[issue13402] Document absoluteness of sys.executable
New submission from Éric Araujo <merwok@netwok.org>: I wanted to know if sys.executable was always an absolute path but the doc does not talk about that. (My use case is that I’d like to reload a process and I wonder if I can call os.execve or if I need to use os.execvpe.) ---------- assignee: docs@python components: Documentation messages: 147609 nosy: docs@python, eric.araujo priority: normal severity: normal status: open title: Document absoluteness of sys.executable versions: Python 2.7, Python 3.2, Python 3.3 _______________________________________ Python tracker <report@bugs.python.org> <http://bugs.python.org/issue13402> _______________________________________
Changes by Petri Lehtinen <petri@digip.org>: ---------- nosy: +petri.lehtinen _______________________________________ Python tracker <report@bugs.python.org> <http://bugs.python.org/issue13402> _______________________________________
Petri Lehtinen <petri@digip.org> added the comment: Attached a patch. ---------- keywords: +patch Added file: http://bugs.python.org/file23738/issue13402.patch _______________________________________ Python tracker <report@bugs.python.org> <http://bugs.python.org/issue13402> _______________________________________
Eli Bendersky <eliben@gmail.com> added the comment: LGTM ---------- nosy: +eli.bendersky _______________________________________ Python tracker <report@bugs.python.org> <http://bugs.python.org/issue13402> _______________________________________
Éric Araujo <merwok@netwok.org> added the comment: Patch looks good, but are you 100% sure that sys.executable is always absolute? (On all OSes, under multiprocessing, etc?) ---------- _______________________________________ Python tracker <report@bugs.python.org> <http://bugs.python.org/issue13402> _______________________________________
Petri Lehtinen <petri@digip.org> added the comment: Éric Araujo wrote:
Patch looks good, but are you 100% sure that sys.executable is always absolute? (On all OSes, under multiprocessing, etc?)
Well, its value is computed by a function named Py_GetProgramFullPath(), so I'm quite sure. I had a quick look on what the function is doing, but it was quite complicated. I can check it better if needed. ---------- _______________________________________ Python tracker <report@bugs.python.org> <http://bugs.python.org/issue13402> _______________________________________
Changes by Eric Snow <ericsnowcurrently@gmail.com>: ---------- nosy: +eric.snow _______________________________________ Python tracker <report@bugs.python.org> <http://bugs.python.org/issue13402> _______________________________________
Éric Araujo <merwok@netwok.org> added the comment: This is the bug I was thinking about: #7774. Adding some people from that discussion to nosy. ---------- nosy: +flox, haypo, pitrou, ronaldoussoren, schmir _______________________________________ Python tracker <report@bugs.python.org> <http://bugs.python.org/issue13402> _______________________________________
Antoine Pitrou <pitrou@free.fr> added the comment: LGTM too. You could also add a test to test_sys ensuring that sys.executable is always executable. ---------- _______________________________________ Python tracker <report@bugs.python.org> <http://bugs.python.org/issue13402> _______________________________________
STINNER Victor <victor.stinner@haypocalc.com> added the comment:
You could also add a test to test_sys ensuring that sys.executable is always executable.
And that sys.executable is absolute? sys.executable is an empty string if sys.argv[0] has been changed and Python is unable to retrieve the real path to the Python executable. See the issue #7774. The fact that sys.executable can be empty should be documented. ---------- _______________________________________ Python tracker <report@bugs.python.org> <http://bugs.python.org/issue13402> _______________________________________
Antoine Pitrou <pitrou@free.fr> added the comment:
You could also add a test to test_sys ensuring that sys.executable is always executable.
And that sys.executable is absolute?
Er, yes, that's what I meant. Sorry. ---------- _______________________________________ Python tracker <report@bugs.python.org> <http://bugs.python.org/issue13402> _______________________________________
Petri Lehtinen <petri@digip.org> added the comment: Attaching an updated patch. The documentation now says that sys.executable may be an empty string. The patch also adds a test to make sure that sys.executable is absolute. ---------- Added file: http://bugs.python.org/file23958/issue13402_v2.patch _______________________________________ Python tracker <report@bugs.python.org> <http://bugs.python.org/issue13402> _______________________________________
Philip Jenvey <pjenvey@underboss.org> added the comment: sys.executable can be None on Jython (and I believe IronPython) when ran in an 'embedded' mode ---------- nosy: +dino.viehland, pjenvey _______________________________________ Python tracker <report@bugs.python.org> <http://bugs.python.org/issue13402> _______________________________________
Petri Lehtinen <petri@digip.org> added the comment:
sys.executable can be None on Jython (and I believe IronPython) when ran in an 'embedded' mode
In CPython, embedding doesn't change the behavior. ---------- _______________________________________ Python tracker <report@bugs.python.org> <http://bugs.python.org/issue13402> _______________________________________
Éric Araujo <merwok@netwok.org> added the comment: This discussion seems relevant: http://mail.python.org/pipermail/python-dev/2006-March/062453.html (it talks about the meaning of sys.executable, special cases such as embedded Python and functions that implement it (so we can have a look at that code)). ---------- _______________________________________ Python tracker <report@bugs.python.org> <http://bugs.python.org/issue13402> _______________________________________
Petri Lehtinen <petri@digip.org> added the comment: Attached one more patch. The documentation now also mentions None as a possible value. ---------- Added file: http://bugs.python.org/file24380/issue13402_v3.patch _______________________________________ Python tracker <report@bugs.python.org> <http://bugs.python.org/issue13402> _______________________________________
Roundup Robot <devnull@psf.upfronthosting.co.za> added the comment: New changeset fdcda5b74317 by Petri Lehtinen in branch '3.2': Document absoluteness of sys.executable http://hg.python.org/cpython/rev/fdcda5b74317 New changeset 8b591a86fc91 by Petri Lehtinen in branch 'default': Merge branch 3.2 http://hg.python.org/cpython/rev/8b591a86fc91 New changeset c351536e804a by Petri Lehtinen in branch '2.7': Document absoluteness of sys.executable http://hg.python.org/cpython/rev/c351536e804a ---------- nosy: +python-dev resolution: -> fixed stage: -> committed/rejected status: open -> closed _______________________________________ Python tracker <report@bugs.python.org> <http://bugs.python.org/issue13402> _______________________________________
Brad Larsen added the comment: It looks like sys.executable is *not* always an absolute path. In Python 2.7: $ which python2.7 /opt/local/bin/python2.7 $ cd /opt/local/bin $ ../bin/python2.7 -m 'import sys; print(sys.executable)' /opt/local/bin/../bin/python2.7 Also in Python 3.5: $ which python3.5 /opt/local/bin/python3.5 $ cd /opt/local/bin $ ../bin/python3.5 -m 'import sys; print(sys.executable)' /opt/local/bin/../bin/python3.5 ---------- nosy: +blarsen _______________________________________ Python tracker <report@bugs.python.org> <http://bugs.python.org/issue13402> _______________________________________
Eryk Sun added the comment: "/opt/local/bin/../bin/python2.7" is an absolute path. >>> os.path.isabs('/opt/local/bin/../bin/python2.7') True It's not relative to the working directory. ---------- nosy: +eryksun _______________________________________ Python tracker <report@bugs.python.org> <http://bugs.python.org/issue13402> _______________________________________
Ned Deily added the comment: @blarsen, and there's something strange about your example. '-m' won't work with a real python executable, presumably you meant '-c'. And judging from the paths involved, I'm guessing you are using MacPorts on OS X. With either MacPorts or any other OS X framework build, you wouldn't see results like that with -c. ../bin/python3.5 -c 'import sys; print(sys.executable)' /opt/local/bin/python3.5 ../bin/python2.7 -c 'import sys; print(sys.executable)' /opt/local/Library/Frameworks/Python.framework/Versions/2.7/Resources/Python.app/Contents/MacOS/Python So I'm guessing what's in ../bin/python3.7 and ../bin/python3.5 is something other than the normal python bootstrap binaries. ---------- nosy: +ned.deily _______________________________________ Python tracker <report@bugs.python.org> <http://bugs.python.org/issue13402> _______________________________________
Brad Larsen added the comment: @eryksun, you are right! The output *is* an absolute path as far as `os.path.isabs` is concerned. @ned.deily, you are right about my example --- I transcribed it wrong, and it should be `-c`. The system in question is not a Mac OS system, but a Linux system with newer Python versions than what come with the system installed to /opt/local. These are compiled from source, but without any modifications. I'm commenting here because I'm seeing the `test_sys` test from the regression suite fail in a slightly modified version of Python that ships with a product. In particular, in Lib/test/test_sys.py around line 640, we have this: @unittest.skipIf(sys.base_prefix != sys.prefix, 'Test is not venv-compatible') def test_executable(self): # sys.executable should be absolute self.assertEqual(os.path.abspath(sys.executable), sys.executable) Yes, /opt/local/bin/../bin/python3.5 is an absolute path as far as `os.path.isabs` is concerned, but `os.path.abspath('/opt/local/bin/../bin/python3.5')` gives '/opt/local/bin/python3.5', and the test fails. So maybe the documentation is fine but the test is wrong? ---------- _______________________________________ Python tracker <report@bugs.python.org> <http://bugs.python.org/issue13402> _______________________________________
Ned Deily added the comment: @blarsen, OK, but if you want to pursue this, please open a new issue against the current versions. This issue was closed years ago and against earlier versions so your comments will not get acted on. ---------- _______________________________________ Python tracker <report@bugs.python.org> <http://bugs.python.org/issue13402> _______________________________________
participants (11)
-
Antoine Pitrou -
Brad Larsen
-
Eli Bendersky
-
Eric Snow
-
Eryk Sun
-
Ned Deily
-
Petri Lehtinen
-
Philip Jenvey
-
Roundup Robot
-
STINNER Victor
-
Éric Araujo