def set_prefixes(html,options) :
def _handle_prefix(node) :
if node.hasAttribute("prefix") :
for pref in node.getAttribute("prefix").strip().split() :
# this should be of the format prefix=uri
spec = pref.split('=')
if len(spec) >= 2 :
node.setAttributeNS("","xmlns:%s" % spec[0],spec[1])
return False
# We have to collect the current bnode id-s from the file to avoid conflicts
traverse_tree(html,_handle_prefix)
def set_prefixes(html, options):
def _handle_prefix(node):
if node.hasAttribute("prefix"):
for pref in node.getAttribute("prefix").strip().split():
# this should be of the format prefix=uri
spec = pref.split('=')
if len(spec) >= 2:
node.setAttributeNS("", "xmlns:%s" % spec[0], spec[1])
return False
# We have to collect the current bnode id-s from the file to avoid conflicts
traverse_tree(html, _handle_prefix)
def decorate_li_s(html, options):
"""
The main transformer entry point. See the module description for details.
@param html: a DOM node for the top level html element
@param options: invocation options
@type options: L{Options<pyRdfa.Options>}
"""
_bids = []
def _collect_BIDs(node):
"""Check and collect the possible bnode id-s in the file that might occur in CURIE-s. The
function is called recursively on each node. The L{_bids} variable is filled with the initial values.
@param node: a DOM element node
"""
def _suspect(val):
if len(val) > 1:
if val[0] == "_" and val[1] == ":":
if not val in _bids: _bids.append(val)
elif val[0] == "[" and val[-1] == "]":
_suspect(val[1:-1])
for a in ["about", "resource", "typeof"]:
if node.hasAttribute(a): _suspect(node.getAttribute(a))
return False
def _give_BID():
"""Generate a new value that can be used as a bnode id...
@return: a string of the form _:XXXX where XXXX is unique (ie, not yet stored in the L{_bids} array).
"""
while True:
i = random.randint(1, 10000)
val = "_:x%s" % i
if not val in _bids:
_bids.append(val)
return val
def _check_if_hit(node, rdfprefix):
"""
Check if the node has one of the C{typeof} values that would trigger the transformation.
@param node: DOM node (standing for a C{<ul>} or a C{<ol>})
@param rdfprefix: prefix to be used for the RDF namespace
@return: the value of _CONTAINER, _COLLECTION, or _NONE
"""
if node.hasAttribute("typeof"):
types = node.getAttribute("typeof").split()
for t in types:
# check if it is a namespaces thing at all...
if t.find(":") != -1:
key = t.split(":", 1)[0]
lname = t.split(":", 1)[1]
if key == rdfprefix:
if lname in ["Seq", "Alt", "Bag"]:
return _CONTAINER
elif lname in ["List"]:
return _COLLECTION
return _NONE
def _decorate_container(node, rdfprefix):
"""Take care of containers (ie, Seq, Alt, and Bag).
@param node: the node for the C{<ul>/<ol>}
@param rdfprefix: the prefix of the RDF namespace
"""
index = 1
originalNodes = [
n for n in node.childNodes
if n.nodeType == node.ELEMENT_NODE and n.tagName == "li"
]
for n in originalNodes:
pr = "%s:_%s" % (rdfprefix, index)
index += 1
if not _has_one_of_attributes(n, "href", "resource", "typeof",
"about", "rel", "rev", "property"):
# the simple case...
n.setAttribute("property", pr)
else:
# the original node should not be changed, but should be reparanted into a new
# enclosure
newEnclosure = node.ownerDocument.createElement("div")
newEnclosure.setAttribute("rel", pr)
node.replaceChild(newEnclosure, n)
newEnclosure.appendChild(n)
def _decorate_collection(node, rdfprefix):
"""Take care of collection (a.k.a. Lists).
@param node: the node for the C{<ul>/<ol>}
@param rdfprefix: the prefix of the RDF namespace
"""
List = "%s:List" % rdfprefix
first = "%s:first" % rdfprefix
rest = "%s:rest" % rdfprefix
nil = "[%s:nil]" % rdfprefix
rtype = "%s:type" % rdfprefix
# the list of 'li'-s is needed in advance (eg, for their numbers)
originalNodes = [
n for n in node.childNodes
if n.nodeType == node.ELEMENT_NODE and n.tagName == "li"
]
# a bnode id should be generated for the top level node
if node.hasAttribute("about"):
currId = node.getAttribute("about")
else:
currId = "[%s]" % _give_BID()
node.setAttribute("about", currId)
index = 1
for i in xrange(0, len(originalNodes)):
n = originalNodes[i]
# first the current <li> must be massaged
if not _has_one_of_attributes(n, "href", "resource", "typeof",
"about", "rel", "rev", "property"):
# the simple case, the node is changed in situ..
n.setAttribute("about", currId)
n.setAttribute("property", first)
else:
# an enclosure for that node should be created, and the original node
# is just reparented
newEnclosure = node.ownerDocument.createElement("div")
newEnclosure.setAttribute("rel", first)
newEnclosure.setAttribute("about", currId)
node.replaceChild(newEnclosure, n)
newEnclosure.appendChild(n)
# An extra <li> is necessary to add some additional info...
newLi = node.ownerDocument.createElement("li")
newLi.setAttribute("about", currId)
newLi.setAttribute("rel", rest)
if i != 0: newLi.setAttribute("typeof", List)
if i == len(originalNodes) - 1:
# This is the last element
newLi.setAttribute("resource", nil)
node.appendChild(newLi)
else:
newId = "[%s]" % _give_BID()
newLi.setAttribute("resource", newId)
currId = newId
node.insertBefore(newLi, originalNodes[i + 1])
uls = [n for n in html.getElementsByTagName("ul")]
ols = [n for n in html.getElementsByTagName("ol")]
if len(uls) == 0 and len(ols) == 0: return
#--------------------------------------------------------------------------------
# We have to extract the prefix used for rdf. It may not be 'rdf'...
rdfprefix = rdf_prefix(html)
#--------------------------------------------------------------------------------
# We have to collect the current bnode id-s from the file to avoid conflicts
traverse_tree(html, _collect_BIDs)
# We will need the random function to generate unique bid-s
random.seed(None)
#--------------------------------------------------------------------------------
for node in uls + ols:
# check if this is one of those guys...
t = _check_if_hit(node, rdfprefix)
if t == _CONTAINER:
_decorate_container(node, rdfprefix)
elif t == _COLLECTION:
_decorate_collection(node, rdfprefix)
def decorate_li_s(html,options) :
"""
The main transformer entry point. See the module description for details.
@param html: a DOM node for the top level html element
@param options: invocation options
@type options: L{Options<pyRdfa.Options>}
"""
_bids = []
def _collect_BIDs(node) :
"""Check and collect the possible bnode id-s in the file that might occur in CURIE-s. The
function is called recursively on each node. The L{_bids} variable is filled with the initial values.
@param node: a DOM element node
"""
def _suspect(val) :
if len(val) > 1 :
if val[0] == "_" and val[1] == ":" :
if not val in _bids : _bids.append(val)
elif val[0] == "[" and val[-1] == "]" :
_suspect(val[1:-1])
for a in ["about","resource","typeof"] :
if node.hasAttribute(a) : _suspect(node.getAttribute(a))
return False
def _give_BID() :
"""Generate a new value that can be used as a bnode id...
@return: a string of the form _:XXXX where XXXX is unique (ie, not yet stored in the L{_bids} array).
"""
while True :
i = random.randint(1,10000)
val = "_:x%s" % i
if not val in _bids :
_bids.append(val)
return val
def _check_if_hit(node,rdfprefix) :
"""
Check if the node has one of the C{typeof} values that would trigger the transformation.
@param node: DOM node (standing for a C{<ul>} or a C{<ol>})
@param rdfprefix: prefix to be used for the RDF namespace
@return: the value of _CONTAINER, _COLLECTION, or _NONE
"""
if node.hasAttribute("typeof") :
types = node.getAttribute("typeof").split()
for t in types :
# check if it is a namespaces thing at all...
if t.find(":") != -1 :
key = t.split(":",1)[0]
lname = t.split(":",1)[1]
if key == rdfprefix :
if lname in ["Seq","Alt","Bag"] :
return _CONTAINER
elif lname in ["List"] :
return _COLLECTION
return _NONE
def _decorate_container(node,rdfprefix) :
"""Take care of containers (ie, Seq, Alt, and Bag).
@param node: the node for the C{<ul>/<ol>}
@param rdfprefix: the prefix of the RDF namespace
"""
index = 1
originalNodes = [n for n in node.childNodes if n.nodeType == node.ELEMENT_NODE and n.tagName == "li" ]
for n in originalNodes :
pr = "%s:_%s" % (rdfprefix,index)
index += 1
if not _has_one_of_attributes(n,"href","resource","typeof","about","rel","rev","property") :
# the simple case...
n.setAttribute("property",pr)
else :
# the original node should not be changed, but should be reparanted into a new
# enclosure
newEnclosure = node.ownerDocument.createElement("div")
newEnclosure.setAttribute("rel",pr)
node.replaceChild(newEnclosure,n)
newEnclosure.appendChild(n)
def _decorate_collection(node,rdfprefix) :
"""Take care of collection (a.k.a. Lists).
@param node: the node for the C{<ul>/<ol>}
@param rdfprefix: the prefix of the RDF namespace
"""
List = "%s:List" % rdfprefix
first = "%s:first" % rdfprefix
rest = "%s:rest" % rdfprefix
nil = "[%s:nil]" % rdfprefix
rtype = "%s:type" % rdfprefix
# the list of 'li'-s is needed in advance (eg, for their numbers)
originalNodes = [ n for n in node.childNodes if n.nodeType == node.ELEMENT_NODE and n.tagName == "li" ]
# a bnode id should be generated for the top level node
if node.hasAttribute("about") :
currId = node.getAttribute("about")
else :
currId = "[%s]" % _give_BID()
node.setAttribute("about",currId)
index = 1
for i in xrange(0,len(originalNodes)) :
n = originalNodes[i]
# first the current <li> must be massaged
if not _has_one_of_attributes(n,"href","resource","typeof","about","rel","rev","property") :
# the simple case, the node is changed in situ..
n.setAttribute("about",currId)
n.setAttribute("property",first)
else :
# an enclosure for that node should be created, and the original node
# is just reparented
newEnclosure = node.ownerDocument.createElement("div")
newEnclosure.setAttribute("rel",first)
newEnclosure.setAttribute("about",currId)
node.replaceChild(newEnclosure,n)
newEnclosure.appendChild(n)
# An extra <li> is necessary to add some additional info...
newLi = node.ownerDocument.createElement("li")
newLi.setAttribute("about",currId)
newLi.setAttribute("rel",rest)
if i != 0 : newLi.setAttribute("typeof",List)
if i == len(originalNodes) - 1 :
# This is the last element
newLi.setAttribute("resource",nil)
node.appendChild(newLi)
else :
newId = "[%s]" % _give_BID()
newLi.setAttribute("resource",newId)
currId = newId
node.insertBefore(newLi,originalNodes[i+1])
uls = [ n for n in html.getElementsByTagName("ul") ]
ols = [ n for n in html.getElementsByTagName("ol") ]
if len(uls) == 0 and len(ols) == 0 : return
#--------------------------------------------------------------------------------
# We have to extract the prefix used for rdf. It may not be 'rdf'...
rdfprefix = rdf_prefix(html)
#--------------------------------------------------------------------------------
# We have to collect the current bnode id-s from the file to avoid conflicts
traverse_tree(html,_collect_BIDs)
# We will need the random function to generate unique bid-s
random.seed(None)
#--------------------------------------------------------------------------------
for node in uls+ols :
# check if this is one of those guys...
t = _check_if_hit(node,rdfprefix)
if t == _CONTAINER :
_decorate_container(node,rdfprefix)
elif t == _COLLECTION :
_decorate_collection(node,rdfprefix)
