1. 26 Jan, 2022 1 commit
  2. 08 Dec, 2021 5 commits
    • Kirill Smelkov's avatar
      pygolang v0.0.9 · e503beb0
      Kirill Smelkov authored
      e503beb0
    • Kirill Smelkov's avatar
      gpython: tests: Adjust test_pymain_syspath to current PyPy state · 57ab5f33
      Kirill Smelkov authored
      See https://foss.heptapod.net/pypy/pypy/-/issues/3610 for stdin case.
      -m started to behave as CPython.
      57ab5f33
    • Kirill Smelkov's avatar
      gpython: tests: test_pymain: Fix -i on PyPy3 · 6ca43f2d
      Kirill Smelkov authored
      On py3 'ps1' and b'ps1' are different keys and it was failing on PyPy3 as follows:
      
                  # -i after stdin (also tests interactive mode as -i forces interactive even on non-tty)
                  d = {
                      b'hellopy': b(hellopy),
                      b'ps1':     b'' # cpython emits prompt to stderr
                  }
                  if is_pypy and not is_gpython:
                      d['ps1'] = b'>>>> ' # native pypy emits prompt to stdout and >>>> instead of >>>
                  _ = pyout(['-i'], stdin=b'import hello\n', cwd=testdata)
          >       assert _ == b"%(ps1)shello\nworld\n['']\n%(ps1)s"           % d
          E       assert b">>>> hello\...\n['']\n>>>> " == b"hello\nworld\n['']\n"
          E         At index 0 diff: b'>' != b'h'
          E         Full diff:
          E         - b"hello\nworld\n['']\n"
          E         + b">>>> hello\nworld\n['']\n>>>> "
          E         ?   +++++                    +++++
      
          gpython/gpython_test.py:200: AssertionError
      
      Fix it.
      Amends e205dbf6 (gpython: Implement -i (interactive inspect after program run) + promised interactive-mode tests)
      6ca43f2d
    • Kirill Smelkov's avatar
      context: Fix deadlock when new context is created from already-canceled parent · 58d4cbfe
      Kirill Smelkov authored
      When _BaseCtx is setting up cancel propagation it locks a parent,
      checks for parent.err != nil, and, if it is, calls
      ctx._cancel(parent.err) _with_ _holding_ parent.mu. Since _cancel
      internally also goes through parents and locks them, this was deadlocking
      on the second call to parent.mu.lock().
      
      -> Fix it by calling ctx._cancel(err) in the constructor outside of
      parent lock.
      
      The bug was there from the beginning - from e9567c7b (context: New
      package that mirrors Go's context).
      
      /trusted-by @jerome
      /reviewed-on nexedi/pygolang!16
      58d4cbfe
    • Kirill Smelkov's avatar
      context: Reorganize the code to make internal logic more clear · d0688e21
      Kirill Smelkov authored
      - _propagateCancel is used only in _BaseCtx constructor -> inline it
        there. Being run in the constructor makes it clear that this code
        works on new _BaseCtx object with empty set of children.
      
      - since _cancelFrom interacts with the code moved from _propagateCancel,
        also move it to be close to cancel propagation setup.
      
      No functional changes, just plain code movement.
      
      /trusted-by @jerome
      /reviewed-on nexedi/pygolang!16
      d0688e21
  3. 16 Aug, 2021 1 commit
  4. 07 Apr, 2021 5 commits
    • Kirill Smelkov's avatar
    • Kirill Smelkov's avatar
      gpython: Run code in carefully prepared __main__ module · 2351dd27
      Kirill Smelkov authored
      Because else sys.modules['__main__'] points to the module of gpython
      itself, not user program. This difference can be important when e.g.
      user defines classes in main module and tries to pickle them. Such
      pickling will fail if gpython is not careful to run user's code in its
      own main.
      
      Without this patch added test fails as
      
            File "check_main.py", line 51, in <module>
              main()
            File "check_main.py", line 39, in main
              assert hasattr(mainmod, 'MyUniqueClassXYZ'),        dir(mainmod)
          AssertionError: ['__builtins__', '__doc__', '__file__', '__name__', '__package__', 'main']
      
      The problem was discovered while trying to run test program from
      https://github.com/zopefoundation/ZEO/issues/166 :
      
          $ gpython zopenrace-zeo.py
          Traceback (most recent call last):
            File "/home/kirr/tmp/trashme/Z/py2.venv/bin/gpython", line 8, in <module>
              sys.exit(main())
            File "/home/kirr/tmp/trashme/Z/py2.venv/lib/python2.7/site-packages/gpython/__init__.py", line 395, in main
              pymain(argv, init)
            File "/home/kirr/tmp/trashme/Z/py2.venv/lib/python2.7/site-packages/gpython/__init__.py", line 266, in pymain
              run()
            File "/home/kirr/tmp/trashme/Z/py2.venv/lib/python2.7/site-packages/gpython/__init__.py", line 172, in run
              _execfile(filepath, g)
            File "/home/kirr/tmp/trashme/Z/py2.venv/lib/python2.7/site-packages/gpython/__init__.py", line 275, in _execfile
              six.exec_(code, globals, locals)
            File "/home/kirr/tmp/trashme/Z/py2.venv/lib/python2.7/site-packages/six.py", line 719, in exec_
              exec("""exec _code_ in _globs_, _locs_""")
            File "<string>", line 1, in <module>
            File "zopenrace-zeo.py", line 205, in <module>
              main()
            File "zopenrace-zeo.py", line 190, in main
              init()
            File "zopenrace-zeo.py", line 126, in init
              transaction.commit()
            File "/home/kirr/tmp/trashme/Z/py2.venv/lib/python2.7/site-packages/transaction/_manager.py", line 257, in commit
              return self.manager.commit()
            File "/home/kirr/tmp/trashme/Z/py2.venv/lib/python2.7/site-packages/transaction/_manager.py", line 134, in commit
              return self.get().commit()
            File "/home/kirr/tmp/trashme/Z/py2.venv/lib/python2.7/site-packages/transaction/_transaction.py", line 282, in commit
              reraise(t, v, tb)
            File "/home/kirr/tmp/trashme/Z/py2.venv/lib/python2.7/site-packages/transaction/_transaction.py", line 273, in commit
              self._commitResources()
            File "/home/kirr/tmp/trashme/Z/py2.venv/lib/python2.7/site-packages/transaction/_transaction.py", line 456, in _commitResources
              reraise(t, v, tb)
            File "/home/kirr/tmp/trashme/Z/py2.venv/lib/python2.7/site-packages/transaction/_transaction.py", line 430, in _commitResources
              rm.commit(self)
            File "/home/kirr/tmp/trashme/Z/py2.venv/lib/python2.7/site-packages/ZODB/Connection.py", line 497, in commit
              self._commit(transaction)
            File "/home/kirr/tmp/trashme/Z/py2.venv/lib/python2.7/site-packages/ZODB/Connection.py", line 546, in _commit
              self._store_objects(ObjectWriter(obj), transaction)
            File "/home/kirr/tmp/trashme/Z/py2.venv/lib/python2.7/site-packages/ZODB/Connection.py", line 578, in _store_objects
              p = writer.serialize(obj)  # This calls __getstate__ of obj
            File "/home/kirr/tmp/trashme/Z/py2.venv/lib/python2.7/site-packages/ZODB/serialize.py", line 430, in serialize
              return self._dump(meta, obj.__getstate__())
            File "/home/kirr/tmp/trashme/Z/py2.venv/lib/python2.7/site-packages/ZODB/serialize.py", line 439, in _dump
              self._p.dump(state)
          PicklingError: Can't pickle <class '__main__.PInt'>: attribute lookup __main__.PInt failed
      
      /reviewed-by @jerome
      /reviewed-on nexedi/pygolang!15
      2351dd27
    • Kirill Smelkov's avatar
      gpython: Fix interactive mode to use stdout or stderr depending on what is a tty · 95c7cce9
      Kirill Smelkov authored
      To match python behaviour:
      
          $ python >/dev/null
          Python 2.7.18 (default, Apr 20 2020, 20:30:41)
          [GCC 9.3.0] on linux2
          Type "help", "copyright", "credits" or "license" for more information.
          >>> 1+1			# NOTE the prompt (printed to stderr)
      
          $ python 2>/dev/null
          >>> 1+1			# NOTE the prompt (printed to stdout)
      
      Tests pending.
      
      /reviewed-by @jerome
      /reviewed-on nexedi/pygolang!15
      95c7cce9
    • Kirill Smelkov's avatar
      gpython: Handle program on stdin through just read+exec · 22fb559a
      Kirill Smelkov authored
      Not through interactive console, because it was printing prompts, while
      `python <prog.py` does not emit anything:
      
          $ cat prog.py
          print 'Hello World'
      
          $ python <prog.py
          Hello World
      
          $ gpython <prog.py
          Python 2.7.18 (default, Apr 20 2020, 20:30:41)
          [GCC 9.3.0] [GPython 0.0.8] [gevent 20.9.0] on linux2
          Type "help", "copyright", "credits" or "license" for more information.
          (InteractiveConsole)
          Hello World
      
      After the patch gpython output is the same as of python:
      
          $ gpython <prog.py
          Hello World
      
      Test coverage for interactive mode is pending.
      We'll add them soon in a follow-up patch after implementing -i.
      
      For _interact - by this patch logic - we should be dropping custom
      raw_input, since now _interact is called only when sys.stdin is tty.
      But we'll soon be invoking _interact when stdin is not a tty (for
      `gpython -i <prog.py`), so leave that logic in place as is.
      
      /reviewed-by @jerome
      /reviewed-on nexedi/pygolang!15
      22fb559a
    • Kirill Smelkov's avatar
      gpython: Factor-out interactive-console functionality into its own function · 6cc4bf32
      Kirill Smelkov authored
      We will soon need this to be run from several places when implementing
      support for -i.
      
      /reviewed-by @jerome
      /reviewed-on nexedi/pygolang!15
      6cc4bf32
  5. 26 Mar, 2021 1 commit
  6. 16 Mar, 2021 1 commit
  7. 14 Dec, 2020 2 commits
    • Kirill Smelkov's avatar
      *: Python3.9 switched __file__ to be always absolute · 4f28dddf
      Kirill Smelkov authored
      https://bugs.python.org/issue20443
      
      This broke test_defer_excchain_dump because
      testprog/golang_test_defer_excchain.txt is prepared with output where
      `python file` shows that file name in traceback as it was specified on
      the command line, e.g.
      
          .../pygolang/golang/testprog$ python golang_test_defer_excchain.py
          Traceback (most recent call last):
            File ".../pygolang/golang/__init__.py", line 103, in _
              return f(*argv, **kw)
            File "golang_test_defer_excchain.py", line 42, in main
              raise RuntimeError("err")
          RuntimeError: err
      
      while with py39 it became
      
          .../pygolang/golang/testprog$ python golang_test_defer_excchain.py
          Traceback (most recent call last):
            File ".../pygolang/golang/__init__.py", line 103, in _
              return f(*argv, **kw)
            File ".../pygolang/golang/testprog/golang_test_defer_excchain.py", line 42, in main
              raise RuntimeError("err")
          RuntimeError: err
      
      (notice the difference related to "line 42")
      
      -> Fix it:
      
      - amend the test to conditionally prefix golang_test_defer_excchain.py
        in expected output with PYGOLANG/golang/testprog/ if it is Python >= 3.9.
      - amend `gpython file` to match behaviour of underlying `python file`,
        so that the test passes unconditionally whether it is run by python or
        gpython.
      
      --------
      
      @jerome also says (nexedi/pygolang!13 (comment 122826)):
      
      FYI, buildout (and many zope packages) essentially use doctests for
      testing and they had to deal with similar differences in output.
      `zope.testing` comes with a doctest checker named "renormalizing" which
      normalize output with regular expressions, see for example setup of some
      buildout test [here](https://github.com/buildout/buildout/blob/db3d6e2fbf5d7ff2cc4b2507253c7a221cfc3e32/src/zc/buildout/tests.py#L3615-L3651).
      We definitely don't need this here for the moment, but maybe one day it
      can be useful.
      
      /reviewed-on nexedi/pygolang!13
      4f28dddf
    • Kirill Smelkov's avatar
      tox += CPython3.9 · 32167853
      Kirill Smelkov authored
      Debian testing recently switched default python3 to be python3.9.
      Let's make sure pygolang works with that python version.
      Currently some tests fail - this will be addressed in the next patch.
      
      /reviewed-on nexedi/pygolang!13
      32167853
  8. 11 Dec, 2020 2 commits
    • Kirill Smelkov's avatar
      sync.WorkGroup: Provide "with" support · 6eb80104
      Kirill Smelkov authored
      So that it becomes possible to write
      
          with WorkGroup(ctx) as wg:
              wg.go(f1)
              wg.go(f2)
      
      instead of
      
          wg = WorkGroup(ctx)
          defer(wg.wait)
          wg.go(f1)
          wg.go(f2)
      
      or
      
          wg = WorkGroup(ctx)
          wg.go(f1)
          wg.go(f2)
          wg.wait()
      
      This is sometimes handy and is referred to as "structured concurrency"
      in Python world.
      
      sync.Sema, sync.Mutex, sync.RWMutex already support "with".
      sync.WaitGroup is imho too low-level, but we might consider adding
      "with" support for it in the future as well.
      
      In general pygolang way is to use defer instead of plugging all classes
      with __enter__/__exit__ "with" support, but for small well-known class of
      concurrency-related things its seems "with" support is worth it:
      
      - having "with" for sync.Mutex+co allows it to be used as a drop-in
        replacement instead of threading.Lock+co, and
      - having "with" for sync.WorkGroup - the most commonly-used tool to
        spawn jobs and wait for their completion - makes it on-par with
        "structured concurrency".
      
      /reviewed-on nexedi/pygolang!12
      6eb80104
    • Kirill Smelkov's avatar
      sync.WorkGroup: Fix typo in wait · 85257b2a
      Kirill Smelkov authored
      WorkGroup methods work on PyWorkGroup, not on PyWaitGroup.
      This probably used to work because Cython ignores (?) provided type of self?
      
      /reviewed-on nexedi/pygolang!12
      85257b2a
  9. 10 Dec, 2020 1 commit
    • Kirill Smelkov's avatar
      gpython: tests: Skip test_pymain_run_via_relpath on SlapOS · f15459b9
      Kirill Smelkov authored
      Similarly to test_pymain_syspath we cannot run this test on SlapOS because with
      buildout raw underlying python interpreter does not have access to eggs with
      which gpython script was generated.
      
      See 0fa9d6e7 and 92bb5bcc for details.
      
      Without this patch running pygolang tests fails on SlapOS as shown below:
      
          (pygolang-env) slapuser34@vifibcloud-rapidspace-hosting-007:~/srv/runner/software/44fe7dd3f13ecd100894c6368a35c055/parts/pygolang-dev$ gpython -m pytest -vsx -k relpath
          ================================================= test session starts ==================================================
          platform linux2 -- Python 2.7.18, pytest-4.6.11, py-1.9.0, pluggy-0.13.1 -- /srv/slapgrid/slappart34/srv/runner/software/44fe7dd3f13ecd100894c6368a35c055/bin/gpython
          cachedir: .pytest_cache
          rootdir: /srv/slapgrid/slappart34/srv/runner/software/44fe7dd3f13ecd100894c6368a35c055/parts/pygolang-dev
          collected 115 items / 114 deselected / 1 selected
      
          gpython/gpython_test.py::test_pymain_run_via_relpath Traceback (most recent call last):
            File "./__init__.py", line 511, in <module>
              main()
            File "./__init__.py", line 395, in main
              pymain(argv, init)
            File "./__init__.py", line 217, in pymain
              init()
            File "./__init__.py", line 364, in init
              import gevent
          ImportError: No module named gevent
          FAILED
      
          ======================================================= FAILURES =======================================================
          _____________________________________________ test_pymain_run_via_relpath ______________________________________________
      
              @gpython_only
              def test_pymain_run_via_relpath():
                  argv = ['-c',  'import sys; print(sys.version)']
                  out1 = pyout(                    argv, pyexe=sys.executable)
          >       out2 = pyout(['./__init__.py'] + argv, pyexe=sys._gpy_underlying_executable, cwd=here)
      
          gpython/gpython_test.py:301:
          _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
          golang/golang_test.py:1842: in pyout
              return pyrun(argv, stdin=stdin, stdout=stdout, stderr=stderr, **kw)
          _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
      
          argv = ['./__init__.py', '-c', 'import sys; print(sys.version)'], stdin = None, stdout = '', stderr = None
          kw = {'cwd': '/srv/slapgrid/slappart34/srv/runner/software/44fe7dd3f13ecd100894c6368a35c055/parts/pygolang-dev/gpython', 'pyexe': '/srv/slapgrid/slappart34/srv//runner//shared/python2.7/44578f5389723084b136574c34bc8023/bin/python2.7'}
          retcode = 1
      
              def pyrun(argv, stdin=None, stdout=None, stderr=None, **kw):
                  retcode, stdout, stderr = _pyrun(argv, stdin=stdin, stdout=stdout, stderr=stderr, **kw)
                  if retcode:
          >           raise RuntimeError(' '.join(argv) + '\n' + (stderr and str(stderr) or '(failed)'))
          E           RuntimeError: ./__init__.py -c import sys; print(sys.version)
          E           (failed)
      
          golang/golang_test.py:1836: RuntimeError
          ======================================= 1 failed, 114 deselected in 0.73 seconds =======================================
      
      Fixes 076cdd8f (gpython: Fix crash when invoked as e.g. ./bin/gpython).
      
      /reviewed-on nexedi/pygolang!11
      f15459b9
  10. 06 Dec, 2020 1 commit
  11. 02 Dec, 2020 2 commits
  12. 01 Dec, 2020 1 commit
  13. 27 Nov, 2020 1 commit
  14. 26 Nov, 2020 7 commits
  15. 02 Nov, 2020 1 commit
  16. 14 Oct, 2020 5 commits
  17. 22 Sep, 2020 3 commits
    • Kirill Smelkov's avatar
      pygolang v0.0.7 · 73fdea50
      Kirill Smelkov authored
      73fdea50
    • Kirill Smelkov's avatar
      gpython: Fix pymain to properly setup sys.path (take 2) · fb98e594
      Kirill Smelkov authored
      This continues 6b4990f6 (gpython: Fix pymain to properly setup
      sys.path[0]) - now we add test and verify that gpython actually sets
      sys.path in the same way as underlying python does:
      
      - don't add directory of gpython to sys.path
      - for `gpython file` add to sys.path[0] realpath(dir(file)) instead of
        dir(file)  (see PySys_SetArgvEx for the reference, and else - if we do
        not make this change - added test breaks)
      - for `gpython -m mod` add to sys.path[0] realpath('') instead of ''.
        This exactly matches PY3 behaviour, while PY2 behaviour is to add ''
        However PY3 behaviour
      
          * is more sane, and
          * fixes test_defer_excchain_traceback on py27-gevent
      
        -> so we stick to it.
      
      Now all tests pass again.
      fb98e594
    • Kirill Smelkov's avatar
      gpython: Preimport golang and gevent from exactly the same place as std python would do · 1f6f31cd
      Kirill Smelkov authored
      GPython preimports golang and gevent, but until now, it was preimporting
      them before adjusting sys.path according to given -c/-m/file/... And
      this was leading to difference in behaviour with standard Python where
      e.g. `import golang` might result in different files loaded if e.g.
      golang/ is there on cwd and -c/-m was given.
      
      -> Rework gpython startup so that golang/gevent preimport happens after
      sys.path initialization.
      
      This fixes `tox -e py27-gevent` which was previously failing on
      test_defer_excchain_dump_pytest / test_defer_excchain_dump_ipython
      because dumped paths were different than expected ones.
      
      It, however, breaks test_defer_excchain_traceback on py27-gevent.  ->
      We'll do more fixups to sys.path handling in the next patch which will
      fix that failure as well.
      1f6f31cd