|
Lurchington posted:I've been writing stuff in python pretty much every day for a year and it's only now coming to my attention how little of the language I'm actually using day-to-day. Context management seems exciting and I'm looking forward to getting better. I've been writing python pretty much full time for about 6-7 years now, and I still don't do it as well, or take advantage of, some of the cooler features/language bits that I see others use. That's one of the big reason why I tend to read through the implementation of core modules in python SVN - they teach you a lot.
|
#
?
Oct 6, 2008 14:25
|
|
|
|
| # ? May 12, 2024 06:50 |
|
|
I'd say learning the internals and standard conventions is pretty helpful, non-obvious stuff which is useful but you might not naturally learn or use. You can do pretty cool stuff fiddling around with anything that comes between two sets of "__". The 'conventions' that I randomly picked up by reading other peoples' code makes me realize why Python OOP is more "pure" than C++ or Java, but why a C++ or Java programmer will think Python OOP is horribly broken and tacked-on. edit: bitprophet posted:See, I just DO NOT GET why everyone has this huge stiffie for developing (sans runserver and the builtin Python, of course -- which is fine for a newbie) on a local OS X workstation. It's loving retarded -- you have to wrestle with 3 or 4 different potential methods of installing the various components and then you end up with 2 versions of just about everything (the Apple default and the new one you installed because the Apple default is too loving old or lovely). I have the same feelings about RHEL / CentOS. To me, "enterprise" seems to mean "really loving old packages which you aren't going to use, and instead compile your own versions of anything important and completely defeat the purpose of using a packaging system". Macs in general just confuse me. Why spend a lot on some alien artifact-looking machine when you can just scrounge up some old hardware, slap Linux on it, and run with it? Seriously, if you are doing web development you aren't really going to need any significant power on your workstation. duck monster posted:Oh god.... One of the angriest emails I ever wrote was something to the effect of "Why the gently caress are you compiling free software with an unfree compiler? USE GCC!!!!!!!!" because I was blocked from using so many libraries due to the need for Visual studio. Hey, that's one of the main reasons I avoid C/C++ these days. People think I'm a pussy, but at least I can get "Hello World" without having to find/compile/link to a zillion dependency libraries. nihilocrat fucked around with this message at 17:58 on Oct 6, 2008 |
|
#
?
Oct 6, 2008 17:46
|
|
|
Using ftplib is there anyway to resume binary file uploads? I've looked and looked, and i can't find anything. REST doesn't seem to work, when uploading to the server. Storing binary on the server, always causes it to start writing from the beginning of the file again.
|
|
#
?
Oct 6, 2008 22:18
|
|
|
nihilocrat posted:Hey, that's one of the main reasons I avoid C/C++ these days. People think I'm a pussy, but at least I can get "Hello World" without having to find/compile/link to a zillion dependency libraries. Actually I agree. I programmed in C++ for years and recently had to return to it. The whole library-link thing is ridiculous: compilers being unable to find a library you have on your disk, or compiling and then complaining about an "unresolved symbol". If nothing else, Java got all that stuff right. m0nk3yz posted:.. on context managers ... Well, huh. Thanks for the explanation.That's actually useful - I actually constructs like that in C++ (set something up in the c'tor, clean it up in the d'tor) and recently tried to do something similar in Python. (It works but only 2 problems - you can't tell when the object will be deleted, and session termination disposes of objects in a weird way). This will be a lot easier.
|
|
#
?
Oct 7, 2008 12:28
|
|
|
My head hurts. What the hell were apple thinking even CONTEMPLATING distributing python without loving database bindings. Without the db bindings, almost the entire spectrum of useful things to do with scripting languages is left a flibbering mess. Lazy Apple. Lazy loving lazy. Anyone know a good way of getting mysqldb of some variety working on the mac. This is making me too angry working it out for myself 
|
|
#
?
Oct 7, 2008 19:09
|
|
|
Macports' version not working? (I forget if we went over this already in the Django thread  )
|
|
#
?
Oct 7, 2008 21:52
|
|
|
duck monster posted:My head hurts. Apple doesn't distribute mysql - why would they give you bindings, especially given those bindings compile against a particular version of MySQL that's local to the machine? Note, I do have instructions, I just need to find them - I'll post them when I have them Edit: Here you go: http://yousefourabi.com/apple/django-on-leopard Note, that when he says edit _mysql.c - you need to change 2 parts, which he omits. Your diff should look like this: code:m0nk3yz fucked around with this message at 22:15 on Oct 7, 2008 |
|
#
?
Oct 7, 2008 22:01
|
|
|
duck monster posted:Anyone know a good way of getting mysqldb of some variety working on the mac. This is making me too angry working it out for myself Getting MySQLdb to work can be a bitch. Sometimes it just works but other times I've spent days trying to get it installed (with occasional periods of it looking like it installed but it actually not working). General advice: use of the precompiled packages out there. Or use Postgres instead. Or follow the tale of woe on my webpages as I encounter every problem under the sun. I've also had problems with MySQLdb on Redhat as well, so it's not just Mac. DPAPI is okay, but by now Python really needs a saner and more standard set of db modules, like Java.
|
|
#
?
Oct 8, 2008 10:09
|
|
|
m0nk3yz posted:Apple doesn't distribute mysql - why would they give you bindings, especially given those bindings compile against a particular version of MySQL that's local to the machine? I'd argue however that Mysql is fundamental enough that apple ought be providing its own package of that as well, or at least providing bindings against the mysql downloadable from the mysql site. Because could you imagine distributing a python app for , say cocoa, that had to do mysql or postgres work? At least with windows, you have mysql.dll or whatever it is, and it all seems to magically kinda work.
|
|
#
?
Oct 8, 2008 13:31
|
|
|
duck monster posted:I'd argue however that Mysql is fundamental enough that apple ought be providing its own package of that as well, or at least providing bindings against the mysql downloadable from the mysql site. Did the instructions I provided help you?
|
|
#
?
Oct 8, 2008 17:41
|
|
|
duck monster posted:I'd argue however that Mysql is fundamental enough that apple ought be providing its own package of that as well, or at least providing bindings against the mysql downloadable from the mysql site. Ever worked with py2app? It would take a little more work than on Windows, sure, but if you use py2app to make the distribution and then use otool to change the _mysql.so file to use a package-local version of libmysqlclient_r.dylib instead of one in /usr/local/lib or whatever, then you'd be fine. I haven't done this myself, but there's no reason it shouldn't work as long as you include both _mysql.so and the client dylib. e: Assuming, of course, that it really is that simple on Windows. Habnabit fucked around with this message at 01:16 on Oct 10, 2008 |
|
#
?
Oct 10, 2008 01:13
|
|
|
Total "RTFM, n00b!!" type question, but I read the manual, and I still have no clue. In fact, you may go so far as to suggest that if I'm asking such a question, I have no business writing code that needs to be documented anyway, but whatever. So I installed Epydoc, as per the manual's suggestion: code:So my problem is this: how do I launch it from the command line? I'm relatively new to using the bash, but I tried looking up how to launch scripts in Unix, and that did nothing for me. I just want to document my code for a class on applied numerical methods! latexenthusiast fucked around with this message at 08:23 on Oct 11, 2008 |
|
#
?
Oct 11, 2008 07:17
|
|
|
The command should be " /Library/Frameworks/Python.framework/Versions/2.5/bin/epydoc"
|
|
#
?
Oct 11, 2008 12:48
|
|
|
Also, do the following - open your ~/.bash_profile or ~/.bashrc in your favorite editor - this file controls your bash environment. For me, my ~/.bashrc file contains:code:Add the following line to your .bash_profile: code:Epydoc is awesome - I just cut work over to using it for our python tests and code. I just can't get the UML charts to render for some odd reason.
|
|
#
?
Oct 11, 2008 13:22
|
|
|
You guys are the best. Thank you so much. I thought there was no way I'd have to type in the full path, given how long it was, but I guess I was wrong. Dead wrong. The tip about .bash_login was really helpful, too, so thanks!
|
|
#
?
Oct 12, 2008 00:01
|
|
|
FWIW, I chucked another side project up on pypi today: http://pypi.python.org/pypi/pyjavaproperties/ Here's the others: http://pypi.python.org/pypi/nose-testconfig/ http://code.google.com/p/testbutler/ I've got two more pending nose plugin projects as well as a multiprocessing/MPI package coming as well, although the last one might turn into an actor-model thing, right now I am trying to pull dramatis (http://dramatis.mischance.net/) apart.
|
|
#
?
Oct 12, 2008 02:54
|
|
|
m0nk3yz posted:FWIW, I chucked another side project up on pypi today: woo testbutler looks nice.
|
|
#
?
Oct 12, 2008 12:54
|
|
|
duck monster posted:woo testbutler looks nice. It's still in the prototype stages, although we have it deployed for actual production use at work. I have a laundry list of things I want to add to it. Right now I am working on a plugin for nose to upload results to the system in a new testbutler.results app.
|
|
#
?
Oct 12, 2008 15:05
|
|
|
Is there an efficient way to remove all non-alphanumeric characters from strings without using regular expressions?
|
|
#
?
Oct 13, 2008 00:12
|
|
|
Yes, but: Why efficient - strings are immutable in python, so often any character by character option won't be that efficient Why not regular expressions - they're in the standard library ?
|
|
#
?
Oct 13, 2008 00:27
|
|
|
One way to do it:code:code:hey mom its 420 fucked around with this message at 00:32 on Oct 13, 2008 |
|
#
?
Oct 13, 2008 00:30
|
|
|
Tomed2000 posted:Is there an efficient way to remove all non-alphanumeric characters from strings without using regular expressions? code:edit: \/\/ this one is better Pivo fucked around with this message at 00:35 on Oct 13, 2008 |
|
#
?
Oct 13, 2008 00:33
|
|
|
code:
|
|
#
?
Oct 13, 2008 00:33
|
|
|
That's definitely a contrived way of using reduce(), and I can't imagine it being efficient. I'd use str.join:code:
|
|
#
?
Oct 13, 2008 00:34
|
|
|
Ugh, I've been in Haskell land for too long, that reduce was uncalled for.
|
|
#
?
Oct 13, 2008 00:36
|
|
|
Pivo posted:
This has O(n2) complexity. Milde posted:
code:
|
|
#
?
Oct 13, 2008 00:39
|
|
|
Scaevolus posted:This has O(n2) complexity. How come? Is it doing string concatenation at every step of the filter? I'm a Python neophyte..
|
|
#
?
Oct 13, 2008 00:47
|
|
|
Pivo posted:How come? Is it doing string concatenation at every step of the filter? I'm a Python neophyte..
|
|
#
?
Oct 13, 2008 00:54
|
|
|
Scaevolus posted:This has O(n2) complexity. Technically the generator expression would be slower, but use less memory, or something like that. The difference is miniscule for that particular string, and it looks like the filter() method is faster than both. The regular expression would probably be the fastest though, and it's a perfect fit.
|
|
#
?
Oct 13, 2008 00:59
|
|
|
Scaevolus posted:character in string is O(n). string.digits and string.letters are constant length though assuming locale doesn't change mid-execution. Our n is the len(s) where s is the input string. So, should still be linear.
|
|
#
?
Oct 13, 2008 01:03
|
|
|
Pivo posted:string.digits and string.letters are constant length though assuming locale doesn't change mid-execution. 
|
|
#
?
Oct 13, 2008 01:04
|
|
|
On the subject of using Regular expressions to remove characters. Is there a good method of removing (or replacing) any characters considered illegal-filename characters? I know those characters are different between Windows and Unix, and so far I manually specified those characters, but I'd prefer something else if I can help it. Some google searches returned that it's impossible to know illegal characters ahead of time, which is surprising to me.
|
|
#
?
Oct 14, 2008 00:34
|
|
|
the problem is that you often don't know the filesystem a directory is using, and so even under unix you can be subject to windows filenames. what's wrong with try/catch instead of trying to guess what might be valid?
|
|
#
?
Oct 14, 2008 01:07
|
|
|
tef posted:the problem is that you often don't know the filesystem a directory is using, and so even under unix you can be subject to windows filenames. are there different exceptions for a name issue vs. permissions vs. out of space? If so, I can try it. It just seemed clunky and wanted to know what other options I had.
|
|
#
?
Oct 14, 2008 03:03
|
|
|
Take a look at OSError and error values in errno. You can compare the exception's errno attribute to the values in the errno library. Also, if you try to call open() with a path containing a null byte, it'll raise a TypeError, at least on OS X. tef is right though, this isn't something you can know with great certainty in advance. For example, HFS+ on OS X will percent-encode invalid characters in filenames. ext3 will allow pretty much anything, even paths invalid for the file system's encoding. And sys.getfilesystemencoding() isn't always accurate when you have multiple file systems. I think what you do depends on your application. If you're presenting a save file dialog, you should try the operation, and if it fails, present an error to the user. If you're trying to save arbitrary filenames for something like a database of files, it might be better to encode the filenames beforehand.
|
|
#
?
Oct 14, 2008 04:22
|
|
|
I'm using MySQLdb to fetch many rows from a table, and all of them contain many fields including DATETIME columns. code:Both cjson and python-json throw an error at it. What I'd need is the string version of it so it can be jsonised properly. I'm looking for a way to either have MySQLdb automatically format DATETIME columns when returning them, or to efficiently go through the tuple of dicts and call .__str__() or str() on the objects. Or have a json writer/encoder do it for me. Is this doable or am I going to need to change the way I store poo poo to use unix timestamps instead?
|
|
#
?
Oct 14, 2008 05:17
|
|
|
There's no standard format for date/times in JSON. You'll need to decide what format to use (ISO 8601 is a good choice) and then override serialization in your JSON encoder for that type. Here's an encoder for simplejson: code:TOO SCSI FOR MY CAT fucked around with this message at 05:58 on Oct 14, 2008 |
|
#
?
Oct 14, 2008 05:55
|
|
|
Milde posted:I think what you do depends on your application. If you're presenting a save file dialog, you should try the operation, and if it fails, present an error to the user. If you're trying to save arbitrary filenames for something like a database of files, it might be better to encode the filenames beforehand. a re-write of my tv-episode bulk renaming app* that will rename files using a user-defined template and information (eg, show name, show title) that has a decent chance of containing illegal characters. And I haven't even accounted for unicode (specifically foreign characters) yet. My original plan was to encode the filenames beforehand, probably with the most common characters removed, and avoid any human interaction as much as possible. It may not be smart to avoid it in this case. I just want need to come up with a smart way of not forcing too much interaction. For example I really don't want the user to manually correct 60 instances of Avatar: The Last Airbender. Thanks for responses, and if there are any suggestions, I'd love to hear it. * When I'm happy that it's not devoid of anything resembling best-practices I'd love to do a code-review type thing. This has been my learn-python project for over a 6 months now.
|
|
#
?
Oct 14, 2008 14:44
|
|
|
Lurchington posted:a re-write of my tv-episode bulk renaming app* that will rename files using a user-defined template and information (eg, show name, show title) that has a decent chance of containing illegal characters. And I haven't even accounted for unicode (specifically foreign characters) yet. A dead simple way is to just replace all illegal characters with _'s with a regex. Who cares about encoding if the result is going to be Avatarx80 the last airbender or whatever. Avatar_the_last_airbender is going to work pretty much anywhere.
|
|
#
?
Oct 14, 2008 14:56
|
|
|
|
| # ? May 12, 2024 06:50 |
|
|
Which I'm fine with doing as long as I know which characters are illegal, which I thought was the issue. I need to investigate OSError, but will that report which characters are illegal? Is there a list of characters that are illegal in at least one filesystem? I'm fine with losing some possibly legit characters in the crossfire for at least the first go.
|
|
#
?
Oct 14, 2008 15:07
|
|