Commit Graph

478 Commits

Author SHA1 Message Date
Antoine Pitrou 8ad982cccf Use PyUnicode_FromFormat() to create the temporary file name.
Also, as in importlib, append the id of an object to make the file name
pseudo-random.
2011-11-15 22:27:32 +01:00
Antoine Pitrou 1d8f3f451c Fix regression under Windows following b75b41237380 (from issue #13392) 2011-11-15 20:40:55 +01:00
Antoine Pitrou 28e401e717 Issue #13392: Writing a pyc file should now be atomic under Windows as well. 2011-11-15 19:15:19 +01:00
Charles-François Natali 0c929d9d39 Issue #13303: Fix bytecode file default permission. 2011-11-10 19:12:29 +01:00
Éric Araujo e64052176d Merge follow-up for #11254 and other changes from 3.2 2011-11-03 05:17:11 +01:00
Éric Araujo 5df1108de2 Add signatures to the docstring of functions added to imp by PEP 3147 2011-11-03 03:38:44 +01:00
Antoine Pitrou 8db076cf8a Issue #10363: Deallocate global locks in Py_Finalize(). 2011-10-30 19:13:55 +01:00
Charles-François Natali e695eec24a Issue #13303: Fix a race condition in the bytecode file creation. 2011-10-31 20:47:31 +01:00
Martin v. Löwis f45dee998f Port import_module_level to Unicode API. 2011-10-30 23:50:02 +01:00
Antoine Pitrou 6fbbc33440 Issue #10363: Deallocate global locks in Py_Finalize(). 2011-10-30 19:14:46 +01:00
Martin v. Löwis 796ea53937 Port PyImport_ReloadModule to Unicode API. 2011-10-30 09:07:07 +01:00
Antoine Pitrou b528fcf954 Fix test_imp failure under Windows 2011-10-25 00:21:02 +02:00
Martin v. Löwis 9af29d39af Rewrite find_module_path using unicode API. 2011-10-23 18:45:16 +02:00
Martin v. Löwis 2cc0cc54a2 Fix off-by-one error. 2011-10-23 18:41:56 +02:00
Martin v. Löwis fadcd317fa Drop unused macros. Use CACHEDIR consistently. 2011-10-23 18:08:20 +02:00
Martin v. Löwis 8a0ef78e8c Rewrite make_source_pathname using Unicode API. 2011-10-23 18:05:06 +02:00
Martin v. Löwis 30260a7fe3 Add ready checks for make_compiled_pathname. 2011-10-23 17:35:46 +02:00
Martin v. Löwis 2db72863fb Reformulate make_compiled_pathname in terms of unicode objects. 2011-10-23 17:29:08 +02:00
Antoine Pitrou 707033a694 Issue #13146: Writing a pyc file is now atomic under POSIX. 2011-10-17 19:28:44 +02:00
Martin v. Löwis bd928fef42 Rename _Py_identifier to _Py_IDENTIFIER. 2011-10-14 10:20:37 +02:00
Victor Stinner 8c9818980b Fix typo in import.c 2011-10-11 22:27:13 +02:00
Victor Stinner beac78bb24 Use PyUnicode_AsUnicodeAndSize() instead of PyUnicode_GET_SIZE() 2011-10-11 21:55:01 +02:00
Martin v. Löwis 1ee1b6fe0d Use identifier API for PyObject_GetAttrString. 2011-10-10 18:11:30 +02:00
Martin v. Löwis afe55bba33 Add API for static strings, primarily good for identifiers.
Thanks to Konrad Schöbel and Jasper Schulz for helping with the mass-editing.
2011-10-09 10:38:36 +02:00
Victor Stinner 46084bad88 Fix find_module_path(): make the string ready 2011-10-06 02:39:42 +02:00
Benjamin Peterson 014cc42db8 remove unused label 2011-10-02 13:19:16 -04:00
Martin v. Löwis 0b1d348990 Issue 13085: Fix some memory leaks. Patch by Stefan Krah. 2011-10-01 16:35:40 +02:00
Georg Brandl 4cb0de246c Rename new macros to conform to naming rules (function macros have "Py" prefix, not "PY"). 2011-09-28 21:49:49 +02:00
Martin v. Löwis d63a3b8beb Implement PEP 393. 2011-09-28 07:41:54 +02:00
Victor Stinner a1fe1f8dcf Merge 3.2: Issue #7732: Don't open a directory as a file anymore while
importing a module. Ignore the direcotry if its name matchs the module name
(e.g.  "__init__.py") and raise a ImportError instead.
2011-09-23 18:59:08 +02:00
Victor Stinner 53ffdc53bf Issue #7732: Don't open a directory as a file anymore while importing a
module. Ignore the direcotry if its name matchs the module name (e.g.
"__init__.py") and raise a ImportError instead.
2011-09-23 18:54:40 +02:00
Victor Stinner 9561d7c526 import.c: remove now useless arbitrary limit 2011-09-15 19:50:01 +02:00
Victor Stinner 84b8e40fd7 Merge 3.2: Fix the import machinery if there is an error on sys.path or sys.meta_path
find_module() now raises a RuntimeError, instead of ImportError, on an error on
sys.path or sys.meta_path because load_package() and import_submodule() returns
None and clear the exception if a ImportError occurred.
2011-09-15 19:38:54 +02:00
Victor Stinner 1619132e5d Fix the import machinery if there is an error on sys.path or sys.meta_path
find_module() now raises a RuntimeError, instead of ImportError, on an error on
sys.path or sys.meta_path because load_package() and import_submodule() returns
None and clear the exception if a ImportError occurred.
2011-09-15 19:28:05 +02:00
Victor Stinner 22d80bcf1b Merge 3.2: Remove unused variable if Python is build without threads 2011-09-02 00:13:16 +02:00
Victor Stinner 0af0306396 Remove unused variable if Python is build without threads 2011-09-02 00:11:43 +02:00
Victor Stinner d417d01ec8 call_find_module() handles dup() failure: raise an OSError exception 2011-06-20 15:16:55 +02:00
Victor Stinner 925ef39949 find_module_path_list() fails if _Py_fopen() failed and raised an exception
(UnicodeEncodeError).
2011-06-20 15:01:10 +02:00
Benjamin Peterson e7c15fa184 bump magic for super closure change 2011-06-19 19:54:45 -05:00
Benjamin Peterson 48deae12d5 some horrible preprocessing tricks to automatically update the tag 2011-06-03 17:50:16 -05:00
Victor Stinner 25095b2be6 Remove useless assignments
Warnings found by the the Clang Static Analyzer.
2011-05-26 13:47:08 +02:00
Benjamin Peterson 04778a8150 make PyImport_ImportModuleLevel's first arg const like similiar functions (closes #12173) 2011-05-25 09:29:00 -05:00
Victor Stinner 783c82c701 Close #11619: write_compiled_module() doesn't encode the filename
Reimplement open_exclusive() using _wopen() to avoid encoding the filename to
the filesystem encoding: use the Unicode version of the Windows API.
2011-04-20 03:27:51 +02:00
Victor Stinner fe7c5b5bdf Issue #9319: Include the filename in "Non-UTF8 code ..." syntax error. 2011-04-05 01:48:03 +02:00
Brett Cannon 442c9b92d8 Make importlib compatible with __import__ by "fixing" code.co_filename
paths.

__import__ does a little trick when importing from bytecode by
back-patching the co_filename paths to point to the file location
where the code object was loaded from, *not* where the code object was
originally created. This allows co_filename to point to a valid path.
Problem is that co_filename is immutable from Python, so a private
function -- imp._fix_co_filename() -- had to be introduced in order to
get things working properly. Originally the plan was to add a file
argument to marshal.loads(), but that failed as the algorithm used by
__import__ is not fully recursive as one might expect, so to be fully
backwards-compatible the code used by __import__ needed to be exposed.

This closes issue #6811 by taking a different approach than outlined
in the issue.
2011-03-23 16:14:42 -07:00
Victor Stinner e9ddbf65a8 Issue #11630, issue #3080: Fix refleak introduced by ef2b6305d395 2011-03-22 10:46:35 +01:00
Victor Stinner 9c61e24be4 Issue #3080: On DJGPP, case_bytes() returns -1 to signal an error if the file
cannot be found.
2011-03-22 01:22:27 +01:00
Victor Stinner 6ae1e7f04e Issue #3080: imp.load_module() accepts None for the module path
imp.find_module() returns None as module path for builtin and frozen builtins.
2011-03-20 22:37:17 +01:00
Victor Stinner cc9564ecd8 Issue #3080: Fix call to case_ok() in find_init_module() 2011-03-20 04:58:29 +01:00
Victor Stinner 1304f2d8a3 Issue #3080: Fix case_ok() using case_bytes()
Invert name and namelen arguments.
2011-03-20 04:28:55 +01:00