Document behaviour of python when launched on symlink to a script

[0d43484e-f3c3-416f-96b8-569e24af7b04]

BPO	14547
Nosy	@ncoghlan, @merwok, @bitdancer
Files	main.py: the script is for a Discord bot hosted on heroku

^{Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.}

Show more details

GitHub fields: ```python assignee = None closed_at = None created_at = labels = ['3.8', 'type-bug', '3.7', 'docs'] title = 'Python symlink to script behaves unexpectedly' updated_at = user = 'https://bugs.python.org/j13r' ``` bugs.python.org fields: ```python activity = actor = 'free.abdo.sh' assignee = 'docs@python' closed = False closed_date = None closer = None components = ['Documentation'] creation = creator = 'j13r' dependencies = [] files = ['50098'] hgrepos = [] issue_num = 14547 keywords = [] message_count = 5.0 messages = ['158039', '158057', '158097', '221842', '395328'] nosy_count = 6.0 nosy_names = ['ncoghlan', 'eric.araujo', 'r.david.murray', 'docs@python', 'j13r', 'free.abdo.sh'] pr_nums = [] priority = 'normal' resolution = None stage = 'needs patch' status = 'open' superseder = None type = 'behavior' url = 'https://bugs.python.org/issue14547' versions = ['Python 3.6', 'Python 3.7', 'Python 3.8'] ```

[0d43484e-f3c3-416f-96b8-569e24af7b04]

If I have a script foo/bar.py import baz

and create a symlink to it, called barhere.py ln -s foo/bar.py barhere.py

when I run it, it behaves unexpectedly, specifically it behaves differently than if I had copied it here. It prefers to import baz from foo/baz, not from the current folder.

Apparently Python (2.7.2-r3) handles symlinks differently than just looking at the content (everything is a file philosophy in UNIX).

[bitdancer]

The content of a symbolic symlink is a symbolic reference to another location in the file system. If you had used a hard link it would certainly work as you expected.

The behavior with respect to symbolic links ought to be documented here:

http://docs.python.org/using/cmdline.html

but doesn't seem to be.

[ncoghlan]

Specifically, we end up calling os.realpath() (or the C level equivalent) when initialising __main.__file and sys.path[0].

Agreed this behaviour should be documented (and tested!) explicitly.

[83d2e70e-e599-4a04-b820-3814bbdb9bef]

I've removed Tests from the components list as I don't think it belongs there.

[1882008a-26cb-4402-bea7-387bfea407d9]

hello guys I am facing problem with my sqlite3 database when i make changes it saves the changes for almost 13 hours and then retakes back all the changes that i made, i am using two threads in my code and i am not committing the data but i set the isolation mode to None, i was using the same queries with mysql database with setting autocommit to on and it was working fine what could be the problem in your opinion? we searched and ask a lot about this issue but nothing was helpful.