Any python HTML generator libs?

[Matt Goodall]

Steve Holden wrote:
> Sullivan WxPyQtKinter wrote:
> 
>>Hi, everyone.  Simply put, what I need most now is a python lib to
>>generate simple HTML.
>>
>>I am now using XML to store my lab report records. I found python
>>really convinient to manipulate XML, so I want to make a small on-line
>>CGI program to help my colleagues to build their lab report records
>>into XML, for storage, HTML display (for others to browse) and search.
>>
>>With python's standard lib, the DOM object could realize the XML
>>storage and search quite easily, but for HTML generation,  it is a
>>great headache.
>>
>>I tried to turn to the in-line server-side python script PSP(something
>>like asp and php) instead of CGI. However, since the report data is
>>always of very complex data structures, it is really hard to write most
>>things in-line. For example, a PCR reaction is expected to be shown in
>>this format or alike on a web browser:
>>
>>PCR
>>        Sample: Sm1032
>>        Operater: SullivanZ
>>        TimeStamp: hh:mm mm-dd-yyyy
>>        Reaction:
>>                       Reagent1:
>>                                         Name:XXXX
>>                                         Concentration:XXXX mM
>>                                         Volumn:XXX uL
>>                       Reagent2:
>>........................
>>........................
>>
>>Since there are hundreds of PCR reaction and other operations in the
>>lab report, in-line PSP is not a suitable solution. But writing HTML
>>directly with print statement is a great pain.
>>
>>Will XSTL be useful? Is my problem somewho related with XML-SIG?
>>Looking forward to your precious suggestion.
>>
> 
> The triple-quoted string with string substitution is your friend. Try 
> writing something(s) like:
> 
> results = {'secnum': 1, 'type': 'string', 'color': 'blue'}
> 
> print """\
> <h1>Section %(secnum)s</h1>
> <p>Elements of type %(type)s should be coloured %(color)s</p>
> """ % results

Don't forget that you may need to escape the application's data for
inclusion in HTML:

    results = {'secnum': 1, 'type': 'string', 'color': 'blue',
            'user':'Matt Goodall <matt at pollenation.net>'}

    print """\
    <h1>Section %(secnum)s</h1>
    <p>Elements of type %(type)s should be coloured %(color)s</p>
    <p>Contributed by: %(user)s</p>
    """ % results

Will print:

<h1>Section 1</h1>
<p>Elements of type string should be coloured blue</p>
<p>Contributed by: Matt Goodall <matt at pollenation.net></p>

The '<' and '>' surrounding my email address breaks the HTML.

To fix that you need to escape results['user'] with cgi.escape or
xml.sax.saxutils.escape. Oh, and don't forget to escape anything
destined for an HTML attribute differently, see sax.saxutils.quoteattr.

A triple-quoted string is beautifully simple but it's not quite as much
a friend as it might initially seem. ;-)

I don't intend to get into a XML- vs text- based templating flame war
;-) but, IMHO, the solution is to use a templating language that
understands where the value is used in the template.

Kid is a great example of an XML-based templating language but there are
many others. Some have probably been mentioned in this thread already.

Another interesting solutions is to use something like Nevow's tags module:

    from nevow import flat, tags as T

    results = {'secnum': 1, 'type': 'string', 'color': 'blue',
            'user':'Matt Goodall <matt at pollenation.net>'}

    doc = T.div[
        T.h1['Section ', results['secnum']],
        T.p['Elements of type ', results['type'],
            ' should be coloured ', results['color']],
        T.p['Contributed by: ', results['user']],
        ]

    print flat.flatten(doc)

This time you get valid HTML with no effort whatsoever:

<div><h1>Section 1</h1><p>Elements of type string should be coloured
blue</p><p>Contributed by: Matt Goodall
<matt at pollenation.net></p></div>

You even get to write HTML in a slightly more Pythonic way (even if it
does abuse Python just a little :wink:), but Nevow will happily load a
template containing actual XHTML from disk if you prefer.

The only real "problem" using Nevow for this is that you will need to
install Twisted too. I suspect you'll find a couple of Nevow tag
implementations that don't need Twisted if you ask Google.

Anyway! This was just to demonstrate an alternate approach than to
evangelise about Nevow. I hope it was at least interesting. :)

Cheers, Matt

Nevow:
	http://divmod.org/trac/wiki/DivmodNevow
Twisted:
	http://twistedmatrix.com/trac

-- 
     __
    /  \__     Matt Goodall, Pollenation Internet Ltd
    \__/  \    w: http://www.pollenation.net
  __/  \__/    e: matt at pollenation.net
 /  \__/  \    t: +44 (0)113 2252500
 \__/  \__/
 /  \          Any views expressed are my own and do not necessarily
 \__/          reflect the views of my employer.