Popular Python projects with lots of pure Python code

[Brett C.]

For my thesis (once the bloody thing stops throwing bugs at me) I am
going to need to collect stats on the frequency that atomic types in
local variables are applied to various opcodes and methods. That means
that I need something to collect stats off of.

So I am trying to come up with a list of projects I can use to get my
stats from. the stdlib is a no-brainer. Also plan to use Docutils.
But beyond those most of the big Python projects I can think of have a
ton of C code which is of no use to me. I don't need to be able to run
it, just going to be compiling to .py files.

Any suggestions?

-Brett

[Larry Bates]

ReportLab?

HTH,
Larry Bates
Syscon, Inc.

"Brett C." <ba*@OCF.Berkel ey.EDU> wrote in message
news:ma******** *************** *************** @python.org...

For my thesis (once the bloody thing stops throwing bugs at me) I am
going to need to collect stats on the frequency that atomic types in
local variables are applied to various opcodes and methods. That means
that I need something to collect stats off of.

So I am trying to come up with a list of projects I can use to get my
stats from. the stdlib is a no-brainer. Also plan to use Docutils.
But beyond those most of the big Python projects I can think of have a
ton of C code which is of no use to me. I don't need to be able to run
it, just going to be compiling to .py files.

Any suggestions?

-Brett

[Troels Therkelsen]

In article <ma************ *************** ***********@pyt hon.org>, Brett C. wrote:

For my thesis (once the bloody thing stops throwing bugs at me) I am
going to need to collect stats on the frequency that atomic types in
local variables are applied to various opcodes and methods. That means
that I need something to collect stats off of.

So I am trying to come up with a list of projects I can use to get my
stats from. the stdlib is a no-brainer. Also plan to use Docutils.
But beyond those most of the big Python projects I can think of have a
ton of C code which is of no use to me. I don't need to be able to run
it, just going to be compiling to .py files.

Any suggestions?

What about MoinMoin and Mailman? Those are fairly large-ish pure-python
projects.

If 100% pure python isn't a must, PIL might also be able to yield useful stats
as all of the abstraction is done in python.

Hope this helps,

Troels Therkelsen

[Roy Smith]

In article <ma************ *************** ***********@pyt hon.org>,
"Brett C." <ba*@OCF.Berkel ey.EDU> wrote:

For my thesis (once the bloody thing stops throwing bugs at me) I am
going to need to collect stats on the frequency that atomic types in
local variables are applied to various opcodes and methods. That means
that I need something to collect stats off of.

So I am trying to come up with a list of projects I can use to get my
stats from. the stdlib is a no-brainer. Also plan to use Docutils.
But beyond those most of the big Python projects I can think of have a
ton of C code which is of no use to me. I don't need to be able to run
it, just going to be compiling to .py files.

Any suggestions?

-Brett

Ilya Etingof's PySNMP package is a pretty large collection of Python
code. See http://pysnmp.sourceforge.net/

[Roger Binns]

Brett C. wrote:

So I am trying to come up with a list of projects I can use to get my
stats from. the stdlib is a no-brainer. Also plan to use Docutils.
But beyond those most of the big Python projects I can think of have a
ton of C code which is of no use to me. I don't need to be able to
run it, just going to be compiling to .py files.

ZOPE/Plone. BitTorrent.

Have a look on SourceForge.

http://sourceforge.net/softwaremap/t...p?form_cat=178

Other than the 10,000 clones of BitTorrent and projects that
use Python very peripherally, there are many appropriate
projects.

My BitPim project has 30,000 lines of Python, 2,500 lines of
description files which are compiled into 41,000 lines of
Python and 500 lines of C/C++. It does GUI stuff, talks on
serial ports to cell phones, has XML-RPC remoting over secure
connections, has installation code, uses external data sources
such as CSV, vCards as well as Outlook, Evolution etc. Consequently
that gives you a nice wide spread of things the code does. If you do
use it, feel free to send me email if you have any questions
or need help. There is also a presentation I did which gives a
good overview:

http://bitpim.org/papers/baypiggies/

Roger

[Fernando Perez]

Brett C. wrote:

For my thesis (once the bloody thing stops throwing bugs at me) I am
going to need to collect stats on the frequency that atomic types in
local variables are applied to various opcodes and methods. That means
that I need something to collect stats off of.

So I am trying to come up with a list of projects I can use to get my
stats from. the stdlib is a no-brainer. Also plan to use Docutils.
But beyond those most of the big Python projects I can think of have a
ton of C code which is of no use to me. I don't need to be able to run
it, just going to be compiling to .py files.

Any suggestions?

well, scipy is not pure python (lots of fortran/C), but it does have a lot of
python in it. Given that it's probably a decent representative of typical
scientific code (and scientific use of python is growing fast, see scipy'04),
it might be an interesting addition to your sample.

Best,

f

[Patrick Ellis]

Brett C. wrote:

For my thesis (once the bloody thing stops throwing bugs at me) I am
going to need to collect stats on the frequency that atomic types in
local variables are applied to various opcodes and methods. That
means that I need something to collect stats off of.

So I am trying to come up with a list of projects I can use to get my
stats from. the stdlib is a no-brainer. Also plan to use Docutils.
But beyond those most of the big Python projects I can think of have a
ton of C code which is of no use to me. I don't need to be able to
run it, just going to be compiling to .py files.

Any suggestions?

-Brett

BoaConstructor

[Dan Sommers]

On Mon, 09 Aug 2004 15:01:29 -0700,
"Brett C." <ba*@OCF.Berkel ey.EDU> wrote:

For my thesis (once the bloody thing stops throwing bugs at me) I am
going to need to collect stats on the frequency that atomic types in
local variables are applied to various opcodes and methods. That means
that I need something to collect stats off of. So I am trying to come up with a list of projects I can use to get my
stats from. the stdlib is a no-brainer. Also plan to use Docutils. But
beyond those most of the big Python projects I can think of have a ton
of C code which is of no use to me. I don't need to be able to run it,
just going to be compiling to .py files. Any suggestions?

It's not a project, as such, but how about Python's standard library?

Tons and tons of python code, written by the best python coders around.

Regards,
Dan

--
Dan Sommers
<http://www.tombstoneze ro.net/dan/>
Never play leapfrog with a unicorn.

[Michael Scarlett]

Troels Therkelsen <t_**********@h otmail.com> wrote in message news:<sl******* *************** **@abattoir.2-10.org>...

In article <ma************ *************** ***********@pyt hon.org>, Brett C. wrote:

[/snipped]

Any suggestions?

yeah, why don't you go to sourcforge and search under python projects,
you'll get a ton

[Brett C.]

"Brett C." <ba*@OCF.Berkel ey.EDU> wrote in message news:

Any suggestions?

Just a quick thanks to everyone who has responded. I truly
appreciate the help on this! This is why the Python community is just
so damn cool!

And here are some quick answers to some questions that have come up in
the thread...

"Big" is relative; not looking for simple scripts, but otherwise I am
open. No real minimum size or anything.

The projects do not have to be devoid of C; just have enough Python
code to warrant using.

"Popular" is also relative. Since Python is not used here at my
school (Cal Poly SLO), I wanted to also use my thesis as a slight PR
piece within the department. Being able to list big, popular Python
projects would help to show the legitimacy of the language to
professors.

Please suggest only open source projects. While I have no issue with
proprietary projects personally I don't want to have to worry about
what I can and cannot discuss in my thesis.

I personally don't need any more web projects listed. Go for it if
this thread just keeps going, but with the ones listed so far that is
plenty. I would love to get a wide swath of projects so as to make
sure my thesis can't be accused of not being representative of non-web
projects.

Something I should have mentioned in the OP is anything that has a
benchmark is an uber-plus! I am going to end up having to benchmark
some code that I did not gather stats on along with some that I did to
try to show that this was all worth it. Already plan to use pystone
and parrotbench.

Anyway, here is the current list that I have going. Ignore the
question marks; just means I have not decided whether I am definitely
using it or not.

----------------

-----------------------------------
Projects for stat collection
-----------------------------------
+ Stdlib (w/o test)
+ Docutils
+ Twisted?
+ Zope?
* Plone?
+ ReportLab?
+ wxPython?
+ BitTorrent?
+ SpamBayes?
+ Mailman?
+ SiGeFi ?
+ Plucker ?
+ Universal Feed Parser ?
+ Shtoom?
+ Divmod Quotient?
+ MoinMoin?
+ PIL?
+ PySNMP ?
+ scipy ?
+ Boa Constructor ?
+ PythonCard ?
+ OpenGLContext ?
+ PyTable ?

------------------
Benchmarks
------------------
+ pystone
+ parrotbench