 0);

	/* %p is ill-defined: ensure leading 0x. */
	if (number[1] == 'X')
	number[1] = 'x';
	else if (number[1] != 'x') {
	memmove(number + 2, number,
	strlen(number) + 1);
	number[0] = '0';
	number[1] = 'x';
	len += 2;
	}

	if (_PyUnicodeWriter_WriteASCIIString(writer, number, len) < 0)
	return NULL;
	break;
	}

gh-94808: Cover %p in PyUnicode_FromFormat #96677

gh-94808: Cover %p in PyUnicode_FromFormat #96677

Conversation

sobolevn commented Sep 8, 2022 • edited by bedevere-bot Loading

sobolevn commented Sep 8, 2022 • edited Loading

kumaraditya303 left a comment

Choose a reason for hiding this comment

sobolevn commented Sep 8, 2022

JelleZijlstra Oct 3, 2022

Choose a reason for hiding this comment

sobolevn Oct 4, 2022

Choose a reason for hiding this comment

JelleZijlstra Oct 6, 2022

Choose a reason for hiding this comment

sobolevn Oct 6, 2022

Choose a reason for hiding this comment

JelleZijlstra Oct 6, 2022

Choose a reason for hiding this comment

sobolevn commented Oct 6, 2022

JelleZijlstra commented Oct 6, 2022

sobolevn commented Oct 6, 2022

JelleZijlstra commented Oct 6, 2022

miss-islington commented Oct 7, 2022

miss-islington commented Oct 7, 2022

bedevere-bot commented Oct 7, 2022

miss-islington commented Oct 7, 2022

bedevere-bot commented Oct 7, 2022

bedevere-bot commented Oct 7, 2022

⚠️⚠️⚠️ Buildbot failure ⚠️⚠️⚠️

JelleZijlstra commented Oct 7, 2022

JelleZijlstra commented Oct 7, 2022

bedevere-bot commented Oct 7, 2022

⚠️⚠️⚠️ Buildbot failure ⚠️⚠️⚠️

bedevere-bot commented Oct 7, 2022

⚠️⚠️⚠️ Buildbot failure ⚠️⚠️⚠️

gh-94808: Cover `%p` in `PyUnicode_FromFormat` #96677

gh-94808: Cover `%p` in `PyUnicode_FromFormat` #96677

sobolevn commented Sep 8, 2022 •

edited by bedevere-bot

Loading

sobolevn commented Sep 8, 2022 •

edited

Loading