[Baypiggies] Discussion for newbies/beginner night talks

Sat Feb 10 18:41:25 CET 2007

On Sat, 10 Feb 2007 08:14:16 -0800, Paul McNett <p at ulmcnett.com> wrote:

> Aahz wrote:
>> On Fri, Feb 09, 2007, Dennis Reinhardt wrote:
>>> You illustration does not show how a long triple quoted string gets
>>> supplied with parameters.  Stylistically, I often prefer substituting
>>> parameters on a line by line basis because there is better locality,  
>>> making
>>> the code easier to understand and read.
>
> How about this contrived example:
>
> def getHeader(self, **kwargs):
> 	html = """<!-- BOF header -->
>
> <html>
> <head>
> 	<title>%(title)s</title>
> 	<meta name="description" value="%(title)s">
> 	<meta name="keywords" value="%(keywords)s">
> </head>
>
> <body>
> %(navigation_bar)s
>
> <!-- EOF header -->
> """
> 	return html % kwargs
>
>
>> Strings like SQL and HTML that do not normally get presented to the user
>> don't need to care about whitespace AT ALL.
>
> I think that with longish triple-quoted strings, ignoring the code
> indentation is the best. But that's me.

I'm with you. I often embed long fragement of SQL or HTML into Python  
source code. I can easily read the embedded text fragement as a whole  
without a lot of quotes inserted if I would have to break them into  
individual lines. Even better, it avoids the obfuscation of escaping the "  
character with \". I think triple quote is one of the feature I like most  
in Python.

One caveat - since you use % to format the arguments, the individual %  
characters in the text block would have to escaped. Something you'd likely  
to encounter when embedding CSS.

Wai Yip