views:

110

answers:

2

Assume that I have some HTML code, like this (generated from Markdown or Textile or something):

<h1>A header</h1>
<p>Foo</p>
<h2>Another header</h2>
<p>More content</p>
<h2>Different header</h2>
<h1>Another toplevel header
<!-- and so on -->

How could I generate a table of contents for it using Python?

+6  A: 

Use an HTML parser such as lxml or BeautifulSoup to find all header elements.

Ignacio Vazquez-Abrams
how do they compare?
van
@van: BeautifulSoup is pure Python but not very 3.x-compatible. lxml is good, but requires a C compiler to build.
Ignacio Vazquez-Abrams
I was really looking more for some example code, but I think I have figured it out. I'll probably use lxml, since (a) it's useful for more stuff, and (b) having a module that looks like a class name disturbs my sense of aesthetics.
LeafStorm
+1  A: 

Here's an example using lxml and xpath.

from lxml import etree
doc = etree.parse("test.xml")
for node in doc.xpath('//h1|//h2|//h3|//h4|//h5'):
    print node.tag, node.text
kloffy