Python CGI: Translate from stdout to html (e.g. replace \n with <BR>)

I have a great command line program that prints to stdout. I want to let people run this through CGI. In the CGI version, I want to have the same output (with "<BR>" instead of "\n").

My cgi_main works by calling my real_main; there are a lot of options to process and it was bad design to do that multiple times. But since real_main calls print, I don't know how to change "\n" to "<BR>" post hoc.

I could easily change all instances of print to go to a tempfile and then dump that at the end, but it seems inelegant to change the existing command line version so much.

One option is to have real_main call another function (passing along argv); that new function returns values. Then real_main prints those values. cgi_main would call the same function, but print the values differently.

Another option is to use a class member/static function everywhere I use print now. This class could be told whether to reformat or not. Then switching between command line and CGI output would just require changing the reform argument.

Here is an example of how I thought this class and function could work: class cgi_tools: def init(self, reform = False): self.reform = reform def myprint(self, *args): if not self.reform: print(args) for i,x in enumerate(args): if type(x) == str: self.myprint(x.replace('\n', '

')), elif type(x) in atoms: # check if x is an atom (i.e. int, float, etc.): print x, else: self.myprint(x), if i != len(args)-1: print ' ', print ''

With this I ran into trouble trying to think of all atoms (i.e. irreducible objects) in Python. I couldn't find a built-in function for this test, and I thought writing it myself was a brittle idea.

I tried to use insepct to print the source code of print, but I couldn't do that! print sys.out.print:

import inspect
print inspect.getsource(print) # this doe开发者_运维百科s not work!

All of this seems like a classic Python CGI question, but I couldn't find a great solution. I know how to patch things together in a sloppy way and get what I want, but I would love your advice on how to be elegant. I want to grow, and I don't want to program sloppily anymore.

Thanks a lot for your thoughts and advice.<br> -Oliver

At the start of your CGI script, put print "Content-type: text/plain\r\n\r\n". The result will show up on a webpage the same as on a console.

In general, if I want to have two variations on a program, one that formats output in HTML and the other that outputs plain text, I abstract as much of the logic as I can into routines or objects that do no output, then have separate output routines for the variations. See for example http://jcomeau.unternet.net/src/colorforth/as/cf2text, which is symlinked to cf2html and cf2ansi; the same program produces 3 different outputs depending on which symlink is called. Note that this is old code and I don't program the same way any more, but maybe you'll get some useful ideas from it; maybe not.

Actually, looking over the link I gave you, I used a different strategy there, using a global to indicate the desired output style. I don't have any recent open-source code with a better example to give you, it's all owned by my clients.

Here's one way to modify your output without changing your print statements:

>>> import sys, StringIO
>>> oldout = sys.stdout
>>> sys.stdout = newout = StringIO.StringIO()
>>> for line in 'this', 'that', 'the other': print line
... 
>>> sys.stdout = oldout
>>> print newout.getvalue().replace('\n', '<br />\n')
this<br />
that<br />
the other<br />

If you use this method, then you can change your content-type to text/html and use any other HTML functionality. For more complex modifications of the output you can use regular expression substitution. And if you want the output as lines, you can:

>>> newout.seek(0)
>>> newout.readlines()
['this\n', 'that\n', 'the other\n']