Google Git
Sign in
chromium / external / github.com / python / cpython / refs/tags/v2.7.4 / . / README
blob: 9b3f16eea496311b3b85a76f6177ca47f9c10c57 [file] [log] [blame]
Benjamin Petersona1f6fab2013-03-23 15:17:29[diff] [blame]1This is Python version 2.7.4
Benjamin Peterson3e0fa422010-11-27 14:40:29[diff] [blame]2============================
Guido van Rossum91447632000-04-11 17:11:09[diff] [blame]3
Benjamin Peterson5f8d6062012-01-01 04:42:26[diff] [blame]4Copyright (c) 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011,
Benjamin Petersonb25d6112013-01-01 03:37:21[diff] [blame]52012, 2013 Python Software Foundation. All rights reserved.
Guido van Rossum67345eb2001-01-18 14:51:12[diff] [blame]6
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]7Copyright (c) 2000 BeOpen.com.
8All rights reserved.
9
Guido van Rossum23f7aed2001-04-12 20:53:31[diff] [blame]10Copyright (c) 1995-2001 Corporation for National Research Initiatives.
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]11All rights reserved.
12
Guido van Rossum6ebd2992000-09-03 04:47:47[diff] [blame]13Copyright (c) 1991-1995 Stichting Mathematisch Centrum.
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]14All rights reserved.
15
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]16
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]17License information
18-------------------
19
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]20See the file "LICENSE" for information on the history of this
21software, terms & conditions for usage, and a DISCLAIMER OF ALL
22WARRANTIES.
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]23
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]24This Python distribution contains no GNU General Public Licensed
25(GPLed) code so it may be used in proprietary projects just like prior
26Python distributions. There are interfaces to some GNU code but these
27are entirely optional.
28
29All trademarks referenced herein are property of their respective
30holders.
Guido van Rossum79808261997-12-11 18:01:47[diff] [blame]31
32
Guido van Rossum8d7d4ed1996-07-30 21:41:07[diff] [blame]33What's new in this release?
34---------------------------
Guido van Rossum3ff96dd1996-07-30 18:05:04[diff] [blame]35
Guido van Rossum67345eb2001-01-18 14:51:12[diff] [blame]36See the file "Misc/NEWS".
Guido van Rossum64773801997-11-26 17:07:02[diff] [blame]37
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]38
Guido van Rossum64773801997-11-26 17:07:02[diff] [blame]39If you don't read instructions
40------------------------------
41
42Congratulations on getting this far. :-)
43
44To start building right away (on UNIX): type "./configure" in the
Guido van Rossum2dcec0c2002-02-04 01:59:23[diff] [blame]45current directory and when it finishes, type "make". This creates an
46executable "./python"; to install in /usr/local, first do "su root"
47and then "make install".
48
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]49The section `Build instructions' below is still recommended reading.
Guido van Rossum3ff96dd1996-07-30 18:05:04[diff] [blame]50
51
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]52What is Python anyway?
53----------------------
54
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]55Python is an interpreted, interactive object-oriented programming
56language suitable (amongst other uses) for distributed application
57development, scripting, numeric computing and system testing. Python
58is often compared to Tcl, Perl, Java, JavaScript, Visual Basic or
59Scheme. To find out more about what Python can do for you, point your
60browser to http://www.python.org/.
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]61
62
Guido van Rossumf501b4e1996-10-25 14:32:48[diff] [blame]63How do I learn Python?
64----------------------
65
Fred Drake0e6444c1999-05-17 19:35:01[diff] [blame]66The official tutorial is still a good place to start; see
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]67http://docs.python.org/ for online and downloadable versions, as well
68as a list of other introductions, and reference documentation.
Guido van Rossumf501b4e1996-10-25 14:32:48[diff] [blame]69
Guido van Rossumd0a42e22000-03-31 20:16:45[diff] [blame]70There's a quickly growing set of books on Python. See
Michael W. Hudsonfad46402005-06-16 15:51:20[diff] [blame]71http://wiki.python.org/moin/PythonBooks for a list.
Guido van Rossum8d90f9d1997-05-22 20:13:25[diff] [blame]72
Guido van Rossumf501b4e1996-10-25 14:32:48[diff] [blame]73
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]74Documentation
75-------------
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]76
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]77All documentation is provided online in a variety of formats. In
78order of importance for new users: Tutorial, Library Reference,
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]79Language Reference, Extending & Embedding, and the Python/C API. The
80Library Reference is especially of immense value since much of
81Python's power is described there, including the built-in data types
82and functions!
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]83
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]84All documentation is also available online at the Python web site
Benjamin Peterson9d0b6042008-03-30 19:35:10[diff] [blame]85(http://docs.python.org/, see below). It is available online for occasional
86reference, or can be downloaded in many formats for faster access. The
Georg Brandl1c88e0f2008-03-30 19:41:39[diff] [blame]87documentation is downloadable in HTML, PostScript, PDF, LaTeX, and
Benjamin Peterson9d0b6042008-03-30 19:35:10[diff] [blame]88reStructuredText (2.6+) formats; the LaTeX and reStructuredText versions are
89primarily for documentation authors, translators, and people with special
90formatting requirements.
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]91
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]92
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]93Web sites
94---------
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]95
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]96New Python releases and related technologies are published at
Guido van Rossum67345eb2001-01-18 14:51:12[diff] [blame]97http://www.python.org/. Come visit us!
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]98
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]99
100Newsgroups and Mailing Lists
101----------------------------
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]102
103Read comp.lang.python, a high-volume discussion newsgroup about
104Python, or comp.lang.python.announce, a low-volume moderated newsgroup
105for Python-related announcements. These are also accessible as
Georg Brandl4492a982009-08-24 17:12:30[diff] [blame]106mailing lists: see http://www.python.org/community/lists/ for an
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]107overview of these and many other Python-related mailing lists.
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]108
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]109Archives are accessible via the Google Groups Usenet archive; see
Michael W. Hudsonbe6e0f42002-06-23 16:29:36[diff] [blame]110http://groups.google.com/. The mailing lists are also archived, see
Georg Brandl4492a982009-08-24 17:12:30[diff] [blame]111http://www.python.org/community/lists/ for details.
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]112
113
114Bug reports
115-----------
116
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]117To report or search for bugs, please use the Python Bug
Georg Brandl4492a982009-08-24 17:12:30[diff] [blame]118Tracker at http://bugs.python.org/.
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]119
120
121Patches and contributions
122-------------------------
123
Fred Drake6caae142000-10-25 17:51:02[diff] [blame]124To submit a patch or other contribution, please use the Python Patch
Georg Brandl4492a982009-08-24 17:12:30[diff] [blame]125Manager at http://bugs.python.org/. Guidelines
Andrew M. Kuchling951300e2008-01-04 14:47:17[diff] [blame]126for patch submission may be found at http://www.python.org/dev/patches/.
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]127
Benjamin Petersonb2355002008-04-02 21:20:35[diff] [blame]128If you have a proposal to change Python, you may want to send an email to the
129comp.lang.python or python-ideas mailing lists for inital feedback. A Python
130Enhancement Proposal (PEP) may be submitted if your idea gains ground. All
131current PEPs, as well as guidelines for submitting a new PEP, are listed at
132http://www.python.org/dev/peps/.
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]133
134
135Questions
136---------
137
138For help, if you can't find it in the manuals or on the web site, it's
139best to post to the comp.lang.python or the Python mailing list (see
140above). If you specifically don't want to involve the newsgroup or
Guido van Rossum67345eb2001-01-18 14:51:12[diff] [blame]141mailing list, send questions to help@python.org (a group of volunteers
142who answer questions as they can). The newsgroup is the most
143efficient way to ask public questions.
Guido van Rossum627b2d71993-12-24 10:39:16[diff] [blame]144
Guido van Rossum901454e2000-06-29 22:28:44[diff] [blame]145
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]146Build instructions
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]147==================
Guido van Rossum627b2d71993-12-24 10:39:16[diff] [blame]148
Guido van Rossum59b0b602002-02-08 22:22:35[diff] [blame]149Before you can build Python, you must first configure it.
150Fortunately, the configuration and build process has been automated
151for Unix and Linux installations, so all you usually have to do is
152type a few commands and sit back. There are some platforms where
153things are not quite as smooth; see the platform specific notes below.
154If you want to build for multiple platforms sharing the same source
155tree, see the section on VPATH below.
Guido van Rossum627b2d71993-12-24 10:39:16[diff] [blame]156
Guido van Rossum59b0b602002-02-08 22:22:35[diff] [blame]157Start by running the script "./configure", which determines your
158system configuration and creates the Makefile. (It takes a minute or
159two -- please be patient!) You may want to pass options to the
160configure script -- see the section below on configuration options and
161variables. When it's done, you are ready to run make.
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]162
Guido van Rossum59b0b602002-02-08 22:22:35[diff] [blame]163To build Python, you normally type "make" in the toplevel directory.
164If you have changed the configuration, the Makefile may have to be
Benjamin Peterson7bee77c2009-02-26 19:05:59[diff] [blame]165rebuilt. In this case, you may have to run make again to correctly
Guido van Rossum59b0b602002-02-08 22:22:35[diff] [blame]166build your desired target. The interpreter executable is built in the
167top level directory.
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]168
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]169Once you have built a Python interpreter, see the subsections below on
Guido van Rossum59b0b602002-02-08 22:22:35[diff] [blame]170testing and installation. If you run into trouble, see the next
171section.
172
173Previous versions of Python used a manual configuration process that
174involved editing the file Modules/Setup. While this file still exists
175and manual configuration is still supported, it is rarely needed any
176more: almost all modules are automatically built as appropriate under
177guidance of the setup.py script, which is run by Make after the
178interpreter has been built.
Guido van Rossum627b2d71993-12-24 10:39:16[diff] [blame]179
Guido van Rossum0a516c91994-09-12 10:58:40[diff] [blame]180
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]181Troubleshooting
182---------------
Guido van Rossum627b2d71993-12-24 10:39:16[diff] [blame]183
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]184See also the platform specific notes in the next section.
185
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]186If you run into other trouble, see the FAQ
Georg Brandl4492a982009-08-24 17:12:30[diff] [blame]187(http://www.python.org/doc/faq/) for hints on what can go wrong, and
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]188how to fix it.
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]189
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]190If you rerun the configure script with different options, remove all
191object files by running "make clean" before rebuilding. Believe it or
192not, "make clean" sometimes helps to clean up other inexplicable
193problems as well. Try it before sending in a bug report!
194
Guido van Rossumd0fe8451996-08-26 03:02:37[diff] [blame]195If the configure script fails or doesn't seem to find things that
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]196should be there, inspect the config.log file.
Guido van Rossumc0be2f51997-10-08 05:05:28[diff] [blame]197
198If you get a warning for every file about the -Olimit option being no
199longer supported, you can ignore it. There's no foolproof way to know
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]200whether this option is needed; all we can do is test whether it is
Guido van Rossumc0be2f51997-10-08 05:05:28[diff] [blame]201accepted without error. On some systems, e.g. older SGI compilers, it
202is essential for performance (specifically when compiling ceval.c,
203which has more basic blocks than the default limit of 1000). If the
204warning bothers you, edit the Makefile to remove "-Olimit 1500" from
205the OPT variable.
Guido van Rossumd0fe8451996-08-26 03:02:37[diff] [blame]206
Guido van Rossum7ac11532000-08-29 21:36:40[diff] [blame]207If you get failures in test_long, or sys.maxint gets set to -1, you
208are probably experiencing compiler bugs, usually related to
Martin v. Löwis9cea5f52002-11-15 07:28:05[diff] [blame]209optimization. This is a common problem with some versions of gcc, and
210some vendor-supplied compilers, which can sometimes be worked around
211by turning off optimization. Consider switching to stable versions
Anthony Baxtere9d719b2004-11-30 01:49:18[diff] [blame]212(gcc 2.95.2, gcc 3.x, or contact your vendor.)
Guido van Rossum7ac11532000-08-29 21:36:40[diff] [blame]213
214From Python 2.0 onward, all Python C code is ANSI C. Compiling using
215old K&R-C-only compilers is no longer possible. ANSI C compilers are
216available for all modern systems, either in the form of updated
Martin v. Löwis9cea5f52002-11-15 07:28:05[diff] [blame]217compilers from the vendor, or one of the free compilers (gcc).
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]218
Georg Brandl250755b2006-11-15 17:42:03[diff] [blame]219If "make install" fails mysteriously during the "compiling the library"
220step, make sure that you don't have any of the PYTHONPATH or PYTHONHOME
221environment variables set, as they may interfere with the newly built
222executable which is compiling the library.
223
Martin v. Löwis779ffc02002-12-02 22:17:01[diff] [blame]224Unsupported systems
225-------------------
226
Andrew M. Kuchlingbffd62e2010-06-12 20:00:55[diff] [blame]227A number of systems are not supported in Python 2.7 anymore. Some
Georg Brandl3e83ff72010-06-12 06:45:33[diff] [blame]228support code is still present, but will be removed in later versions.
Martin v. Löwis779ffc02002-12-02 22:17:01[diff] [blame]229If you still need to use current Python versions on these systems,
230please send a message to python-dev@python.org indicating that you
Martin v. Löwis8cab8b02006-07-24 05:05:22[diff] [blame]231volunteer to support this system. For a more detailed discussion
232regarding no-longer-supported and resupporting platforms, as well
233as a list of platforms that became or will be unsupported, see PEP 11.
Martin v. Löwis779ffc02002-12-02 22:17:01[diff] [blame]234
235More specifically, the following systems are not supported any
236longer:
237- SunOS 4
238- DYNIX
239- dgux
240- Minix
Martin v. Löwis8cab8b02006-07-24 05:05:22[diff] [blame]241- NeXT
Martin v. Löwis779ffc02002-12-02 22:17:01[diff] [blame]242- Irix 4 and --with-sgi-dl
243- Linux 1
Matthias Klose3cef2a92012-03-14 22:39:33[diff] [blame]244- Systems defining __d6_pthread_create (configure.ac)
Martin v. Löwis779ffc02002-12-02 22:17:01[diff] [blame]245- Systems defining PY_PTHREAD_D4, PY_PTHREAD_D6,
246 or PY_PTHREAD_D7 in thread_pthread.h
247- Systems using --with-dl-dld
Martin v. Löwis7e4cfcb2002-12-19 16:21:49[diff] [blame]248- Systems using --without-universal-newlines
Martin v. Löwis8cab8b02006-07-24 05:05:22[diff] [blame]249- MacOS 9
Martin v. Löwis8cab8b02006-07-24 05:05:22[diff] [blame]250- Systems using --with-wctype-functions
251- Win9x, WinME
252
Martin v. Löwis779ffc02002-12-02 22:17:01[diff] [blame]253
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]254Platform specific notes
255-----------------------
256
Guido van Rossum0447a321995-10-08 01:22:33[diff] [blame]257(Some of these may no longer apply. If you find you can build Python
Guido van Rossum67345eb2001-01-18 14:51:12[diff] [blame]258on these platforms without the special directions mentioned here,
259submit a documentation bug report to SourceForge (see Bug Reports
260above) so we can remove them!)
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]261
Skip Montanaro6dc4a8e2003-05-02 15:27:59[diff] [blame]262Unix platforms: If your vendor still ships (and you still use) Berkeley DB
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]263 1.85 you will need to edit Modules/Setup to build the bsddb185
264 module and add a line to sitecustomize.py which makes it the
265 default. In Modules/Setup a line like
Skip Montanaro6dc4a8e2003-05-02 15:27:59[diff] [blame]266
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]267 bsddb185 bsddbmodule.c
Skip Montanaro6dc4a8e2003-05-02 15:27:59[diff] [blame]268
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]269 should work. (You may need to add -I, -L or -l flags to direct the
270 compiler and linker to your include files and libraries.)
Skip Montanaro6dc4a8e2003-05-02 15:27:59[diff] [blame]271
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]272XXX I think this next bit is out of date:
Skip Montanaro6dc4a8e2003-05-02 15:27:59[diff] [blame]273
Brett Cannondc48b742007-05-20 07:09:50[diff] [blame]27464-bit platforms: The modules audioop, and imageop don't work.
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]275 The setup.py script disables them on 64-bit installations.
276 Don't try to enable them in the Modules/Setup file. They
277 contain code that is quite wordsize sensitive. (If you have a
278 fix, let us know!)
Guido van Rossumc0be2f51997-10-08 05:05:28[diff] [blame]279
Guido van Rossum4462e931997-01-22 21:00:32[diff] [blame]280Solaris: When using Sun's C compiler with threads, at least on Solaris
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]281 2.5.1, you need to add the "-mt" compiler option (the simplest
282 way is probably to specify the compiler with this option as
283 the "CC" environment variable when running the configure
284 script).
Guido van Rossum4462e931997-01-22 21:00:32[diff] [blame]285
Guido van Rossum42992312002-09-23 14:02:18[diff] [blame]286 When using GCC on Solaris, beware of binutils 2.13 or GCC
287 versions built using it. This mistakenly enables the
288 -zcombreloc option which creates broken shared libraries on
289 Solaris. binutils 2.12 works, and the binutils maintainers
Skip Montanaro4de9cba2003-01-03 16:26:23[diff] [blame]290 are aware of the problem. Binutils 2.13.1 only partially
291 fixed things. It appears that 2.13.2 solves the problem
292 completely. This problem is known to occur with Solaris 2.7
293 and 2.8, but may also affect earlier and later versions of the
294 OS.
Guido van Rossum42992312002-09-23 14:02:18[diff] [blame]295
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]296 When the dynamic loader complains about errors finding shared
297 libraries, such as
Martin v. Löwis9cea5f52002-11-15 07:28:05[diff] [blame]298
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]299 ld.so.1: ./python: fatal: libstdc++.so.5: open failed:
300 No such file or directory
Martin v. Löwis9cea5f52002-11-15 07:28:05[diff] [blame]301
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]302 you need to first make sure that the library is available on
303 your system. Then, you need to instruct the dynamic loader how
304 to find it. You can choose any of the following strategies:
Martin v. Löwis9cea5f52002-11-15 07:28:05[diff] [blame]305
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]306 1. When compiling Python, set LD_RUN_PATH to the directories
307 containing missing libraries.
308 2. When running Python, set LD_LIBRARY_PATH to these directories.
309 3. Use crle(8) to extend the search path of the loader.
310 4. Modify the installed GCC specs file, adding -R options into the
311 *link: section.
Martin v. Löwis9cea5f52002-11-15 07:28:05[diff] [blame]312
Skip Montanaro9f6606a2005-11-17 18:19:39[diff] [blame]313 The complex object fails to compile on Solaris 10 with gcc 3.4 (at
314 least up to 3.4.3). To work around it, define Py_HUGE_VAL as
315 HUGE_VAL(), e.g.:
316
317 make CPPFLAGS='-D"Py_HUGE_VAL=HUGE_VAL()" -I. -I$(srcdir)/Include'
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]318 ./python setup.py CPPFLAGS='-D"Py_HUGE_VAL=HUGE_VAL()"'
Skip Montanaro9f6606a2005-11-17 18:19:39[diff] [blame]319
Guido van Rossum7ac11532000-08-29 21:36:40[diff] [blame]320Linux: A problem with threads and fork() was tracked down to a bug in
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]321 the pthreads code in glibc version 2.0.5; glibc version 2.0.7
322 solves the problem. This causes the popen2 test to fail;
323 problem and solution reported by Pablo Bleyer.
Guido van Rossum4462e931997-01-22 21:00:32[diff] [blame]324
Martin v. Löwis3efa04a2003-06-14 05:29:27[diff] [blame]325Red Hat Linux: Red Hat 9 built Python2.2 in UCS-4 mode and hacked
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]326 Tcl to support it. To compile Python2.3 with Tkinter, you will
327 need to pass --enable-unicode=ucs4 flag to ./configure.
Martin v. Löwis3efa04a2003-06-14 05:29:27[diff] [blame]328
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]329 There's an executable /usr/bin/python which is Python
330 1.5.2 on most older Red Hat installations; several key Red Hat tools
331 require this version. Python 2.1.x may be installed as
332 /usr/bin/python2. The Makefile installs Python as
333 /usr/local/bin/python, which may or may not take precedence
334 over /usr/bin/python, depending on how you have set up $PATH.
Guido van Rossum2dcec0c2002-02-04 01:59:23[diff] [blame]335
Guido van Rossum59230501999-06-21 20:51:46[diff] [blame]336FreeBSD 3.x and probably platforms with NCurses that use libmytinfo or
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]337 similar: When using cursesmodule, the linking is not done in
338 the correct order with the defaults. Remove "-ltermcap" from
339 the readline entry in Setup, and use as curses entry: "curses
340 cursesmodule.c -lmytinfo -lncurses -ltermcap" - "mytinfo" (so
341 called on FreeBSD) should be the name of the auxiliary library
342 required on your platform. Normally, it would be linked
343 automatically, but not necessarily in the correct order.
Guido van Rossum59230501999-06-21 20:51:46[diff] [blame]344
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]345BSDI: BSDI versions before 4.1 have known problems with threads,
346 which can cause strange errors in a number of modules (for
347 instance, the 'test_signal' test script will hang forever.)
348 Turning off threads (with --with-threads=no) or upgrading to
349 BSDI 4.1 solves this problem.
Guido van Rossum7ac11532000-08-29 21:36:40[diff] [blame]350
351DEC Unix: Run configure with --with-dec-threads, or with
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]352 --with-threads=no if no threads are desired (threads are on by
353 default). When using GCC, it is possible to get an internal
354 compiler error if optimization is used. This was reported for
355 GCC 2.7.2.3 on selectmodule.c. Manually compile the affected
356 file without optimization to solve the problem.
Guido van Rossum8eca2c21996-02-14 18:37:46[diff] [blame]357
Guido van Rossume1cdb161999-04-10 16:04:07[diff] [blame]358DEC Ultrix: compile with GCC to avoid bugs in the native compiler,
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]359 and pass SHELL=/bin/sh5 to Make when installing.
Guido van Rossume1cdb161999-04-10 16:04:07[diff] [blame]360
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]361AIX: A complete overhaul of the shared library support is now in
362 place. See Misc/AIX-NOTES for some notes on how it's done.
363 (The optimizer bug reported at this place in previous releases
364 has been worked around by a minimal code change.) If you get
365 errors about pthread_* functions, during compile or during
366 testing, try setting CC to a thread-safe (reentrant) compiler,
367 like "cc_r". For full C++ module support, set CC="xlC_r" (or
368 CC="xlC" without thread support).
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]369
Guido van Rossuma2534c82005-09-20 19:06:23[diff] [blame]370AIX 5.3: To build a 64-bit version with IBM's compiler, I used the
371 following:
372
373 export PATH=/usr/bin:/usr/vacpp/bin
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]374 ./configure --with-gcc="xlc_r -q64" --with-cxx="xlC_r -q64" \
Guido van Rossuma2534c82005-09-20 19:06:23[diff] [blame]375 --disable-ipv6 AR="ar -X64"
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]376 make
Guido van Rossuma2534c82005-09-20 19:06:23[diff] [blame]377
Anthony Baxter7c497472003-10-01 07:24:54[diff] [blame]378HP-UX: When using threading, you may have to add -D_REENTRANT to the
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]379 OPT variable in the top-level Makefile; reported by Pat Knight,
380 this seems to make a difference (at least for HP-UX 10.20)
381 even though pyconfig.h defines it. This seems unnecessary when
382 using HP/UX 11 and later - threading seems to work "out of the
383 box".
Anthony Baxter7c497472003-10-01 07:24:54[diff] [blame]384
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]385HP-UX ia64: When building on the ia64 (Itanium) platform using HP's
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]386 compiler, some experience has shown that the compiler's
387 optimiser produces a completely broken version of python
Georg Brandl4492a982009-08-24 17:12:30[diff] [blame]388 (see http://bugs.python.org/814976). To work around this,
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]389 edit the Makefile and remove -O from the OPT line.
Guido van Rossum539c6622005-09-14 17:49:54[diff] [blame]390
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]391 To build a 64-bit executable on an Itanium 2 system using HP's
392 compiler, use these environment variables:
Guido van Rossum539c6622005-09-14 17:49:54[diff] [blame]393
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]394 CC=cc
395 CXX=aCC
396 BASECFLAGS="+DD64"
397 LDFLAGS="+DD64 -lxnet"
Guido van Rossum539c6622005-09-14 17:49:54[diff] [blame]398
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]399 and call configure as:
Guido van Rossum539c6622005-09-14 17:49:54[diff] [blame]400
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]401 ./configure --without-gcc
Guido van Rossum539c6622005-09-14 17:49:54[diff] [blame]402
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]403 then *unset* the environment variables again before running
404 make. (At least one of these flags causes the build to fail
405 if it remains set.) You still have to edit the Makefile and
406 remove -O from the OPT line.
Guido van Rossum2094e041998-05-14 15:47:35[diff] [blame]407
Georg Brandl4492a982009-08-24 17:12:30[diff] [blame]408HP PA-RISC 2.0: A recent bug report (http://bugs.python.org/546117)
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]409 suggests that the C compiler in this 64-bit system has bugs
410 in the optimizer that break Python. Compiling without
411 optimization solves the problems.
Guido van Rossum4e1dd7d2002-04-23 13:06:07[diff] [blame]412
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]413SCO: The following apply to SCO 3 only; Python builds out of the box
414 on SCO 5 (or so we've heard).
Guido van Rossuma16e2751998-04-13 20:14:05[diff] [blame]415
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]416 1) Everything works much better if you add -U__STDC__ to the
417 defs. This is because all the SCO header files are broken.
418 Anything that isn't mentioned in the C standard is
419 conditionally excluded when __STDC__ is defined.
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]420
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]421 2) Due to the U.S. export restrictions, SCO broke the crypt
422 stuff out into a separate library, libcrypt_i.a so the LIBS
423 needed be set to:
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]424
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]425 LIBS=' -lsocket -lcrypt_i'
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]426
Martin v. Löwis387c5472001-09-06 08:16:17[diff] [blame]427UnixWare: There are known bugs in the math library of the system, as well as
428 problems in the handling of threads (calling fork in one
429 thread may interrupt system calls in others). Therefore, test_math and
430 tests involving threads will fail until those problems are fixed.
431
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]432QNX: Chris Herborth (chrish@qnx.com) writes:
433 configure works best if you use GNU bash; a port is available on
434 ftp.qnx.com in /usr/free. I used the following process to build,
435 test and install Python 1.5.x under QNX:
Guido van Rossumb7f454d1997-12-02 19:44:31[diff] [blame]436
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]437 1) CONFIG_SHELL=/usr/local/bin/bash CC=cc RANLIB=: \
438 ./configure --verbose --without-gcc --with-libm=""
Guido van Rossumb7f454d1997-12-02 19:44:31[diff] [blame]439
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]440 2) edit Modules/Setup to activate everything that makes sense for
441 your system... tested here at QNX with the following modules:
Guido van Rossumb7f454d1997-12-02 19:44:31[diff] [blame]442
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]443 array, audioop, binascii, cPickle, cStringIO, cmath,
444 crypt, curses, errno, fcntl, gdbm, grp, imageop,
445 _locale, math, md5, new, operator, parser, pcre,
Neal Norwitz90340a12007-05-20 18:06:27[diff] [blame]446 posix, pwd, readline, regex, reop,
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]447 select, signal, socket, soundex, strop, struct,
Brett Cannondc48b742007-05-20 07:09:50[diff] [blame]448 syslog, termios, time, timing, zlib, audioop, imageop
Guido van Rossumb7f454d1997-12-02 19:44:31[diff] [blame]449
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]450 3) make SHELL=/usr/local/bin/bash
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]451
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]452 or, if you feel the need for speed:
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]453
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]454 make SHELL=/usr/local/bin/bash OPT="-5 -Oil+nrt"
Guido van Rossumb7f454d1997-12-02 19:44:31[diff] [blame]455
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]456 4) make SHELL=/usr/local/bin/bash test
Guido van Rossumec95c7b1998-08-04 17:59:56[diff] [blame]457
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]458 Using GNU readline 2.2 seems to behave strangely, but I
459 think that's a problem with my readline 2.2 port. :-\
Guido van Rossumec95c7b1998-08-04 17:59:56[diff] [blame]460
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]461 5) make SHELL=/usr/local/bin/bash install
Guido van Rossumb7f454d1997-12-02 19:44:31[diff] [blame]462
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]463 If you get SIGSEGVs while running Python (I haven't yet, but
464 I've only run small programs and the test cases), you're
465 probably running out of stack; the default 32k could be a
466 little tight. To increase the stack size, edit the Makefile
467 to read: LDFLAGS = -N 48k
Guido van Rossum40d63581997-08-14 19:45:30[diff] [blame]468
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]469BeOS: See Misc/BeOS-NOTES for notes about compiling/installing
470 Python on BeOS R3 or later. Note that only the PowerPC
471 platform is supported for R3; both PowerPC and x86 are
472 supported for R4.
Guido van Rossumec95c7b1998-08-04 17:59:56[diff] [blame]473
Michael W. Hudson33876f52002-07-31 09:55:25[diff] [blame]474Cray T3E: Mark Hadfield (m.hadfield@niwa.co.nz) writes:
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]475 Python can be built satisfactorily on a Cray T3E but based on
476 my experience with the NIWA T3E (2002-05-22, version 2.2.1)
477 there are a few bugs and gotchas. For more information see a
478 thread on comp.lang.python in May 2002 entitled "Building
479 Python on Cray T3E".
Michael W. Hudson33876f52002-07-31 09:55:25[diff] [blame]480
481 1) Use Cray's cc and not gcc. The latter was reported not to
482 work by Konrad Hinsen. It may work now, but it may not.
483
484 2) To set sys.platform to something sensible, pass the
485 following environment variable to the configure script:
486
487 MACHDEP=unicosmk
488
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]489 2) Run configure with option "--enable-unicode=ucs4".
Michael W. Hudson33876f52002-07-31 09:55:25[diff] [blame]490
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]491 3) The Cray T3E does not support dynamic linking, so extension
492 modules have to be built by adding (or uncommenting) lines
493 in Modules/Setup. The minimum set of modules is
Michael W. Hudson33876f52002-07-31 09:55:25[diff] [blame]494
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]495 posix, new, _sre, unicodedata
Michael W. Hudson33876f52002-07-31 09:55:25[diff] [blame]496
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]497 On NIWA's vanilla T3E system the following have also been
498 included successfully:
Michael W. Hudson33876f52002-07-31 09:55:25[diff] [blame]499
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]500 _codecs, _locale, _socket, _symtable, _testcapi, _weakref
501 array, binascii, cmath, cPickle, crypt, cStringIO, dbm
502 errno, fcntl, grp, math, md5, operator, parser, pcre, pwd
503 regex, rotor, select, struct, strop, syslog, termios
504 time, timing, xreadlines
Michael W. Hudson33876f52002-07-31 09:55:25[diff] [blame]505
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]506 4) Once the python executable and library have been built, make
507 will execute setup.py, which will attempt to build remaining
508 extensions and link them dynamically. Each of these attempts
509 will fail but should not halt the make process. This is
510 normal.
Michael W. Hudson33876f52002-07-31 09:55:25[diff] [blame]511
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]512 5) Running "make test" uses a lot of resources and causes
513 problems on our system. You might want to try running tests
514 singly or in small groups.
Guido van Rossum1bf0bf41997-08-20 23:50:51[diff] [blame]515
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]516SGI: SGI's standard "make" utility (/bin/make or /usr/bin/make)
517 does not check whether a command actually changed the file it
518 is supposed to build. This means that whenever you say "make"
519 it will redo the link step. The remedy is to use SGI's much
520 smarter "smake" utility (/usr/sbin/smake), or GNU make. If
521 you set the first line of the Makefile to #!/usr/sbin/smake
522 smake will be invoked by make (likewise for GNU make).
Guido van Rossum0078aaf1997-08-21 03:05:11[diff] [blame]523
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]524 WARNING: There are bugs in the optimizer of some versions of
525 SGI's compilers that can cause bus errors or other strange
526 behavior, especially on numerical operations. To avoid this,
527 try building with "make OPT=".
Fred Drake0b5fb2b2000-09-29 17:45:05[diff] [blame]528
Guido van Rossum64773801997-11-26 17:07:02[diff] [blame]529OS/2: If you are running Warp3 or Warp4 and have IBM's VisualAge C/C++
530 compiler installed, just change into the pc\os2vacpp directory
531 and type NMAKE. Threading and sockets are supported by default
532 in the resulting binaries of PYTHON15.DLL and PYTHON.EXE.
533
Fred Drake36ed5602000-10-06 01:58:48[diff] [blame]534Reliant UNIX: The thread support does not compile on Reliant UNIX, and
535 there is a (minor) problem in the configure script for that
536 platform as well. This should be resolved in time for a
537 future release.
Trent Mick635f6fb2000-08-23 21:33:05[diff] [blame]538
Barry Warsaw6a8557d2002-10-14 18:04:39[diff] [blame]539MacOSX: The tests will crash on both 10.1 and 10.2 with SEGV in
Barry Warsaw6e753642002-10-14 18:15:35[diff] [blame]540 test_re and test_sre due to the small default stack size. If
541 you set the stack size to 2048 before doing a "make test" the
Georg Brandl667eb7c2007-03-06 18:47:31[diff] [blame]542 failure can be avoided. If you're using the tcsh or csh shells,
543 use "limit stacksize 2048" and for the bash shell (the default
544 as of OSX 10.3), use "ulimit -s 2048".
Barry Warsaw06557452001-12-14 20:47:12[diff] [blame]545
Jack Jansen6a77cf12001-12-06 21:49:02[diff] [blame]546 On naked Darwin you may want to add the configure option
Jack Jansen4c398fd2001-10-08 13:21:15[diff] [blame]547 "--disable-toolbox-glue" to disable the glue code for the Carbon
Jack Jansen6a77cf12001-12-06 21:49:02[diff] [blame]548 interface modules. The modules themselves are currently only built
549 if you add the --enable-framework option, see below.
Barry Warsaw06557452001-12-14 20:47:12[diff] [blame]550
551 On a clean OSX /usr/local does not exist. Do a
Jack Jansen61fec302002-01-04 15:59:57[diff] [blame]552 "sudo mkdir -m 775 /usr/local"
Jack Jansen84a01642002-08-04 21:19:55[diff] [blame]553 before you do a make install. It is probably not a good idea to
554 do "sudo make install" which installs everything as superuser,
555 as this may later cause problems when installing distutils-based
556 additions.
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]557
Jack Jansen84a01642002-08-04 21:19:55[diff] [blame]558 Some people have reported problems building Python after using "fink"
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]559 to install additional unix software. Disabling fink (remove all
560 references to /sw from your .profile or .login) should solve this.
Barry Warsaw06557452001-12-14 20:47:12[diff] [blame]561
562 You may want to try the configure option "--enable-framework"
563 which installs Python as a framework. The location can be set
564 as argument to the --enable-framework option (default
Jack Jansen84a01642002-08-04 21:19:55[diff] [blame]565 /Library/Frameworks). A framework install is probably needed if you
566 want to use any Aqua-based GUI toolkit (whether Tkinter, wxPython,
567 Carbon, Cocoa or anything else).
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]568
Georg Brandlb0885b72007-12-09 08:59:45[diff] [blame]569 You may also want to try the configure option "--enable-universalsdk"
570 which builds Python as a universal binary with support for the
571 i386 and PPC architetures. This requires Xcode 2.1 or later to build.
Ronald Oussoren988117f2006-04-29 11:31:35[diff] [blame]572
Georg Brandl3fc792b2007-12-09 08:54:02[diff] [blame]573 See Mac/README for more information on framework and
Georg Brandlb0885b72007-12-09 08:59:45[diff] [blame]574 universal builds.
Fred Drakedabed752001-02-01 19:41:13[diff] [blame]575
Michael W. Hudson1c4523f2001-12-19 19:49:58[diff] [blame]576Cygwin: With recent (relative to the time of writing, 2001-12-19)
577 Cygwin installations, there are problems with the interaction
578 of dynamic linking and fork(). This manifests itself in build
579 failures during the execution of setup.py.
Tim Peters42fc7cc2001-04-04 18:35:19[diff] [blame]580
Michael W. Hudson1c4523f2001-12-19 19:49:58[diff] [blame]581 There are two workarounds that both enable Python (albeit
582 without threading support) to build and pass all tests on
583 NT/2000 (and most likely XP as well, though reports of testing
584 on XP would be appreciated).
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]585
Michael W. Hudson1c4523f2001-12-19 19:49:58[diff] [blame]586 The workarounds:
587
588 (a) the band-aid fix is to link the _socket module statically
589 rather than dynamically (which is the default).
590
591 To do this, run "./configure --with-threads=no" including any
592 other options you need (--prefix, etc.). Then in Modules/Setup
Walter Dörwald7dc65d62003-02-02 23:37:05[diff] [blame]593 uncomment the lines:
Tim Peters42fc7cc2001-04-04 18:35:19[diff] [blame]594
Michael W. Hudsonbe5ce182001-12-12 11:46:52[diff] [blame]595 #SSL=/usr/local/ssl
596 #_socket socketmodule.c \
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]597 # -DUSE_SSL -I$(SSL)/include -I$(SSL)/include/openssl \
598 # -L$(SSL)/lib -lssl -lcrypto
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]599
Michael W. Hudson1c4523f2001-12-19 19:49:58[diff] [blame]600 and remove "local/" from the SSL variable. Finally, just run
601 "make"!
Tim Peters42fc7cc2001-04-04 18:35:19[diff] [blame]602
Michael W. Hudsonb6554642001-12-19 22:09:09[diff] [blame]603 (b) The "proper" fix is to rebase the Cygwin DLLs to prevent
Michael W. Hudson1c4523f2001-12-19 19:49:58[diff] [blame]604 base address conflicts. Details on how to do this can be
605 found in the following mail:
606
607 http://sources.redhat.com/ml/cygwin/2001-12/msg00894.html
608
609 It is hoped that a version of this solution will be
Michael W. Hudsonb6554642001-12-19 22:09:09[diff] [blame]610 incorporated into the Cygwin distribution fairly soon.
Michael W. Hudson1c4523f2001-12-19 19:49:58[diff] [blame]611
612 Two additional problems:
613
614 (1) Threading support should still be disabled due to a known
615 bug in Cygwin pthreads that causes test_threadedtempfile to
616 hang.
617
Neal Norwitz4ebde092001-12-19 20:44:13[diff] [blame]618 (2) The _curses module does not build. This is a known
Michael W. Hudson1c4523f2001-12-19 19:49:58[diff] [blame]619 Cygwin ncurses problem that should be resolved the next time
620 that this package is released.
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]621
Michael W. Hudsonbe5ce182001-12-12 11:46:52[diff] [blame]622 On older versions of Cygwin, test_poll may hang and test_strftime
623 may fail.
Tim Peters42fc7cc2001-04-04 18:35:19[diff] [blame]624
Michael W. Hudson1c4523f2001-12-19 19:49:58[diff] [blame]625 The situation on 9X/Me is not accurately known at present.
626 Some time ago, there were reports that the following
627 regression tests failed:
Tim Peters42fc7cc2001-04-04 18:35:19[diff] [blame]628
629 test_pwd
630 test_select (hang)
631 test_socket
632
Michael W. Hudson1c4523f2001-12-19 19:49:58[diff] [blame]633 Due to the test_select hang on 9X/Me, one should run the
634 regression test using the following:
Tim Peters42fc7cc2001-04-04 18:35:19[diff] [blame]635
Michael W. Hudsonbe5ce182001-12-12 11:46:52[diff] [blame]636 make TESTOPTS='-l -x test_select' test
Tim Peters42fc7cc2001-04-04 18:35:19[diff] [blame]637
Michael W. Hudson1c4523f2001-12-19 19:49:58[diff] [blame]638 News regarding these platforms with more recent Cygwin
639 versions would be appreciated!
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]640
Georg Brandl69036212009-03-31 21:43:03[diff] [blame]641Windows: When executing Python scripts on the command line using file type
642 associations (i.e. starting "script.py" instead of "python script.py"),
643 redirects may not work unless you set a specific registry key. See
644 the Knowledge Base article <http://support.microsoft.com/kb/321788>.
Skip Montanaro56d83012003-01-04 04:05:51[diff] [blame]645
646
Skip Montanaro57454e52002-06-14 20:30:31[diff] [blame]647Configuring the bsddb and dbm modules
648-------------------------------------
649
Skip Montanaro56d83012003-01-04 04:05:51[diff] [blame]650Beginning with Python version 2.3, the PyBsddb package
651<http://pybsddb.sf.net/> was adopted into Python as the bsddb package,
652exposing a set of package-level functions which provide
Gregory P. Smith3adc4aa2006-04-13 19:19:01[diff] [blame]653backwards-compatible behavior. Only versions 3.3 through 4.4 of
Skip Montanaro56d83012003-01-04 04:05:51[diff] [blame]654Sleepycat's libraries provide the necessary API, so older versions
655aren't supported through this interface. The old bsddb module has
656been retained as bsddb185, though it is not built by default. Users
657wishing to use it will have to tweak Modules/Setup to build it. The
658dbm module will still be built against the Sleepycat libraries if
Gregory P. Smith3adc4aa2006-04-13 19:19:01[diff] [blame]659other preferred alternatives (ndbm, gdbm) are not found.
Guido van Rossum62d45c02002-12-02 15:49:13[diff] [blame]660
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]661Building the sqlite3 module
662---------------------------
663
664To build the sqlite3 module, you'll need the sqlite3 or libsqlite3
665packages installed, including the header files. Many modern operating
666systems distribute the headers in a separate package to the library -
667often it will be the same name as the main package, but with a -dev or
668-devel suffix.
669
670The version of pysqlite2 that's including in Python needs sqlite3 3.0.8
671or later. setup.py attempts to check that it can find a correct version.
Skip Montanaro57454e52002-06-14 20:30:31[diff] [blame]672
Guido van Rossumdc8a3cb1998-05-12 15:29:18[diff] [blame]673Configuring threads
674-------------------
675
Guido van Rossum7ac11532000-08-29 21:36:40[diff] [blame]676As of Python 2.0, threads are enabled by default. If you wish to
677compile without threads, or if your thread support is broken, pass the
678--with-threads=no switch to configure. Unfortunately, on some
679platforms, additional compiler and/or linker options are required for
680threads to work properly. Below is a table of those options,
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]681collected by Bill Janssen. We would love to automate this process
Guido van Rossum7ac11532000-08-29 21:36:40[diff] [blame]682more, but the information below is not enough to write a patch for the
Matthias Klose3cef2a92012-03-14 22:39:33[diff] [blame]683configure.ac file, so manual intervention is required. If you patch
684the configure.ac file and are confident that the patch works, please
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]685send in the patch. (Don't bother patching the configure script itself
Matthias Klose3cef2a92012-03-14 22:39:33[diff] [blame]686-- it is regenerated each time the configure.ac file changes.)
Guido van Rossumdc8a3cb1998-05-12 15:29:18[diff] [blame]687
688Compiler switches for threads
689.............................
690
Martin v. Löwis1c24f9c2001-03-06 12:07:27[diff] [blame]691The definition of _REENTRANT should be configured automatically, if
692that does not work on your system, or if _REENTRANT is defined
693incorrectly, please report that as a bug.
Guido van Rossumdc8a3cb1998-05-12 15:29:18[diff] [blame]694
Martin v. Löwis1c24f9c2001-03-06 12:07:27[diff] [blame]695 OS/Compiler/threads Switches for use with threads
696 (POSIX is draft 10, DCE is draft 4) compile & link
697
698 SunOS 5.{1-5}/{gcc,SunPro cc}/solaris -mt
699 SunOS 5.5/{gcc,SunPro cc}/POSIX (nothing)
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]700 DEC OSF/1 3.x/cc/DCE -threads
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]701 (butenhof@zko.dec.com)
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]702 Digital UNIX 4.x/cc/DCE -threads
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]703 (butenhof@zko.dec.com)
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]704 Digital UNIX 4.x/cc/POSIX -pthread
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]705 (butenhof@zko.dec.com)
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]706 AIX 4.1.4/cc_r/d7 (nothing)
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]707 (buhrt@iquest.net)
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]708 AIX 4.1.4/cc_r4/DCE (nothing)
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]709 (buhrt@iquest.net)
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]710 IRIX 6.2/cc/POSIX (nothing)
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]711 (robertl@cwi.nl)
Guido van Rossumdc8a3cb1998-05-12 15:29:18[diff] [blame]712
713
714Linker (ld) libraries and flags for threads
715...........................................
716
717 OS/threads Libraries/switches for use with threads
718
719 SunOS 5.{1-5}/solaris -lthread
720 SunOS 5.5/POSIX -lpthread
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]721 DEC OSF/1 3.x/DCE -lpthreads -lmach -lc_r -lc
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]722 (butenhof@zko.dec.com)
Guido van Rossumdc8a3cb1998-05-12 15:29:18[diff] [blame]723 Digital UNIX 4.x/DCE -lpthreads -lpthread -lmach -lexc -lc
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]724 (butenhof@zko.dec.com)
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]725 Digital UNIX 4.x/POSIX -lpthread -lmach -lexc -lc
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]726 (butenhof@zko.dec.com)
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]727 AIX 4.1.4/{draft7,DCE} (nothing)
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]728 (buhrt@iquest.net)
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]729 IRIX 6.2/POSIX -lpthread
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]730 (jph@emilia.engr.sgi.com)
Guido van Rossumdc8a3cb1998-05-12 15:29:18[diff] [blame]731
732
Martin v. Löwis1142de32002-03-29 16:28:31[diff] [blame]733Building a shared libpython
734---------------------------
735
736Starting with Python 2.3, the majority of the interpreter can be built
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]737into a shared library, which can then be used by the interpreter
Martin v. Löwis1142de32002-03-29 16:28:31[diff] [blame]738executable, and by applications embedding Python. To enable this feature,
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]739configure with --enable-shared.
Martin v. Löwis65069672002-08-03 21:38:27[diff] [blame]740
Fred Drake55512dc2003-07-21 16:01:39[diff] [blame]741If you enable this feature, the same object files will be used to create
742a static library. In particular, the static library will contain object
743files using position-independent code (PIC) on platforms where PIC flags
744are needed for the shared library.
Martin v. Löwis1142de32002-03-29 16:28:31[diff] [blame]745
746
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]747Configuring additional built-in modules
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]748---------------------------------------
Guido van Rossum19e0c261995-01-17 16:36:34[diff] [blame]749
Andrew M. Kuchlingebe30452001-02-27 03:24:24[diff] [blame]750Starting with Python 2.1, the setup.py script at the top of the source
751distribution attempts to detect which modules can be built and
752automatically compiles them. Autodetection doesn't always work, so
Guido van Rossum59b0b602002-02-08 22:22:35[diff] [blame]753you can still customize the configuration by editing the Modules/Setup
754file; but this should be considered a last resort. The rest of this
755section only applies if you decide to edit the Modules/Setup file.
756You also need this to enable static linking of certain modules (which
757is needed to enable profiling on some systems).
758
Andrew M. Kuchlingebe30452001-02-27 03:24:24[diff] [blame]759This file is initially copied from Setup.dist by the configure script;
760if it does not exist yet, create it by copying Modules/Setup.dist
761yourself (configure will never overwrite it). Never edit Setup.dist
762-- always edit Setup or Setup.local (see below). Read the comments in
763the file for information on what kind of edits are allowed. When you
764have edited Setup in the Modules directory, the interpreter will
Guido van Rossumef67ded2001-10-05 17:04:42[diff] [blame]765automatically be rebuilt the next time you run make (in the toplevel
766directory).
Guido van Rossum627b2d71993-12-24 10:39:16[diff] [blame]767
Andrew M. Kuchlingebe30452001-02-27 03:24:24[diff] [blame]768Many useful modules can be built on any Unix system, but some optional
769modules can't be reliably autodetected. Often the quickest way to
770determine whether a particular module works or not is to see if it
771will build: enable it in Setup, then if you get compilation or link
Fred Drake6caae142000-10-25 17:51:02[diff] [blame]772errors, disable it -- you're either missing support or need to adjust
773the compilation and linking parameters for that module.
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]774
775On SGI IRIX, there are modules that interface to many SGI specific
Andrew M. Kuchlingebe30452001-02-27 03:24:24[diff] [blame]776system libraries, e.g. the GL library and the audio hardware. These
777modules will not be built by the setup.py script.
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]778
Guido van Rossum4462e931997-01-22 21:00:32[diff] [blame]779In addition to the file Setup, you can also edit the file Setup.local.
780(the makesetup script processes both). You may find it more
781convenient to edit Setup.local and leave Setup alone. Then, when
782installing a new Python version, you can copy your old Setup.local
783file.
784
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]785
786Setting the optimization/debugging options
787------------------------------------------
788
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]789If you want or need to change the optimization/debugging options for
790the C compiler, assign to the OPT variable on the toplevel make
791command; e.g. "make OPT=-g" will build a debugging version of Python
792on most platforms. The default is OPT=-O; a value for OPT in the
793environment when the configure script is run overrides this default
794(likewise for CC; and the initial value for LIBS is used as the base
795set of libraries to link with).
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]796
Fred Drake6caae142000-10-25 17:51:02[diff] [blame]797When compiling with GCC, the default value of OPT will also include
798the -Wall and -Wstrict-prototypes options.
799
800Additional debugging code to help debug memory management problems can
801be enabled by using the --with-pydebug option to the configure script.
802
Brett Cannon08cd5982005-04-24 22:26:38[diff] [blame]803For flags that change binary compatibility, use the EXTRA_CFLAGS
804variable.
805
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]806
Guido van Rossuma68ca352001-10-04 17:00:07[diff] [blame]807Profiling
808---------
809
810If you want C profiling turned on, the easiest way is to run configure
811with the CC environment variable to the necessary compiler
812invocation. For example, on Linux, this works for profiling using
813gprof(1):
814
815 CC="gcc -pg" ./configure
816
817Note that on Linux, gprof apparently does not work for shared
818libraries. The Makefile/Setup mechanism can be used to compile and
Walter Dörwald7dc65d62003-02-02 23:37:05[diff] [blame]819link most extension modules statically.
Guido van Rossuma68ca352001-10-04 17:00:07[diff] [blame]820
821
Georg Brandl405b5f32008-05-18 11:46:51[diff] [blame]822Coverage checking
823-----------------
824
825For C coverage checking using gcov, run "make coverage". This will
826build a Python binary with profiling activated, and a ".gcno" and
827".gcda" file for every source file compiled with that option. With
828the built binary, now run the code whose coverage you want to check.
829Then, you can see coverage statistics for each individual source file
830by running gcov, e.g.
831
832 gcov -o Modules zlibmodule
833
834This will create a "zlibmodule.c.gcov" file in the current directory
835containing coverage info for that source file.
836
837This works only for source files statically compiled into the
838executable; use the Makefile/Setup mechanism to compile and link
839extension modules you want to coverage-check statically.
840
841
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]842Testing
843-------
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]844
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]845To test the interpreter, type "make test" in the top-level directory.
Guido van Rossum6ae5d3d1997-05-14 21:39:05[diff] [blame]846This runs the test set twice (once with no compiled files, once with
847the compiled files left by the previous test run). The test set
848produces some output. You can generally ignore the messages about
Guido van Rossum59b0b602002-02-08 22:22:35[diff] [blame]849skipped tests due to optional features which can't be imported.
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]850If a message is printed about a failed test or a traceback or core
851dump is produced, something is wrong. On some Linux systems (those
Guido van Rossum24df6841997-12-30 04:32:30[diff] [blame]852that are not yet using glibc 6), test_strftime fails due to a
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]853non-standard implementation of strftime() in the C library. Please
854ignore this, or upgrade to glibc version 6.
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]855
Benjamin Petersona18736b2009-02-24 02:45:35[diff] [blame]856By default, tests are prevented from overusing resources like disk space and
857memory. To enable these tests, run "make testall".
858
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]859IMPORTANT: If the tests fail and you decide to mail a bug report,
860*don't* include the output of "make test". It is useless. Run the
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]861failing test manually, as follows:
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]862
Benjamin Petersona18736b2009-02-24 02:45:35[diff] [blame]863 ./python Lib/test/regrtest.py -v test_whatever
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]864
Andrew M. Kuchling83d042d2002-03-21 23:52:20[diff] [blame]865(substituting the top of the source tree for '.' if you built in a
Guido van Rossum6ae5d3d1997-05-14 21:39:05[diff] [blame]866different directory). This runs the test in verbose mode.
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]867
868
869Installing
870----------
871
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]872To install the Python binary, library modules, shared library modules
873(see below), include files, configuration files, and the manual page,
Guido van Rossum64773801997-11-26 17:07:02[diff] [blame]874just type
875
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]876 make install
Guido van Rossum64773801997-11-26 17:07:02[diff] [blame]877
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]878This will install all platform-independent files in subdirectories of
879the directory given with the --prefix option to configure or to the
880`prefix' Make variable (default /usr/local). All binary and other
881platform-specific files will be installed in subdirectories if the
882directory given by --exec-prefix or the `exec_prefix' Make variable
883(defaults to the --prefix directory) is given.
Guido van Rossum64773801997-11-26 17:07:02[diff] [blame]884
Martin v. Löwis3b8ee082003-05-11 20:25:35[diff] [blame]885If DESTDIR is set, it will be taken as the root directory of the
886installation, and files will be installed into $(DESTDIR)$(prefix),
887$(DESTDIR)$(exec_prefix), etc.
888
Guido van Rossum64773801997-11-26 17:07:02[diff] [blame]889All subdirectories created will have Python's version number in their
890name, e.g. the library modules are installed in
Guido van Rossum67345eb2001-01-18 14:51:12[diff] [blame]891"/usr/local/lib/python<version>/" by default, where <version> is the
892<major>.<minor> release number (e.g. "2.1"). The Python binary is
893installed as "python<version>" and a hard link named "python" is
894created. The only file not installed with a version number in its
895name is the manual page, installed as "/usr/local/man/man1/python.1"
896by default.
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]897
Skip Montanaro86d4e7a2008-03-05 16:41:09[diff] [blame]898If you want to install multiple versions of Python see the section below
899entitled "Installing multiple versions".
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]900
901The only thing you may have to install manually is the Python mode for
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]902Emacs found in Misc/python-mode.el. (But then again, more recent
903versions of Emacs may already have it.) Follow the instructions that
904came with Emacs for installation of site-specific files.
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]905
Jack Jansena39ef862001-08-19 21:17:03[diff] [blame]906On Mac OS X, if you have configured Python with --enable-framework, you
907should use "make frameworkinstall" to do the installation. Note that this
908installs the Python executable in a place that is not normally on your
909PATH, you may want to set up a symlink in /usr/local/bin.
910
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]911
Skip Montanaro86d4e7a2008-03-05 16:41:09[diff] [blame]912Installing multiple versions
913----------------------------
914
915On Unix and Mac systems if you intend to install multiple versions of Python
916using the same installation prefix (--prefix argument to the configure
917script) you must take care that your primary python executable is not
Ezio Melotti41c75702009-07-04 01:18:08[diff] [blame]918overwritten by the installation of a different version. All files and
Skip Montanaro86d4e7a2008-03-05 16:41:09[diff] [blame]919directories installed using "make altinstall" contain the major and minor
920version and can thus live side-by-side. "make install" also creates
921${prefix}/bin/python which refers to ${prefix}/bin/pythonX.Y. If you intend
922to install multiple versions using the same prefix you must decide which
923version (if any) is your "primary" version. Install that version using
924"make install". Install all other versions using "make altinstall".
925
926For example, if you want to install Python 2.5, 2.6 and 3.0 with 2.6 being
927the primary version, you would execute "make install" in your 2.6 build
928directory and "make altinstall" in the others.
929
930
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]931Configuration options and variables
932-----------------------------------
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]933
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]934Some special cases are handled by passing options to the configure
935script.
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]936
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]937WARNING: if you rerun the configure script with different options, you
938must run "make clean" before rebuilding. Exceptions to this rule:
939after changing --prefix or --exec-prefix, all you need to do is remove
Guido van Rossumb06df271997-08-05 21:50:20[diff] [blame]940Modules/getpath.o.
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]941
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]942--with(out)-gcc: The configure script uses gcc (the GNU C compiler) if
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]943 it finds it. If you don't want this, or if this compiler is
944 installed but broken on your platform, pass the option
945 --without-gcc. You can also pass "CC=cc" (or whatever the
946 name of the proper C compiler is) in the environment, but the
947 advantage of using --without-gcc is that this option is
948 remembered by the config.status script for its --recheck
949 option.
Guido van Rossum76be6ed1995-01-02 18:33:54[diff] [blame]950
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]951--prefix, --exec-prefix: If you want to install the binaries and the
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]952 Python library somewhere else than in /usr/local/{bin,lib},
953 you can pass the option --prefix=DIRECTORY; the interpreter
954 binary will be installed as DIRECTORY/bin/python and the
955 library files as DIRECTORY/lib/python/*. If you pass
956 --exec-prefix=DIRECTORY (as well) this overrides the
957 installation prefix for architecture-dependent files (like the
958 interpreter binary). Note that --prefix=DIRECTORY also
959 affects the default module search path (sys.path), when
960 Modules/config.c is compiled. Passing make the option
961 prefix=DIRECTORY (and/or exec_prefix=DIRECTORY) overrides the
962 prefix set at configuration time; this may be more convenient
963 than re-running the configure script if you change your mind
964 about the install prefix.
Guido van Rossum76be6ed1995-01-02 18:33:54[diff] [blame]965
Guido van Rossum59b0b602002-02-08 22:22:35[diff] [blame]966--with-readline: This option is no longer supported. GNU
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]967 readline is automatically enabled by setup.py when present.
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]968
Guido van Rossum7ac11532000-08-29 21:36:40[diff] [blame]969--with-threads: On most Unix systems, you can now use multiple
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]970 threads, and support for this is enabled by default. To
971 disable this, pass --with-threads=no. If the library required
972 for threads lives in a peculiar place, you can use
973 --with-thread=DIRECTORY. IMPORTANT: run "make clean" after
974 changing (either enabling or disabling) this option, or you
975 will get link errors! Note: for DEC Unix use
976 --with-dec-threads instead.
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]977
978--with-sgi-dl: On SGI IRIX 4, dynamic loading of extension modules is
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]979 supported by the "dl" library by Jack Jansen, which is
980 ftp'able from ftp://ftp.cwi.nl/pub/dynload/dl-1.6.tar.Z.
981 This is enabled (after you've ftp'ed and compiled the dl
982 library) by passing --with-sgi-dl=DIRECTORY where DIRECTORY
983 is the absolute pathname of the dl library. (Don't bother on
984 IRIX 5, it already has dynamic linking using SunOS style
985 shared libraries.) THIS OPTION IS UNSUPPORTED.
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]986
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]987--with-dl-dld: Dynamic loading of modules is rumored to be supported
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]988 on some other systems: VAX (Ultrix), Sun3 (SunOS 3.4), Sequent
989 Symmetry (Dynix), and Atari ST. This is done using a
990 combination of the GNU dynamic loading package
991 (ftp://ftp.cwi.nl/pub/dynload/dl-dld-1.1.tar.Z) and an
992 emulation of the SGI dl library mentioned above (the emulation
993 can be found at
994 ftp://ftp.cwi.nl/pub/dynload/dld-3.2.3.tar.Z). To
995 enable this, ftp and compile both libraries, then call
996 configure, passing it the option
997 --with-dl-dld=DL_DIRECTORY,DLD_DIRECTORY where DL_DIRECTORY is
998 the absolute pathname of the dl emulation library and
999 DLD_DIRECTORY is the absolute pathname of the GNU dld library.
1000 (Don't bother on SunOS 4 or 5, they already have dynamic
1001 linking using shared libraries.) THIS OPTION IS UNSUPPORTED.
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]1002
1003--with-libm, --with-libc: It is possible to specify alternative
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1004 versions for the Math library (default -lm) and the C library
1005 (default the empty string) using the options
1006 --with-libm=STRING and --with-libc=STRING, respectively. For
1007 example, if your system requires that you pass -lc_s to the C
1008 compiler to use the shared C library, you can pass
1009 --with-libc=-lc_s. These libraries are passed after all other
1010 libraries, the C library last.
Guido van Rossum16e809e2001-04-16 02:07:08[diff] [blame]1011
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]1012--with-libs='libs': Add 'libs' to the LIBS that the python interpreter
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1013 is linked against.
Guido van Rossumd02ba451996-07-31 17:36:01[diff] [blame]1014
Martin v. Löwis0f48d982006-04-14 14:34:26[diff] [blame]1015--with-cxx-main=<compiler>: If you plan to use C++ extension modules,
1016 then -- on some platforms -- you need to compile python's main()
1017 function with the C++ compiler. With this option, make will use
1018 <compiler> to compile main() *and* to link the python executable.
1019 It is likely that the resulting executable depends on the C++
1020 runtime library of <compiler>. (The default is --without-cxx-main.)
1021
1022 There are platforms that do not require you to build Python
1023 with a C++ compiler in order to use C++ extension modules.
1024 E.g., x86 Linux with ELF shared binaries and GCC 3.x, 4.x is such
1025 a platform. We recommend that you configure Python
1026 --without-cxx-main on those platforms because a mismatch
1027 between the C++ compiler version used to build Python and to
1028 build a C++ extension module is likely to cause a crash at
1029 runtime.
1030
1031 The Python installation also stores the variable CXX that
1032 determines, e.g., the C++ compiler distutils calls by default
1033 to build C++ extensions. If you set CXX on the configure command
1034 line to any string of non-zero length, then configure won't
1035 change CXX. If you do not preset CXX but pass
1036 --with-cxx-main=<compiler>, then configure sets CXX=<compiler>.
1037 In all other cases, configure looks for a C++ compiler by
1038 some common names (c++, g++, gcc, CC, cxx, cc++, cl) and sets
1039 CXX to the first compiler it finds. If it does not find any
1040 C++ compiler, then it sets CXX="".
1041
1042 Similarly, if you want to change the command used to link the
1043 python executable, then set LINKCC on the configure command line.
Martin v. Löwis2df66942000-12-13 14:14:32[diff] [blame]1044
1045
Fred Drake6caae142000-10-25 17:51:02[diff] [blame]1046--with-pydebug: Enable additional debugging code to help track down
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1047 memory management problems. This allows printing a list of all
1048 live objects when the interpreter terminates.
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]1049
Jack Jansen7b8c7542002-04-14 20:12:41[diff] [blame]1050--with(out)-universal-newlines: enable reading of text files with
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1051 foreign newline convention (default: enabled). In other words,
1052 any of \r, \n or \r\n is acceptable as end-of-line character.
1053 If enabled import and execfile will automatically accept any newline
1054 in files. Python code can open a file with open(file, 'U') to
1055 read it in universal newline mode. THIS OPTION IS UNSUPPORTED.
Fred Drake6caae142000-10-25 17:51:02[diff] [blame]1056
Martin v. Löwisf30d60e2004-06-08 08:17:44[diff] [blame]1057--with-tsc: Profile using the Pentium timestamping counter (TSC).
Guido van Rossum76be6ed1995-01-02 18:33:54[diff] [blame]1058
Martin v. Löwis9176fc12006-04-11 11:12:43[diff] [blame]1059--with-system-ffi: Build the _ctypes extension module using an ffi
Georg Brandlb0885b72007-12-09 08:59:45[diff] [blame]1060 library installed on the system.
Martin v. Löwis9176fc12006-04-11 11:12:43[diff] [blame]1061
Matthias Klose10cbe482009-04-29 17:18:19[diff] [blame]1062--with-dbmliborder=db1:db2:...: Specify the order that backends for the
1063 dbm extension are checked. Valid value is a colon separated string
1064 with the backend names `ndbm', `gdbm' and `bdb'.
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]1065
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1066Building for multiple architectures (using the VPATH feature)
1067-------------------------------------------------------------
1068
1069If your file system is shared between multiple architectures, it
1070usually is not necessary to make copies of the sources for each
1071architecture you want to support. If the make program supports the
1072VPATH feature, you can create an empty build directory for each
1073architecture, and in each directory run the configure script (on the
1074appropriate machine with the appropriate options). This creates the
1075necessary subdirectories and the Makefiles therein. The Makefiles
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]1076contain a line VPATH=... which points to a directory containing the
Guido van Rossumd0fe8451996-08-26 03:02:37[diff] [blame]1077actual sources. (On SGI systems, use "smake -J1" instead of "make" if
1078you use VPATH -- don't try gnumake.)
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1079
1080For example, the following is all you need to build a minimal Python
1081in /usr/tmp/python (assuming ~guido/src/python is the toplevel
1082directory and you want to build in /usr/tmp/python):
1083
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1084 $ mkdir /usr/tmp/python
1085 $ cd /usr/tmp/python
1086 $ ~guido/src/python/configure
1087 [...]
1088 $ make
1089 [...]
1090 $
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1091
Neil Schemenauer73f8ab22001-01-26 22:18:55[diff] [blame]1092Note that configure copies the original Setup file to the build
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1093directory if it finds no Setup file there. This means that you can
1094edit the Setup file for each architecture independently. For this
1095reason, subsequent changes to the original Setup file are not tracked
1096automatically, as they might overwrite local changes. To force a copy
1097of a changed original Setup file, delete the target Setup file. (The
1098makesetup script supports multiple input files, so if you want to be
1099fancy you can change the rules to create an empty Setup.local if it
1100doesn't exist and run it with arguments $(srcdir)/Setup Setup.local;
1101however this assumes that you only need to add modules.)
1102
Christian Heimes0a0e5832007-12-13 19:23:16[diff] [blame]1103Also note that you can't use a workspace for VPATH and non VPATH builds. The
1104object files left behind by one version confuses the other.
1105
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1106
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1107Building on non-UNIX systems
1108----------------------------
1109
Martin v. Löwis2eba0d62004-12-24 08:28:28[diff] [blame]1110For Windows (2000/NT/ME/98/95), assuming you have MS VC++ 7.1, the
Tim Peters03444242000-09-19 00:38:35[diff] [blame]1111project files are in PCbuild, the workspace is pcbuild.dsw. See
1112PCbuild\readme.txt for detailed instructions.
Guido van Rossum8d7d4ed1996-07-30 21:41:07[diff] [blame]1113
Martin v. Löwis2eba0d62004-12-24 08:28:28[diff] [blame]1114For other non-Unix Windows compilers, in particular MS VC++ 6.0 and
Guido van Rossum31ae2071999-04-12 14:47:30[diff] [blame]1115for OS/2, enter the directory "PC" and read the file "readme.txt".
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1116
1117For the Mac, a separate source distribution will be made available,
1118for use with the CodeWarrior compiler. If you are interested in Mac
1119development, join the PythonMac Special Interest Group
1120(http://www.python.org/sigs/pythonmac-sig/, or send email to
1121pythonmac-sig-request@python.org).
1122
1123Of course, there are also binary distributions available for these
Guido van Rossum67345eb2001-01-18 14:51:12[diff] [blame]1124platforms -- see http://www.python.org/.
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1125
1126To port Python to a new non-UNIX system, you will have to fake the
1127effect of running the configure script manually (for Mac and PC, this
1128has already been done for you). A good start is to copy the file
Michael W. Hudsonbe6e0f42002-06-23 16:29:36[diff] [blame]1129pyconfig.h.in to pyconfig.h and edit the latter to reflect the actual
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1130configuration of your system. Most symbols must simply be defined as
11311 only if the corresponding feature is present and can be left alone
Michael W. Hudsonbe6e0f42002-06-23 16:29:36[diff] [blame]1132otherwise; however the *_t type symbols must be defined as some
1133variant of int if they need to be defined at all.
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1134
Barry Warsaw06557452001-12-14 20:47:12[diff] [blame]1135For all platforms, it's important that the build arrange to define the
1136preprocessor symbol NDEBUG on the compiler command line in a release
1137build of Python (else assert() calls remain in the code, hurting
1138release-build performance). The Unix, Windows and Mac builds already
1139do this.
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1140
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]1141
1142Miscellaneous issues
1143====================
1144
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]1145Emacs mode
1146----------
1147
Guido van Rossumc0be2f51997-10-08 05:05:28[diff] [blame]1148There's an excellent Emacs editing mode for Python code; see the file
Benjamin Petersonc733b322010-06-20 15:12:04[diff] [blame]1149Misc/python-mode.el. Originally written by the famous Tim Peters, it is now
1150maintained by the equally famous Barry Warsaw. The latest version, along with
1151various other contributed Python-related Emacs goodies, is online at
1152http://launchpad.net/python-mode/.
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]1153
1154
Guido van Rossum59b0b602002-02-08 22:22:35[diff] [blame]1155Tkinter
1156-------
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]1157
Guido van Rossum59b0b602002-02-08 22:22:35[diff] [blame]1158The setup.py script automatically configures this when it detects a
1159usable Tcl/Tk installation. This requires Tcl/Tk version 8.0 or
Fred Drake8179a9d2000-09-28 23:28:30[diff] [blame]1160higher.
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1161
Guido van Rossum973e4dc2000-07-01 00:34:39[diff] [blame]1162For more Tkinter information, see the Tkinter Resource page:
1163http://www.python.org/topics/tkinter/
Guido van Rossumc0be2f51997-10-08 05:05:28[diff] [blame]1164
Michael W. Hudson71dcc3e2005-02-22 15:33:26[diff] [blame]1165There are demos in the Demo/tkinter directory.
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]1166
1167Note that there's a Python module called "Tkinter" (capital T) which
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]1168lives in Lib/lib-tk/Tkinter.py, and a C module called "_tkinter"
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1169(lower case t and leading underscore) which lives in
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]1170Modules/_tkinter.c. Demos and normal Tk applications import only the
Guido van Rossum59b0b602002-02-08 22:22:35[diff] [blame]1171Python Tkinter module -- only the latter imports the C _tkinter
1172module. In order to find the C _tkinter module, it must be compiled
1173and linked into the Python interpreter -- the setup.py script does
1174this. In order to find the Python Tkinter module, sys.path must be
1175set correctly -- normal installation takes care of this.
Guido van Rossum84c8c7f1995-08-28 02:44:24[diff] [blame]1176
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]1177
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1178Distribution structure
1179----------------------
1180
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]1181Most subdirectories have their own README files. Most files have
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1182comments.
1183
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1184Demo/ Demonstration scripts, modules and programs
Georg Brandlc207c712008-02-09 21:38:54[diff] [blame]1185Doc/ Documentation sources (reStructuredText)
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1186Grammar/ Input for the parser generator
1187Include/ Public header files
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1188LICENSE Licensing information
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1189Lib/ Python library modules
Michael W. Hudsonbe6e0f42002-06-23 16:29:36[diff] [blame]1190Mac/ Macintosh specific resources
Neil Schemenauer3f52da52001-01-24 17:49:59[diff] [blame]1191Makefile.pre.in Source from which config.status creates the Makefile.pre
Guido van Rossumc0be2f51997-10-08 05:05:28[diff] [blame]1192Misc/ Miscellaneous useful files
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1193Modules/ Implementation of most built-in modules
1194Objects/ Implementation of most built-in object types
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]1195PC/ Files specific to PC ports (DOS, Windows, OS/2)
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1196PCbuild/ Build directory for Microsoft Visual C++
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1197Parser/ The parser and tokenizer and their input handling
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]1198Python/ The byte-compiler and interpreter
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1199README The file you're reading now
Neal Norwitzf8519c42007-05-20 18:43:00[diff] [blame]1200RISCOS/ Files specific to RISC OS port
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1201Tools/ Some useful programs written in Python
Michael W. Hudsonbe6e0f42002-06-23 16:29:36[diff] [blame]1202pyconfig.h.in Source from which pyconfig.h is created (GNU autoheader output)
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1203configure Configuration shell script (GNU autoconf output)
Matthias Klose3cef2a92012-03-14 22:39:33[diff] [blame]1204configure.ac Configuration specification (input for GNU autoconf)
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1205install-sh Shell script used to install files
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1206setup.py Python script used to build extension modules
Guido van Rossum91cb9d21995-04-10 11:47:38[diff] [blame]1207
1208The following files will (may) be created in the toplevel directory by
1209the configuration and build processes:
1210
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1211Makefile Build rules
Neil Schemenauer3f52da52001-01-24 17:49:59[diff] [blame]1212Makefile.pre Build rules before running Modules/makesetup
Michael W. Hudsonbe6e0f42002-06-23 16:29:36[diff] [blame]1213buildno Keeps track of the build number
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]1214config.cache Cache of configuration variables
Michael W. Hudsonbe6e0f42002-06-23 16:29:36[diff] [blame]1215pyconfig.h Configuration header
Guido van Rossumc0be2f51997-10-08 05:05:28[diff] [blame]1216config.log Log from last configure run
Guido van Rossum4f40c2d2000-09-05 16:46:22[diff] [blame]1217config.status Status from last run of the configure script
Anthony Baxter70e07962006-04-03 14:16:27[diff] [blame]1218getbuildinfo.o Object file from Modules/getbuildinfo.c
1219libpython<version>.a The library archive
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1220python The executable interpreter
Neal Norwitzf8519c42007-05-20 18:43:00[diff] [blame]1221reflog.txt Output from running the regression suite with the -R flag
Guido van Rossum6d9cc801996-08-01 17:31:22[diff] [blame]1222tags, TAGS Tags files for vi and Emacs
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1223
1224
Guido van Rossumc07d5fa2000-09-01 22:50:02[diff] [blame]1225That's all, folks!
1226------------------
Guido van Rossum433c8ad1994-08-01 12:07:07[diff] [blame]1227
1228
Guido van Rossum67345eb2001-01-18 14:51:12[diff] [blame]1229--Guido van Rossum (home page: http://www.python.org/~guido/)