mirror of
https://github.com/python/cpython.git
synced 2025-08-04 00:48:58 +00:00

svn+ssh://pythondev@svn.python.org/python/trunk ........ r53954 | jeremy.hylton | 2007-02-26 10:41:18 -0800 (Mon, 26 Feb 2007) | 10 lines Do not copy free variables to locals in class namespaces. Fixes bug 1569356, but at the cost of a minor incompatibility in locals(). Add test that verifies that the class namespace is not polluted. Also clarify the behavior in the library docs. Along the way, cleaned up the dict_to_map and map_to_dict implementations and added some comments that explain what they do. ........ r53955 | jeremy.hylton | 2007-02-26 11:00:20 -0800 (Mon, 26 Feb 2007) | 2 lines Fix assertion. ........ r53969 | neal.norwitz | 2007-02-26 14:41:45 -0800 (Mon, 26 Feb 2007) | 3 lines When printing an unraisable error, don't print exceptions. before the name. This duplicates the behavior whening normally printing exceptions. ........ r53970 | andrew.kuchling | 2007-02-26 15:02:47 -0800 (Mon, 26 Feb 2007) | 1 line Markup fix ........ r53975 | neal.norwitz | 2007-02-26 15:48:27 -0800 (Mon, 26 Feb 2007) | 3 lines SF #1669182, 2.5 was already fixed. Just assert in 2.6 since string exceptions are gone. ........ r53976 | andrew.kuchling | 2007-02-26 15:54:17 -0800 (Mon, 26 Feb 2007) | 1 line Add some items ........ r53981 | jeremy.hylton | 2007-02-26 17:01:59 -0800 (Mon, 26 Feb 2007) | 4 lines Fix long-standing bug in name mangling for package imports Reported by Mike Verdone. ........ r53993 | jeremy.hylton | 2007-02-27 08:00:06 -0800 (Tue, 27 Feb 2007) | 2 lines tabify ........ r53994 | jeremy.hylton | 2007-02-27 08:13:23 -0800 (Tue, 27 Feb 2007) | 5 lines tabify Note that ast.c still has a mix of tabs and spaces, because it attempts to use four-space indents for more of the new code. ........ r53996 | jeremy.hylton | 2007-02-27 09:24:48 -0800 (Tue, 27 Feb 2007) | 2 lines whitespace normalization ........ r53997 | jeremy.hylton | 2007-02-27 10:29:45 -0800 (Tue, 27 Feb 2007) | 24 lines Add checking for a number of metaclass error conditions. We add some new rules that are required for preserving internal invariants of types. 1. If type (or a subclass of type) appears in bases, it must appear before any non-type bases. If a non-type base (like a regular new-style class) occurred first, it could trick type into allocating the new class an __dict__ which must be impossible. 2. There are several checks that are made of bases when creating a type. Those checks are now repeated when assigning to __bases__. We also add the restriction that assignment to __bases__ may not change the metaclass of the type. Add new tests for these cases and for a few other oddball errors that were no previously tested. Remove a crasher test that was fixed. Also some internal refactoring: Extract the code to find the most derived metaclass of a type and its bases. It is now needed in two places. Rewrite the TypeError checks in test_descr to use doctest. The tests now clearly show what exception they expect to see. ........ r53998 | jeremy.hylton | 2007-02-27 10:33:31 -0800 (Tue, 27 Feb 2007) | 2 lines Add news about changes to metaclasses and __bases__ error checking. ........ r54016 | armin.rigo | 2007-02-28 01:25:29 -0800 (Wed, 28 Feb 2007) | 3 lines Modify the segfaulting example to show why r53997 is not a solution to it. ........ r54022 | brett.cannon | 2007-02-28 10:15:00 -0800 (Wed, 28 Feb 2007) | 2 lines Add a test for instantiating SyntaxError with no arguments. ........ r54026 | raymond.hettinger | 2007-02-28 10:27:41 -0800 (Wed, 28 Feb 2007) | 1 line Docstring nit. ........ r54033 | raymond.hettinger | 2007-02-28 10:37:52 -0800 (Wed, 28 Feb 2007) | 1 line Prepare collections module for pure python code entries. ........ r54053 | raymond.hettinger | 2007-02-28 22:16:43 -0800 (Wed, 28 Feb 2007) | 1 line Add collections.NamedTuple ........ r54054 | neal.norwitz | 2007-02-28 23:04:41 -0800 (Wed, 28 Feb 2007) | 1 line Add Pat and Eric for work on PEP 3101 in the sandbox ........ r54061 | andrew.kuchling | 2007-03-01 06:36:12 -0800 (Thu, 01 Mar 2007) | 1 line Add NamedTuple ........ r54080 | georg.brandl | 2007-03-02 06:37:12 -0800 (Fri, 02 Mar 2007) | 2 lines Bug #1628895: some better tries to find HTML documentation in pydoc. ........ r54086 | raymond.hettinger | 2007-03-02 11:20:46 -0800 (Fri, 02 Mar 2007) | 1 line Fix embarrassing typo and fix constantification of None ........ r54088 | georg.brandl | 2007-03-02 12:30:14 -0800 (Fri, 02 Mar 2007) | 2 lines Bugs #1668032, #1668036, #1669304: clarify behavior of PyMem_Realloc and _Resize. ........ r54114 | georg.brandl | 2007-03-04 09:18:54 -0800 (Sun, 04 Mar 2007) | 2 lines Fix a bug in test_dict and test_userdict, found at the PyPy sprint. ........ r54124 | skip.montanaro | 2007-03-04 12:52:28 -0800 (Sun, 04 Mar 2007) | 2 lines Teach setup.py how to find Berkeley DB on Macs using MacPorts. ........ r54125 | skip.montanaro | 2007-03-04 12:54:12 -0800 (Sun, 04 Mar 2007) | 1 line note MacPorts/BerkDB change in setup.py ........ r54136 | neal.norwitz | 2007-03-04 23:52:01 -0800 (Sun, 04 Mar 2007) | 1 line Added Pete for 3101 too ........ r54138 | facundo.batista | 2007-03-05 08:31:54 -0800 (Mon, 05 Mar 2007) | 1 line Minor corrections to docs, and an explanation comentary ........ r54139 | georg.brandl | 2007-03-05 14:28:08 -0800 (Mon, 05 Mar 2007) | 3 lines Patch #1674228: when assigning a slice (old-style), check for the sq_ass_slice instead of the sq_slice slot. ........ r54149 | georg.brandl | 2007-03-06 01:33:01 -0800 (Tue, 06 Mar 2007) | 3 lines Nit: a struct field is set to GenericAlloc, not GenericAlloc(). ........ r54150 | georg.brandl | 2007-03-06 02:02:47 -0800 (Tue, 06 Mar 2007) | 3 lines Patch #1671450: add a section about subclassing builtin types to the "extending and embedding" tutorial. ........ r54152 | martin.v.loewis | 2007-03-06 02:41:24 -0800 (Tue, 06 Mar 2007) | 2 lines Patch #1121142: Implement ZipFile.open. ........ r54154 | georg.brandl | 2007-03-06 03:51:14 -0800 (Tue, 06 Mar 2007) | 2 lines A test case for the fix in #1674228. ........ r54156 | georg.brandl | 2007-03-06 03:52:24 -0800 (Tue, 06 Mar 2007) | 2 lines Patch #1672481: fix bug in idlelib.MultiCall. ........ r54159 | georg.brandl | 2007-03-06 04:17:50 -0800 (Tue, 06 Mar 2007) | 1 line Bug #1674503: close the file opened by execfile() in an error condition. ........ r54160 | georg.brandl | 2007-03-06 05:32:52 -0800 (Tue, 06 Mar 2007) | 3 lines Fix another reincarnation of bug #1576657 in defaultdict. ........ r54162 | georg.brandl | 2007-03-06 05:35:00 -0800 (Tue, 06 Mar 2007) | 2 lines A test case for the defaultdict KeyError bug. ........ r54164 | georg.brandl | 2007-03-06 05:37:45 -0800 (Tue, 06 Mar 2007) | 3 lines Patch #1663234: you can now run doctest on test files and modules using "python -m doctest [-v] filename ...". ........ r54165 | martin.v.loewis | 2007-03-06 06:43:00 -0800 (Tue, 06 Mar 2007) | 3 lines Patch #912410: Replace HTML entity references for attribute values in HTMLParser. ........ r54166 | skip.montanaro | 2007-03-06 07:41:38 -0800 (Tue, 06 Mar 2007) | 1 line patch 1673619 - identify extension modules which cannot be built ........ r54167 | guido.van.rossum | 2007-03-06 07:50:01 -0800 (Tue, 06 Mar 2007) | 5 lines Patch #1646728: datetime.fromtimestamp fails with negative fractional times. With unittest. Somebody please backport to 2.5. ........ r54169 | georg.brandl | 2007-03-06 09:49:14 -0800 (Tue, 06 Mar 2007) | 2 lines Fix cmp vs. key argument for list.sort. ........ r54170 | georg.brandl | 2007-03-06 10:21:32 -0800 (Tue, 06 Mar 2007) | 2 lines Small nit, found by Neal. ........ r54171 | georg.brandl | 2007-03-06 10:29:58 -0800 (Tue, 06 Mar 2007) | 3 lines Patch #1602128: clarify that richcmp methods can return NotImplemented and should return True or False otherwise. ........ r54173 | georg.brandl | 2007-03-06 10:41:12 -0800 (Tue, 06 Mar 2007) | 2 lines Patch #1638879: don't accept strings with embedded NUL bytes in long(). ........ r54175 | georg.brandl | 2007-03-06 10:47:31 -0800 (Tue, 06 Mar 2007) | 2 lines Patch #1673121: update README wrt. OSX default shell. ........ r54177 | georg.brandl | 2007-03-06 10:59:11 -0800 (Tue, 06 Mar 2007) | 3 lines Patch #1654417: make operator.{get,set,del}slice use the full range of Py_ssize_t. ........ r54180 | walter.doerwald | 2007-03-06 12:38:57 -0800 (Tue, 06 Mar 2007) | 4 lines Patch for bug #1633621: if curses.resizeterm() or curses.resize_term() is called, update _curses.LINES, _curses.COLS, curses.LINES and curses.COLS. ........ r54182 | walter.doerwald | 2007-03-06 13:15:24 -0800 (Tue, 06 Mar 2007) | 2 lines Document change to curses. ........ r54188 | georg.brandl | 2007-03-06 16:34:46 -0800 (Tue, 06 Mar 2007) | 5 lines Variant of patch #697613: don't exit the interpreter on a SystemExit exception if the -i command line option or PYTHONINSPECT environment variable is given, but break into the interactive interpreter just like on other exceptions or normal program exit. (backport) ........ r54189 | georg.brandl | 2007-03-06 16:40:28 -0800 (Tue, 06 Mar 2007) | 4 lines Patch #703779: unset __file__ in __main__ after running a file. This makes the filenames the warning module prints much more sensible when a PYTHONSTARTUP file is used. ........ r54192 | george.yoshida | 2007-03-06 20:21:18 -0800 (Tue, 06 Mar 2007) | 2 lines add versionadded info ........ r54195 | georg.brandl | 2007-03-06 23:39:06 -0800 (Tue, 06 Mar 2007) | 2 lines Patch #812285: allow multiple auth schemes in AbstractBasicAuthHandler. ........ r54197 | georg.brandl | 2007-03-07 00:31:51 -0800 (Wed, 07 Mar 2007) | 3 lines Patch #1001604: glob.glob() now returns unicode filenames if it was given a unicode argument and os.listdir() returns unicode filenames. ........ r54199 | georg.brandl | 2007-03-07 01:09:40 -0800 (Wed, 07 Mar 2007) | 3 lines Patches #1550273, #1550272: fix a few bugs in unittest and add a comprehensive test suite for the module. ........ r54201 | georg.brandl | 2007-03-07 01:21:06 -0800 (Wed, 07 Mar 2007) | 3 lines Patch #787789: allow to pass custom TestRunner instances to unittest's main() function. ........ r54202 | georg.brandl | 2007-03-07 01:34:45 -0800 (Wed, 07 Mar 2007) | 2 lines Patch #1669331: clarify shutil.copyfileobj() behavior wrt. file position. ........ r54204 | martin.v.loewis | 2007-03-07 03:04:33 -0800 (Wed, 07 Mar 2007) | 2 lines Bug #1115886: os.path.splitext('.cshrc') gives now ('.cshrc', ''). ........ r54206 | georg.brandl | 2007-03-07 03:37:42 -0800 (Wed, 07 Mar 2007) | 2 lines Patch #1675471: convert test_pty to unittest. ........ r54207 | georg.brandl | 2007-03-07 03:54:49 -0800 (Wed, 07 Mar 2007) | 4 lines Add some sanity checks to unittest.TestSuite's addTest(s) methods. Fixes #878275. ........ r54209 | guido.van.rossum | 2007-03-07 07:16:29 -0800 (Wed, 07 Mar 2007) | 3 lines Windows doesn't support negative timestamps. Skip the tests involving them if os.name == "nt". ........ r54219 | martin.v.loewis | 2007-03-08 05:42:43 -0800 (Thu, 08 Mar 2007) | 2 lines Add missing ) in parenthical remark. ........ r54220 | georg.brandl | 2007-03-08 09:49:06 -0800 (Thu, 08 Mar 2007) | 2 lines Fix #1676656: \em is different from \emph... ........ r54222 | georg.brandl | 2007-03-08 10:37:31 -0800 (Thu, 08 Mar 2007) | 2 lines Add a NEWS entry for rev. 54207,8. ........ r54225 | raymond.hettinger | 2007-03-08 11:24:27 -0800 (Thu, 08 Mar 2007) | 1 line SF 1676321: empty() returned wrong result ........ r54227 | collin.winter | 2007-03-08 11:58:14 -0800 (Thu, 08 Mar 2007) | 1 line Backported r54226 from p3yk: Move test_unittest, test_doctest and test_doctest2 higher up in the testing order. ........ r54230 | raymond.hettinger | 2007-03-08 13:33:47 -0800 (Thu, 08 Mar 2007) | 1 line SF #1637850: make_table in difflib did not work with unicode ........ r54232 | collin.winter | 2007-03-08 14:16:25 -0800 (Thu, 08 Mar 2007) | 1 line Patch #1668482: don't use '-' in mkstemp ........ r54233 | brett.cannon | 2007-03-08 15:58:11 -0800 (Thu, 08 Mar 2007) | 10 lines Introduce test.test_support.TransientResource. It's a context manager to surround calls to resources that may or may not be available. Specifying the expected exception and attributes to be raised if the resource is not available prevents overly broad catching of exceptions. This is meant to help suppress spurious failures by raising test.test_support.ResourceDenied if the exception matches. It would probably be good to go through the various network tests and surround the calls to catch connection timeouts (as done with test_socket_ssl in this commit). ........ r54234 | collin.winter | 2007-03-08 19:15:56 -0800 (Thu, 08 Mar 2007) | 1 line Patch #1481079: Support of HTTP_REFERER in CGIHTTPServer.py ........ r54235 | collin.winter | 2007-03-08 19:26:32 -0800 (Thu, 08 Mar 2007) | 1 line Add NEWS item for patch #1481079 (r54234). ........ r54237 | neal.norwitz | 2007-03-08 21:59:01 -0800 (Thu, 08 Mar 2007) | 1 line Fix SF #1676971, Complex OverflowError has a typo ........ r54239 | georg.brandl | 2007-03-09 04:58:41 -0800 (Fri, 09 Mar 2007) | 2 lines Typo. ........ r54240 | martin.v.loewis | 2007-03-09 07:35:55 -0800 (Fri, 09 Mar 2007) | 2 lines Patch #957003: Implement smtplib.LMTP. ........ r54243 | collin.winter | 2007-03-09 10:09:10 -0800 (Fri, 09 Mar 2007) | 2 lines Bug #1629566: clarify the docs on the return values of parsedate() and parsedate_tz() in email.utils and rfc822. ........ r54244 | thomas.heller | 2007-03-09 11:21:28 -0800 (Fri, 09 Mar 2007) | 3 lines Fix bug #1646630: ctypes.string_at(buf, 0) and ctypes.wstring_at(buf, 0) returned string up to the first NUL character. ........ r54245 | martin.v.loewis | 2007-03-09 11:36:01 -0800 (Fri, 09 Mar 2007) | 2 lines Add Ziga Seilnacht. ........ r54247 | collin.winter | 2007-03-09 12:33:07 -0800 (Fri, 09 Mar 2007) | 2 lines Patch #1491866: change the complex() constructor to allow parthensized forms. This means complex(repr(x)) now works instead of raising a ValueError. ........ r54248 | thomas.heller | 2007-03-09 12:39:22 -0800 (Fri, 09 Mar 2007) | 7 lines Bug #1651235: When a tuple was passed to a ctypes function call, Python would crash instead of raising an error. The crash was caused by a section of code that should have been removed long ago, at that time ctypes had other ways to pass parameters to function calls. ........ r54250 | collin.winter | 2007-03-09 15:30:39 -0800 (Fri, 09 Mar 2007) | 1 line Hashing simplification pointed out by Thomas Wouters. ........ r54252 | collin.winter | 2007-03-09 18:23:40 -0800 (Fri, 09 Mar 2007) | 5 lines * Unlink test files before and after each test; hopefully this will cut down on recent buildbot failures in test_islink. * Drop safe_remove() in favor of test_support.unlink(). * Fix the indentation of test_samefile so that it runs. ........ r54253 | collin.winter | 2007-03-09 18:51:26 -0800 (Fri, 09 Mar 2007) | 3 lines Bug #1531963: Make SocketServer.TCPServer's server_address always be equal to calling getsockname() on the server's socket. Will backport. ........ r54254 | neal.norwitz | 2007-03-09 19:19:18 -0800 (Fri, 09 Mar 2007) | 4 lines Simplify a little by handling the TCP case first. Update to use predominant style of spaces around = in args list and print to stderr if debugging. ........ r54256 | collin.winter | 2007-03-09 19:35:34 -0800 (Fri, 09 Mar 2007) | 1 line Add proper attribution for a bug fix. ........ r54257 | georg.brandl | 2007-03-09 23:38:14 -0800 (Fri, 09 Mar 2007) | 2 lines Typos. ........ r54260 | collin.winter | 2007-03-10 06:33:32 -0800 (Sat, 10 Mar 2007) | 1 line Convert an assert to a raise so it works even in the presence of -O. ........ r54262 | collin.winter | 2007-03-10 06:41:48 -0800 (Sat, 10 Mar 2007) | 2 lines Patch #1599845: Add an option to disable the implicit calls to server_bind() and server_activate() in the constructors for TCPServer, SimpleXMLRPCServer and DocXMLRPCServer. ........ r54268 | georg.brandl | 2007-03-11 00:28:46 -0800 (Sun, 11 Mar 2007) | 2 lines Add missing "return" statements in exception handler. ........ r54270 | ziga.seilnacht | 2007-03-11 08:54:54 -0700 (Sun, 11 Mar 2007) | 3 lines Patch #1675981: remove unreachable code from type.__new__() method. __dict__ and __weakref__ are removed from the slots tuple earlier in the code, in the loop that mangles slot names. Will backport. ........ r54271 | collin.winter | 2007-03-11 09:00:20 -0700 (Sun, 11 Mar 2007) | 3 lines Patch #1192590: Fix pdb's "ignore" and "condition" commands so they trap the IndexError caused by passing in an invalid breakpoint number. Will backport. ........ r54274 | vinay.sajip | 2007-03-11 11:32:07 -0700 (Sun, 11 Mar 2007) | 1 line Fix resource leak reported in SF #1516995. ........ r54278 | collin.winter | 2007-03-11 18:55:54 -0700 (Sun, 11 Mar 2007) | 4 lines Patch #1678662: ftp.python.org does not exist. So the testcode in urllib.py must use a more stable FTP. Will backport. ........ r54280 | barry.warsaw | 2007-03-11 20:20:01 -0700 (Sun, 11 Mar 2007) | 8 lines Tokio Kikuchi's fix for SF bug #1629369; folding whitespace allowed in the display name of an email address, e.g. Foo \tBar <foo@example.com> Test case added by Barry. ........ r54282 | skip.montanaro | 2007-03-11 20:30:50 -0700 (Sun, 11 Mar 2007) | 4 lines Sane humans would call these invalid tests, but Andrew McNamara pointed out that given the inputs in these tests Excel does indeed produce the output these tests expect. Document that for future confused folks. ........ r54283 | martin.v.loewis | 2007-03-12 03:50:39 -0700 (Mon, 12 Mar 2007) | 2 lines Bug #1675511: Use -Kpic instead of -xcode=pic32 on Solaris/x86. ........ r54285 | martin.v.loewis | 2007-03-12 04:01:10 -0700 (Mon, 12 Mar 2007) | 2 lines Patch #1677862: Require a space or tab after import in .pth files. ........ r54287 | georg.brandl | 2007-03-12 06:17:36 -0700 (Mon, 12 Mar 2007) | 8 lines Backport from Py3k branch: Patch #1591665: implement the __dir__() special function lookup in PyObject_Dir. Had to change a few bits of the patch because classobjs and __methods__ are still in Py2.6. ........ r54288 | georg.brandl | 2007-03-12 07:30:05 -0700 (Mon, 12 Mar 2007) | 3 lines Bug #1678647: write a newline after printing an exception in any case, even when converting the value to a string failed. ........ r54290 | collin.winter | 2007-03-12 08:57:19 -0700 (Mon, 12 Mar 2007) | 1 line Patch #1678088: convert test_operations to use unittest, fold the result into test_dict. ........ r54291 | collin.winter | 2007-03-12 09:11:39 -0700 (Mon, 12 Mar 2007) | 3 lines Bug #742342: make Python stop segfaulting on infinitely-recursive reload()s. Fixed by patch #922167. Will backport. ........ r54292 | georg.brandl | 2007-03-12 09:15:09 -0700 (Mon, 12 Mar 2007) | 2 lines Typo fix. ........ r54295 | collin.winter | 2007-03-12 10:24:07 -0700 (Mon, 12 Mar 2007) | 1 line Patch #1670993: Refactor test_threadedtempfile.py to use unittest. ........ r54296 | tim.peters | 2007-03-12 11:07:52 -0700 (Mon, 12 Mar 2007) | 2 lines Whitespace normalization. ........ r54297 | tim.peters | 2007-03-12 11:09:22 -0700 (Mon, 12 Mar 2007) | 2 lines Set missing svn:eol-style property on text files. ........ r54315 | brett.cannon | 2007-03-12 19:34:09 -0700 (Mon, 12 Mar 2007) | 8 lines Add test.test_support.transient_internet . Returns a context manager that nests test.test_support.TransientResource context managers that capture exceptions raised when the Internet connection is flaky. Initially using in test_socket_ssl but should probably be expanded to cover any test that should not raise the captured exceptions if the Internet connection works. ........ r54316 | brett.cannon | 2007-03-12 20:05:40 -0700 (Mon, 12 Mar 2007) | 2 lines Fix a typo where the variable name was not updated. ........ r54318 | neal.norwitz | 2007-03-12 21:59:58 -0700 (Mon, 12 Mar 2007) | 1 line Add Jerry Seutter for a bunch of his recent patches refactoring tests ........ r54319 | neal.norwitz | 2007-03-12 22:07:14 -0700 (Mon, 12 Mar 2007) | 7 lines Add some other acks for recent checkins: Brian Leair - 922167 Tomer Filiba - 1591665 Jeremy Jones - 1192590 ........ r54321 | neal.norwitz | 2007-03-12 22:31:38 -0700 (Mon, 12 Mar 2007) | 9 lines Fix some style nits: * lines too long * wrong indentation * space after a function name * wrong function name in error string * simplifying some logic Also add an error check to PyDict_SetItemString. ........ r54322 | georg.brandl | 2007-03-13 00:23:16 -0700 (Tue, 13 Mar 2007) | 2 lines Typo and grammar fixes. ........ r54323 | georg.brandl | 2007-03-13 00:50:57 -0700 (Tue, 13 Mar 2007) | 2 lines Patch #1679379: add documentation for fnmatch.translate(). ........ r54325 | georg.brandl | 2007-03-13 00:57:51 -0700 (Tue, 13 Mar 2007) | 2 lines Patch #1642844: comments to clarify the complexobject constructor. ........ r54326 | georg.brandl | 2007-03-13 01:14:27 -0700 (Tue, 13 Mar 2007) | 3 lines Patch #1668100: urllib2 now correctly raises URLError instead of OSError if accessing a local file via the file:// protocol fails. ........ r54327 | georg.brandl | 2007-03-13 02:32:11 -0700 (Tue, 13 Mar 2007) | 4 lines Patch #1635454: the csv.DictWriter class now includes the offending field names in its exception message if you try to write a record with a dictionary containing fields not in the CSV field names list. ........ r54328 | georg.brandl | 2007-03-13 02:41:31 -0700 (Tue, 13 Mar 2007) | 3 lines Patch #1555098: use str.join() instead of repeated string concatenation in robotparser. ........ r54329 | georg.brandl | 2007-03-13 03:06:48 -0700 (Tue, 13 Mar 2007) | 3 lines Patch #1542681: add entries for "with", "as" and "CONTEXTMANAGERS" to pydoc's help keywords. ........ r54331 | georg.brandl | 2007-03-13 03:19:22 -0700 (Tue, 13 Mar 2007) | 3 lines Patch #1569798: fix a bug in distutils when building Python from a directory within sys.exec_prefix. ........ r54333 | martin.v.loewis | 2007-03-13 03:24:00 -0700 (Tue, 13 Mar 2007) | 4 lines Patch #1449244: Support Unicode strings in email.message.Message.{set_charset,get_content_charset}. Will backport. ........ r54335 | lars.gustaebel | 2007-03-13 03:47:19 -0700 (Tue, 13 Mar 2007) | 34 lines This is the implementation of POSIX.1-2001 (pax) format read/write support. The TarInfo class now contains all necessary logic to process and create tar header data which has been moved there from the TarFile class. The fromtarfile() method was added. The new path and linkpath properties are aliases for the name and linkname attributes in correspondence to the pax naming scheme. The TarFile constructor and classmethods now accept a number of keyword arguments which could only be set as attributes before (e.g. dereference, ignore_zeros). The encoding and pax_headers arguments were added for pax support. There is a new tarinfo keyword argument that allows using subclassed TarInfo objects in TarFile. The boolean TarFile.posix attribute is deprecated, because now three tar formats are supported. Instead, the desired format for writing is specified using the constants USTAR_FORMAT, GNU_FORMAT and PAX_FORMAT as the format keyword argument. This change affects TarInfo.tobuf() as well. The test suite has been heavily reorganized and partially rewritten. A new testtar.tar was added that contains sample data in many formats from 4 different tar programs. Some bugs and quirks that also have been fixed: Directory names do no longer have a trailing slash in TarInfo.name or TarFile.getnames(). Adding the same file twice does not create a hardlink file member. The TarFile constructor does no longer need a name argument. The TarFile._mode attribute was renamed to mode and contains either 'r', 'w' or 'a'. ........ r54336 | georg.brandl | 2007-03-13 05:34:25 -0700 (Tue, 13 Mar 2007) | 3 lines Bug #1622896: fix a rare corner case where the bz2 module raised an error in spite of a succesful compression. ........ r54338 | lars.gustaebel | 2007-03-13 08:47:07 -0700 (Tue, 13 Mar 2007) | 3 lines Quick fix for tests that fail on systems with an encoding other than 'iso8859-1'. ........ r54339 | georg.brandl | 2007-03-13 10:43:32 -0700 (Tue, 13 Mar 2007) | 4 lines Patch #1603688: ConfigParser.SafeConfigParser now checks values that are set for invalid interpolation sequences that would lead to errors on reading back those values. ........ r54341 | georg.brandl | 2007-03-13 11:15:41 -0700 (Tue, 13 Mar 2007) | 3 lines Patch #1581073: add a flag to textwrap that prevents the dropping of whitespace while wrapping. ........ r54343 | georg.brandl | 2007-03-13 11:24:40 -0700 (Tue, 13 Mar 2007) | 2 lines Patch #1605192: list allowed states in error messages for imaplib. ........ r54344 | georg.brandl | 2007-03-13 11:31:49 -0700 (Tue, 13 Mar 2007) | 4 lines Patch #1537850: tempfile.NamedTemporaryFile now has a "delete" parameter which can be set to False to prevent the default delete-on-close behavior. ........ r54345 | collin.winter | 2007-03-13 11:53:04 -0700 (Tue, 13 Mar 2007) | 9 lines Add acks for recent patch checkins: Arvin Schnell - 1668482 S?\195?\169bastien Martini - 1481079 Heiko Wundram - 1491866 Damon Kohler - 1545011 Peter Parente - 1599845 Bjorn Lindqvist - 1678662 ........ r54346 | georg.brandl | 2007-03-13 12:00:36 -0700 (Tue, 13 Mar 2007) | 2 lines Acks for recent patches. ........ r54347 | georg.brandl | 2007-03-13 12:18:18 -0700 (Tue, 13 Mar 2007) | 3 lines Fix a tab. ........ r54348 | georg.brandl | 2007-03-13 12:32:21 -0700 (Tue, 13 Mar 2007) | 4 lines Patch #1533909: the timeit module now accepts callables in addition to strings for the code to time and the setup code. Also added two convenience functions for instantiating a Timer and calling its methods. ........ r54352 | georg.brandl | 2007-03-13 13:02:57 -0700 (Tue, 13 Mar 2007) | 3 lines Patch #1530482: add pydoc.render_doc() which returns the documentation for a thing instead of paging it to stdout, which pydoc.doc() does. ........ r54357 | thomas.heller | 2007-03-13 13:42:52 -0700 (Tue, 13 Mar 2007) | 1 line Patch #1649190: Adding support for _Bool to ctypes as c_bool, by David Remahl. ........ r54358 | georg.brandl | 2007-03-13 13:46:32 -0700 (Tue, 13 Mar 2007) | 2 lines Patch #1444529: the builtin compile() now accepts keyword arguments. (backport) ........ r54359 | thomas.heller | 2007-03-13 14:01:39 -0700 (Tue, 13 Mar 2007) | 1 line Add versionadded marker for ctypes.c_bool. ........ r54360 | georg.brandl | 2007-03-13 14:08:15 -0700 (Tue, 13 Mar 2007) | 3 lines Patch #1393667: pdb now has a "run" command which restarts the debugged Python program, optionally with different arguments. ........ r54361 | georg.brandl | 2007-03-13 14:32:01 -0700 (Tue, 13 Mar 2007) | 3 lines Deprecate commands.getstatus(). ........ r54362 | georg.brandl | 2007-03-13 14:32:56 -0700 (Tue, 13 Mar 2007) | 2 lines NEWS entry for getstatus() deprecation. ........ r54363 | georg.brandl | 2007-03-13 14:58:44 -0700 (Tue, 13 Mar 2007) | 4 lines Patch #1429539: pdb now correctly initializes the __main__ module for the debugged script, which means that imports from __main__ work correctly now. ........ r54364 | georg.brandl | 2007-03-13 15:07:36 -0700 (Tue, 13 Mar 2007) | 4 lines Patch #957650: "%var%" environment variable references are now properly expanded in ntpath.expandvars(), also "~user" home directory references are recognized and handled on Windows. ........ r54365 | georg.brandl | 2007-03-13 15:16:30 -0700 (Tue, 13 Mar 2007) | 2 lines Patch #1194449: correctly detect unbound methods in pydoc. ........ r54367 | georg.brandl | 2007-03-13 15:49:43 -0700 (Tue, 13 Mar 2007) | 5 lines Patch #1185447: binascii.b2a_qp() now correctly quotes binary characters with ASCII value less than 32. Also, it correctly quotes dots only if they occur on a single line, as opposed to the previous behavior of quoting dots if they are the second character of any line. ........ r54368 | collin.winter | 2007-03-13 16:02:15 -0700 (Tue, 13 Mar 2007) | 1 line Inline PyImport_GetModulesReloading(). ........ r54371 | barry.warsaw | 2007-03-13 21:59:50 -0700 (Tue, 13 Mar 2007) | 6 lines SF bug #1582282; decode_header() incorrectly splits not-conformant RFC 2047-like headers where there is no whitespace between encoded words. This fix changes the matching regexp to include a trailing lookahead assertion that the closing ?= must be followed by whitespace, newline, or end-of-string. This also changes the regexp to add the MULTILINE flag. ........ r54372 | gregory.p.smith | 2007-03-14 00:17:40 -0700 (Wed, 14 Mar 2007) | 2 lines correct order and names of the less often used keyword parameters. ........ r54373 | gregory.p.smith | 2007-03-14 00:19:50 -0700 (Wed, 14 Mar 2007) | 5 lines Its time to stop listing (Unix, Windows) when we really mean "everything but Mac OS 9" now that nobody is likely to use Python on Mac OS 9 and most of the (Mac) platform items are all OS X special API specific since OS X is unixy enough for these modules to be available out of the box. ........ r54376 | georg.brandl | 2007-03-14 01:27:52 -0700 (Wed, 14 Mar 2007) | 4 lines Bug #767111: fix long-standing bug in urllib which caused an AttributeError instead of an IOError when the server's response didn't contain a valid HTTP status line. ........ r54378 | ziga.seilnacht | 2007-03-14 05:24:09 -0700 (Wed, 14 Mar 2007) | 4 lines Patch #1680015: Don't modify __slots__ tuple if it contains an unicode name. Remove a reference leak that happened if the name could not be converted to string. Will backport. ........ r54386 | martin.v.loewis | 2007-03-14 13:02:31 -0700 (Wed, 14 Mar 2007) | 3 lines Patch #1559413: Fix test_cmd_line if sys.executable contains a space. Will backport. ........ r54389 | brett.cannon | 2007-03-14 14:40:13 -0700 (Wed, 14 Mar 2007) | 3 lines Note how test_socket_ssl has various exceptions that deal with a flaky Net connection are silenced. ........ r54390 | brett.cannon | 2007-03-14 14:44:15 -0700 (Wed, 14 Mar 2007) | 2 lines Raise ResourceDenied in test_urllib2net when the Net connection goes bad. ........ r54391 | neal.norwitz | 2007-03-14 21:41:20 -0700 (Wed, 14 Mar 2007) | 1 line Wrap a long line and fix a typo (is -> if) ........ r54392 | georg.brandl | 2007-03-15 00:38:14 -0700 (Thu, 15 Mar 2007) | 3 lines Patch #1680978: consistently use "alive" instead of "active" in the thread lib doc. ........ r54394 | georg.brandl | 2007-03-15 00:41:30 -0700 (Thu, 15 Mar 2007) | 3 lines Patch #1681153: the wave module now closes a file object it opened if initialization failed. ........ r54397 | ziga.seilnacht | 2007-03-15 04:44:55 -0700 (Thu, 15 Mar 2007) | 3 lines Patch #1462488: prevent a segfault in object_reduce_ex() by splitting the implementation for __reduce__ and __reduce_ex__ into two separate functions. Fixes bug #931877. Will backport. ........ r54404 | collin.winter | 2007-03-15 21:11:30 -0700 (Thu, 15 Mar 2007) | 3 lines Patch #1642547: Fix an error/crash when encountering syntax errors in complex if statements. Will backport. ........ r54406 | georg.brandl | 2007-03-16 00:55:09 -0700 (Fri, 16 Mar 2007) | 5 lines Bug #1681228: the webbrowser module now correctly uses the default GNOME or KDE browser, depending on whether there is a session of one of those present. Also, it tries the Windows default browser before trying Mozilla variants. (backport) ........ r54407 | georg.brandl | 2007-03-16 01:22:40 -0700 (Fri, 16 Mar 2007) | 4 lines Patch #1273829: os.walk() now has a "followlinks" parameter. If set to True (which is not the default), it visits symlinks pointing to directories. ........ r54408 | georg.brandl | 2007-03-16 01:24:21 -0700 (Fri, 16 Mar 2007) | 2 lines Add \versionadded tag. ........ r54409 | georg.brandl | 2007-03-16 01:33:47 -0700 (Fri, 16 Mar 2007) | 2 lines RFE #1670167: fix in isinstance() docs. ........ r54412 | ziga.seilnacht | 2007-03-16 04:59:38 -0700 (Fri, 16 Mar 2007) | 3 lines Patch #1623563: allow __class__ assignment for classes with __slots__. The old and the new class are still required to have the same slot names, but the order in which they are specified is not relevant. ........ r54413 | ziga.seilnacht | 2007-03-16 05:11:11 -0700 (Fri, 16 Mar 2007) | 2 lines Whitespace cleanup. Also remove the empty lines from the previous check in. ........ r54414 | jeremy.hylton | 2007-03-16 07:49:11 -0700 (Fri, 16 Mar 2007) | 2 lines Remove warning: funcion declaration isn't a prototype ........ r54415 | jeremy.hylton | 2007-03-16 08:59:47 -0700 (Fri, 16 Mar 2007) | 11 lines Clean up formatting of this file. The file should now follow PEP 7, except that it uses 4 space indents (in the style of Py3k). This particular code would be really hard to read with the regular tab idents. Other changes: - reflow long lines - change multi-line conditionals to have test at end of line ........ r54417 | collin.winter | 2007-03-16 14:13:35 -0700 (Fri, 16 Mar 2007) | 1 line Patch #1676994: Refactor test_popen2 to use unittest. ........ r54418 | collin.winter | 2007-03-16 14:15:35 -0700 (Fri, 16 Mar 2007) | 1 line Remove test/output/test_popen2 (missed in r54417). ........ r54419 | collin.winter | 2007-03-16 15:16:08 -0700 (Fri, 16 Mar 2007) | 1 line Patch 1339796: add a relpath() function to os.path. ........ r54421 | georg.brandl | 2007-03-17 09:08:45 -0700 (Sat, 17 Mar 2007) | 5 lines Patch #1675423: PyComplex_AsCComplex() now tries to convert an object to complex using its __complex__() method before falling back to the __float__() method. Therefore, the functions in the cmath module now can operate on objects that define a __complex__() method. (backport) ........ r54423 | gregory.p.smith | 2007-03-17 15:33:35 -0700 (Sat, 17 Mar 2007) | 2 lines move note to the correct section ........ r54426 | georg.brandl | 2007-03-18 01:25:00 -0700 (Sun, 18 Mar 2007) | 2 lines Patch #1682878: the new socket methods are recv_into and recvfrom_into, not *_buf. ........ r54432 | georg.brandl | 2007-03-18 11:28:25 -0700 (Sun, 18 Mar 2007) | 2 lines Patch #1678339: test case for bug in difflib. ........ r54439 | collin.winter | 2007-03-19 11:52:08 -0700 (Mon, 19 Mar 2007) | 1 line Patch #1630118: add a SpooledTemporaryFile class to tempfile. ........ r54441 | georg.brandl | 2007-03-19 12:02:48 -0700 (Mon, 19 Mar 2007) | 2 lines Patch #1683328: fixes and enhancements for "unparse" demo. ........ r54456 | neal.norwitz | 2007-03-19 22:07:28 -0700 (Mon, 19 Mar 2007) | 1 line Add some doc that was left out from some change to platform.py ........ r54457 | neal.norwitz | 2007-03-19 22:08:23 -0700 (Mon, 19 Mar 2007) | 1 line Add a comment about 3k migration ........ r54458 | neal.norwitz | 2007-03-19 22:21:21 -0700 (Mon, 19 Mar 2007) | 1 line Get rid of deprecation warning when testing commands.getstatus() ........ r54459 | neal.norwitz | 2007-03-19 22:23:09 -0700 (Mon, 19 Mar 2007) | 4 lines Try backing out 54407 to see if it corrects the problems on the Windows buildbots. This rev was backported, so we will need to keep both branches in sync, pending the outcome of the test after this checkin. ........ r54460 | neal.norwitz | 2007-03-19 23:13:25 -0700 (Mon, 19 Mar 2007) | 1 line Try to make this test more resistant to dropping from previous runs (ie, files that may exist but cause the test to fail). Should be backported (assuming it works :-) ........ r54461 | neal.norwitz | 2007-03-19 23:16:26 -0700 (Mon, 19 Mar 2007) | 1 line Try to make this test more resistant to dropping from previous runs (ie, files that may exist but cause the test to fail). Should be backported (assuming it works :-) ........ r54462 | neal.norwitz | 2007-03-19 23:53:17 -0700 (Mon, 19 Mar 2007) | 5 lines Try to be a little more resilient to errors. This might help the test pass, but my guess is that it won't. I'm guessing that some other test is leaving this file open which means it can't be removed under Windows AFAIK. ........ r54463 | neal.norwitz | 2007-03-20 01:14:57 -0700 (Tue, 20 Mar 2007) | 8 lines Try to get test_urllib to pass on Windows by closing the file. I'm guessing that's the problem. h.getfile() must be called *after* h.getreply() and the fp can be None. I'm not entirely convinced this is the best fix (or even correct). The buildbots will tell us if things improve or not. I don't know if this needs to be backported (assuming it actually works). ........ r54465 | raymond.hettinger | 2007-03-20 14:27:24 -0700 (Tue, 20 Mar 2007) | 1 line Extend work on rev 52962 and 53829 eliminating redundant PyObject_Hash() calls and fixing set/dict interoperability. ........ r54468 | georg.brandl | 2007-03-20 16:05:14 -0700 (Tue, 20 Mar 2007) | 2 lines Fix for glob.py if filesystem encoding is None. ........ r54479 | neal.norwitz | 2007-03-20 23:39:48 -0700 (Tue, 20 Mar 2007) | 1 line Remove unused file spotted by Paul Hankin ........ r54480 | georg.brandl | 2007-03-21 02:00:39 -0700 (Wed, 21 Mar 2007) | 3 lines Patch #1682205: a TypeError while unpacking an iterable is no longer masked by a generic one with the message "unpack non-sequence". ........ r54482 | georg.brandl | 2007-03-21 02:10:29 -0700 (Wed, 21 Mar 2007) | 2 lines New test for rev. 54407 which only uses directories under TESTFN. ........ r54483 | georg.brandl | 2007-03-21 02:16:53 -0700 (Wed, 21 Mar 2007) | 2 lines Patch #1684834: document some utility C API functions. ........ r54485 | georg.brandl | 2007-03-21 04:51:25 -0700 (Wed, 21 Mar 2007) | 2 lines Fix #1684254: split BROWSER contents with shlex to avoid displaying 'URL'. ........ r54487 | andrew.kuchling | 2007-03-21 07:32:43 -0700 (Wed, 21 Mar 2007) | 1 line Add comments on maintenance of this file ........ r54489 | andrew.kuchling | 2007-03-21 09:57:32 -0700 (Wed, 21 Mar 2007) | 1 line Fix sentence, and fix typo in example ........ r54490 | andrew.kuchling | 2007-03-21 09:59:20 -0700 (Wed, 21 Mar 2007) | 1 line Put code examples at left margin instead of indenting them ........ r54491 | facundo.batista | 2007-03-21 12:41:24 -0700 (Wed, 21 Mar 2007) | 1 line Minor clarification, saying that blocking means no timeout (from bug #882297) ........ r54492 | ziga.seilnacht | 2007-03-21 13:07:56 -0700 (Wed, 21 Mar 2007) | 2 lines Bug #1675967: re patterns pickled with older Python versions can now be unpickled. Will backport. ........ r54495 | raymond.hettinger | 2007-03-21 13:33:57 -0700 (Wed, 21 Mar 2007) | 1 line Add test and fix for fromkeys() optional argument. ........ r54524 | georg.brandl | 2007-03-22 01:05:45 -0700 (Thu, 22 Mar 2007) | 2 lines Bug #1685704: use -m switch in timeit docs. ........ r54533 | thomas.heller | 2007-03-22 12:44:31 -0700 (Thu, 22 Mar 2007) | 5 lines Back out "Patch #1643874: memory leak in ctypes fixed." The code in this patch leaves no way to give up the ownership of a BSTR instance. ........ r54538 | thomas.heller | 2007-03-22 13:34:37 -0700 (Thu, 22 Mar 2007) | 2 lines Explain the purpose of the b_needsfree flag (forward ported from release25-maint). ........ r54539 | guido.van.rossum | 2007-03-22 21:58:42 -0700 (Thu, 22 Mar 2007) | 12 lines - Bug #1683368: The object.__init__() and object.__new__() methods are now stricter in rejecting excess arguments. The only time when either allows excess arguments is when it is not overridden and the other one is. For backwards compatibility, when both are overridden, it is a deprecation warning (for now; maybe a Py3k warning later). When merging this into 3.0, the warnings should become errors. Note: without the change to string.py, lots of spurious warnings happen. What's going on there? ........ r54540 | neal.norwitz | 2007-03-22 22:17:23 -0700 (Thu, 22 Mar 2007) | 1 line Add Mark Dickinson for SF # 1675423. ........ r54541 | martin.v.loewis | 2007-03-23 03:35:49 -0700 (Fri, 23 Mar 2007) | 3 lines Patch #1686451: Fix return type for PySequence_{Count,Index,Fast_GET_SIZE}. Will backport. ........ r54543 | martin.v.loewis | 2007-03-23 06:27:15 -0700 (Fri, 23 Mar 2007) | 3 lines Bug #978833: Revert r50844, as it broke _socketobject.dup. Will backport. ........ r54545 | guido.van.rossum | 2007-03-23 11:53:03 -0700 (Fri, 23 Mar 2007) | 8 lines Add a type.__init__() method that enforces the same signature as type.__new__(), and then calls object.__init__(cls), just to be anal. This allows us to restore the code in string.py's _TemplateMetaclass that called super(...).__init__(name, bases, dct), which I commented out yesterday since it broke due to the stricter argument checking added to object.__init__(). ........ r54546 | facundo.batista | 2007-03-23 11:54:07 -0700 (Fri, 23 Mar 2007) | 4 lines Added a 'create_connect()' function to socket.py, which creates a connection with an optional timeout, and modified httplib.py to use this function in HTTPConnection. Applies patch 1676823. ........ r54547 | guido.van.rossum | 2007-03-23 12:39:01 -0700 (Fri, 23 Mar 2007) | 2 lines Add note about type.__init__(). ........ r54553 | thomas.heller | 2007-03-23 12:55:27 -0700 (Fri, 23 Mar 2007) | 5 lines Prevent creation (followed by a segfault) of array types when the size overflows the valid Py_ssize_t range. Check return values of PyMem_Malloc. Will backport to release25-maint. ........ r54555 | facundo.batista | 2007-03-23 13:23:08 -0700 (Fri, 23 Mar 2007) | 6 lines Surrounded with try/finally to socket's default timeout setting changes in the tests, so failing one test won't produce strange results in others. Also relaxed the timeout settings in the test (where actually the value didn't mean anything). ........ r54556 | collin.winter | 2007-03-23 15:24:39 -0700 (Fri, 23 Mar 2007) | 1 line Make test_relpath() pass on Windows. ........ r54559 | ziga.seilnacht | 2007-03-24 07:24:26 -0700 (Sat, 24 Mar 2007) | 6 lines Patch #1489771: update syntax rules in Python Reference Manual. Python 2.5 added support for explicit relative import statements and yield expressions, which were missing in the manual. Also fix grammar productions that used the names from the Grammar file, markup that broke the generated grammar.txt, and wrap some lines that broke the pdf output. Will backport. ........ r54565 | georg.brandl | 2007-03-24 15:20:34 -0700 (Sat, 24 Mar 2007) | 2 lines Remove typo accent. ........ r54566 | georg.brandl | 2007-03-24 15:27:56 -0700 (Sat, 24 Mar 2007) | 2 lines Revert accidental change. ........ r54567 | brett.cannon | 2007-03-24 18:32:36 -0700 (Sat, 24 Mar 2007) | 3 lines Change the docs to no longer claim that unittest is preferred over doctest for regression tests. ........ r54568 | facundo.batista | 2007-03-24 18:53:21 -0700 (Sat, 24 Mar 2007) | 4 lines Redone the tests, using the infrastructure already present for threading and socket serving. ........ r54570 | facundo.batista | 2007-03-24 20:20:05 -0700 (Sat, 24 Mar 2007) | 3 lines Closing the HTTP connection after each test, and listening more. ........ r54572 | georg.brandl | 2007-03-25 11:44:35 -0700 (Sun, 25 Mar 2007) | 2 lines Markup fix. ........ r54573 | georg.brandl | 2007-03-25 12:04:55 -0700 (Sun, 25 Mar 2007) | 2 lines Markup fix. ........ r54580 | facundo.batista | 2007-03-26 13:18:31 -0700 (Mon, 26 Mar 2007) | 5 lines Added an optional timeout to FTP class. Also I started a test_ftplib.py file to test the ftp lib (right now I included a basic test, the timeout one, and nothing else). ........ r54581 | georg.brandl | 2007-03-26 13:28:28 -0700 (Mon, 26 Mar 2007) | 2 lines Some nits. ........ r54582 | facundo.batista | 2007-03-26 13:56:09 -0700 (Mon, 26 Mar 2007) | 4 lines Forgot to add the file before the previous commit, here go the ftplib tests. ........ r54585 | facundo.batista | 2007-03-27 11:23:21 -0700 (Tue, 27 Mar 2007) | 5 lines Added an optional timeout to poplib.POP3. Also created a test_poplib.py file with a basic test and the timeout ones. Docs are also updated. ........ r54586 | facundo.batista | 2007-03-27 11:50:29 -0700 (Tue, 27 Mar 2007) | 3 lines The basic test cases of poplib.py. ........ r54594 | facundo.batista | 2007-03-27 20:45:20 -0700 (Tue, 27 Mar 2007) | 4 lines Bug 1688393. Adds a control of negative values in socket.recvfrom, which caused an ugly crash. ........ r54599 | facundo.batista | 2007-03-28 11:25:54 -0700 (Wed, 28 Mar 2007) | 5 lines Added timeout to smtplib (to SMTP and SMTP_SSL). Also created the test_smtplib.py file, with a basic test and the timeout ones. Docs are updated too. ........ r54603 | collin.winter | 2007-03-28 16:34:06 -0700 (Wed, 28 Mar 2007) | 3 lines Consolidate patches #1690164, 1683397, and 1690169, all of which refactor XML-related test suites. The patches are applied together because they use a common output/xmltests file. Thanks to Jerry Seutter for all three patches. ........ r54604 | collin.winter | 2007-03-28 19:28:16 -0700 (Wed, 28 Mar 2007) | 1 line Make test_zipfile clean up its temporary files properly. ........ r54605 | georg.brandl | 2007-03-29 00:41:32 -0700 (Thu, 29 Mar 2007) | 2 lines These are actually methods. ........ r54606 | georg.brandl | 2007-03-29 05:42:07 -0700 (Thu, 29 Mar 2007) | 4 lines In Windows' time.clock(), when QueryPerformanceFrequency() fails, the C lib's clock() is used, but it must be divided by CLOCKS_PER_SEC as for the POSIX implementation (thanks to #pypy). ........ r54608 | facundo.batista | 2007-03-29 11:22:35 -0700 (Thu, 29 Mar 2007) | 5 lines Added timout parameter to telnetlib.Telnet. Also created test_telnetlib.py with a basic test and timeout ones. Docs are also updated. ........ r54613 | facundo.batista | 2007-03-30 06:00:35 -0700 (Fri, 30 Mar 2007) | 4 lines Added the posibility to pass the timeout to FTP.connect, not only when instantiating the class. Docs and tests are updated. ........ r54614 | collin.winter | 2007-03-30 07:01:25 -0700 (Fri, 30 Mar 2007) | 1 line Bug #1688274: add documentation for C-level class objects. ........ r54615 | marc-andre.lemburg | 2007-03-30 08:01:42 -0700 (Fri, 30 Mar 2007) | 4 lines Bump the patch level version of distutils since there were a few bug fixes since the 2.5.0 release. ........ r54617 | georg.brandl | 2007-03-30 08:49:05 -0700 (Fri, 30 Mar 2007) | 2 lines Markup fix. ........ r54618 | georg.brandl | 2007-03-30 10:39:39 -0700 (Fri, 30 Mar 2007) | 2 lines Label name fix. ........ r54619 | georg.brandl | 2007-03-30 10:47:21 -0700 (Fri, 30 Mar 2007) | 2 lines Duplicate label fix. ........ r54620 | georg.brandl | 2007-03-30 10:48:39 -0700 (Fri, 30 Mar 2007) | 2 lines Markup fix. ........ r54623 | andrew.kuchling | 2007-03-30 11:00:15 -0700 (Fri, 30 Mar 2007) | 1 line Add item. (Oops, accidentally checked this in on my branch) ........ r54624 | georg.brandl | 2007-03-30 12:01:38 -0700 (Fri, 30 Mar 2007) | 2 lines Duplicate label fix. ........ r54625 | georg.brandl | 2007-03-30 12:14:02 -0700 (Fri, 30 Mar 2007) | 2 lines Markup fix. ........ r54629 | georg.brandl | 2007-03-31 03:17:31 -0700 (Sat, 31 Mar 2007) | 2 lines repair string literal. ........ r54630 | georg.brandl | 2007-03-31 04:54:58 -0700 (Sat, 31 Mar 2007) | 2 lines Markup fix. ........ r54631 | georg.brandl | 2007-03-31 04:58:36 -0700 (Sat, 31 Mar 2007) | 2 lines Duplicate label fix. ........ r54632 | georg.brandl | 2007-03-31 04:59:54 -0700 (Sat, 31 Mar 2007) | 2 lines Typo fix. ........ r54633 | neal.norwitz | 2007-03-31 11:54:18 -0700 (Sat, 31 Mar 2007) | 1 line Fix method names. Will backport. ........ r54634 | georg.brandl | 2007-03-31 11:56:11 -0700 (Sat, 31 Mar 2007) | 4 lines Bug #1655392: don't add -L/usr/lib/pythonX.Y/config to the LDFLAGS returned by python-config if Python was built with --enable-shared because that prevented the shared library from being used. ........ r54637 | collin.winter | 2007-03-31 12:31:34 -0700 (Sat, 31 Mar 2007) | 1 line Shut up an occaisonal buildbot error due to test files being left around. ........ r54644 | neal.norwitz | 2007-04-01 11:24:22 -0700 (Sun, 01 Apr 2007) | 11 lines SF #1685563, MSVCCompiler creates redundant and long PATH strings If MSVCCompiler.initialize() was called multiple times, the path would get duplicated. On Windows, this is a problem because the path is limited to 4k. There's no benefit in adding a path multiple times, so prevent that from occuring. We also normalize the path before checking for duplicates so things like /a and /a/ won't both be stored. Will backport. ........ r54646 | brett.cannon | 2007-04-01 11:47:27 -0700 (Sun, 01 Apr 2007) | 8 lines time.strptime's caching of its locale object was being recreated when the locale changed but not used during the function call it was recreated during. The test in this checkin is untested (OS X does not have the proper locale support for me to test), although the fix for the bug this deals with was tested by the OP (#1290505). Once the buildbots verify the test at least doesn't fail it becomes a backport candidate. ........ r54647 | brett.cannon | 2007-04-01 12:46:19 -0700 (Sun, 01 Apr 2007) | 3 lines Fix the test for recreating the locale cache object by not worrying about if one of the test locales cannot be set. ........ r54649 | georg.brandl | 2007-04-01 14:29:15 -0700 (Sun, 01 Apr 2007) | 2 lines Fix a lot of markup and meta-information glitches. ........ r54650 | georg.brandl | 2007-04-01 14:39:52 -0700 (Sun, 01 Apr 2007) | 2 lines Another fix. ........ r54651 | georg.brandl | 2007-04-01 15:39:10 -0700 (Sun, 01 Apr 2007) | 2 lines Lots of explicit class names for method and member descs. ........ r54652 | georg.brandl | 2007-04-01 15:40:12 -0700 (Sun, 01 Apr 2007) | 2 lines Explicit class names. ........ r54653 | georg.brandl | 2007-04-01 15:47:31 -0700 (Sun, 01 Apr 2007) | 2 lines Some semantic fixes. ........ r54654 | georg.brandl | 2007-04-01 16:29:10 -0700 (Sun, 01 Apr 2007) | 2 lines Remove bogus entry. ........ r54655 | georg.brandl | 2007-04-01 16:31:30 -0700 (Sun, 01 Apr 2007) | 2 lines Fix the class name of strings. ........ r54658 | raymond.hettinger | 2007-04-02 10:29:30 -0700 (Mon, 02 Apr 2007) | 1 line SF #1693079: Cannot save empty array in shelve ........ r54663 | raymond.hettinger | 2007-04-02 15:54:21 -0700 (Mon, 02 Apr 2007) | 3 lines Array module's buffer interface can now handle empty arrays. ........ r54664 | guido.van.rossum | 2007-04-02 16:55:37 -0700 (Mon, 02 Apr 2007) | 5 lines Fix warnings about object.__init__() signature. Two (test_array and test_descr) were bug IMO; the third (copy_reg) is a work-around which recognizes that object.__init__() doesn't do anything. ........ r54666 | raymond.hettinger | 2007-04-02 17:02:11 -0700 (Mon, 02 Apr 2007) | 1 line SF 1602378 Clarify docstrings for bisect ........ r54668 | raymond.hettinger | 2007-04-02 18:39:43 -0700 (Mon, 02 Apr 2007) | 3 lines SF #1382213: Tutorial section 9.5.1 ignores MRO for new-style classes ........ r54669 | matthias.klose | 2007-04-02 21:35:59 -0700 (Mon, 02 Apr 2007) | 4 lines - Fix an off-by-one bug in locale.strxfrm(). patch taken from http://bugs.debian.org/416934. ........ r54671 | georg.brandl | 2007-04-03 00:04:27 -0700 (Tue, 03 Apr 2007) | 9 lines Fix the strange case of \begin{methoddesc}[NNTP]{...} where \ifx#1\@undefined ended up comparing N and N, therefore executing the true part of the conditional, blowing up at \@undefined. ........ r54672 | facundo.batista | 2007-04-03 07:05:08 -0700 (Tue, 03 Apr 2007) | 4 lines Now using unittest for the tests infraestructure. Also split the tests in those who need the network, and that who doesn't. ........ r54673 | walter.doerwald | 2007-04-03 09:08:10 -0700 (Tue, 03 Apr 2007) | 4 lines Move the functionality for catching warnings in test_warnings.py into a separate class to that reusing the functionality in test_structmembers.py doesn't rerun the tests from test_warnings.py. ........ r54674 | walter.doerwald | 2007-04-03 09:16:24 -0700 (Tue, 03 Apr 2007) | 2 lines Document that CatchWarningTests is reused by test_structmembers.py. ........ r54675 | walter.doerwald | 2007-04-03 09:53:43 -0700 (Tue, 03 Apr 2007) | 4 lines Add tests for the filename. Test that the stacklevel is handled correctly. ........ r54676 | facundo.batista | 2007-04-03 10:29:48 -0700 (Tue, 03 Apr 2007) | 6 lines Added a SSL server to test_socket_ssl.py to be able to test locally. Now, it checks if have openssl available and run those specific tests (it starts openssl at the beggining of all the tests and then kills it at the end). ........ r54677 | walter.doerwald | 2007-04-03 11:33:29 -0700 (Tue, 03 Apr 2007) | 6 lines Implement a contextmanager test.test_support.catch_warning that can be used to catch the last warning issued by the warning framework. Change test_warnings.py and test_structmembers.py to use this new contextmanager. ........ r54678 | facundo.batista | 2007-04-03 14:15:34 -0700 (Tue, 03 Apr 2007) | 4 lines Changed the whole structure of startup and checking if the server is available. Hope to not get more false alarms. ........ r54681 | facundo.batista | 2007-04-04 07:10:40 -0700 (Wed, 04 Apr 2007) | 4 lines Fixed the way that the .pem files are looked for, and changed how to kill the process in win32 to use the _handle attribute. ........ r54682 | guido.van.rossum | 2007-04-04 10:43:02 -0700 (Wed, 04 Apr 2007) | 4 lines Fix a race condition in this test -- instead of assuming that it will take the test server thread at most 0.5 seconds to get ready, use an event variable. ........ r54683 | collin.winter | 2007-04-04 11:14:17 -0700 (Wed, 04 Apr 2007) | 1 line Clean up imports. ........ r54684 | collin.winter | 2007-04-04 11:16:24 -0700 (Wed, 04 Apr 2007) | 1 line Stop using test_support.verify(). ........ r54685 | martin.v.loewis | 2007-04-04 11:30:36 -0700 (Wed, 04 Apr 2007) | 2 lines Bug #1686475: Support stat'ing open files on Windows again. Will backport to 2.5. ........ r54687 | collin.winter | 2007-04-04 11:33:40 -0700 (Wed, 04 Apr 2007) | 1 line Make test_getopt use unittest. ........ r54688 | collin.winter | 2007-04-04 11:36:30 -0700 (Wed, 04 Apr 2007) | 1 line Make test_softspace use unittest. ........ r54689 | ziga.seilnacht | 2007-04-04 11:38:47 -0700 (Wed, 04 Apr 2007) | 2 lines Fix WalkTests.test_traversal() on Windows. The cleanup in MakedirTests.setUp() can now be removed. ........ r54695 | raymond.hettinger | 2007-04-05 11:00:03 -0700 (Thu, 05 Apr 2007) | 3 lines Bug #1563759: struct.unpack doens't support buffer protocol objects ........ r54697 | collin.winter | 2007-04-05 13:05:07 -0700 (Thu, 05 Apr 2007) | 1 line Convert test_long_future to use unittest. ........ r54698 | collin.winter | 2007-04-05 13:08:56 -0700 (Thu, 05 Apr 2007) | 1 line Convert test_normalization to use unittest. ........ r54699 | andrew.kuchling | 2007-04-05 18:11:58 -0700 (Thu, 05 Apr 2007) | 1 line Some grammar fixes ........ r54704 | collin.winter | 2007-04-06 12:27:40 -0700 (Fri, 06 Apr 2007) | 1 line Convert test_stringprep to use unittest. ........ r54705 | collin.winter | 2007-04-06 12:32:32 -0700 (Fri, 06 Apr 2007) | 1 line Import cleanup in test_crypt. ........ r54706 | collin.winter | 2007-04-06 13:00:05 -0700 (Fri, 06 Apr 2007) | 1 line Convert test_gc to use unittest. ........ r54707 | collin.winter | 2007-04-06 13:03:11 -0700 (Fri, 06 Apr 2007) | 1 line Convert test_module to use unittest. ........ r54711 | collin.winter | 2007-04-06 21:40:43 -0700 (Fri, 06 Apr 2007) | 1 line Convert test_fileinput to use unittest. ........ r54712 | brett.cannon | 2007-04-07 21:29:32 -0700 (Sat, 07 Apr 2007) | 5 lines Doc that file.next() has undefined behaviour when called on a file opened with 'w'. Closes bug #1569057. To be backported once 2.5 branch is unfrozen. ........ r54726 | vinay.sajip | 2007-04-09 09:16:10 -0700 (Mon, 09 Apr 2007) | 1 line Added optional timeout to SocketHandler.makeSocket (SF #1695948) ........ r54727 | ziga.seilnacht | 2007-04-09 12:10:29 -0700 (Mon, 09 Apr 2007) | 3 lines Patch #1695862: remove old test directory that causes test_urllib failures on Windows buildbots. The change is a one time fix and will be removed after a successful buildbot run. ........ r54729 | facundo.batista | 2007-04-09 20:00:37 -0700 (Mon, 09 Apr 2007) | 3 lines Minor fix to the tests pass ok even with -O. ........ r54730 | collin.winter | 2007-04-09 21:44:49 -0700 (Mon, 09 Apr 2007) | 1 line Typo fix. ........ r54732 | facundo.batista | 2007-04-10 05:58:45 -0700 (Tue, 10 Apr 2007) | 5 lines General clean-up. Lot of margin corrections, comments, some typos. Exceptions now are raised in the new style. And a mockup class is now also new style. Thanks Santiago Pereson. ........ r54741 | georg.brandl | 2007-04-10 14:39:38 -0700 (Tue, 10 Apr 2007) | 2 lines Repair a duplicate label and some obsolete uses of \setindexsubitem. ........ r54746 | andrew.kuchling | 2007-04-11 06:39:00 -0700 (Wed, 11 Apr 2007) | 1 line Add window.chgat() method, submitted via e-mail by Fabian Kreutz ........ r54747 | andrew.kuchling | 2007-04-11 06:42:25 -0700 (Wed, 11 Apr 2007) | 1 line Point readers at the patch submission instructions ........ r54748 | andrew.kuchling | 2007-04-11 06:47:13 -0700 (Wed, 11 Apr 2007) | 1 line Describe undocumented third argument to touchline() ........ r54757 | georg.brandl | 2007-04-11 10:16:24 -0700 (Wed, 11 Apr 2007) | 3 lines Add some missing NULL checks which trigger crashes on low-memory conditions. Found by Victor Stinner. Will backport when 2.5 branch is unfrozen. ........ r54760 | raymond.hettinger | 2007-04-11 11:40:58 -0700 (Wed, 11 Apr 2007) | 1 line SF 1191699: Make slices picklable ........ r54762 | georg.brandl | 2007-04-11 12:25:11 -0700 (Wed, 11 Apr 2007) | 2 lines Exceptions are no longer old-style instances. Fix accordingly. ........ r54763 | georg.brandl | 2007-04-11 16:28:44 -0700 (Wed, 11 Apr 2007) | 2 lines Repair missing spaces after \UNIX. ........ r54772 | raymond.hettinger | 2007-04-11 21:10:00 -0700 (Wed, 11 Apr 2007) | 1 line SF 1193128: Let str.translate(None) be an identity transformation ........ r54784 | georg.brandl | 2007-04-12 00:01:19 -0700 (Thu, 12 Apr 2007) | 2 lines Patch #1698951: clarify deprecation message in rexec and Bastion ........ r54785 | ziga.seilnacht | 2007-04-12 01:46:51 -0700 (Thu, 12 Apr 2007) | 2 lines Patch #1695862: remove the cleanup code, now that Windows buildbots are green again. ........ r54786 | walter.doerwald | 2007-04-12 03:35:00 -0700 (Thu, 12 Apr 2007) | 3 lines Fix utf-8-sig incremental decoder, which didn't recognise a BOM when the first chunk fed to the decoder started with a BOM, but was longer than 3 bytes. ........ r54807 | barry.warsaw | 2007-04-13 11:47:14 -0700 (Fri, 13 Apr 2007) | 8 lines Port r54805 from python25-maint branch: Add code to read from master_fd in the parent, breaking when we get an OSError (EIO can occur on Linux) or there's no more data to read. Without this, test_pty.py can hang on the waitpid() because the child is blocking on the stdout write. This will definitely happen on Mac OS X and could potentially happen on other platforms. See the comment for details. ........ r54812 | kristjan.jonsson | 2007-04-13 15:07:33 -0700 (Fri, 13 Apr 2007) | 1 line Fix a bug when using the __lltrace__ opcode tracer, and a problem sith signed chars in frameobject.c which can occur with opcodes > 127 ........ r54814 | kristjan.jonsson | 2007-04-13 15:20:13 -0700 (Fri, 13 Apr 2007) | 1 line Fix potential crash in path manipulation on windows ........ r54816 | trent.mick | 2007-04-13 16:22:05 -0700 (Fri, 13 Apr 2007) | 4 lines Add the necessary dependency for the Windows VC6 build to ensure 'pythoncore' is built before '_ctypes' is attempted. Will backport to 2.5 once it is unfrozen for 2.5.1. ........ r54825 | neal.norwitz | 2007-04-13 22:25:50 -0700 (Fri, 13 Apr 2007) | 3 lines When __slots__ are set to a unicode string, make it work the same as setting a plain string, ie don't expand to single letter identifiers. ........ r54841 | neal.norwitz | 2007-04-16 00:37:55 -0700 (Mon, 16 Apr 2007) | 1 line SF #1701207, Fix bogus assertion (and test it!) ........ r54844 | collin.winter | 2007-04-16 15:10:32 -0700 (Mon, 16 Apr 2007) | 1 line Check the availability of the urlfetch resource earlier than before. ........ r54849 | martin.v.loewis | 2007-04-16 22:02:01 -0700 (Mon, 16 Apr 2007) | 2 lines Add Travis Oliphant. ........ r54873 | brett.cannon | 2007-04-18 20:44:17 -0700 (Wed, 18 Apr 2007) | 2 lines Silence a compiler warning about incompatible pointer types. ........ r54874 | neal.norwitz | 2007-04-18 22:52:37 -0700 (Wed, 18 Apr 2007) | 2 lines SF #1703270, add missing declaration in readline.c to avoid compiler warning. ........ r54875 | armin.rigo | 2007-04-19 07:44:48 -0700 (Thu, 19 Apr 2007) | 8 lines Revert r53997 as per http://mail.python.org/pipermail/python-dev/2007-March/071796.html . I've kept a couple of still-valid extra tests in test_descr, but didn't bother to sort through the new comments and refactorings added in r53997 to see if some of them could be kept. If so, they could go in a follow-up check-in. ........ r54876 | armin.rigo | 2007-04-19 07:56:48 -0700 (Thu, 19 Apr 2007) | 2 lines Fix a usage of the dangerous pattern decref - modify field - incref. ........ r54884 | neal.norwitz | 2007-04-19 22:20:38 -0700 (Thu, 19 Apr 2007) | 9 lines Add an optional address to copy the failure mails to. Detect a conflict in the only file that should have outstanding changes when this script is run. This doesn't matter on the trunk, but does when run on a branch. Trunk always has the date set to today in boilerplate.tex. Each time a release is cut with a different date, a conflict occurs. (We could copy a known good version, but then we would lose changes to this file.) ........ r54918 | georg.brandl | 2007-04-21 13:35:38 -0700 (Sat, 21 Apr 2007) | 3 lines Bug #1704790: bind name "sys" locally in __del__ method so that it is not cleared before __del__ is run. ........ r54920 | facundo.batista | 2007-04-21 18:18:56 -0700 (Sat, 21 Apr 2007) | 5 lines Added tests for other methods of SSL object. Now we cover all the object methods. This is the final step to close the #451607 bug. ........ r54927 | facundo.batista | 2007-04-23 10:08:31 -0700 (Mon, 23 Apr 2007) | 5 lines As specified in RFC 2616, 2xx code indicates that the client's request was successfully received, understood, and accepted. Now in these cases no error is raised. Also fixed tests. ........ r54929 | collin.winter | 2007-04-23 20:43:46 -0700 (Mon, 23 Apr 2007) | 1 line Convert PyUnit -> unittest. ........ r54931 | collin.winter | 2007-04-23 21:09:52 -0700 (Mon, 23 Apr 2007) | 1 line Remove code that hasn't been called in years. ........ r54932 | neal.norwitz | 2007-04-23 21:53:12 -0700 (Mon, 23 Apr 2007) | 1 line Fix SF #1703110, Incorrect example for add_password() (use uri, not host) ........ r54934 | georg.brandl | 2007-04-24 03:36:42 -0700 (Tue, 24 Apr 2007) | 2 lines Some new year updates. ........ r54938 | facundo.batista | 2007-04-24 06:54:38 -0700 (Tue, 24 Apr 2007) | 4 lines Added a comment about last change in urllib2.py (all 2xx responses are ok now). ........ r54939 | georg.brandl | 2007-04-24 08:10:09 -0700 (Tue, 24 Apr 2007) | 2 lines Bug #1705717: error in sys.argv docs. ........ r54941 | georg.brandl | 2007-04-24 08:27:13 -0700 (Tue, 24 Apr 2007) | 4 lines Bug #1706381: Specifying the SWIG option "-c++" in the setup.py file (as opposed to the command line) will now write file names ending in ".cpp" too. ........ r54944 | raymond.hettinger | 2007-04-24 15:13:43 -0700 (Tue, 24 Apr 2007) | 1 line Fix markup ........ r54945 | kristjan.jonsson | 2007-04-24 17:10:50 -0700 (Tue, 24 Apr 2007) | 1 line Merge change 54909 from release25-maint: Fix several minor issues discovered using code analysis in VisualStudio 2005 Team Edition ........ r54947 | kristjan.jonsson | 2007-04-24 17:17:39 -0700 (Tue, 24 Apr 2007) | 1 line Make pythoncore compile cleanly with VisualStudio 2005. Used an explicit typecast to get a 64 bit integer, and undefined the Yield macro that conflicts with winbase.h ........ r54948 | kristjan.jonsson | 2007-04-24 17:19:26 -0700 (Tue, 24 Apr 2007) | 1 line Remove obsolete comment. Importing of .dll files has been discontinued, only .pyd files supported on windows now. ........ r54949 | georg.brandl | 2007-04-24 23:24:59 -0700 (Tue, 24 Apr 2007) | 2 lines Patch #1698768: updated the "using Python on the Mac" intro. ........ r54951 | georg.brandl | 2007-04-24 23:25:55 -0700 (Tue, 24 Apr 2007) | 2 lines Markup fix. ........ r54953 | neal.norwitz | 2007-04-24 23:30:05 -0700 (Tue, 24 Apr 2007) | 3 lines Whitespace normalization. Ugh, we really need to do this more often. You might want to review this change as it's my first time. Be gentle. :-) ........ r54956 | collin.winter | 2007-04-25 10:29:52 -0700 (Wed, 25 Apr 2007) | 1 line Standardize on test.test_support.run_unittest() (as opposed to a mix of run_unittest() and run_suite()). Also, add functionality to run_unittest() that admits usage of unittest.TestLoader.loadTestsFromModule(). ........ r54957 | collin.winter | 2007-04-25 10:37:35 -0700 (Wed, 25 Apr 2007) | 1 line Remove functionality from test_datetime.test_main() that does reference count checking; 'regrtest.py -R' is the way to do this kind of testing. ........ r54958 | collin.winter | 2007-04-25 10:57:53 -0700 (Wed, 25 Apr 2007) | 1 line Change test_support.have_unicode to use True/False instead of 1/0. ........ r54959 | tim.peters | 2007-04-25 11:47:18 -0700 (Wed, 25 Apr 2007) | 2 lines Whitespace normalization. ........ r54960 | tim.peters | 2007-04-25 11:48:35 -0700 (Wed, 25 Apr 2007) | 2 lines Set missing svn:eol-style property on text files. ........ r54961 | collin.winter | 2007-04-25 11:54:36 -0700 (Wed, 25 Apr 2007) | 1 line Import and raise statement cleanup. ........ r54969 | collin.winter | 2007-04-25 13:41:34 -0700 (Wed, 25 Apr 2007) | 1 line Convert test_ossaudiodev to use unittest. ........ r54974 | collin.winter | 2007-04-25 14:50:25 -0700 (Wed, 25 Apr 2007) | 1 line Fix an issue related to the unittest conversion. ........ r54979 | fred.drake | 2007-04-25 21:42:19 -0700 (Wed, 25 Apr 2007) | 1 line fix some markup errors ........ r54982 | kristjan.jonsson | 2007-04-26 02:15:08 -0700 (Thu, 26 Apr 2007) | 1 line Export function sanitize_the_mode from fileobject.c as _PyFile_SanitizeMode(). Use this function in posixmodule.c when implementing fdopen(). This fixes test_subprocess.py for a VisualStudio 2005 compile. ........ r54983 | kristjan.jonsson | 2007-04-26 06:44:16 -0700 (Thu, 26 Apr 2007) | 1 line The locale "En" appears not to be valid on windows underi VisualStudio.2005. Added "English" to the test_locale.py to make the testsuite pass for that build ........ r54984 | steve.holden | 2007-04-26 07:23:12 -0700 (Thu, 26 Apr 2007) | 1 line Minor wording change on slicing aide-memoire. ........ r54985 | kristjan.jonsson | 2007-04-26 08:24:54 -0700 (Thu, 26 Apr 2007) | 1 line Accomodate 64 bit time_t in the _bsddb module. ........
2454 lines
62 KiB
C
2454 lines
62 KiB
C
/* File object implementation */
|
|
|
|
#define PY_SSIZE_T_CLEAN
|
|
#include "Python.h"
|
|
#include "structmember.h"
|
|
|
|
#ifdef HAVE_SYS_TYPES_H
|
|
#include <sys/types.h>
|
|
#endif /* HAVE_SYS_TYPES_H */
|
|
|
|
#ifdef MS_WINDOWS
|
|
#define fileno _fileno
|
|
/* can simulate truncate with Win32 API functions; see file_truncate */
|
|
#define HAVE_FTRUNCATE
|
|
#define WIN32_LEAN_AND_MEAN
|
|
#include <windows.h>
|
|
#endif
|
|
|
|
#ifdef _MSC_VER
|
|
/* Need GetVersion to see if on NT so safe to use _wfopen */
|
|
#define WIN32_LEAN_AND_MEAN
|
|
#include <windows.h>
|
|
#endif /* _MSC_VER */
|
|
|
|
#if defined(PYOS_OS2) && defined(PYCC_GCC)
|
|
#include <io.h>
|
|
#endif
|
|
|
|
#define BUF(v) PyString_AS_STRING((PyStringObject *)v)
|
|
|
|
#ifndef DONT_HAVE_ERRNO_H
|
|
#include <errno.h>
|
|
#endif
|
|
|
|
#ifdef HAVE_GETC_UNLOCKED
|
|
#define GETC(f) getc_unlocked(f)
|
|
#define FLOCKFILE(f) flockfile(f)
|
|
#define FUNLOCKFILE(f) funlockfile(f)
|
|
#else
|
|
#define GETC(f) getc(f)
|
|
#define FLOCKFILE(f)
|
|
#define FUNLOCKFILE(f)
|
|
#endif
|
|
|
|
/* Bits in f_newlinetypes */
|
|
#define NEWLINE_UNKNOWN 0 /* No newline seen, yet */
|
|
#define NEWLINE_CR 1 /* \r newline seen */
|
|
#define NEWLINE_LF 2 /* \n newline seen */
|
|
#define NEWLINE_CRLF 4 /* \r\n newline seen */
|
|
|
|
#ifdef __cplusplus
|
|
extern "C" {
|
|
#endif
|
|
|
|
FILE *
|
|
PyFile_AsFile(PyObject *f)
|
|
{
|
|
if (f == NULL || !PyFile_Check(f))
|
|
return NULL;
|
|
else
|
|
return ((PyFileObject *)f)->f_fp;
|
|
}
|
|
|
|
PyObject *
|
|
PyFile_Name(PyObject *f)
|
|
{
|
|
if (f == NULL || !PyFile_Check(f))
|
|
return NULL;
|
|
else
|
|
return ((PyFileObject *)f)->f_name;
|
|
}
|
|
|
|
/* On Unix, fopen will succeed for directories.
|
|
In Python, there should be no file objects referring to
|
|
directories, so we need a check. */
|
|
|
|
static PyFileObject*
|
|
dircheck(PyFileObject* f)
|
|
{
|
|
#if defined(HAVE_FSTAT) && defined(S_IFDIR) && defined(EISDIR)
|
|
struct stat buf;
|
|
if (f->f_fp == NULL)
|
|
return f;
|
|
if (fstat(fileno(f->f_fp), &buf) == 0 &&
|
|
S_ISDIR(buf.st_mode)) {
|
|
#ifdef HAVE_STRERROR
|
|
char *msg = strerror(EISDIR);
|
|
#else
|
|
char *msg = "Is a directory";
|
|
#endif
|
|
PyObject *exc = PyObject_CallFunction(PyExc_IOError, "(is)",
|
|
EISDIR, msg);
|
|
PyErr_SetObject(PyExc_IOError, exc);
|
|
Py_XDECREF(exc);
|
|
return NULL;
|
|
}
|
|
#endif
|
|
return f;
|
|
}
|
|
|
|
|
|
static PyObject *
|
|
fill_file_fields(PyFileObject *f, FILE *fp, PyObject *name, char *mode,
|
|
int (*close)(FILE *))
|
|
{
|
|
assert(name != NULL);
|
|
assert(f != NULL);
|
|
assert(PyFile_Check(f));
|
|
assert(f->f_fp == NULL);
|
|
|
|
Py_DECREF(f->f_name);
|
|
Py_DECREF(f->f_mode);
|
|
Py_DECREF(f->f_encoding);
|
|
|
|
Py_INCREF(name);
|
|
f->f_name = name;
|
|
|
|
f->f_mode = PyString_FromString(mode);
|
|
|
|
f->f_close = close;
|
|
f->f_binary = strchr(mode,'b') != NULL;
|
|
f->f_buf = NULL;
|
|
f->f_univ_newline = (strchr(mode, 'U') != NULL);
|
|
f->f_newlinetypes = NEWLINE_UNKNOWN;
|
|
f->f_skipnextlf = 0;
|
|
Py_INCREF(Py_None);
|
|
f->f_encoding = Py_None;
|
|
|
|
if (f->f_mode == NULL)
|
|
return NULL;
|
|
f->f_fp = fp;
|
|
f = dircheck(f);
|
|
return (PyObject *) f;
|
|
}
|
|
|
|
/* check for known incorrect mode strings - problem is, platforms are
|
|
free to accept any mode characters they like and are supposed to
|
|
ignore stuff they don't understand... write or append mode with
|
|
universal newline support is expressly forbidden by PEP 278.
|
|
Additionally, remove the 'U' from the mode string as platforms
|
|
won't know what it is. Non-zero return signals an exception */
|
|
int
|
|
_PyFile_SanitizeMode(char *mode)
|
|
{
|
|
char *upos;
|
|
size_t len = strlen(mode);
|
|
|
|
if (!len) {
|
|
PyErr_SetString(PyExc_ValueError, "empty mode string");
|
|
return -1;
|
|
}
|
|
|
|
upos = strchr(mode, 'U');
|
|
if (upos) {
|
|
memmove(upos, upos+1, len-(upos-mode)); /* incl null char */
|
|
|
|
if (mode[0] == 'w' || mode[0] == 'a') {
|
|
PyErr_Format(PyExc_ValueError, "universal newline "
|
|
"mode can only be used with modes "
|
|
"starting with 'r'");
|
|
return -1;
|
|
}
|
|
|
|
if (mode[0] != 'r') {
|
|
memmove(mode+1, mode, strlen(mode)+1);
|
|
mode[0] = 'r';
|
|
}
|
|
|
|
if (!strchr(mode, 'b')) {
|
|
memmove(mode+2, mode+1, strlen(mode));
|
|
mode[1] = 'b';
|
|
}
|
|
} else if (mode[0] != 'r' && mode[0] != 'w' && mode[0] != 'a') {
|
|
PyErr_Format(PyExc_ValueError, "mode string must begin with "
|
|
"one of 'r', 'w', 'a' or 'U', not '%.200s'", mode);
|
|
return -1;
|
|
}
|
|
|
|
return 0;
|
|
}
|
|
|
|
static PyObject *
|
|
open_the_file(PyFileObject *f, char *name, char *mode)
|
|
{
|
|
char *newmode;
|
|
assert(f != NULL);
|
|
assert(PyFile_Check(f));
|
|
#ifdef MS_WINDOWS
|
|
/* windows ignores the passed name in order to support Unicode */
|
|
assert(f->f_name != NULL);
|
|
#else
|
|
assert(name != NULL);
|
|
#endif
|
|
assert(mode != NULL);
|
|
assert(f->f_fp == NULL);
|
|
|
|
/* probably need to replace 'U' by 'rb' */
|
|
newmode = PyMem_MALLOC(strlen(mode) + 3);
|
|
if (!newmode) {
|
|
PyErr_NoMemory();
|
|
return NULL;
|
|
}
|
|
strcpy(newmode, mode);
|
|
|
|
if (_PyFile_SanitizeMode(newmode)) {
|
|
f = NULL;
|
|
goto cleanup;
|
|
}
|
|
|
|
/* rexec.py can't stop a user from getting the file() constructor --
|
|
all they have to do is get *any* file object f, and then do
|
|
type(f). Here we prevent them from doing damage with it. */
|
|
if (PyEval_GetRestricted()) {
|
|
PyErr_SetString(PyExc_IOError,
|
|
"file() constructor not accessible in restricted mode");
|
|
f = NULL;
|
|
goto cleanup;
|
|
}
|
|
errno = 0;
|
|
|
|
#ifdef MS_WINDOWS
|
|
if (PyUnicode_Check(f->f_name)) {
|
|
PyObject *wmode;
|
|
wmode = PyUnicode_DecodeASCII(newmode, strlen(newmode), NULL);
|
|
if (f->f_name && wmode) {
|
|
Py_BEGIN_ALLOW_THREADS
|
|
/* PyUnicode_AS_UNICODE OK without thread
|
|
lock as it is a simple dereference. */
|
|
f->f_fp = _wfopen(PyUnicode_AS_UNICODE(f->f_name),
|
|
PyUnicode_AS_UNICODE(wmode));
|
|
Py_END_ALLOW_THREADS
|
|
}
|
|
Py_XDECREF(wmode);
|
|
}
|
|
#endif
|
|
if (NULL == f->f_fp && NULL != name) {
|
|
Py_BEGIN_ALLOW_THREADS
|
|
f->f_fp = fopen(name, newmode);
|
|
Py_END_ALLOW_THREADS
|
|
}
|
|
|
|
if (f->f_fp == NULL) {
|
|
#if defined _MSC_VER && (_MSC_VER < 1400 || !defined(__STDC_SECURE_LIB__))
|
|
/* MSVC 6 (Microsoft) leaves errno at 0 for bad mode strings,
|
|
* across all Windows flavors. When it sets EINVAL varies
|
|
* across Windows flavors, the exact conditions aren't
|
|
* documented, and the answer lies in the OS's implementation
|
|
* of Win32's CreateFile function (whose source is secret).
|
|
* Seems the best we can do is map EINVAL to ENOENT.
|
|
* Starting with Visual Studio .NET 2005, EINVAL is correctly
|
|
* set by our CRT error handler (set in exceptions.c.)
|
|
*/
|
|
if (errno == 0) /* bad mode string */
|
|
errno = EINVAL;
|
|
else if (errno == EINVAL) /* unknown, but not a mode string */
|
|
errno = ENOENT;
|
|
#endif
|
|
if (errno == EINVAL)
|
|
PyErr_Format(PyExc_IOError, "invalid mode: %s",
|
|
mode);
|
|
else
|
|
PyErr_SetFromErrnoWithFilenameObject(PyExc_IOError, f->f_name);
|
|
f = NULL;
|
|
}
|
|
if (f != NULL)
|
|
f = dircheck(f);
|
|
|
|
cleanup:
|
|
PyMem_FREE(newmode);
|
|
|
|
return (PyObject *)f;
|
|
}
|
|
|
|
PyObject *
|
|
PyFile_FromFile(FILE *fp, char *name, char *mode, int (*close)(FILE *))
|
|
{
|
|
PyFileObject *f = (PyFileObject *)PyFile_Type.tp_new(&PyFile_Type,
|
|
NULL, NULL);
|
|
if (f != NULL) {
|
|
PyObject *o_name = PyString_FromString(name);
|
|
if (o_name == NULL)
|
|
return NULL;
|
|
if (fill_file_fields(f, fp, o_name, mode, close) == NULL) {
|
|
Py_DECREF(f);
|
|
f = NULL;
|
|
}
|
|
Py_DECREF(o_name);
|
|
}
|
|
return (PyObject *) f;
|
|
}
|
|
|
|
PyObject *
|
|
PyFile_FromString(char *name, char *mode)
|
|
{
|
|
extern int fclose(FILE *);
|
|
PyFileObject *f;
|
|
|
|
f = (PyFileObject *)PyFile_FromFile((FILE *)NULL, name, mode, fclose);
|
|
if (f != NULL) {
|
|
if (open_the_file(f, name, mode) == NULL) {
|
|
Py_DECREF(f);
|
|
f = NULL;
|
|
}
|
|
}
|
|
return (PyObject *)f;
|
|
}
|
|
|
|
void
|
|
PyFile_SetBufSize(PyObject *f, int bufsize)
|
|
{
|
|
PyFileObject *file = (PyFileObject *)f;
|
|
if (bufsize >= 0) {
|
|
int type;
|
|
switch (bufsize) {
|
|
case 0:
|
|
type = _IONBF;
|
|
break;
|
|
#ifdef HAVE_SETVBUF
|
|
case 1:
|
|
type = _IOLBF;
|
|
bufsize = BUFSIZ;
|
|
break;
|
|
#endif
|
|
default:
|
|
type = _IOFBF;
|
|
#ifndef HAVE_SETVBUF
|
|
bufsize = BUFSIZ;
|
|
#endif
|
|
break;
|
|
}
|
|
fflush(file->f_fp);
|
|
if (type == _IONBF) {
|
|
PyMem_Free(file->f_setbuf);
|
|
file->f_setbuf = NULL;
|
|
} else {
|
|
file->f_setbuf = (char *)PyMem_Realloc(file->f_setbuf,
|
|
bufsize);
|
|
}
|
|
#ifdef HAVE_SETVBUF
|
|
setvbuf(file->f_fp, file->f_setbuf, type, bufsize);
|
|
#else /* !HAVE_SETVBUF */
|
|
setbuf(file->f_fp, file->f_setbuf);
|
|
#endif /* !HAVE_SETVBUF */
|
|
}
|
|
}
|
|
|
|
/* Set the encoding used to output Unicode strings.
|
|
Returh 1 on success, 0 on failure. */
|
|
|
|
int
|
|
PyFile_SetEncoding(PyObject *f, const char *enc)
|
|
{
|
|
PyFileObject *file = (PyFileObject*)f;
|
|
PyObject *str = PyString_FromString(enc);
|
|
|
|
assert(PyFile_Check(f));
|
|
if (!str)
|
|
return 0;
|
|
Py_DECREF(file->f_encoding);
|
|
file->f_encoding = str;
|
|
return 1;
|
|
}
|
|
|
|
static PyObject *
|
|
err_closed(void)
|
|
{
|
|
PyErr_SetString(PyExc_ValueError, "I/O operation on closed file");
|
|
return NULL;
|
|
}
|
|
|
|
/* Refuse regular file I/O if there's data in the iteration-buffer.
|
|
* Mixing them would cause data to arrive out of order, as the read*
|
|
* methods don't use the iteration buffer. */
|
|
static PyObject *
|
|
err_iterbuffered(void)
|
|
{
|
|
PyErr_SetString(PyExc_ValueError,
|
|
"Mixing iteration and read methods would lose data");
|
|
return NULL;
|
|
}
|
|
|
|
static void drop_readahead(PyFileObject *);
|
|
|
|
/* Methods */
|
|
|
|
static void
|
|
file_dealloc(PyFileObject *f)
|
|
{
|
|
int sts = 0;
|
|
if (f->weakreflist != NULL)
|
|
PyObject_ClearWeakRefs((PyObject *) f);
|
|
if (f->f_fp != NULL && f->f_close != NULL) {
|
|
Py_BEGIN_ALLOW_THREADS
|
|
sts = (*f->f_close)(f->f_fp);
|
|
Py_END_ALLOW_THREADS
|
|
if (sts == EOF)
|
|
#ifdef HAVE_STRERROR
|
|
PySys_WriteStderr("close failed: [Errno %d] %s\n", errno, strerror(errno));
|
|
#else
|
|
PySys_WriteStderr("close failed: [Errno %d]\n", errno);
|
|
#endif
|
|
}
|
|
PyMem_Free(f->f_setbuf);
|
|
Py_XDECREF(f->f_name);
|
|
Py_XDECREF(f->f_mode);
|
|
Py_XDECREF(f->f_encoding);
|
|
drop_readahead(f);
|
|
f->ob_type->tp_free((PyObject *)f);
|
|
}
|
|
|
|
static PyObject *
|
|
file_repr(PyFileObject *f)
|
|
{
|
|
if (PyUnicode_Check(f->f_name)) {
|
|
#ifdef Py_USING_UNICODE
|
|
PyObject *ret = NULL;
|
|
PyObject *name = PyUnicode_AsUnicodeEscapeString(f->f_name);
|
|
const char *name_str = name ? PyString_AsString(name) : "?";
|
|
ret = PyString_FromFormat("<%s file u'%s', mode '%s' at %p>",
|
|
f->f_fp == NULL ? "closed" : "open",
|
|
name_str,
|
|
PyString_AsString(f->f_mode),
|
|
f);
|
|
Py_XDECREF(name);
|
|
return ret;
|
|
#endif
|
|
} else {
|
|
return PyString_FromFormat("<%s file '%s', mode '%s' at %p>",
|
|
f->f_fp == NULL ? "closed" : "open",
|
|
PyString_AsString(f->f_name),
|
|
PyString_AsString(f->f_mode),
|
|
f);
|
|
}
|
|
}
|
|
|
|
static PyObject *
|
|
file_close(PyFileObject *f)
|
|
{
|
|
int sts = 0;
|
|
if (f->f_fp != NULL) {
|
|
if (f->f_close != NULL) {
|
|
Py_BEGIN_ALLOW_THREADS
|
|
errno = 0;
|
|
sts = (*f->f_close)(f->f_fp);
|
|
Py_END_ALLOW_THREADS
|
|
}
|
|
f->f_fp = NULL;
|
|
}
|
|
PyMem_Free(f->f_setbuf);
|
|
f->f_setbuf = NULL;
|
|
if (sts == EOF)
|
|
return PyErr_SetFromErrno(PyExc_IOError);
|
|
if (sts != 0)
|
|
return PyInt_FromLong((long)sts);
|
|
Py_INCREF(Py_None);
|
|
return Py_None;
|
|
}
|
|
|
|
|
|
/* Our very own off_t-like type, 64-bit if possible */
|
|
#if !defined(HAVE_LARGEFILE_SUPPORT)
|
|
typedef off_t Py_off_t;
|
|
#elif SIZEOF_OFF_T >= 8
|
|
typedef off_t Py_off_t;
|
|
#elif SIZEOF_FPOS_T >= 8
|
|
typedef fpos_t Py_off_t;
|
|
#else
|
|
#error "Large file support, but neither off_t nor fpos_t is large enough."
|
|
#endif
|
|
|
|
|
|
/* a portable fseek() function
|
|
return 0 on success, non-zero on failure (with errno set) */
|
|
static int
|
|
_portable_fseek(FILE *fp, Py_off_t offset, int whence)
|
|
{
|
|
#if !defined(HAVE_LARGEFILE_SUPPORT)
|
|
return fseek(fp, offset, whence);
|
|
#elif defined(HAVE_FSEEKO) && SIZEOF_OFF_T >= 8
|
|
return fseeko(fp, offset, whence);
|
|
#elif defined(HAVE_FSEEK64)
|
|
return fseek64(fp, offset, whence);
|
|
#elif defined(__BEOS__)
|
|
return _fseek(fp, offset, whence);
|
|
#elif SIZEOF_FPOS_T >= 8
|
|
/* lacking a 64-bit capable fseek(), use a 64-bit capable fsetpos()
|
|
and fgetpos() to implement fseek()*/
|
|
fpos_t pos;
|
|
switch (whence) {
|
|
case SEEK_END:
|
|
#ifdef MS_WINDOWS
|
|
fflush(fp);
|
|
if (_lseeki64(fileno(fp), 0, 2) == -1)
|
|
return -1;
|
|
#else
|
|
if (fseek(fp, 0, SEEK_END) != 0)
|
|
return -1;
|
|
#endif
|
|
/* fall through */
|
|
case SEEK_CUR:
|
|
if (fgetpos(fp, &pos) != 0)
|
|
return -1;
|
|
offset += pos;
|
|
break;
|
|
/* case SEEK_SET: break; */
|
|
}
|
|
return fsetpos(fp, &offset);
|
|
#else
|
|
#error "Large file support, but no way to fseek."
|
|
#endif
|
|
}
|
|
|
|
|
|
/* a portable ftell() function
|
|
Return -1 on failure with errno set appropriately, current file
|
|
position on success */
|
|
static Py_off_t
|
|
_portable_ftell(FILE* fp)
|
|
{
|
|
#if !defined(HAVE_LARGEFILE_SUPPORT)
|
|
return ftell(fp);
|
|
#elif defined(HAVE_FTELLO) && SIZEOF_OFF_T >= 8
|
|
return ftello(fp);
|
|
#elif defined(HAVE_FTELL64)
|
|
return ftell64(fp);
|
|
#elif SIZEOF_FPOS_T >= 8
|
|
fpos_t pos;
|
|
if (fgetpos(fp, &pos) != 0)
|
|
return -1;
|
|
return pos;
|
|
#else
|
|
#error "Large file support, but no way to ftell."
|
|
#endif
|
|
}
|
|
|
|
|
|
static PyObject *
|
|
file_seek(PyFileObject *f, PyObject *args)
|
|
{
|
|
int whence;
|
|
int ret;
|
|
Py_off_t offset;
|
|
PyObject *offobj, *off_index;
|
|
|
|
if (f->f_fp == NULL)
|
|
return err_closed();
|
|
drop_readahead(f);
|
|
whence = 0;
|
|
if (!PyArg_ParseTuple(args, "O|i:seek", &offobj, &whence))
|
|
return NULL;
|
|
off_index = PyNumber_Index(offobj);
|
|
if (!off_index) {
|
|
if (!PyFloat_Check(offobj))
|
|
return NULL;
|
|
/* Deprecated in 2.6 */
|
|
PyErr_Clear();
|
|
if (PyErr_Warn(PyExc_DeprecationWarning,
|
|
"integer argument expected, got float"))
|
|
return NULL;
|
|
off_index = offobj;
|
|
Py_INCREF(offobj);
|
|
}
|
|
#if !defined(HAVE_LARGEFILE_SUPPORT)
|
|
offset = PyInt_AsLong(off_index);
|
|
#else
|
|
offset = PyLong_Check(off_index) ?
|
|
PyLong_AsLongLong(off_index) : PyInt_AsLong(off_index);
|
|
#endif
|
|
Py_DECREF(off_index);
|
|
if (PyErr_Occurred())
|
|
return NULL;
|
|
|
|
Py_BEGIN_ALLOW_THREADS
|
|
errno = 0;
|
|
ret = _portable_fseek(f->f_fp, offset, whence);
|
|
Py_END_ALLOW_THREADS
|
|
|
|
if (ret != 0) {
|
|
PyErr_SetFromErrno(PyExc_IOError);
|
|
clearerr(f->f_fp);
|
|
return NULL;
|
|
}
|
|
f->f_skipnextlf = 0;
|
|
Py_INCREF(Py_None);
|
|
return Py_None;
|
|
}
|
|
|
|
|
|
#ifdef HAVE_FTRUNCATE
|
|
static PyObject *
|
|
file_truncate(PyFileObject *f, PyObject *args)
|
|
{
|
|
Py_off_t newsize;
|
|
PyObject *newsizeobj = NULL;
|
|
Py_off_t initialpos;
|
|
int ret;
|
|
|
|
if (f->f_fp == NULL)
|
|
return err_closed();
|
|
if (!PyArg_UnpackTuple(args, "truncate", 0, 1, &newsizeobj))
|
|
return NULL;
|
|
|
|
/* Get current file position. If the file happens to be open for
|
|
* update and the last operation was an input operation, C doesn't
|
|
* define what the later fflush() will do, but we promise truncate()
|
|
* won't change the current position (and fflush() *does* change it
|
|
* then at least on Windows). The easiest thing is to capture
|
|
* current pos now and seek back to it at the end.
|
|
*/
|
|
Py_BEGIN_ALLOW_THREADS
|
|
errno = 0;
|
|
initialpos = _portable_ftell(f->f_fp);
|
|
Py_END_ALLOW_THREADS
|
|
if (initialpos == -1)
|
|
goto onioerror;
|
|
|
|
/* Set newsize to current postion if newsizeobj NULL, else to the
|
|
* specified value.
|
|
*/
|
|
if (newsizeobj != NULL) {
|
|
#if !defined(HAVE_LARGEFILE_SUPPORT)
|
|
newsize = PyInt_AsLong(newsizeobj);
|
|
#else
|
|
newsize = PyLong_Check(newsizeobj) ?
|
|
PyLong_AsLongLong(newsizeobj) :
|
|
PyInt_AsLong(newsizeobj);
|
|
#endif
|
|
if (PyErr_Occurred())
|
|
return NULL;
|
|
}
|
|
else /* default to current position */
|
|
newsize = initialpos;
|
|
|
|
/* Flush the stream. We're mixing stream-level I/O with lower-level
|
|
* I/O, and a flush may be necessary to synch both platform views
|
|
* of the current file state.
|
|
*/
|
|
Py_BEGIN_ALLOW_THREADS
|
|
errno = 0;
|
|
ret = fflush(f->f_fp);
|
|
Py_END_ALLOW_THREADS
|
|
if (ret != 0)
|
|
goto onioerror;
|
|
|
|
#ifdef MS_WINDOWS
|
|
/* MS _chsize doesn't work if newsize doesn't fit in 32 bits,
|
|
so don't even try using it. */
|
|
{
|
|
HANDLE hFile;
|
|
|
|
/* Have to move current pos to desired endpoint on Windows. */
|
|
Py_BEGIN_ALLOW_THREADS
|
|
errno = 0;
|
|
ret = _portable_fseek(f->f_fp, newsize, SEEK_SET) != 0;
|
|
Py_END_ALLOW_THREADS
|
|
if (ret)
|
|
goto onioerror;
|
|
|
|
/* Truncate. Note that this may grow the file! */
|
|
Py_BEGIN_ALLOW_THREADS
|
|
errno = 0;
|
|
hFile = (HANDLE)_get_osfhandle(fileno(f->f_fp));
|
|
ret = hFile == (HANDLE)-1;
|
|
if (ret == 0) {
|
|
ret = SetEndOfFile(hFile) == 0;
|
|
if (ret)
|
|
errno = EACCES;
|
|
}
|
|
Py_END_ALLOW_THREADS
|
|
if (ret)
|
|
goto onioerror;
|
|
}
|
|
#else
|
|
Py_BEGIN_ALLOW_THREADS
|
|
errno = 0;
|
|
ret = ftruncate(fileno(f->f_fp), newsize);
|
|
Py_END_ALLOW_THREADS
|
|
if (ret != 0)
|
|
goto onioerror;
|
|
#endif /* !MS_WINDOWS */
|
|
|
|
/* Restore original file position. */
|
|
Py_BEGIN_ALLOW_THREADS
|
|
errno = 0;
|
|
ret = _portable_fseek(f->f_fp, initialpos, SEEK_SET) != 0;
|
|
Py_END_ALLOW_THREADS
|
|
if (ret)
|
|
goto onioerror;
|
|
|
|
Py_INCREF(Py_None);
|
|
return Py_None;
|
|
|
|
onioerror:
|
|
PyErr_SetFromErrno(PyExc_IOError);
|
|
clearerr(f->f_fp);
|
|
return NULL;
|
|
}
|
|
#endif /* HAVE_FTRUNCATE */
|
|
|
|
static PyObject *
|
|
file_tell(PyFileObject *f)
|
|
{
|
|
Py_off_t pos;
|
|
|
|
if (f->f_fp == NULL)
|
|
return err_closed();
|
|
Py_BEGIN_ALLOW_THREADS
|
|
errno = 0;
|
|
pos = _portable_ftell(f->f_fp);
|
|
Py_END_ALLOW_THREADS
|
|
if (pos == -1) {
|
|
PyErr_SetFromErrno(PyExc_IOError);
|
|
clearerr(f->f_fp);
|
|
return NULL;
|
|
}
|
|
if (f->f_skipnextlf) {
|
|
int c;
|
|
c = GETC(f->f_fp);
|
|
if (c == '\n') {
|
|
pos++;
|
|
f->f_skipnextlf = 0;
|
|
} else if (c != EOF) ungetc(c, f->f_fp);
|
|
}
|
|
#if !defined(HAVE_LARGEFILE_SUPPORT)
|
|
return PyInt_FromLong(pos);
|
|
#else
|
|
return PyLong_FromLongLong(pos);
|
|
#endif
|
|
}
|
|
|
|
static PyObject *
|
|
file_fileno(PyFileObject *f)
|
|
{
|
|
if (f->f_fp == NULL)
|
|
return err_closed();
|
|
return PyInt_FromLong((long) fileno(f->f_fp));
|
|
}
|
|
|
|
static PyObject *
|
|
file_flush(PyFileObject *f)
|
|
{
|
|
int res;
|
|
|
|
if (f->f_fp == NULL)
|
|
return err_closed();
|
|
Py_BEGIN_ALLOW_THREADS
|
|
errno = 0;
|
|
res = fflush(f->f_fp);
|
|
Py_END_ALLOW_THREADS
|
|
if (res != 0) {
|
|
PyErr_SetFromErrno(PyExc_IOError);
|
|
clearerr(f->f_fp);
|
|
return NULL;
|
|
}
|
|
Py_INCREF(Py_None);
|
|
return Py_None;
|
|
}
|
|
|
|
static PyObject *
|
|
file_isatty(PyFileObject *f)
|
|
{
|
|
long res;
|
|
if (f->f_fp == NULL)
|
|
return err_closed();
|
|
Py_BEGIN_ALLOW_THREADS
|
|
res = isatty((int)fileno(f->f_fp));
|
|
Py_END_ALLOW_THREADS
|
|
return PyBool_FromLong(res);
|
|
}
|
|
|
|
|
|
#if BUFSIZ < 8192
|
|
#define SMALLCHUNK 8192
|
|
#else
|
|
#define SMALLCHUNK BUFSIZ
|
|
#endif
|
|
|
|
#if SIZEOF_INT < 4
|
|
#define BIGCHUNK (512 * 32)
|
|
#else
|
|
#define BIGCHUNK (512 * 1024)
|
|
#endif
|
|
|
|
static size_t
|
|
new_buffersize(PyFileObject *f, size_t currentsize)
|
|
{
|
|
#ifdef HAVE_FSTAT
|
|
off_t pos, end;
|
|
struct stat st;
|
|
if (fstat(fileno(f->f_fp), &st) == 0) {
|
|
end = st.st_size;
|
|
/* The following is not a bug: we really need to call lseek()
|
|
*and* ftell(). The reason is that some stdio libraries
|
|
mistakenly flush their buffer when ftell() is called and
|
|
the lseek() call it makes fails, thereby throwing away
|
|
data that cannot be recovered in any way. To avoid this,
|
|
we first test lseek(), and only call ftell() if lseek()
|
|
works. We can't use the lseek() value either, because we
|
|
need to take the amount of buffered data into account.
|
|
(Yet another reason why stdio stinks. :-) */
|
|
pos = lseek(fileno(f->f_fp), 0L, SEEK_CUR);
|
|
if (pos >= 0) {
|
|
pos = ftell(f->f_fp);
|
|
}
|
|
if (pos < 0)
|
|
clearerr(f->f_fp);
|
|
if (end > pos && pos >= 0)
|
|
return currentsize + end - pos + 1;
|
|
/* Add 1 so if the file were to grow we'd notice. */
|
|
}
|
|
#endif
|
|
if (currentsize > SMALLCHUNK) {
|
|
/* Keep doubling until we reach BIGCHUNK;
|
|
then keep adding BIGCHUNK. */
|
|
if (currentsize <= BIGCHUNK)
|
|
return currentsize + currentsize;
|
|
else
|
|
return currentsize + BIGCHUNK;
|
|
}
|
|
return currentsize + SMALLCHUNK;
|
|
}
|
|
|
|
#if defined(EWOULDBLOCK) && defined(EAGAIN) && EWOULDBLOCK != EAGAIN
|
|
#define BLOCKED_ERRNO(x) ((x) == EWOULDBLOCK || (x) == EAGAIN)
|
|
#else
|
|
#ifdef EWOULDBLOCK
|
|
#define BLOCKED_ERRNO(x) ((x) == EWOULDBLOCK)
|
|
#else
|
|
#ifdef EAGAIN
|
|
#define BLOCKED_ERRNO(x) ((x) == EAGAIN)
|
|
#else
|
|
#define BLOCKED_ERRNO(x) 0
|
|
#endif
|
|
#endif
|
|
#endif
|
|
|
|
static PyObject *
|
|
file_read(PyFileObject *f, PyObject *args)
|
|
{
|
|
long bytesrequested = -1;
|
|
size_t bytesread, buffersize, chunksize;
|
|
PyObject *v;
|
|
|
|
if (f->f_fp == NULL)
|
|
return err_closed();
|
|
/* refuse to mix with f.next() */
|
|
if (f->f_buf != NULL &&
|
|
(f->f_bufend - f->f_bufptr) > 0 &&
|
|
f->f_buf[0] != '\0')
|
|
return err_iterbuffered();
|
|
if (!PyArg_ParseTuple(args, "|l:read", &bytesrequested))
|
|
return NULL;
|
|
if (bytesrequested < 0)
|
|
buffersize = new_buffersize(f, (size_t)0);
|
|
else
|
|
buffersize = bytesrequested;
|
|
if (buffersize > PY_SSIZE_T_MAX) {
|
|
PyErr_SetString(PyExc_OverflowError,
|
|
"requested number of bytes is more than a Python string can hold");
|
|
return NULL;
|
|
}
|
|
v = PyString_FromStringAndSize((char *)NULL, buffersize);
|
|
if (v == NULL)
|
|
return NULL;
|
|
bytesread = 0;
|
|
for (;;) {
|
|
Py_BEGIN_ALLOW_THREADS
|
|
errno = 0;
|
|
chunksize = Py_UniversalNewlineFread(BUF(v) + bytesread,
|
|
buffersize - bytesread, f->f_fp, (PyObject *)f);
|
|
Py_END_ALLOW_THREADS
|
|
if (chunksize == 0) {
|
|
if (!ferror(f->f_fp))
|
|
break;
|
|
clearerr(f->f_fp);
|
|
/* When in non-blocking mode, data shouldn't
|
|
* be discarded if a blocking signal was
|
|
* received. That will also happen if
|
|
* chunksize != 0, but bytesread < buffersize. */
|
|
if (bytesread > 0 && BLOCKED_ERRNO(errno))
|
|
break;
|
|
PyErr_SetFromErrno(PyExc_IOError);
|
|
Py_DECREF(v);
|
|
return NULL;
|
|
}
|
|
bytesread += chunksize;
|
|
if (bytesread < buffersize) {
|
|
clearerr(f->f_fp);
|
|
break;
|
|
}
|
|
if (bytesrequested < 0) {
|
|
buffersize = new_buffersize(f, buffersize);
|
|
if (_PyString_Resize(&v, buffersize) < 0)
|
|
return NULL;
|
|
} else {
|
|
/* Got what was requested. */
|
|
break;
|
|
}
|
|
}
|
|
if (bytesread != buffersize)
|
|
_PyString_Resize(&v, bytesread);
|
|
return v;
|
|
}
|
|
|
|
static PyObject *
|
|
file_readinto(PyFileObject *f, PyObject *args)
|
|
{
|
|
char *ptr;
|
|
Py_ssize_t ntodo;
|
|
Py_ssize_t ndone, nnow;
|
|
|
|
if (f->f_fp == NULL)
|
|
return err_closed();
|
|
if (!f->f_binary) {
|
|
PyErr_SetString(PyExc_TypeError,
|
|
"readinto() requires binary mode");
|
|
return NULL;
|
|
}
|
|
/* refuse to mix with f.next() */
|
|
if (f->f_buf != NULL &&
|
|
(f->f_bufend - f->f_bufptr) > 0 &&
|
|
f->f_buf[0] != '\0')
|
|
return err_iterbuffered();
|
|
if (!PyArg_ParseTuple(args, "w#", &ptr, &ntodo))
|
|
return NULL;
|
|
ndone = 0;
|
|
while (ntodo > 0) {
|
|
Py_BEGIN_ALLOW_THREADS
|
|
errno = 0;
|
|
nnow = Py_UniversalNewlineFread(ptr+ndone, ntodo, f->f_fp,
|
|
(PyObject *)f);
|
|
Py_END_ALLOW_THREADS
|
|
if (nnow == 0) {
|
|
if (!ferror(f->f_fp))
|
|
break;
|
|
PyErr_SetFromErrno(PyExc_IOError);
|
|
clearerr(f->f_fp);
|
|
return NULL;
|
|
}
|
|
ndone += nnow;
|
|
ntodo -= nnow;
|
|
}
|
|
return PyInt_FromSsize_t(ndone);
|
|
}
|
|
|
|
/**************************************************************************
|
|
Routine to get next line using platform fgets().
|
|
|
|
Under MSVC 6:
|
|
|
|
+ MS threadsafe getc is very slow (multiple layers of function calls before+
|
|
after each character, to lock+unlock the stream).
|
|
+ The stream-locking functions are MS-internal -- can't access them from user
|
|
code.
|
|
+ There's nothing Tim could find in the MS C or platform SDK libraries that
|
|
can worm around this.
|
|
+ MS fgets locks/unlocks only once per line; it's the only hook we have.
|
|
|
|
So we use fgets for speed(!), despite that it's painful.
|
|
|
|
MS realloc is also slow.
|
|
|
|
Reports from other platforms on this method vs getc_unlocked (which MS doesn't
|
|
have):
|
|
Linux a wash
|
|
Solaris a wash
|
|
Tru64 Unix getline_via_fgets significantly faster
|
|
|
|
CAUTION: The C std isn't clear about this: in those cases where fgets
|
|
writes something into the buffer, can it write into any position beyond the
|
|
required trailing null byte? MSVC 6 fgets does not, and no platform is (yet)
|
|
known on which it does; and it would be a strange way to code fgets. Still,
|
|
getline_via_fgets may not work correctly if it does. The std test
|
|
test_bufio.py should fail if platform fgets() routinely writes beyond the
|
|
trailing null byte. #define DONT_USE_FGETS_IN_GETLINE to disable this code.
|
|
**************************************************************************/
|
|
|
|
/* Use this routine if told to, or by default on non-get_unlocked()
|
|
* platforms unless told not to. Yikes! Let's spell that out:
|
|
* On a platform with getc_unlocked():
|
|
* By default, use getc_unlocked().
|
|
* If you want to use fgets() instead, #define USE_FGETS_IN_GETLINE.
|
|
* On a platform without getc_unlocked():
|
|
* By default, use fgets().
|
|
* If you don't want to use fgets(), #define DONT_USE_FGETS_IN_GETLINE.
|
|
*/
|
|
#if !defined(USE_FGETS_IN_GETLINE) && !defined(HAVE_GETC_UNLOCKED)
|
|
#define USE_FGETS_IN_GETLINE
|
|
#endif
|
|
|
|
#if defined(DONT_USE_FGETS_IN_GETLINE) && defined(USE_FGETS_IN_GETLINE)
|
|
#undef USE_FGETS_IN_GETLINE
|
|
#endif
|
|
|
|
#ifdef USE_FGETS_IN_GETLINE
|
|
static PyObject*
|
|
getline_via_fgets(FILE *fp)
|
|
{
|
|
/* INITBUFSIZE is the maximum line length that lets us get away with the fast
|
|
* no-realloc, one-fgets()-call path. Boosting it isn't free, because we have
|
|
* to fill this much of the buffer with a known value in order to figure out
|
|
* how much of the buffer fgets() overwrites. So if INITBUFSIZE is larger
|
|
* than "most" lines, we waste time filling unused buffer slots. 100 is
|
|
* surely adequate for most peoples' email archives, chewing over source code,
|
|
* etc -- "regular old text files".
|
|
* MAXBUFSIZE is the maximum line length that lets us get away with the less
|
|
* fast (but still zippy) no-realloc, two-fgets()-call path. See above for
|
|
* cautions about boosting that. 300 was chosen because the worst real-life
|
|
* text-crunching job reported on Python-Dev was a mail-log crawler where over
|
|
* half the lines were 254 chars.
|
|
*/
|
|
#define INITBUFSIZE 100
|
|
#define MAXBUFSIZE 300
|
|
char* p; /* temp */
|
|
char buf[MAXBUFSIZE];
|
|
PyObject* v; /* the string object result */
|
|
char* pvfree; /* address of next free slot */
|
|
char* pvend; /* address one beyond last free slot */
|
|
size_t nfree; /* # of free buffer slots; pvend-pvfree */
|
|
size_t total_v_size; /* total # of slots in buffer */
|
|
size_t increment; /* amount to increment the buffer */
|
|
size_t prev_v_size;
|
|
|
|
/* Optimize for normal case: avoid _PyString_Resize if at all
|
|
* possible via first reading into stack buffer "buf".
|
|
*/
|
|
total_v_size = INITBUFSIZE; /* start small and pray */
|
|
pvfree = buf;
|
|
for (;;) {
|
|
Py_BEGIN_ALLOW_THREADS
|
|
pvend = buf + total_v_size;
|
|
nfree = pvend - pvfree;
|
|
memset(pvfree, '\n', nfree);
|
|
assert(nfree < INT_MAX); /* Should be atmost MAXBUFSIZE */
|
|
p = fgets(pvfree, (int)nfree, fp);
|
|
Py_END_ALLOW_THREADS
|
|
|
|
if (p == NULL) {
|
|
clearerr(fp);
|
|
if (PyErr_CheckSignals())
|
|
return NULL;
|
|
v = PyString_FromStringAndSize(buf, pvfree - buf);
|
|
return v;
|
|
}
|
|
/* fgets read *something* */
|
|
p = memchr(pvfree, '\n', nfree);
|
|
if (p != NULL) {
|
|
/* Did the \n come from fgets or from us?
|
|
* Since fgets stops at the first \n, and then writes
|
|
* \0, if it's from fgets a \0 must be next. But if
|
|
* that's so, it could not have come from us, since
|
|
* the \n's we filled the buffer with have only more
|
|
* \n's to the right.
|
|
*/
|
|
if (p+1 < pvend && *(p+1) == '\0') {
|
|
/* It's from fgets: we win! In particular,
|
|
* we haven't done any mallocs yet, and can
|
|
* build the final result on the first try.
|
|
*/
|
|
++p; /* include \n from fgets */
|
|
}
|
|
else {
|
|
/* Must be from us: fgets didn't fill the
|
|
* buffer and didn't find a newline, so it
|
|
* must be the last and newline-free line of
|
|
* the file.
|
|
*/
|
|
assert(p > pvfree && *(p-1) == '\0');
|
|
--p; /* don't include \0 from fgets */
|
|
}
|
|
v = PyString_FromStringAndSize(buf, p - buf);
|
|
return v;
|
|
}
|
|
/* yuck: fgets overwrote all the newlines, i.e. the entire
|
|
* buffer. So this line isn't over yet, or maybe it is but
|
|
* we're exactly at EOF. If we haven't already, try using the
|
|
* rest of the stack buffer.
|
|
*/
|
|
assert(*(pvend-1) == '\0');
|
|
if (pvfree == buf) {
|
|
pvfree = pvend - 1; /* overwrite trailing null */
|
|
total_v_size = MAXBUFSIZE;
|
|
}
|
|
else
|
|
break;
|
|
}
|
|
|
|
/* The stack buffer isn't big enough; malloc a string object and read
|
|
* into its buffer.
|
|
*/
|
|
total_v_size = MAXBUFSIZE << 1;
|
|
v = PyString_FromStringAndSize((char*)NULL, (int)total_v_size);
|
|
if (v == NULL)
|
|
return v;
|
|
/* copy over everything except the last null byte */
|
|
memcpy(BUF(v), buf, MAXBUFSIZE-1);
|
|
pvfree = BUF(v) + MAXBUFSIZE - 1;
|
|
|
|
/* Keep reading stuff into v; if it ever ends successfully, break
|
|
* after setting p one beyond the end of the line. The code here is
|
|
* very much like the code above, except reads into v's buffer; see
|
|
* the code above for detailed comments about the logic.
|
|
*/
|
|
for (;;) {
|
|
Py_BEGIN_ALLOW_THREADS
|
|
pvend = BUF(v) + total_v_size;
|
|
nfree = pvend - pvfree;
|
|
memset(pvfree, '\n', nfree);
|
|
assert(nfree < INT_MAX);
|
|
p = fgets(pvfree, (int)nfree, fp);
|
|
Py_END_ALLOW_THREADS
|
|
|
|
if (p == NULL) {
|
|
clearerr(fp);
|
|
if (PyErr_CheckSignals()) {
|
|
Py_DECREF(v);
|
|
return NULL;
|
|
}
|
|
p = pvfree;
|
|
break;
|
|
}
|
|
p = memchr(pvfree, '\n', nfree);
|
|
if (p != NULL) {
|
|
if (p+1 < pvend && *(p+1) == '\0') {
|
|
/* \n came from fgets */
|
|
++p;
|
|
break;
|
|
}
|
|
/* \n came from us; last line of file, no newline */
|
|
assert(p > pvfree && *(p-1) == '\0');
|
|
--p;
|
|
break;
|
|
}
|
|
/* expand buffer and try again */
|
|
assert(*(pvend-1) == '\0');
|
|
increment = total_v_size >> 2; /* mild exponential growth */
|
|
prev_v_size = total_v_size;
|
|
total_v_size += increment;
|
|
/* check for overflow */
|
|
if (total_v_size <= prev_v_size ||
|
|
total_v_size > PY_SSIZE_T_MAX) {
|
|
PyErr_SetString(PyExc_OverflowError,
|
|
"line is longer than a Python string can hold");
|
|
Py_DECREF(v);
|
|
return NULL;
|
|
}
|
|
if (_PyString_Resize(&v, (int)total_v_size) < 0)
|
|
return NULL;
|
|
/* overwrite the trailing null byte */
|
|
pvfree = BUF(v) + (prev_v_size - 1);
|
|
}
|
|
if (BUF(v) + total_v_size != p)
|
|
_PyString_Resize(&v, p - BUF(v));
|
|
return v;
|
|
#undef INITBUFSIZE
|
|
#undef MAXBUFSIZE
|
|
}
|
|
#endif /* ifdef USE_FGETS_IN_GETLINE */
|
|
|
|
/* Internal routine to get a line.
|
|
Size argument interpretation:
|
|
> 0: max length;
|
|
<= 0: read arbitrary line
|
|
*/
|
|
|
|
static PyObject *
|
|
get_line(PyFileObject *f, int n)
|
|
{
|
|
FILE *fp = f->f_fp;
|
|
int c;
|
|
char *buf, *end;
|
|
size_t total_v_size; /* total # of slots in buffer */
|
|
size_t used_v_size; /* # used slots in buffer */
|
|
size_t increment; /* amount to increment the buffer */
|
|
PyObject *v;
|
|
int newlinetypes = f->f_newlinetypes;
|
|
int skipnextlf = f->f_skipnextlf;
|
|
int univ_newline = f->f_univ_newline;
|
|
|
|
#if defined(USE_FGETS_IN_GETLINE)
|
|
if (n <= 0 && !univ_newline )
|
|
return getline_via_fgets(fp);
|
|
#endif
|
|
total_v_size = n > 0 ? n : 100;
|
|
v = PyString_FromStringAndSize((char *)NULL, total_v_size);
|
|
if (v == NULL)
|
|
return NULL;
|
|
buf = BUF(v);
|
|
end = buf + total_v_size;
|
|
|
|
for (;;) {
|
|
Py_BEGIN_ALLOW_THREADS
|
|
FLOCKFILE(fp);
|
|
if (univ_newline) {
|
|
c = 'x'; /* Shut up gcc warning */
|
|
while ( buf != end && (c = GETC(fp)) != EOF ) {
|
|
if (skipnextlf ) {
|
|
skipnextlf = 0;
|
|
if (c == '\n') {
|
|
/* Seeing a \n here with
|
|
* skipnextlf true means we
|
|
* saw a \r before.
|
|
*/
|
|
newlinetypes |= NEWLINE_CRLF;
|
|
c = GETC(fp);
|
|
if (c == EOF) break;
|
|
} else {
|
|
newlinetypes |= NEWLINE_CR;
|
|
}
|
|
}
|
|
if (c == '\r') {
|
|
skipnextlf = 1;
|
|
c = '\n';
|
|
} else if ( c == '\n')
|
|
newlinetypes |= NEWLINE_LF;
|
|
*buf++ = c;
|
|
if (c == '\n') break;
|
|
}
|
|
if ( c == EOF && skipnextlf )
|
|
newlinetypes |= NEWLINE_CR;
|
|
} else /* If not universal newlines use the normal loop */
|
|
while ((c = GETC(fp)) != EOF &&
|
|
(*buf++ = c) != '\n' &&
|
|
buf != end)
|
|
;
|
|
FUNLOCKFILE(fp);
|
|
Py_END_ALLOW_THREADS
|
|
f->f_newlinetypes = newlinetypes;
|
|
f->f_skipnextlf = skipnextlf;
|
|
if (c == '\n')
|
|
break;
|
|
if (c == EOF) {
|
|
if (ferror(fp)) {
|
|
PyErr_SetFromErrno(PyExc_IOError);
|
|
clearerr(fp);
|
|
Py_DECREF(v);
|
|
return NULL;
|
|
}
|
|
clearerr(fp);
|
|
if (PyErr_CheckSignals()) {
|
|
Py_DECREF(v);
|
|
return NULL;
|
|
}
|
|
break;
|
|
}
|
|
/* Must be because buf == end */
|
|
if (n > 0)
|
|
break;
|
|
used_v_size = total_v_size;
|
|
increment = total_v_size >> 2; /* mild exponential growth */
|
|
total_v_size += increment;
|
|
if (total_v_size > PY_SSIZE_T_MAX) {
|
|
PyErr_SetString(PyExc_OverflowError,
|
|
"line is longer than a Python string can hold");
|
|
Py_DECREF(v);
|
|
return NULL;
|
|
}
|
|
if (_PyString_Resize(&v, total_v_size) < 0)
|
|
return NULL;
|
|
buf = BUF(v) + used_v_size;
|
|
end = BUF(v) + total_v_size;
|
|
}
|
|
|
|
used_v_size = buf - BUF(v);
|
|
if (used_v_size != total_v_size)
|
|
_PyString_Resize(&v, used_v_size);
|
|
return v;
|
|
}
|
|
|
|
/* External C interface */
|
|
|
|
PyObject *
|
|
PyFile_GetLine(PyObject *f, int n)
|
|
{
|
|
PyObject *result;
|
|
|
|
if (f == NULL) {
|
|
PyErr_BadInternalCall();
|
|
return NULL;
|
|
}
|
|
|
|
if (PyFile_Check(f)) {
|
|
PyFileObject *fo = (PyFileObject *)f;
|
|
if (fo->f_fp == NULL)
|
|
return err_closed();
|
|
/* refuse to mix with f.next() */
|
|
if (fo->f_buf != NULL &&
|
|
(fo->f_bufend - fo->f_bufptr) > 0 &&
|
|
fo->f_buf[0] != '\0')
|
|
return err_iterbuffered();
|
|
result = get_line(fo, n);
|
|
}
|
|
else {
|
|
PyObject *reader;
|
|
PyObject *args;
|
|
|
|
reader = PyObject_GetAttrString(f, "readline");
|
|
if (reader == NULL)
|
|
return NULL;
|
|
if (n <= 0)
|
|
args = PyTuple_New(0);
|
|
else
|
|
args = Py_BuildValue("(i)", n);
|
|
if (args == NULL) {
|
|
Py_DECREF(reader);
|
|
return NULL;
|
|
}
|
|
result = PyEval_CallObject(reader, args);
|
|
Py_DECREF(reader);
|
|
Py_DECREF(args);
|
|
if (result != NULL && !PyString_Check(result) &&
|
|
!PyUnicode_Check(result)) {
|
|
Py_DECREF(result);
|
|
result = NULL;
|
|
PyErr_SetString(PyExc_TypeError,
|
|
"object.readline() returned non-string");
|
|
}
|
|
}
|
|
|
|
if (n < 0 && result != NULL && PyString_Check(result)) {
|
|
char *s = PyString_AS_STRING(result);
|
|
Py_ssize_t len = PyString_GET_SIZE(result);
|
|
if (len == 0) {
|
|
Py_DECREF(result);
|
|
result = NULL;
|
|
PyErr_SetString(PyExc_EOFError,
|
|
"EOF when reading a line");
|
|
}
|
|
else if (s[len-1] == '\n') {
|
|
if (result->ob_refcnt == 1)
|
|
_PyString_Resize(&result, len-1);
|
|
else {
|
|
PyObject *v;
|
|
v = PyString_FromStringAndSize(s, len-1);
|
|
Py_DECREF(result);
|
|
result = v;
|
|
}
|
|
}
|
|
}
|
|
#ifdef Py_USING_UNICODE
|
|
if (n < 0 && result != NULL && PyUnicode_Check(result)) {
|
|
Py_UNICODE *s = PyUnicode_AS_UNICODE(result);
|
|
Py_ssize_t len = PyUnicode_GET_SIZE(result);
|
|
if (len == 0) {
|
|
Py_DECREF(result);
|
|
result = NULL;
|
|
PyErr_SetString(PyExc_EOFError,
|
|
"EOF when reading a line");
|
|
}
|
|
else if (s[len-1] == '\n') {
|
|
if (result->ob_refcnt == 1)
|
|
PyUnicode_Resize(&result, len-1);
|
|
else {
|
|
PyObject *v;
|
|
v = PyUnicode_FromUnicode(s, len-1);
|
|
Py_DECREF(result);
|
|
result = v;
|
|
}
|
|
}
|
|
}
|
|
#endif
|
|
return result;
|
|
}
|
|
|
|
/* Python method */
|
|
|
|
static PyObject *
|
|
file_readline(PyFileObject *f, PyObject *args)
|
|
{
|
|
int n = -1;
|
|
|
|
if (f->f_fp == NULL)
|
|
return err_closed();
|
|
/* refuse to mix with f.next() */
|
|
if (f->f_buf != NULL &&
|
|
(f->f_bufend - f->f_bufptr) > 0 &&
|
|
f->f_buf[0] != '\0')
|
|
return err_iterbuffered();
|
|
if (!PyArg_ParseTuple(args, "|i:readline", &n))
|
|
return NULL;
|
|
if (n == 0)
|
|
return PyString_FromString("");
|
|
if (n < 0)
|
|
n = 0;
|
|
return get_line(f, n);
|
|
}
|
|
|
|
static PyObject *
|
|
file_readlines(PyFileObject *f, PyObject *args)
|
|
{
|
|
long sizehint = 0;
|
|
PyObject *list;
|
|
PyObject *line;
|
|
char small_buffer[SMALLCHUNK];
|
|
char *buffer = small_buffer;
|
|
size_t buffersize = SMALLCHUNK;
|
|
PyObject *big_buffer = NULL;
|
|
size_t nfilled = 0;
|
|
size_t nread;
|
|
size_t totalread = 0;
|
|
char *p, *q, *end;
|
|
int err;
|
|
int shortread = 0;
|
|
|
|
if (f->f_fp == NULL)
|
|
return err_closed();
|
|
/* refuse to mix with f.next() */
|
|
if (f->f_buf != NULL &&
|
|
(f->f_bufend - f->f_bufptr) > 0 &&
|
|
f->f_buf[0] != '\0')
|
|
return err_iterbuffered();
|
|
if (!PyArg_ParseTuple(args, "|l:readlines", &sizehint))
|
|
return NULL;
|
|
if ((list = PyList_New(0)) == NULL)
|
|
return NULL;
|
|
for (;;) {
|
|
if (shortread)
|
|
nread = 0;
|
|
else {
|
|
Py_BEGIN_ALLOW_THREADS
|
|
errno = 0;
|
|
nread = Py_UniversalNewlineFread(buffer+nfilled,
|
|
buffersize-nfilled, f->f_fp, (PyObject *)f);
|
|
Py_END_ALLOW_THREADS
|
|
shortread = (nread < buffersize-nfilled);
|
|
}
|
|
if (nread == 0) {
|
|
sizehint = 0;
|
|
if (!ferror(f->f_fp))
|
|
break;
|
|
PyErr_SetFromErrno(PyExc_IOError);
|
|
clearerr(f->f_fp);
|
|
error:
|
|
Py_DECREF(list);
|
|
list = NULL;
|
|
goto cleanup;
|
|
}
|
|
totalread += nread;
|
|
p = (char *)memchr(buffer+nfilled, '\n', nread);
|
|
if (p == NULL) {
|
|
/* Need a larger buffer to fit this line */
|
|
nfilled += nread;
|
|
buffersize *= 2;
|
|
if (buffersize > PY_SSIZE_T_MAX) {
|
|
PyErr_SetString(PyExc_OverflowError,
|
|
"line is longer than a Python string can hold");
|
|
goto error;
|
|
}
|
|
if (big_buffer == NULL) {
|
|
/* Create the big buffer */
|
|
big_buffer = PyString_FromStringAndSize(
|
|
NULL, buffersize);
|
|
if (big_buffer == NULL)
|
|
goto error;
|
|
buffer = PyString_AS_STRING(big_buffer);
|
|
memcpy(buffer, small_buffer, nfilled);
|
|
}
|
|
else {
|
|
/* Grow the big buffer */
|
|
if ( _PyString_Resize(&big_buffer, buffersize) < 0 )
|
|
goto error;
|
|
buffer = PyString_AS_STRING(big_buffer);
|
|
}
|
|
continue;
|
|
}
|
|
end = buffer+nfilled+nread;
|
|
q = buffer;
|
|
do {
|
|
/* Process complete lines */
|
|
p++;
|
|
line = PyString_FromStringAndSize(q, p-q);
|
|
if (line == NULL)
|
|
goto error;
|
|
err = PyList_Append(list, line);
|
|
Py_DECREF(line);
|
|
if (err != 0)
|
|
goto error;
|
|
q = p;
|
|
p = (char *)memchr(q, '\n', end-q);
|
|
} while (p != NULL);
|
|
/* Move the remaining incomplete line to the start */
|
|
nfilled = end-q;
|
|
memmove(buffer, q, nfilled);
|
|
if (sizehint > 0)
|
|
if (totalread >= (size_t)sizehint)
|
|
break;
|
|
}
|
|
if (nfilled != 0) {
|
|
/* Partial last line */
|
|
line = PyString_FromStringAndSize(buffer, nfilled);
|
|
if (line == NULL)
|
|
goto error;
|
|
if (sizehint > 0) {
|
|
/* Need to complete the last line */
|
|
PyObject *rest = get_line(f, 0);
|
|
if (rest == NULL) {
|
|
Py_DECREF(line);
|
|
goto error;
|
|
}
|
|
PyString_Concat(&line, rest);
|
|
Py_DECREF(rest);
|
|
if (line == NULL)
|
|
goto error;
|
|
}
|
|
err = PyList_Append(list, line);
|
|
Py_DECREF(line);
|
|
if (err != 0)
|
|
goto error;
|
|
}
|
|
cleanup:
|
|
Py_XDECREF(big_buffer);
|
|
return list;
|
|
}
|
|
|
|
static PyObject *
|
|
file_write(PyFileObject *f, PyObject *args)
|
|
{
|
|
char *s;
|
|
Py_ssize_t n, n2;
|
|
if (f->f_fp == NULL)
|
|
return err_closed();
|
|
if (!PyArg_ParseTuple(args, f->f_binary ? "s#" : "t#", &s, &n))
|
|
return NULL;
|
|
Py_BEGIN_ALLOW_THREADS
|
|
errno = 0;
|
|
n2 = fwrite(s, 1, n, f->f_fp);
|
|
Py_END_ALLOW_THREADS
|
|
if (n2 != n) {
|
|
PyErr_SetFromErrno(PyExc_IOError);
|
|
clearerr(f->f_fp);
|
|
return NULL;
|
|
}
|
|
Py_INCREF(Py_None);
|
|
return Py_None;
|
|
}
|
|
|
|
static PyObject *
|
|
file_writelines(PyFileObject *f, PyObject *seq)
|
|
{
|
|
#define CHUNKSIZE 1000
|
|
PyObject *list, *line;
|
|
PyObject *it; /* iter(seq) */
|
|
PyObject *result;
|
|
int index, islist;
|
|
Py_ssize_t i, j, nwritten, len;
|
|
|
|
assert(seq != NULL);
|
|
if (f->f_fp == NULL)
|
|
return err_closed();
|
|
|
|
result = NULL;
|
|
list = NULL;
|
|
islist = PyList_Check(seq);
|
|
if (islist)
|
|
it = NULL;
|
|
else {
|
|
it = PyObject_GetIter(seq);
|
|
if (it == NULL) {
|
|
PyErr_SetString(PyExc_TypeError,
|
|
"writelines() requires an iterable argument");
|
|
return NULL;
|
|
}
|
|
/* From here on, fail by going to error, to reclaim "it". */
|
|
list = PyList_New(CHUNKSIZE);
|
|
if (list == NULL)
|
|
goto error;
|
|
}
|
|
|
|
/* Strategy: slurp CHUNKSIZE lines into a private list,
|
|
checking that they are all strings, then write that list
|
|
without holding the interpreter lock, then come back for more. */
|
|
for (index = 0; ; index += CHUNKSIZE) {
|
|
if (islist) {
|
|
Py_XDECREF(list);
|
|
list = PyList_GetSlice(seq, index, index+CHUNKSIZE);
|
|
if (list == NULL)
|
|
goto error;
|
|
j = PyList_GET_SIZE(list);
|
|
}
|
|
else {
|
|
for (j = 0; j < CHUNKSIZE; j++) {
|
|
line = PyIter_Next(it);
|
|
if (line == NULL) {
|
|
if (PyErr_Occurred())
|
|
goto error;
|
|
break;
|
|
}
|
|
PyList_SetItem(list, j, line);
|
|
}
|
|
}
|
|
if (j == 0)
|
|
break;
|
|
|
|
/* Check that all entries are indeed strings. If not,
|
|
apply the same rules as for file.write() and
|
|
convert the results to strings. This is slow, but
|
|
seems to be the only way since all conversion APIs
|
|
could potentially execute Python code. */
|
|
for (i = 0; i < j; i++) {
|
|
PyObject *v = PyList_GET_ITEM(list, i);
|
|
if (!PyString_Check(v)) {
|
|
const char *buffer;
|
|
if (((f->f_binary &&
|
|
PyObject_AsReadBuffer(v,
|
|
(const void**)&buffer,
|
|
&len)) ||
|
|
PyObject_AsCharBuffer(v,
|
|
&buffer,
|
|
&len))) {
|
|
PyErr_SetString(PyExc_TypeError,
|
|
"writelines() argument must be a sequence of strings");
|
|
goto error;
|
|
}
|
|
line = PyString_FromStringAndSize(buffer,
|
|
len);
|
|
if (line == NULL)
|
|
goto error;
|
|
Py_DECREF(v);
|
|
PyList_SET_ITEM(list, i, line);
|
|
}
|
|
}
|
|
|
|
/* Since we are releasing the global lock, the
|
|
following code may *not* execute Python code. */
|
|
Py_BEGIN_ALLOW_THREADS
|
|
errno = 0;
|
|
for (i = 0; i < j; i++) {
|
|
line = PyList_GET_ITEM(list, i);
|
|
len = PyString_GET_SIZE(line);
|
|
nwritten = fwrite(PyString_AS_STRING(line),
|
|
1, len, f->f_fp);
|
|
if (nwritten != len) {
|
|
Py_BLOCK_THREADS
|
|
PyErr_SetFromErrno(PyExc_IOError);
|
|
clearerr(f->f_fp);
|
|
goto error;
|
|
}
|
|
}
|
|
Py_END_ALLOW_THREADS
|
|
|
|
if (j < CHUNKSIZE)
|
|
break;
|
|
}
|
|
|
|
Py_INCREF(Py_None);
|
|
result = Py_None;
|
|
error:
|
|
Py_XDECREF(list);
|
|
Py_XDECREF(it);
|
|
return result;
|
|
#undef CHUNKSIZE
|
|
}
|
|
|
|
static PyObject *
|
|
file_self(PyFileObject *f)
|
|
{
|
|
if (f->f_fp == NULL)
|
|
return err_closed();
|
|
Py_INCREF(f);
|
|
return (PyObject *)f;
|
|
}
|
|
|
|
static PyObject *
|
|
file_exit(PyFileObject *f, PyObject *args)
|
|
{
|
|
PyObject *ret = file_close(f);
|
|
if (!ret)
|
|
/* If error occurred, pass through */
|
|
return NULL;
|
|
Py_DECREF(ret);
|
|
/* We cannot return the result of close since a true
|
|
* value will be interpreted as "yes, swallow the
|
|
* exception if one was raised inside the with block". */
|
|
Py_RETURN_NONE;
|
|
}
|
|
|
|
PyDoc_STRVAR(readline_doc,
|
|
"readline([size]) -> next line from the file, as a string.\n"
|
|
"\n"
|
|
"Retain newline. A non-negative size argument limits the maximum\n"
|
|
"number of bytes to return (an incomplete line may be returned then).\n"
|
|
"Return an empty string at EOF.");
|
|
|
|
PyDoc_STRVAR(read_doc,
|
|
"read([size]) -> read at most size bytes, returned as a string.\n"
|
|
"\n"
|
|
"If the size argument is negative or omitted, read until EOF is reached.\n"
|
|
"Notice that when in non-blocking mode, less data than what was requested\n"
|
|
"may be returned, even if no size parameter was given.");
|
|
|
|
PyDoc_STRVAR(write_doc,
|
|
"write(str) -> None. Write string str to file.\n"
|
|
"\n"
|
|
"Note that due to buffering, flush() or close() may be needed before\n"
|
|
"the file on disk reflects the data written.");
|
|
|
|
PyDoc_STRVAR(fileno_doc,
|
|
"fileno() -> integer \"file descriptor\".\n"
|
|
"\n"
|
|
"This is needed for lower-level file interfaces, such os.read().");
|
|
|
|
PyDoc_STRVAR(seek_doc,
|
|
"seek(offset[, whence]) -> None. Move to new file position.\n"
|
|
"\n"
|
|
"Argument offset is a byte count. Optional argument whence defaults to\n"
|
|
"0 (offset from start of file, offset should be >= 0); other values are 1\n"
|
|
"(move relative to current position, positive or negative), and 2 (move\n"
|
|
"relative to end of file, usually negative, although many platforms allow\n"
|
|
"seeking beyond the end of a file). If the file is opened in text mode,\n"
|
|
"only offsets returned by tell() are legal. Use of other offsets causes\n"
|
|
"undefined behavior."
|
|
"\n"
|
|
"Note that not all file objects are seekable.");
|
|
|
|
#ifdef HAVE_FTRUNCATE
|
|
PyDoc_STRVAR(truncate_doc,
|
|
"truncate([size]) -> None. Truncate the file to at most size bytes.\n"
|
|
"\n"
|
|
"Size defaults to the current file position, as returned by tell().");
|
|
#endif
|
|
|
|
PyDoc_STRVAR(tell_doc,
|
|
"tell() -> current file position, an integer (may be a long integer).");
|
|
|
|
PyDoc_STRVAR(readinto_doc,
|
|
"readinto() -> Undocumented. Don't use this; it may go away.");
|
|
|
|
PyDoc_STRVAR(readlines_doc,
|
|
"readlines([size]) -> list of strings, each a line from the file.\n"
|
|
"\n"
|
|
"Call readline() repeatedly and return a list of the lines so read.\n"
|
|
"The optional size argument, if given, is an approximate bound on the\n"
|
|
"total number of bytes in the lines returned.");
|
|
|
|
PyDoc_STRVAR(writelines_doc,
|
|
"writelines(sequence_of_strings) -> None. Write the strings to the file.\n"
|
|
"\n"
|
|
"Note that newlines are not added. The sequence can be any iterable object\n"
|
|
"producing strings. This is equivalent to calling write() for each string.");
|
|
|
|
PyDoc_STRVAR(flush_doc,
|
|
"flush() -> None. Flush the internal I/O buffer.");
|
|
|
|
PyDoc_STRVAR(close_doc,
|
|
"close() -> None or (perhaps) an integer. Close the file.\n"
|
|
"\n"
|
|
"Sets data attribute .closed to True. A closed file cannot be used for\n"
|
|
"further I/O operations. close() may be called more than once without\n"
|
|
"error. Some kinds of file objects (for example, opened by popen())\n"
|
|
"may return an exit status upon closing.");
|
|
|
|
PyDoc_STRVAR(isatty_doc,
|
|
"isatty() -> true or false. True if the file is connected to a tty device.");
|
|
|
|
PyDoc_STRVAR(enter_doc,
|
|
"__enter__() -> self.");
|
|
|
|
PyDoc_STRVAR(exit_doc,
|
|
"__exit__(*excinfo) -> None. Closes the file.");
|
|
|
|
static PyMethodDef file_methods[] = {
|
|
{"readline", (PyCFunction)file_readline, METH_VARARGS, readline_doc},
|
|
{"read", (PyCFunction)file_read, METH_VARARGS, read_doc},
|
|
{"write", (PyCFunction)file_write, METH_VARARGS, write_doc},
|
|
{"fileno", (PyCFunction)file_fileno, METH_NOARGS, fileno_doc},
|
|
{"seek", (PyCFunction)file_seek, METH_VARARGS, seek_doc},
|
|
#ifdef HAVE_FTRUNCATE
|
|
{"truncate", (PyCFunction)file_truncate, METH_VARARGS, truncate_doc},
|
|
#endif
|
|
{"tell", (PyCFunction)file_tell, METH_NOARGS, tell_doc},
|
|
{"readinto", (PyCFunction)file_readinto, METH_VARARGS, readinto_doc},
|
|
{"readlines", (PyCFunction)file_readlines,METH_VARARGS, readlines_doc},
|
|
{"writelines",(PyCFunction)file_writelines, METH_O, writelines_doc},
|
|
{"flush", (PyCFunction)file_flush, METH_NOARGS, flush_doc},
|
|
{"close", (PyCFunction)file_close, METH_NOARGS, close_doc},
|
|
{"isatty", (PyCFunction)file_isatty, METH_NOARGS, isatty_doc},
|
|
{"__enter__", (PyCFunction)file_self, METH_NOARGS, enter_doc},
|
|
{"__exit__", (PyCFunction)file_exit, METH_VARARGS, exit_doc},
|
|
{NULL, NULL} /* sentinel */
|
|
};
|
|
|
|
#define OFF(x) offsetof(PyFileObject, x)
|
|
|
|
static PyMemberDef file_memberlist[] = {
|
|
{"mode", T_OBJECT, OFF(f_mode), RO,
|
|
"file mode ('r', 'U', 'w', 'a', possibly with 'b' or '+' added)"},
|
|
{"name", T_OBJECT, OFF(f_name), RO,
|
|
"file name"},
|
|
{"encoding", T_OBJECT, OFF(f_encoding), RO,
|
|
"file encoding"},
|
|
/* getattr(f, "closed") is implemented without this table */
|
|
{NULL} /* Sentinel */
|
|
};
|
|
|
|
static PyObject *
|
|
get_closed(PyFileObject *f, void *closure)
|
|
{
|
|
return PyBool_FromLong((long)(f->f_fp == 0));
|
|
}
|
|
static PyObject *
|
|
get_newlines(PyFileObject *f, void *closure)
|
|
{
|
|
switch (f->f_newlinetypes) {
|
|
case NEWLINE_UNKNOWN:
|
|
Py_INCREF(Py_None);
|
|
return Py_None;
|
|
case NEWLINE_CR:
|
|
return PyString_FromString("\r");
|
|
case NEWLINE_LF:
|
|
return PyString_FromString("\n");
|
|
case NEWLINE_CR|NEWLINE_LF:
|
|
return Py_BuildValue("(ss)", "\r", "\n");
|
|
case NEWLINE_CRLF:
|
|
return PyString_FromString("\r\n");
|
|
case NEWLINE_CR|NEWLINE_CRLF:
|
|
return Py_BuildValue("(ss)", "\r", "\r\n");
|
|
case NEWLINE_LF|NEWLINE_CRLF:
|
|
return Py_BuildValue("(ss)", "\n", "\r\n");
|
|
case NEWLINE_CR|NEWLINE_LF|NEWLINE_CRLF:
|
|
return Py_BuildValue("(sss)", "\r", "\n", "\r\n");
|
|
default:
|
|
PyErr_Format(PyExc_SystemError,
|
|
"Unknown newlines value 0x%x\n",
|
|
f->f_newlinetypes);
|
|
return NULL;
|
|
}
|
|
}
|
|
|
|
static PyGetSetDef file_getsetlist[] = {
|
|
{"closed", (getter)get_closed, NULL, "True if the file is closed"},
|
|
{"newlines", (getter)get_newlines, NULL,
|
|
"end-of-line convention used in this file"},
|
|
{0},
|
|
};
|
|
|
|
static void
|
|
drop_readahead(PyFileObject *f)
|
|
{
|
|
if (f->f_buf != NULL) {
|
|
PyMem_Free(f->f_buf);
|
|
f->f_buf = NULL;
|
|
}
|
|
}
|
|
|
|
/* Make sure that file has a readahead buffer with at least one byte
|
|
(unless at EOF) and no more than bufsize. Returns negative value on
|
|
error, will set MemoryError if bufsize bytes cannot be allocated. */
|
|
static int
|
|
readahead(PyFileObject *f, int bufsize)
|
|
{
|
|
Py_ssize_t chunksize;
|
|
|
|
if (f->f_buf != NULL) {
|
|
if( (f->f_bufend - f->f_bufptr) >= 1)
|
|
return 0;
|
|
else
|
|
drop_readahead(f);
|
|
}
|
|
if ((f->f_buf = (char *)PyMem_Malloc(bufsize)) == NULL) {
|
|
PyErr_NoMemory();
|
|
return -1;
|
|
}
|
|
Py_BEGIN_ALLOW_THREADS
|
|
errno = 0;
|
|
chunksize = Py_UniversalNewlineFread(
|
|
f->f_buf, bufsize, f->f_fp, (PyObject *)f);
|
|
Py_END_ALLOW_THREADS
|
|
if (chunksize == 0) {
|
|
if (ferror(f->f_fp)) {
|
|
PyErr_SetFromErrno(PyExc_IOError);
|
|
clearerr(f->f_fp);
|
|
drop_readahead(f);
|
|
return -1;
|
|
}
|
|
}
|
|
f->f_bufptr = f->f_buf;
|
|
f->f_bufend = f->f_buf + chunksize;
|
|
return 0;
|
|
}
|
|
|
|
/* Used by file_iternext. The returned string will start with 'skip'
|
|
uninitialized bytes followed by the remainder of the line. Don't be
|
|
horrified by the recursive call: maximum recursion depth is limited by
|
|
logarithmic buffer growth to about 50 even when reading a 1gb line. */
|
|
|
|
static PyStringObject *
|
|
readahead_get_line_skip(PyFileObject *f, int skip, int bufsize)
|
|
{
|
|
PyStringObject* s;
|
|
char *bufptr;
|
|
char *buf;
|
|
Py_ssize_t len;
|
|
|
|
if (f->f_buf == NULL)
|
|
if (readahead(f, bufsize) < 0)
|
|
return NULL;
|
|
|
|
len = f->f_bufend - f->f_bufptr;
|
|
if (len == 0)
|
|
return (PyStringObject *)
|
|
PyString_FromStringAndSize(NULL, skip);
|
|
bufptr = (char *)memchr(f->f_bufptr, '\n', len);
|
|
if (bufptr != NULL) {
|
|
bufptr++; /* Count the '\n' */
|
|
len = bufptr - f->f_bufptr;
|
|
s = (PyStringObject *)
|
|
PyString_FromStringAndSize(NULL, skip+len);
|
|
if (s == NULL)
|
|
return NULL;
|
|
memcpy(PyString_AS_STRING(s)+skip, f->f_bufptr, len);
|
|
f->f_bufptr = bufptr;
|
|
if (bufptr == f->f_bufend)
|
|
drop_readahead(f);
|
|
} else {
|
|
bufptr = f->f_bufptr;
|
|
buf = f->f_buf;
|
|
f->f_buf = NULL; /* Force new readahead buffer */
|
|
assert(skip+len < INT_MAX);
|
|
s = readahead_get_line_skip(
|
|
f, (int)(skip+len), bufsize + (bufsize>>2) );
|
|
if (s == NULL) {
|
|
PyMem_Free(buf);
|
|
return NULL;
|
|
}
|
|
memcpy(PyString_AS_STRING(s)+skip, bufptr, len);
|
|
PyMem_Free(buf);
|
|
}
|
|
return s;
|
|
}
|
|
|
|
/* A larger buffer size may actually decrease performance. */
|
|
#define READAHEAD_BUFSIZE 8192
|
|
|
|
static PyObject *
|
|
file_iternext(PyFileObject *f)
|
|
{
|
|
PyStringObject* l;
|
|
|
|
if (f->f_fp == NULL)
|
|
return err_closed();
|
|
|
|
l = readahead_get_line_skip(f, 0, READAHEAD_BUFSIZE);
|
|
if (l == NULL || PyString_GET_SIZE(l) == 0) {
|
|
Py_XDECREF(l);
|
|
return NULL;
|
|
}
|
|
return (PyObject *)l;
|
|
}
|
|
|
|
|
|
static PyObject *
|
|
file_new(PyTypeObject *type, PyObject *args, PyObject *kwds)
|
|
{
|
|
PyObject *self;
|
|
static PyObject *not_yet_string;
|
|
|
|
assert(type != NULL && type->tp_alloc != NULL);
|
|
|
|
if (not_yet_string == NULL) {
|
|
not_yet_string = PyString_FromString("<uninitialized file>");
|
|
if (not_yet_string == NULL)
|
|
return NULL;
|
|
}
|
|
|
|
self = type->tp_alloc(type, 0);
|
|
if (self != NULL) {
|
|
/* Always fill in the name and mode, so that nobody else
|
|
needs to special-case NULLs there. */
|
|
Py_INCREF(not_yet_string);
|
|
((PyFileObject *)self)->f_name = not_yet_string;
|
|
Py_INCREF(not_yet_string);
|
|
((PyFileObject *)self)->f_mode = not_yet_string;
|
|
Py_INCREF(Py_None);
|
|
((PyFileObject *)self)->f_encoding = Py_None;
|
|
((PyFileObject *)self)->weakreflist = NULL;
|
|
}
|
|
return self;
|
|
}
|
|
|
|
static int
|
|
file_init(PyObject *self, PyObject *args, PyObject *kwds)
|
|
{
|
|
PyFileObject *foself = (PyFileObject *)self;
|
|
int ret = 0;
|
|
static char *kwlist[] = {"name", "mode", "buffering", 0};
|
|
char *name = NULL;
|
|
char *mode = "r";
|
|
int bufsize = -1;
|
|
int wideargument = 0;
|
|
|
|
assert(PyFile_Check(self));
|
|
if (foself->f_fp != NULL) {
|
|
/* Have to close the existing file first. */
|
|
PyObject *closeresult = file_close(foself);
|
|
if (closeresult == NULL)
|
|
return -1;
|
|
Py_DECREF(closeresult);
|
|
}
|
|
|
|
#ifdef Py_WIN_WIDE_FILENAMES
|
|
if (GetVersion() < 0x80000000) { /* On NT, so wide API available */
|
|
PyObject *po;
|
|
if (PyArg_ParseTupleAndKeywords(args, kwds, "U|si:file",
|
|
kwlist, &po, &mode, &bufsize)) {
|
|
wideargument = 1;
|
|
if (fill_file_fields(foself, NULL, po, mode,
|
|
fclose) == NULL)
|
|
goto Error;
|
|
} else {
|
|
/* Drop the argument parsing error as narrow
|
|
strings are also valid. */
|
|
PyErr_Clear();
|
|
}
|
|
}
|
|
#endif
|
|
|
|
if (!wideargument) {
|
|
PyObject *o_name;
|
|
|
|
if (!PyArg_ParseTupleAndKeywords(args, kwds, "et|si:file", kwlist,
|
|
Py_FileSystemDefaultEncoding,
|
|
&name,
|
|
&mode, &bufsize))
|
|
return -1;
|
|
|
|
/* We parse again to get the name as a PyObject */
|
|
if (!PyArg_ParseTupleAndKeywords(args, kwds, "O|si:file",
|
|
kwlist, &o_name, &mode,
|
|
&bufsize))
|
|
goto Error;
|
|
|
|
if (fill_file_fields(foself, NULL, o_name, mode,
|
|
fclose) == NULL)
|
|
goto Error;
|
|
}
|
|
if (open_the_file(foself, name, mode) == NULL)
|
|
goto Error;
|
|
foself->f_setbuf = NULL;
|
|
PyFile_SetBufSize(self, bufsize);
|
|
goto Done;
|
|
|
|
Error:
|
|
ret = -1;
|
|
/* fall through */
|
|
Done:
|
|
PyMem_Free(name); /* free the encoded string */
|
|
return ret;
|
|
}
|
|
|
|
PyDoc_VAR(file_doc) =
|
|
PyDoc_STR(
|
|
"file(name[, mode[, buffering]]) -> file object\n"
|
|
"\n"
|
|
"Open a file. The mode can be 'r', 'w' or 'a' for reading (default),\n"
|
|
"writing or appending. The file will be created if it doesn't exist\n"
|
|
"when opened for writing or appending; it will be truncated when\n"
|
|
"opened for writing. Add a 'b' to the mode for binary files.\n"
|
|
"Add a '+' to the mode to allow simultaneous reading and writing.\n"
|
|
"If the buffering argument is given, 0 means unbuffered, 1 means line\n"
|
|
"buffered, and larger numbers specify the buffer size.\n"
|
|
)
|
|
PyDoc_STR(
|
|
"Add a 'U' to mode to open the file for input with universal newline\n"
|
|
"support. Any line ending in the input file will be seen as a '\\n'\n"
|
|
"in Python. Also, a file so opened gains the attribute 'newlines';\n"
|
|
"the value for this attribute is one of None (no newline read yet),\n"
|
|
"'\\r', '\\n', '\\r\\n' or a tuple containing all the newline types seen.\n"
|
|
"\n"
|
|
"'U' cannot be combined with 'w' or '+' mode.\n"
|
|
);
|
|
|
|
PyTypeObject PyFile_Type = {
|
|
PyObject_HEAD_INIT(&PyType_Type)
|
|
0,
|
|
"file",
|
|
sizeof(PyFileObject),
|
|
0,
|
|
(destructor)file_dealloc, /* tp_dealloc */
|
|
0, /* tp_print */
|
|
0, /* tp_getattr */
|
|
0, /* tp_setattr */
|
|
0, /* tp_compare */
|
|
(reprfunc)file_repr, /* tp_repr */
|
|
0, /* tp_as_number */
|
|
0, /* tp_as_sequence */
|
|
0, /* tp_as_mapping */
|
|
0, /* tp_hash */
|
|
0, /* tp_call */
|
|
0, /* tp_str */
|
|
PyObject_GenericGetAttr, /* tp_getattro */
|
|
0, /* tp_setattro */
|
|
0, /* tp_as_buffer */
|
|
Py_TPFLAGS_DEFAULT | Py_TPFLAGS_BASETYPE, /* tp_flags */
|
|
file_doc, /* tp_doc */
|
|
0, /* tp_traverse */
|
|
0, /* tp_clear */
|
|
0, /* tp_richcompare */
|
|
offsetof(PyFileObject, weakreflist), /* tp_weaklistoffset */
|
|
(getiterfunc)file_self, /* tp_iter */
|
|
(iternextfunc)file_iternext, /* tp_iternext */
|
|
file_methods, /* tp_methods */
|
|
file_memberlist, /* tp_members */
|
|
file_getsetlist, /* tp_getset */
|
|
0, /* tp_base */
|
|
0, /* tp_dict */
|
|
0, /* tp_descr_get */
|
|
0, /* tp_descr_set */
|
|
0, /* tp_dictoffset */
|
|
file_init, /* tp_init */
|
|
PyType_GenericAlloc, /* tp_alloc */
|
|
file_new, /* tp_new */
|
|
PyObject_Del, /* tp_free */
|
|
};
|
|
|
|
/* Interfaces to write objects/strings to file-like objects */
|
|
|
|
int
|
|
PyFile_WriteObject(PyObject *v, PyObject *f, int flags)
|
|
{
|
|
PyObject *writer, *value, *args, *result;
|
|
if (f == NULL) {
|
|
PyErr_SetString(PyExc_TypeError, "writeobject with NULL file");
|
|
return -1;
|
|
}
|
|
else if (PyFile_Check(f)) {
|
|
FILE *fp = PyFile_AsFile(f);
|
|
#ifdef Py_USING_UNICODE
|
|
PyObject *enc = ((PyFileObject*)f)->f_encoding;
|
|
int result;
|
|
#endif
|
|
if (fp == NULL) {
|
|
err_closed();
|
|
return -1;
|
|
}
|
|
#ifdef Py_USING_UNICODE
|
|
if ((flags & Py_PRINT_RAW) &&
|
|
PyUnicode_Check(v) && enc != Py_None) {
|
|
char *cenc = PyString_AS_STRING(enc);
|
|
value = PyUnicode_AsEncodedString(v, cenc, "strict");
|
|
if (value == NULL)
|
|
return -1;
|
|
} else {
|
|
value = v;
|
|
Py_INCREF(value);
|
|
}
|
|
result = PyObject_Print(value, fp, flags);
|
|
Py_DECREF(value);
|
|
return result;
|
|
#else
|
|
return PyObject_Print(v, fp, flags);
|
|
#endif
|
|
}
|
|
writer = PyObject_GetAttrString(f, "write");
|
|
if (writer == NULL)
|
|
return -1;
|
|
if (flags & Py_PRINT_RAW) {
|
|
if (PyUnicode_Check(v)) {
|
|
value = v;
|
|
Py_INCREF(value);
|
|
} else
|
|
value = PyObject_Str(v);
|
|
}
|
|
else
|
|
value = PyObject_Repr(v);
|
|
if (value == NULL) {
|
|
Py_DECREF(writer);
|
|
return -1;
|
|
}
|
|
args = PyTuple_Pack(1, value);
|
|
if (args == NULL) {
|
|
Py_DECREF(value);
|
|
Py_DECREF(writer);
|
|
return -1;
|
|
}
|
|
result = PyEval_CallObject(writer, args);
|
|
Py_DECREF(args);
|
|
Py_DECREF(value);
|
|
Py_DECREF(writer);
|
|
if (result == NULL)
|
|
return -1;
|
|
Py_DECREF(result);
|
|
return 0;
|
|
}
|
|
|
|
int
|
|
PyFile_WriteString(const char *s, PyObject *f)
|
|
{
|
|
if (f == NULL) {
|
|
/* Should be caused by a pre-existing error */
|
|
if (!PyErr_Occurred())
|
|
PyErr_SetString(PyExc_SystemError,
|
|
"null file for PyFile_WriteString");
|
|
return -1;
|
|
}
|
|
else if (PyFile_Check(f)) {
|
|
FILE *fp = PyFile_AsFile(f);
|
|
if (fp == NULL) {
|
|
err_closed();
|
|
return -1;
|
|
}
|
|
fputs(s, fp);
|
|
return 0;
|
|
}
|
|
else if (!PyErr_Occurred()) {
|
|
PyObject *v = PyString_FromString(s);
|
|
int err;
|
|
if (v == NULL)
|
|
return -1;
|
|
err = PyFile_WriteObject(v, f, Py_PRINT_RAW);
|
|
Py_DECREF(v);
|
|
return err;
|
|
}
|
|
else
|
|
return -1;
|
|
}
|
|
|
|
/* Try to get a file-descriptor from a Python object. If the object
|
|
is an integer or long integer, its value is returned. If not, the
|
|
object's fileno() method is called if it exists; the method must return
|
|
an integer or long integer, which is returned as the file descriptor value.
|
|
-1 is returned on failure.
|
|
*/
|
|
|
|
int PyObject_AsFileDescriptor(PyObject *o)
|
|
{
|
|
int fd;
|
|
PyObject *meth;
|
|
|
|
if (PyInt_Check(o)) {
|
|
fd = PyInt_AsLong(o);
|
|
}
|
|
else if (PyLong_Check(o)) {
|
|
fd = PyLong_AsLong(o);
|
|
}
|
|
else if ((meth = PyObject_GetAttrString(o, "fileno")) != NULL)
|
|
{
|
|
PyObject *fno = PyEval_CallObject(meth, NULL);
|
|
Py_DECREF(meth);
|
|
if (fno == NULL)
|
|
return -1;
|
|
|
|
if (PyInt_Check(fno)) {
|
|
fd = PyInt_AsLong(fno);
|
|
Py_DECREF(fno);
|
|
}
|
|
else if (PyLong_Check(fno)) {
|
|
fd = PyLong_AsLong(fno);
|
|
Py_DECREF(fno);
|
|
}
|
|
else {
|
|
PyErr_SetString(PyExc_TypeError,
|
|
"fileno() returned a non-integer");
|
|
Py_DECREF(fno);
|
|
return -1;
|
|
}
|
|
}
|
|
else {
|
|
PyErr_SetString(PyExc_TypeError,
|
|
"argument must be an int, or have a fileno() method.");
|
|
return -1;
|
|
}
|
|
|
|
if (fd == -1 && PyErr_Occurred())
|
|
return -1;
|
|
if (fd < 0) {
|
|
PyErr_Format(PyExc_ValueError,
|
|
"file descriptor cannot be a negative integer (%i)",
|
|
fd);
|
|
return -1;
|
|
}
|
|
return fd;
|
|
}
|
|
|
|
/* From here on we need access to the real fgets and fread */
|
|
#undef fgets
|
|
#undef fread
|
|
|
|
/*
|
|
** Py_UniversalNewlineFgets is an fgets variation that understands
|
|
** all of \r, \n and \r\n conventions.
|
|
** The stream should be opened in binary mode.
|
|
** If fobj is NULL the routine always does newline conversion, and
|
|
** it may peek one char ahead to gobble the second char in \r\n.
|
|
** If fobj is non-NULL it must be a PyFileObject. In this case there
|
|
** is no readahead but in stead a flag is used to skip a following
|
|
** \n on the next read. Also, if the file is open in binary mode
|
|
** the whole conversion is skipped. Finally, the routine keeps track of
|
|
** the different types of newlines seen.
|
|
** Note that we need no error handling: fgets() treats error and eof
|
|
** identically.
|
|
*/
|
|
char *
|
|
Py_UniversalNewlineFgets(char *buf, int n, FILE *stream, PyObject *fobj)
|
|
{
|
|
char *p = buf;
|
|
int c;
|
|
int newlinetypes = 0;
|
|
int skipnextlf = 0;
|
|
int univ_newline = 1;
|
|
|
|
if (fobj) {
|
|
if (!PyFile_Check(fobj)) {
|
|
errno = ENXIO; /* What can you do... */
|
|
return NULL;
|
|
}
|
|
univ_newline = ((PyFileObject *)fobj)->f_univ_newline;
|
|
if ( !univ_newline )
|
|
return fgets(buf, n, stream);
|
|
newlinetypes = ((PyFileObject *)fobj)->f_newlinetypes;
|
|
skipnextlf = ((PyFileObject *)fobj)->f_skipnextlf;
|
|
}
|
|
FLOCKFILE(stream);
|
|
c = 'x'; /* Shut up gcc warning */
|
|
while (--n > 0 && (c = GETC(stream)) != EOF ) {
|
|
if (skipnextlf ) {
|
|
skipnextlf = 0;
|
|
if (c == '\n') {
|
|
/* Seeing a \n here with skipnextlf true
|
|
** means we saw a \r before.
|
|
*/
|
|
newlinetypes |= NEWLINE_CRLF;
|
|
c = GETC(stream);
|
|
if (c == EOF) break;
|
|
} else {
|
|
/*
|
|
** Note that c == EOF also brings us here,
|
|
** so we're okay if the last char in the file
|
|
** is a CR.
|
|
*/
|
|
newlinetypes |= NEWLINE_CR;
|
|
}
|
|
}
|
|
if (c == '\r') {
|
|
/* A \r is translated into a \n, and we skip
|
|
** an adjacent \n, if any. We don't set the
|
|
** newlinetypes flag until we've seen the next char.
|
|
*/
|
|
skipnextlf = 1;
|
|
c = '\n';
|
|
} else if ( c == '\n') {
|
|
newlinetypes |= NEWLINE_LF;
|
|
}
|
|
*p++ = c;
|
|
if (c == '\n') break;
|
|
}
|
|
if ( c == EOF && skipnextlf )
|
|
newlinetypes |= NEWLINE_CR;
|
|
FUNLOCKFILE(stream);
|
|
*p = '\0';
|
|
if (fobj) {
|
|
((PyFileObject *)fobj)->f_newlinetypes = newlinetypes;
|
|
((PyFileObject *)fobj)->f_skipnextlf = skipnextlf;
|
|
} else if ( skipnextlf ) {
|
|
/* If we have no file object we cannot save the
|
|
** skipnextlf flag. We have to readahead, which
|
|
** will cause a pause if we're reading from an
|
|
** interactive stream, but that is very unlikely
|
|
** unless we're doing something silly like
|
|
** execfile("/dev/tty").
|
|
*/
|
|
c = GETC(stream);
|
|
if ( c != '\n' )
|
|
ungetc(c, stream);
|
|
}
|
|
if (p == buf)
|
|
return NULL;
|
|
return buf;
|
|
}
|
|
|
|
/*
|
|
** Py_UniversalNewlineFread is an fread variation that understands
|
|
** all of \r, \n and \r\n conventions.
|
|
** The stream should be opened in binary mode.
|
|
** fobj must be a PyFileObject. In this case there
|
|
** is no readahead but in stead a flag is used to skip a following
|
|
** \n on the next read. Also, if the file is open in binary mode
|
|
** the whole conversion is skipped. Finally, the routine keeps track of
|
|
** the different types of newlines seen.
|
|
*/
|
|
size_t
|
|
Py_UniversalNewlineFread(char *buf, size_t n,
|
|
FILE *stream, PyObject *fobj)
|
|
{
|
|
char *dst = buf;
|
|
PyFileObject *f = (PyFileObject *)fobj;
|
|
int newlinetypes, skipnextlf;
|
|
|
|
assert(buf != NULL);
|
|
assert(stream != NULL);
|
|
|
|
if (!fobj || !PyFile_Check(fobj)) {
|
|
errno = ENXIO; /* What can you do... */
|
|
return 0;
|
|
}
|
|
if (!f->f_univ_newline)
|
|
return fread(buf, 1, n, stream);
|
|
newlinetypes = f->f_newlinetypes;
|
|
skipnextlf = f->f_skipnextlf;
|
|
/* Invariant: n is the number of bytes remaining to be filled
|
|
* in the buffer.
|
|
*/
|
|
while (n) {
|
|
size_t nread;
|
|
int shortread;
|
|
char *src = dst;
|
|
|
|
nread = fread(dst, 1, n, stream);
|
|
assert(nread <= n);
|
|
if (nread == 0)
|
|
break;
|
|
|
|
n -= nread; /* assuming 1 byte out for each in; will adjust */
|
|
shortread = n != 0; /* true iff EOF or error */
|
|
while (nread--) {
|
|
char c = *src++;
|
|
if (c == '\r') {
|
|
/* Save as LF and set flag to skip next LF. */
|
|
*dst++ = '\n';
|
|
skipnextlf = 1;
|
|
}
|
|
else if (skipnextlf && c == '\n') {
|
|
/* Skip LF, and remember we saw CR LF. */
|
|
skipnextlf = 0;
|
|
newlinetypes |= NEWLINE_CRLF;
|
|
++n;
|
|
}
|
|
else {
|
|
/* Normal char to be stored in buffer. Also
|
|
* update the newlinetypes flag if either this
|
|
* is an LF or the previous char was a CR.
|
|
*/
|
|
if (c == '\n')
|
|
newlinetypes |= NEWLINE_LF;
|
|
else if (skipnextlf)
|
|
newlinetypes |= NEWLINE_CR;
|
|
*dst++ = c;
|
|
skipnextlf = 0;
|
|
}
|
|
}
|
|
if (shortread) {
|
|
/* If this is EOF, update type flags. */
|
|
if (skipnextlf && feof(stream))
|
|
newlinetypes |= NEWLINE_CR;
|
|
break;
|
|
}
|
|
}
|
|
f->f_newlinetypes = newlinetypes;
|
|
f->f_skipnextlf = skipnextlf;
|
|
return dst - buf;
|
|
}
|
|
|
|
#ifdef __cplusplus
|
|
}
|
|
#endif
|