Cria um novo elemento ou substitui um existente,
com o texto de outro elemento, truncado com X caracteres.
Baseado no filtro excerpt.py e alterado por Duarte Carreira em 16/Julho/2009.

Parameters:
  width:  maximum number of characters in the excerpt.  Default: 500
  omit:   whitespace delimited list of html tags to remove.  Default: none
  target: name of element created.  Default: content
  source: name of element to get data from. Default: summary
  replace: yes to delete duplicate target. Default: yes

Example to test:
python tests/reconstitute.py http://localhost/feedorig.xml
--filters "planetaSIG.py?width=500&source=content&target=content&replace=yes">tes
te3.xml

Notes:
* if you want to avoid duplicate entries use replace=yes.
* Venus does not expose summary in the feeds to tmpl templates. With this filter,
   you cant replace the text inside content with the text from summary. This is
   what the default values do.
* if 'img' is in the list of tags to be omitted <img> tags are replaced with
   hypertext links associated with the value of the 'alt' attribute.  If there
   is no alt attribute value, <img> is used instead.  If the parent element
   of the img tag is already an <a> tag, no additional hypertext links are
   added.
"""

import sys, xml.dom.minidom, textwrap
from xml.dom import Node, minidom

atomNS = 'http://www.w3.org/2005/Atom'
planetNS = 'http://planet.intertwingly.net/'

args = dict(zip([name.lstrip('-') for name in sys.argv[1::2]], sys.argv[2::2]))

wrapper = textwrap.TextWrapper(width=int(args.get('width','500')))
omit = args.get('omit', '').split()
target = args.get('target', 'content')
original = args.get('source', 'summary')
replace = args.get('replace','yes')

class copy:
    """ recursively copy a source to a target, up to a given width """

    def __init__(self, dom, source, target):
        self.dom = dom
        self.full = False
        self.text = []
        self.textlen = 0
        self.copyChildren(source, target)

    def copyChildren(self, source, target):
        """ copy child nodes of a source to the target """
        for child in source.childNodes:
            if child.nodeType == Node.ELEMENT_NODE:
                 self.copyElement(child, target)
            elif child.nodeType == Node.TEXT_NODE:
                 self.copyText(child.data, target)
            if self.full: break

    def copyElement(self, source, target):
        """ copy source element to the target """

        # check the omit list
        if source.nodeName in omit:
            if source.nodeName == 'img':
               return self.elideImage(source, target)
            return self.copyChildren(source, target)

        # copy element, attributes, and children
        child = self.dom.createElementNS(source.namespaceURI, source.nodeName)
        target.appendChild(child)
        for i in range(0, source.attributes.length):
            attr = source.attributes.item(i)
            child.setAttributeNS(attr.namespaceURI, attr.name, attr.value)
        self.copyChildren(source, child)

    def elideImage(self, source, target):
        """ copy an elided form of the image element to the target """
        alt = source.getAttribute('alt') or '<img>'
        src = source.getAttribute('src')

        if target.nodeName == 'a' or not src:
            self.copyText(alt, target)
        else:
            child = self.dom.createElement('a')
            child.setAttribute('href', src)
            self.copyText(alt, child)
            target.appendChild(child)

    def copyText(self, source, target):
        """ copy text to the target, until the point where it would wrap """
        if not source.isspace() and source.strip():
            self.text.append(source.strip())
        lines = wrapper.wrap(' '.join(self.text))
        if len(lines) == 1:
            target.appendChild(self.dom.createTextNode(source))
            self.textlen = len(lines[0])
        elif lines:
            excerpt = source[:len(lines[0])-self.textlen] + u' \u2026'
            target.appendChild(dom.createTextNode(excerpt))
            self.full = True

# select summary or content element
dom = minidom.parse(sys.stdin)

#source = dom.getElementsByTagNameNS(atomNS, 'summary')
#if not source:
#    source = dom.getElementsByTagNameNS(atomNS, 'content')
source = dom.getElementsByTagNameNS(atomNS, original)

# if present, recursively copy it to a planet:excerpt element
if source:
    fonteelem = source[0]
    if target.startswith('planet:'):
        dom.documentElement.setAttribute('xmlns:planet', planetNS)
    if target.startswith('atom:'): target = target.split(':',1)[1]
    excerpt = dom.createElementNS(planetNS, target)
    source[0].parentNode.appendChild(excerpt)
    copy(dom, source[0], excerpt)
    #source[0].parentNode.replaceChild(excerpt, source[0])
    #if source[0].nodeName == excerpt.nodeName:
    #  source[0].parentNode.removeChild(source[0])

#apagar o original
if replace == 'yes':
    source = dom.getElementsByTagName(target)
    fonteelem = source[0]
    if len(source)>1:
        source[0].parentNode.removeChild(source[0])

# print out results
print dom.toxml('utf-8')