github/pelican
1
0
Fork 1
mirror of https://github.com/getpelican/pelican.git synced 2025-10-15 20:28:56 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

Added support for HTML as input format

Browse source
This commit is contained in:
skami 2011-02-14 19:10:01 +01:00
commit c019443026
1 changed files with 19 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

20
pelican/readers.py
View file

@ -59,7 +59,25 @@ class MarkdownReader(object):
)(value[0]) )(value[0])
return content, metadatas return content, metadatas
_EXTENSIONS = {'rst': RstReader, 'md': MarkdownReader} # supported formats
class HtmlReader(object):
_re = re.compile('\<\!\-\-\#\s?[A-z0-9_-]*\s?\:s?[A-z0-9\s_-]*\s?\-\-\>')
def read(self, filename):
"""Parse content and metadata of (x)HTML files"""
content = open(filename)
metadatas = {'title':'unnamed'}
for i in self._re.findall(content):
key = i.split(':')[0][5:].strip()
value = i.split(':')[-1][:-3].strip()
print [key,value]
metadatas[key.lower()] = value
return content, metadatas
_EXTENSIONS = {'rst': RstReader, 'md': MarkdownReader, 'html': HtmlReader} # supported formats
def read_file(filename, fmt=None): def read_file(filename, fmt=None):