#!/usr/bin/python
"""CGI script to produce RSS feeds of New York Times headlines

Usage:
- Install as CGI script (e.g. http://mydomain/cgi-bin/nyt.cgi)
- In your RSS readers, subscribe to individual feeds like this:
  http://mydomain/cgi-bin/nyt.cgi?section=technology
  http://mydomain/cgi-bin/nyt.cgi?section=business
  etc.
- Other sections: national, arts, opinion, many others
"""

from xml.dom.minidom import Document, Element, Text

class xElement(Element):
    def __init__(self, *children):
        Element.__init__(self, str(self.__class__).split(".")[-1])
        apply(self.appendChildren, children)

    def appendChildren(self, *children):
        if children:
            for child in children:
                try:
                    self.appendChild(child)
                except AttributeError:
                    for node in child:
                        self.appendChild(node)

class RSSDocument(Document):
    def __init__(self, root):
        Document.__init__(self)
        self.appendChild(root)
        
class rss(xElement):
    def __init__(self, *channels):
        apply(xElement.__init__, (self,) + channels)
        self.attributes["version"] = "0.92"

class channel(xElement): pass
class item(xElement): pass

class xSimpleText(xElement):
    def __init__(self, text):
        xElement.__init__(self, Text(text))

class title(xSimpleText): pass
class link(xSimpleText): pass
class description(xSimpleText): pass

def getURLData(url):
    import urllib
    usock = urllib.urlopen(url)
    html = usock.read()
    usock.close()
    return html
    
def readChannelAsXML(name):
    return getURLData('http://radiouser:Csm!]-tvMm@partners.userland.com/nyt/%s.xml' % name)

def parseRawChannelData(rawxml):
    rawxml = readChannelAsXML(name)
    from xml.dom import minidom
    return minidom.parseString(rawxml)

def getNodeText(node):
    rc = ""
    for child in node.childNodes:
        rc = rc + child.data
    return rc

def extractRelevantBitsFromParsedChannelData(xmldoc):
    return [(getNodeText(node.getElementsByTagName("headline")[0]),
             getNodeText(node.getElementsByTagName("url")[0]),
             getNodeText(node.getElementsByTagName("abstract")[0]))
            for node in xmldoc.getElementsByTagName("article")]

if __name__ == '__main__':
    import cgi
    form = cgi.FieldStorage()
    name = form["section"].value
    bits = extractRelevantBitsFromParsedChannelData(parseRawChannelData(readChannelAsXML(name)))
    channelNode = channel(title('New York Times %s' % name.capitalize()))
    for t, l, d in bits:
        channelNode.appendChild(item(title(t), link(l), description(d)))
    rssDocument = RSSDocument(rss(channelNode))
    output = rssDocument.toxml()
    print "Content-type: text/xml"
    print "Content-length: %s" % len(output)
    print
    print output