Commit Graph

2156 Commits

Author SHA1 Message Date
Guido van Rossum 09c8b6c3e4 OpenSSL support. This is based on patches for a version of SSLeay by
Brian E Gallew, which were improved and adapted to OpenSSL 0.9.4 by
Laszlo Kovacs of HP.  Both have kindly given permission to include
the patches in the Python distribution.  Final formatting by GvR.
1999-12-07 21:37:17 +00:00
Guido van Rossum 5274c336f5 According to Craig H Rowland, openbsd2 is yet another BSD variant that
uses the BSD version of the lock structure.  Sigh, @!%$.
1999-12-06 14:51:05 +00:00
Greg Ward 631e6a0c07 [from 1999-11-04]
Bunch of little bug fixes that appeared in building non-packagized
distributions.  Mainly:
  - brain-slip typo in 'get_package_dir()'
  - don't try to os.path.join() an empty path tuple -- it doesn't like it
  - more type-safety in 'build_module()'
1999-12-03 16:18:56 +00:00
Guido van Rossum 6dfc792fea In abspath(), always use normpath(), even when win32api is available
(and even when it fails).  This avoids the problem where a trailing
separator is not removed when win32api.GetFullPathName() is used.
1999-11-30 15:00:00 +00:00
Barry Warsaw d25c1b73d2 A bunch of docstring fixes. 1999-11-28 17:11:06 +00:00
Greg Stein 32efef33ae look for builtins before stuff on the path. 1999-11-24 02:38:37 +00:00
Greg Stein 2b23413ce2 add loading of dynamic library modules. 1999-11-24 02:37:05 +00:00
Greg Stein 7ec28d298d turn SysPathImporter into PathImporter. 1999-11-20 12:31:07 +00:00
Greg Stein 72ee43527d remove the __version__ global. 1999-11-20 11:39:56 +00:00
Greg Stein 6d3165a5a1 only put __path__ into package modules. 1999-11-20 11:39:00 +00:00
Greg Stein 63faa01538 shift code from DirectoryImporter out to a common area.
remove use of "os" module (bootstrap issues) and go to the underlying
  platform-specific modules
fix problem in _compile() (trapped wrong error on permission issues)
add SysPathImporter and BuiltinImporter
put __file__ into modules imported from the filesystem. [backwards compat]
put __path__ into modules [backwards compat]
  oops: it is doing this for all modules, not just packages.
comment and tweak to the PackageArchiveImporter
1999-11-20 11:22:37 +00:00
Guido van Rossum d7b147b040 Moshe Zadka writes: When deploying SimpleHTTPServer, I noticed a
problem: it does not encode/decode the urls, which is wrong.
1999-11-16 19:04:32 +00:00
Guido van Rossum 77980a731c Correct typo in module doc string doscovered by Jonathan Giddy. 1999-11-15 14:19:15 +00:00
Guido van Rossum 768d2271a8 Oops. Remove some garbage from the doc string that was accidentally
checked in due to a patching mishap.  Reported by Detlef Lannert;
thanks!
1999-11-09 16:36:45 +00:00
Greg Stein d4c64ba0f9 Add some header comments to all the files. 1999-11-07 13:14:58 +00:00
Greg Stein 281b8d82f7 initial checkin for my Python stuff. 1999-11-07 12:54:45 +00:00
Fred Drake e4f13660f8 split() docstring: Made signature and description for the first
parameter match.  Error pointed out by François
                    Pinard <pinard@iro.umontreal.ca> on c.l.py.
1999-11-04 19:19:48 +00:00
Guido van Rossum 19878f58fe Sjoerd Mullender writes:
I regularly find that pdb sets the breakpoint on the wrong line when I
try to set a breakpoint on a function.  This fixes the problem
somewhat.
The real problem is that pdb tries to parse the Python source code to
find the first executable line.  A better way might be to inspect the
code object, or even have a variable in the code object
co_firstexecutablelineno, but that's too much work.

The patch fixes the problem when the first code line after the def
statement contains the start *and* end of a triple-quoted string.  The
code assumed that the end of a triple-quoted string is not on the same
line as the start, and so it would skip to the end of the *next*
triple-quoted string.
1999-11-03 13:10:07 +00:00
Guido van Rossum dd7cbbf4d3 Oops. spawnl() and spawnle() should be implemented on Windows too.
Also added a comment that the 'p' variants (spawnvp() etc.) are *not*
supported on Windows.  (They could be by adding them to posixmodule.c)
1999-11-02 20:44:07 +00:00
Guido van Rossum f618a48d11 Correct typo in walk.__doc__ reported by Francois Pinard. 1999-11-02 13:29:08 +00:00
Guido van Rossum 5a2ca9328d Checking in a bunch of spawn functions. These are only defined if we
have fork and execv (and friends) but not spawnv.  They operate
exactly like the spawn functions on Windows.  A limited set of needed
constants is also defined (P_WAIT, P_NOWAIT etc.).

Also add getenv() as a familiar alias for environ.get().
1999-11-02 13:27:32 +00:00
Guido van Rossum 2d72687006 New module by Moshe Zadka (submitted on Sept. 25). This unifies the
functionality of cmp.py and cmpcache.py, which are hereby declared
obsolescent.
1999-10-26 14:02:01 +00:00
Guido van Rossum 4e20de59dd Patch by Michael Hudson: when the object of attribute expansion is a
class instance, include the class attributes in the list of possible
expansions.
1999-10-26 13:09:08 +00:00
Guido van Rossum d65b53923e Fix by Moshe Zadka (cleaned up and documented by GvR) to break out the
request handling into separate parse_request() and handle_request()
methods.
1999-10-26 13:01:36 +00:00
Greg Ward e1ada50559 Don't assume GNU tar -- generate tar file and compress in separate steps.
Now supports the full range of intended formats (tar, ztar, gztar, zip).
"-f" no longer a short option for "--formats" -- conflicts with new
  global option "--force"!
1999-10-23 19:25:05 +00:00
Greg Ward d6c30f66c7 Removed massive comment speculating about needlessly complex variations
on the manifest file syntax.
1999-10-23 19:10:59 +00:00
Greg Ward 346e320c6e Qualified use of 'newer_group' function. 1999-10-23 19:06:56 +00:00
Greg Ward 6bad86462b Fix how we run 'zip' -- give explicit .zip extension. 1999-10-23 19:06:20 +00:00
Guido van Rossum bda10c81d0 In helo() and ehlo(), Don't fail when gethostbyaddr() fails -- just
keep whatever gethostname() returns.  After a suggestion by Doug Wyatt.
1999-10-22 13:09:20 +00:00
Guido van Rossum 8fa42af978 Fix PR#107: wm_colormapwindows() did the wrong thing when presented
more than one window argument.
1999-10-20 12:29:56 +00:00
Guido van Rossum 910d9a0634 Test output.
(XXX perhaps a bit too verbose; in particular it is sensitive to
all the doc strings.)
1999-10-19 19:09:00 +00:00
Guido van Rossum fdecda0123 Rewritten -- this now tests the binascii *except* for the binhex
module, which is tested by test_binhex.py.
1999-10-19 19:08:13 +00:00
Guido van Rossum a0e85b241d Test output for test_binhex.py. 1999-10-19 19:07:33 +00:00
Guido van Rossum 7b8f1abfca This test really only tests the binhex module.
Renamed it and adapted a comment and an error message.
1999-10-19 17:48:54 +00:00
Jeremy Hylton 88d23309e9 print a warning if the password will be echoed.
At import time, getpass will be bound to the appropriate
platform-specific function.  If the platform's echo-disabler is not
available, default_getpass, which prints the warning, will be used
1999-10-18 22:25:22 +00:00
Guido van Rossum f8d8e07601 Fixed PR#106: winfo_visualsavailable() with the includeids=1 option
didn't properly handle the hex numbers returned.
1999-10-18 22:06:38 +00:00
Guido van Rossum dcb8583c18 Fix for PR#111: when using the inplace option, give the new file the
same permissions as the old file, plugging a security hole.
(Not using exactly the suggested bugfix.)
1999-10-18 21:41:43 +00:00
Fred Drake 14bb71d553 os.fork raises AttributeError, not NameError, if fork() isn't
supported.  Pointed out by Moshe Zadka <moshez@math.huji.ac.il>.
1999-10-18 13:43:44 +00:00
Fred Drake 40e84db0f4 Based on comments from Paul Prescod:
If os.fork() doesn't exist, raise SystemError with an explanation at
the top of the module.  Added a note to the module docstring.
1999-10-16 02:07:50 +00:00
Barry Warsaw 226ae6ca12 Mainlining the string_methods branch. See branch revision log
messages for specific changes.
1999-10-12 19:54:53 +00:00
Jeremy Hylton 75260275fe update to use threading module instead of thread. 1999-10-12 16:20:13 +00:00
Barry Warsaw 2539451fb9 fixed a typo in a docstring, and slightly expanded the module
docstring info for readfp().
1999-10-12 16:12:48 +00:00
Guido van Rossum 1b7aec35c4 Fix PR#31 -- zfill() mishandles empty string. 1999-10-11 22:15:41 +00:00
Guido van Rossum 08a92cb568 Jeremy writes:
I found the following patch helpful in tracking down a bug in some
code.  I had appended time, the module, instead of time.time().  Not
sure if it is generally true that printing the repr of the object is
good, but I expect that most unpicklable things will have fairly
information and concise reprs (like files or sockets or modules).
1999-10-10 21:14:25 +00:00
Guido van Rossum d8957d6802 Fix PR#3, submitted by Skip Montanaro: if no space appears after the
colon, the first character of the value is lost.
1999-10-06 15:19:19 +00:00
Guido van Rossum a5a24b76f4 Added has_option(); fix bug in get() which botched interpolation if
'%(' was found in first position (found by Fred Drake).
1999-10-04 19:58:22 +00:00
Guido van Rossum 6a8d84b0c1 Urmpfh!
Withdraw the change that Fred just checked in -- it was a poorly
documented feature, not a bug, to ignore I/O errors in read().

The new docstring explains the reason for the feature:
"""
this is designed so that you can specifiy a list of potential
configuration file locations (e.g. current directory, user's home
directory, systemwide directory), and all existing configuration files
in the list will be read.
"""

Also add a lower-level function, readfp(), which takes an open file
object (and optionally a filename).

XXX There are some other problems with this module, but I don't have
time to dig into these; in particular, there are complaints that the
%(name)s substitution from the [DEFAULTS] section doesn't work
correctly.
1999-10-04 18:57:27 +00:00
Fred Drake 2438a485c8 ConfigParser.read(): Don't mask IOError exceptions. 1999-10-04 18:11:56 +00:00
Greg Ward ef93095adb Filter 'glob()' results so we only look at regular files. 1999-10-03 21:09:14 +00:00
Greg Ward 3c6204a87f Pass 'force' flag to 'new_compiler()'. 1999-10-03 21:08:42 +00:00