PEP: 42 Title: Feature Requests Version: $Revision$ Author: Jeremy Hylton Status: Active Type: Informational Created: 12-Sep-2000 Introduction This PEP contains a list of feature requests that may be considered for future versions of Python. Large feature requests should not be included here, but should be described in separate PEPs; however a large feature request that doesn't have its own PEP can be listed here until its own PEP is created. See pep-0000.txt for details. This PEP was created to allow us to close bug reports that are really feature requests. Marked as Open, they distract from the list of real bugs (which should ideally be less than a page). Marked as Closed, they tend to be forgotten. The procedure now is: if a bug report is really a feature request, add the feature request to this PEP; mark the bug as "feature request", "later", and "closed"; and add a comment to the bug saying that this is the case (mentioning the PEP explicitly). It is also acceptable to move large feature requests directly from the bugs database to a separate PEP. Core Language / Builtins - A builtin function that returns the number of bytes an object uses internally. Apparently mxTools has a sizeof function that returns the size of the object struct itself. http://sourceforge.net/bugs/?func=detailbug&bug_id=110835&group_id=5470 - Add C API functions to help Windows users who are building embedded applications where the FILE * structure does not match the FILE * the interpreter was compiled with. http://sourceforge.net/bugs/?func=detailbug&bug_id=110821&group_id=5470 See this bug report for a specific suggestion that will allow a Borland C++ builder application to interact with a python.dll build with MSVC. - The parser should handle more deeply nested parse trees. The following will fail -- eval("["*50 + "]"*50) -- because the parser has a hard-coded limit on stack size. This limit should be raised or removed. Removal would be hard because the current compiler can overflow the C stack if the nesting is too deep. http://sourceforge.net/bugs/?func=detailbug&bug_id=115555&group_id=5470 - The buffer interface could be smarter when a buffer object is created that depends on another buffer object -- if the original buffer object has no base, the depending object will have no base either. It could be argued that the depending object should have the original object as a base. Or not. http://sourceforge.net/bugs/?func=detailbug&bug_id=116405&group_id=5470 Standard Library - The test suite is incomplete (and probably always will be). This is a reminder to people that more regression tests are needed. http://sourceforge.net/bugs/?func=detailbug&group_id=5470&bug_id=110819 - The urllib module should support proxies which require authenication. See SourceForge bug #110619 for information: http://sourceforge.net/bugs/?func=detailbug&bug_id=110619&group_id=5470 - Use objects with attributes in place of tuples of values for return values in several places in the library. Two specific APIs which could use this treatment include os.stat() and os.statvfs(); see SourceForge bug #111481: http://sourceforge.net/bugs/?func=detailbug&bug_id=111481&group_id=5470 which shows very specifically why this approach is better than the tuple approach: additional attributes can be added as needed without having as detrimental an affect on existing code. - os.rename() should be modified to handle EXDEV errors on platforms that don't allow rename() to operate across filesystem boundaries by copying the file over and removing the original. Linux is one system that requires this treatment. http://sourceforge.net/bugs/?func=detailbug&bug_id=112317&group_id=5470 - signal handling doesn't always work as expected. E.g. if sys.stdin.readline() is interrupted by a (returning) signal handler, it returns "". It would be better to make it raise an exception (corresponding to EINTR) or to restart. But these changes would have to applied to all places that can do blocking interruptable I/O. So it's a big project. http://sourceforge.net/bugs/?func=detailbug&bug_id=110599&group_id=5470 - Ensure that all .py files in the std library use 4-space indents and no hard tabs. This was actually a PEP200 precondition for the release of 2.0b1, but got misinterpreted as the weaker condition that tabnanny not complain. Tim Peters will do this now, but, since about 250 files are affected, will wait until after 2.0final is released. http://sourceforge.net/bugs/?func=detailbug&bug_id=114557&group_id=5470 - Port the Python SSL code to Windows. http://sourceforge.net/bugs/?func=detailbug&bug_id=110683&group_id=5470 - Extend Windows utime to accept directory paths. http://sourceforge.net/bugs/?func=detailbug&bug_id=114245&group_id=5470 - Extend copy.py to class, module & function types. http://sourceforge.net/bugs/?func=detailbug&bug_id=114553&group_id=5470 - Better checking for bad input to marshal.load*(). http://sourceforge.net/bugs/?func=detailbug&bug_id=114754&group_id=5470 - Generalize eval to accept any mapping objects for locals and globals. http://sourceforge.net/bugs/?func=detailbug&bug_id=115126&group_id=5470 - Add a portable implementation of time.strptime() that works in clearly defined ways on all platforms. http://sourceforge.net/bugs/?func=detailbug&bug_id=115146&group_id=5470 http://sourceforge.net/bugs/?func=detailbug&bug_id=112244&group_id=5470 - rfc822.py should be more lenient than the spec in the types of address fields it parses. Specifically, an invalid address of the form "From: Amazon.com " should be parsed correctly. http://sourceforge.net/bugs/?func=detailbug&bug_id=110678&group_id=5470 - cgi.py's FieldStorage class should be more conservative with memory in the face of large binary file uploads. http://sourceforge.net/bugs/?func=detailbug&bug_id=110674&group_id=5470 There are two issues here: first, because read_lines_to_outerboundary() uses readline() it is possible that a large amount of data will be read into memory for a binary file upload. This should probably look at the Content-Type header of the section and do a chunked read if it's a binary type. The second issue was related to the self.lines attribute, which was removed in revision 1.56 of cgi.py (see also): http://sourceforge.net/bugs/?func=detailbug&bug_id=119806&group_id=5470 - urllib should support proxy definitions that contain just the host and port http://sourceforge.net/bugs/?func=detailbug&bug_id=110849&group_id=5470 - urlparse should be updated to comply with RFC 2396, which defines optional parameters for each segment of the page. http://sourceforge.net/bugs/?func=detailbug&bug_id=110834&group_id=5470 - The exceptions raised by pickle and cPickle are currently different; these should be unified (probably the exceptions should be defined in a helper module that's imported by both). [No bug report; I just thought of this.] - More standard library routines should support Unicode. For example, urllib.quote() could convert Unicode strings to UTF-8 and then do the usual %HH conversion. But this is not the only one! http://sourceforge.net/bugs/?func=detailbug&bug_id=116716&group_id=5470 - There should be a way to say that you don't mind if str() or __str__() return a Unicode string object. Or a different function -- ustr() has been proposed. Or something... http://sourceforge.net/patch/?func=detailpatch&patch_id=101527&group_id=5470 - A UTF-7 codec. It's not a great encoding, but it's required in some places (e.g. IMAP mailbox names with Asian language support). [No bug report; private email from Seo Junwon .] - Killing a thread from another thread. Or maybe sending a signal. Or maybe raising an asynchronous exception. http://sourceforge.net/bugs/?func=detailbug&bug_id=121115&group_id=5470 - The debugger (pdb) should understand packages. http://sourceforge.net/bugs/?func=detailbug&bug_id=110631&group_id=5470 - The cmath library should be rewritten in Python. http://sourceforge.net/bugs/?func=detailbug&bug_id=110838&group_id=5470 - Could use a more comprehensive email module. (But then again, that may be better done as a 3rd party project.) http://sourceforge.net/bugs/?func=detailbug&bug_id=121208&group_id=5470 - every built-in function or method (including all core extensions) that accepts a string, dict, or list, should also accept a UserString, UserDict, or UserList. (The latter two should more generally accept all mappings, all sequences.) http://sourceforge.net/bugs/?func=detailbug&bug_id=132493&group_id=5470 Tools - IDLE should reload & recompile modules changed externally. To be done properly, scripts will have to be run in a separate process. http://sourceforge.net/bugs/?func=detailbug&bug_id=110841&group_id=5470 - Python could use a GUI builder. http://sourceforge.net/bugs/?func=detailbug&bug_id=110820&group_id=5470 - IDLE's key bindings should be revised. Some of the advertised bindings don't even work! http://sourceforge.net/bugs/?func=detailbug&bug_id=110659&group_id=5470 - IDLE has deep problems running threaded programs. Re-architect. http://sourceforge.net/bugs/?func=detailbug&bug_id=121963&group_id=5470 Building and Installing - You should be able to configure and build Python with a cross-compiler. http://sourceforge.net/bugs/?func=detailbug&bug_id=110836&group_id=5470 - Modules/makesetup should make sure the 'config.c' file it generates from the various Setup files, is valid C. It currently accepts module names with characters that are not allowable in Python or C identifiers. http://sourceforge.net/bugs/?func=detailbug&bug_id=116326&group_id=5470 - We should rename and rationalize the API getopt.c provides, and always use this, instead of trying to rely on the platform getopt and fighting with its prototype. http://www.python.org/pipermail/python-dev/2000-October/016620.html - There's a name conflict on case insensitive filesystems (in particular Mac OSX) between the directory "Python" and the key build target "python". That's currently solved by abusing the --with-suffix option, but that's not ideal (since that also causes the binary to be *installed* as python.exe). What should be the solution? http://sourceforge.net/bugs/?func=detailbug&bug_id=122215&group_id=5470 - Building from source should not attempt to overwrite the Include/graminit.h and Parser/graminit.c files, at least for people downloading a source release rather than working from CVS or snapshots. Some people find this a problem in unusual build environments. http://sourceforge.net/bugs/?func=detailbug&bug_id=119221&group_id=5470 Local Variables: mode: indented-text indent-tabs-mode: nil End: