--- documentViewer/MpdlXmlTextServer.py	2011/06/14 09:57:11	1.238
+++ documentViewer/MpdlXmlTextServer.py	2011/07/15 19:34:41	1.238.2.2
@@ -1,15 +1,83 @@
 
 from OFS.SimpleItem import SimpleItem
 from Products.PageTemplates.PageTemplateFile import PageTemplateFile 
+
 from Ft.Xml import EMPTY_NAMESPACE, Parse
 from Ft.Xml.Domlette import NonvalidatingReader
+import Ft.Xml.Domlette
+import cStringIO
+
+import xml.etree.ElementTree as ET
 
-import md5
-import sys
+import re
 import logging
 import urllib
 import documentViewer
-from documentViewer import getTextFromNode, serializeNode
+#from documentViewer import getTextFromNode, serializeNode
+
+def intOr0(s, default=0):
+    """convert s to int or return default"""
+    try:
+        return int(s)
+    except:
+        return default
+
+def getText(node):
+    """get the cdata content of a node"""
+    if node is None:
+        return ""
+    # ET:
+    text = node.text or ""
+    for e in node:
+        text += gettext(e)
+        if e.tail:
+            text += e.tail
+
+    return text
+
+def serialize(node):
+    """returns a string containing an XML snippet of node"""
+    s = ET.tostring(node, 'UTF-8')
+    # snip off XML declaration
+    if s.startswith('<?xml'):
+        i = s.find('?>')
+        return s[i+3:]
+
+    return s
+
+
+def getTextFromNode(node):
+    """get the cdata content of a node"""
+    if node is None:
+        return ""
+    # ET:
+#    text = node.text or ""
+#    for e in node:
+#        text += gettext(e)
+#        if e.tail:
+#            text += e.tail
+
+    # 4Suite:
+    nodelist=node.childNodes
+    text = ""
+    for n in nodelist:
+        if n.nodeType == node.TEXT_NODE:
+           text = text + n.data
+    
+    return text
+
+def serializeNode(node, encoding="utf-8"):
+    """returns a string containing node as XML"""
+    #s = ET.tostring(node)
+    
+    # 4Suite:
+    stream = cStringIO.StringIO()
+    Ft.Xml.Domlette.Print(node, stream=stream, encoding=encoding)
+    s = stream.getvalue()
+    stream.close()
+
+    return s
+
 
 class MpdlXmlTextServer(SimpleItem):
     """TextServer implementation for MPDL-XML eXist server"""
@@ -21,8 +89,7 @@ class MpdlXmlTextServer(SimpleItem):
     
     manage_changeMpdlXmlTextServerForm = PageTemplateFile("zpt/manage_changeMpdlXmlTextServer", globals())
         
-    def __init__(self,id,title="",serverUrl="http://mpdl-system.mpiwg-berlin.mpg.de/mpdl/interface/", serverName=None, timeout=40):
-    #def __init__(self,id,title="",serverUrl="http://mpdl-text.mpiwg-berlin.mpg.de:30030/mpdl/interface/", serverName=None, timeout=40):    
+    def __init__(self,id,title="",serverUrl="http://mpdl-text.mpiwg-berlin.mpg.de/mpdl/interface/", serverName=None, timeout=40):
         
         """constructor"""
         self.id=id
@@ -42,8 +109,10 @@ class MpdlXmlTextServer(SimpleItem):
         url = self.serverUrl+method
         return documentViewer.getHttpData(url,data,timeout=self.timeout)
 
+    # WTF: what does this really do? can it be integrated in getPage?
     def getSearch(self, pageinfo=None,  docinfo=None):
         """get search list"""
+        logging.debug("getSearch()")
         docpath = docinfo['textURLPath'] 
         url = docinfo['url']
         pagesize = pageinfo['queryPageSize']
@@ -146,12 +215,12 @@ class MpdlXmlTextServer(SimpleItem):
         hrefList=[]
         myList= ""
         text=self.getServerData("xpath.xql", "document=%s&xpath=%s&pn=%s"%(docinfo['textURLPath'],xpath,pn))
-        dom = Parse(text)
-        result = dom.xpath("//result/resultPage/place")
+        dom = ET.fromstring(text)
+        result = dom.findall(".//result/resultPage/place")
         for l in result:
-            hrefNode= l.getAttributeNodeNS(None, u"id")
-            href= hrefNode.nodeValue
+            href = l.get("id")
             hrefList.append(href)
+            # WTF: what does this do?
             myList = ",".join(hrefList)
         #logging.debug("getGisPlaces :%s"%(myList))                             
         return myList
@@ -166,264 +235,223 @@ class MpdlXmlTextServer(SimpleItem):
         hrefList=[]
         myList=""
         text=self.getServerData("xpath.xql", "document=%s&xpath=%s"%(docinfo['textURLPath'],xpath))
-        dom =Parse(text)
-        result = dom.xpath("//result/resultPage/place")
+        dom = ET.fromstring(text)
+        result = dom.findall(".//result/resultPage/place")
         
         for l in result:
-            hrefNode = l.getAttributeNodeNS(None, u"id")
-            href= hrefNode.nodeValue
+            href = l.get("id")
             hrefList.append(href)
+            # WTF: what does this do?
             myList = ",".join(hrefList)
             #logging.debug("getALLGisPlaces :%s"%(myList))
         return myList
           
+    def processPageInfo(self, dom, docinfo, pageinfo):
+        """processes page info divs from dom and stores in docinfo and pageinfo"""
+        # process all toplevel divs
+        alldivs = dom.findall(".//div")
+        pagediv = None
+        for div in alldivs:
+            dc = div.get('class')
+            
+            # page content div
+            if dc == 'pageContent':
+                pagediv = div
+            
+            # pageNumberOrig  
+            elif dc == 'pageNumberOrig':
+                pageinfo['pageNumberOrig'] = div.text
+                
+            # pageNumberOrigNorm
+            elif dc == 'pageNumberOrigNorm':
+                pageinfo['pageNumberOrigNorm'] = div.text
+                
+            # pageNumberOrigNorm
+            elif dc == 'countFigureEntries':
+                docinfo['countFigureEntries'] = intOr0(div.text)
+                
+            # pageNumberOrigNorm
+            elif dc == 'countTocEntries':
+                # WTF: s1 = int(s)/30+1
+                docinfo['countTocEntries'] = intOr0(div.text)
+                
+            # numTextPages
+            elif dc == 'countPages':
+                np = intOr0(div.text)                    
+                if np > 0:
+                    docinfo['numTextPages'] = np
+                    if docinfo.get('numPages', 0) == 0:
+                        # seems to be text-only
+                        docinfo['numTextPages'] = np
+                        pageinfo['end'] = min(pageinfo['end'], np)
+                        pageinfo['numgroups'] = int(np / pageinfo['groupsize'])
+                        if np % pageinfo['groupsize'] > 0:
+                            pageinfo['numgroups'] += 1
+                            
+        return
+         
            
     def getTextPage(self, mode="text_dict", pn=1, docinfo=None, pageinfo=None):
         """returns single page from fulltext"""
+        logging.debug("getTextPage mode=%s, pn=%s"%(mode,pn))
+        # check for cached text -- but this shouldn't be called twice
+        if pageinfo.has_key('textPage'):
+            logging.debug("getTextPage: using cached text")
+            return pageinfo['textPage']
+        
         docpath = docinfo['textURLPath']
-        path = docinfo['textURLPath']
-        url = docinfo.get('url',None)
-        name = docinfo.get('name',None)
-        pn =pageinfo['current']
-        sn = pageinfo['sn']
-        #optionToggle =pageinfo ['optionToggle']
-        highlightQuery = pageinfo['highlightQuery']
-        #mode = pageinfo ['viewMode']
-        tocMode = pageinfo['tocMode']
-        characterNormalization=pageinfo['characterNormalization']
-        tocPN = pageinfo['tocPN']
-        selfurl = self.absolute_url()   
+        # just checking
+        if pageinfo['current'] != pn:
+            logging.warning("getTextPage: current!=pn!")
+            
+        # stuff for constructing full urls
+        url = docinfo['url']
+        urlmode = docinfo['mode']
+        sn = pageinfo.get('sn', None)
+        highlightQuery = pageinfo.get('highlightQuery', None)
+        tocMode = pageinfo.get('tocMode', None)
+        tocPN = pageinfo.get('tocPN',None)
+        characterNormalization = pageinfo.get('characterNormalization', None)
+        selfurl = docinfo['viewerUrl']   
+        
         if mode == "text_dict":
+            # text_dict is called textPollux in the backend
             textmode = "textPollux"
         else:
             textmode = mode
         
         textParam = "document=%s&mode=%s&pn=%s&characterNormalization=%s"%(docpath,textmode,pn,characterNormalization)
-        if highlightQuery is not None:
+        if highlightQuery:
             textParam +="&highlightQuery=%s&sn=%s"%(urllib.quote(highlightQuery),sn)           
         
+        # fetch the page
         pagexml = self.getServerData("page-fragment.xql",textParam)
-        dom = Parse(pagexml)
-        #dom = NonvalidatingReader.parseStream(pagexml)
-        
-        #original Pages
-        pagedivs = dom.xpath("//div[@class='pageNumberOrig']")       
-        
-        """if pagedivs == dom.xpath("//div[@class='pageNumberOrig']"):
-            if len(pagedivs)>0:
-                docinfo['pageNumberOrig']= getTextFromNode(pagedivs[0])
-                logging.debug("ORIGINAL PAGE: %s"%(docinfo['pageNumberOrig']))
-        
-        #original Pages Norm
-        pagedivs = dom.xpath("//div[@class='pageNumberOrigNorm']")
-        if pagedivs == dom.xpath("//div[@class='pageNumberOrigNorm']"):
-            if len(pagedivs)>0:
-                docinfo['pageNumberOrigNorm']= getTextFromNode(pagedivs[0])
-                logging.debug("ORIGINAL PAGE NORM: %s"%(docinfo['pageNumberOrigNorm']))
-        """
-        #figureEntries
-        pagedivs = dom.xpath("//div[@class='countFigureEntries']")
-        if pagedivs == dom.xpath("//div[@class='countFigureEntries']"):
-            if len(pagedivs)>0:
-                docinfo['countFigureEntries'] = getTextFromNode(pagedivs[0])
-                s = getTextFromNode(pagedivs[0])
-                if s=='0':
-                    try:
-                        docinfo['countFigureEntries'] = int(s)
-                    except:
-                        docinfo['countFigureEntries'] = 0
-                else:
-                    s1 = int(s)/30+1
-                    try:
-                        docinfo['countFigureEntries'] = int(s1)
-                    except:
-                        docinfo['countFigureEntries'] = 0      
-        
-        #allPlaces
-        pagedivs = dom.xpath("//div[@class='countPlaces']")
-        if pagedivs == dom.xpath("//div[@class='countPlaces']"):
-            if len(pagedivs)>0:
-                docinfo['countPlaces']= getTextFromNode(pagedivs[0])
-                s = getTextFromNode(pagedivs[0])
-                try:
-                    docinfo['countPlaces'] = int(s)
-                except:
-                    docinfo['countPlaces'] = 0
-        
-        #tocEntries
-        pagedivs = dom.xpath("//div[@class='countTocEntries']")
-        if pagedivs == dom.xpath("//div[@class='countTocEntries']"):
-            if len(pagedivs)>0:
-                docinfo['countTocEntries'] = int(getTextFromNode(pagedivs[0]))
-                s = getTextFromNode(pagedivs[0])
-                if s=='0':
-                    try:
-                        docinfo['countTocEntries'] = int(s)
-                    except:
-                        docinfo['countTocEntries'] = 0
-                else:
-                    s1 = int(s)/30+1
-                    try:
-                        docinfo['countTocEntries'] = int(s1)
-                    except:
-                        docinfo['countTocEntries'] = 0
-        
-        #numTextPages
-        pagedivs = dom.xpath("//div[@class='countPages']")
-        if pagedivs == dom.xpath("//div[@class='countPages']"):
-            if len(pagedivs)>0:
-                docinfo['numPages'] = getTextFromNode(pagedivs[0])
-                s = getTextFromNode(pagedivs[0])
-                
-                try:
-                    docinfo['numPages'] = int(s)
-                    #logging.debug("PAGE NUMBER: %s"%(s))
-                
-                    np = docinfo['numPages']
-                    pageinfo['end'] = min(pageinfo['end'], np)
-                    pageinfo['numgroups'] = int(np / pageinfo['groupsize'])
-                    if np % pageinfo['groupsize'] > 0:
-                        pageinfo['numgroups'] += 1                
-                except:
-                    docinfo['numPages'] = 0
-                  
-        else:
-         #no full text -- init to 0
-            docinfo['pageNumberOrig'] = 0
-            docinfo['countFigureEntries'] = 0
-            docinfo['countPlaces'] = 0
-            docinfo['countTocEntries'] = 0
-            docinfo['numPages'] = 0
-            docinfo['pageNumberOrigNorm'] = 0
-        #return docinfo
+        dom = ET.fromstring(pagexml)
+        # extract additional info
+        self.processPageInfo(dom, docinfo, pageinfo)
+        # page content is in <div class="pageContent">
+        pagediv = None
+        # ElementTree 1.2 in Python 2.6 can't do div[@class='pageContent']
+        alldivs = dom.findall(".//div")
+        for div in alldivs:
+            dc = div.get('class')
+            # page content div
+            if dc == 'pageContent':
+                pagediv = div
+                break
         
         # plain text mode
         if mode == "text":
-            # first div contains text
-            pagedivs = dom.xpath("/div")
-            if len(pagedivs) > 0:      
-                pagenode = pagedivs[0]
-                links = pagenode.xpath("//a")
+            if pagediv:
+                links = pagediv.findall(".//a")
                 for l in links:
-                    hrefNode = l.getAttributeNodeNS(None, u"href")
-                    if hrefNode:
-                        href= hrefNode.nodeValue
-                        if href.startswith('#note-'):
-                            hrefNode.nodeValue = href.replace('#note-',"?url=%s&viewMode=text&tocMode=%s&tocPN=%s&pn=%s#note-"%(url,tocMode,tocPN,pn))
-                return serializeNode(pagenode)
-        if mode == "xml":
-              # first div contains text
-              pagedivs = dom.xpath("/div")
-              if len(pagedivs) > 0:
-                  pagenode = pagedivs[0]
-                  return serializeNode(pagenode)
-        if mode == "gis":
-              # first div contains text
-              pagedivs = dom.xpath("/div")
-              if len(pagedivs) > 0:
-                  pagenode = pagedivs[0]
-                  links =pagenode.xpath("//a")
-                  for l in links:
-                      hrefNode =l.getAttributeNodeNS(None, u"href")
-                      if hrefNode:
-                          href=hrefNode.nodeValue
-                          if href.startswith('http://chinagis.mpiwg-berlin.mpg.de'):
-                              hrefNode.nodeValue =href.replace('chinagis_REST/REST/db/chgis/mpdl','chinagis/REST/db/mpdl/%s'%name)
-                              l.setAttributeNS(None, 'target', '_blank') 
-                  return serializeNode(pagenode)
-                    
-        if mode == "pureXml":
-              # first div contains text
-              pagedivs = dom.xpath("/div")
-              if len(pagedivs) > 0:
-                  pagenode = pagedivs[0]
-                  return serializeNode(pagenode)      
+                    href = l.get('href')
+                    if href and href.startswith('#note-'):
+                        href = href.replace('#note-',"?mode=%s&url=%s&viewMode=text&tocMode=%s&tocPN=%s&pn=%s#note-"%(urlmode,url,tocMode,tocPN,pn))
+                        l.set('href', href)
+
+                return serialize(pagediv)
+            
         # text-with-links mode
-        if mode == "text_dict":
-            # first div contains text
-            #mode = pageinfo ['viewMode']
-            pagedivs = dom.xpath("/div")
-            if len(pagedivs) > 0:
-                pagenode = pagedivs[0]
+        elif mode == "text_dict":
+            if pagediv:
                 # check all a-tags
-                links = pagenode.xpath("//a")
-                
+                links = pagediv.findall(".//a")
                 for l in links:
-                    hrefNode = l.getAttributeNodeNS(None, u"href")
+                    href = l.get('href')
                     
-                    if hrefNode:
+                    if href:
                         # is link with href
-                        href = hrefNode.nodeValue
                         if href.startswith('http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/lt/wordInfo.xql'):
                             # is pollux link
                             selfurl = self.absolute_url()
                             # change href
-                            hrefNode.nodeValue = href.replace('http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/lt/wordInfo.xql','%s/head_main_voc'%selfurl)
+                            l.set('href', href.replace('http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/interface/lt/wordInfo.xql','%s/head_main_voc'%selfurl))
                             # add target
-                            l.setAttributeNS(None, 'target', '_blank')
-                            #l.setAttributeNS(None, 'onclick',"popupWin = window.open(this.href, 'InfoWindow', 'menubar=no, location,width=500,height=600,top=180, left=700, toolbar=no, scrollbars=1'); return false;")
-                            #l.setAttributeNS(None, "ondblclick", "popupWin.focus();")
-                            #window.open("this.href, 'InfoWindow', 'menubar=no, location,width=500,height=600,top=180, left=700, toolbar=yes, scrollbars=1'"); return false;") 
+                            l.set('target', '_blank')
                                                           
                         if href.startswith('http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/lt/lemma.xql'):    
                             selfurl = self.absolute_url()
-                            hrefNode.nodeValue = href.replace('http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/lt/lemma.xql','%s/head_main_lemma'%selfurl)
-                            l.setAttributeNS(None, 'target', '_blank')
-                            l.setAttributeNS(None, 'onclick',"popupWin = window.open(this.href, 'InfoWindow', 'menubar=no, location,width=500,height=600,top=180, left=700, toolbar=no, scrollbars=1'); return false;")
-                            l.setAttributeNS(None, 'ondblclick', 'popupWin.focus();')   
+                            l.set('href', href.replace('http://mpdl-proto.mpiwg-berlin.mpg.de/mpdl/lt/lemma.xql','%s/head_main_lemma'%selfurl))
+                            l.set('target', '_blank')
+                            l.set('onclick',"popupWin = window.open(this.href, 'InfoWindow', 'menubar=no, location,width=500,height=600,top=180, left=700, toolbar=no, scrollbars=1'); return false;")
+                            l.set('ondblclick', 'popupWin.focus();')   
                     
                         if href.startswith('#note-'):
-                            hrefNode.nodeValue = href.replace('#note-',"?url=%s&viewMode=text_dict&tocMode=%s&tocPN=%s&pn=%s#note-"%(url,tocMode,tocPN,pn))  
+                            l.set('href', href.replace('#note-',"?mode=%s&url=%s&viewMode=text_dict&tocMode=%s&tocPN=%s&pn=%s#note-"%(urlmode,url,tocMode,tocPN,pn)))
                               
-                return serializeNode(pagenode)
+                return serialize(pagediv)
+            
+        # xml mode
+        elif mode == "xml":
+            if pagediv:
+                return serialize(pagediv)
+            
+        # pureXml mode
+        elif mode == "pureXml":
+            if pagediv:
+                return serialize(pagediv)
+                  
+        # gis mode
+        elif mode == "gis":
+            name = docinfo['name']
+            if pagediv:
+                # check all a-tags
+                links = pagediv.findall(".//a")
+                for l in links:
+                    href = l.get('href')
+                    if href:
+                        if href.startswith('http://chinagis.mpiwg-berlin.mpg.de'):
+                            l.set('href', href.replace('chinagis_REST/REST/db/chgis/mpdl','chinagis/REST/db/mpdl/%s'%name))
+                            l.set('target', '_blank') 
+                            
+                return serialize(pagediv)
+                    
         return "no text here"
     
+    # WTF: is this needed?
     def getOrigPages(self, docinfo=None, pageinfo=None):
-        docpath = docinfo['textURLPath']
-        pn =pageinfo['current']
-        selfurl = self.absolute_url()   
-        pagexml = self.getServerData("page-fragment.xql","document=%s&pn=%s"%(docpath, pn))
-        dom = Parse(pagexml)
-        pagedivs = dom.xpath("//div[@class='pageNumberOrig']")
-        if pagedivs == dom.xpath("//div[@class='pageNumberOrig']"):
-            if len(pagedivs)>0:
-                docinfo['pageNumberOrig']= getTextFromNode(pagedivs[0])        
-                return docinfo['pageNumberOrig']
+        logging.debug("CALLED: getOrigPages!")
+        if not pageinfo.has_key('pageNumberOrig'):
+            logging.warning("getOrigPages: not in pageinfo!")
+            return None
+        
+        return pageinfo['pageNumberOrig']
     
+    # WTF: is this needed?
     def getOrigPagesNorm(self, docinfo=None, pageinfo=None):
-        docpath = docinfo['textURLPath']
-        pn =pageinfo['current']
-        selfurl = self.absolute_url()   
-        pagexml = self.getServerData("page-fragment.xql","document=%s&pn=%s"%(docpath, pn))
-        dom = Parse(pagexml)
-        pagedivs = dom.xpath("//div[@class='pageNumberOrigNorm']")
-        if pagedivs == dom.xpath("//div[@class='pageNumberOrigNorm']"):
-            if len(pagedivs)>0:
-                docinfo['pageNumberOrigNorm']= getTextFromNode(pagedivs[0])        
-                return docinfo['pageNumberOrigNorm']
-
+        logging.debug("CALLED: getOrigPagesNorm!")
+        if not pageinfo.has_key('pageNumberOrigNorm'):
+            logging.warning("getOrigPagesNorm: not in pageinfo!")
+            return None
+        
+        return pageinfo['pageNumberOrigNorm']
                 
+    # TODO: should be getWordInfo
     def getTranslate(self, word=None, language=None):
         """translate into another languages"""
         data = self.getServerData("lt/wordInfo.xql","language="+str(language)+"&word="+urllib.quote(word)+"&output=html")
-        #pagexml=self.template.fulltextclient.eval("/mpdl/interface/lt/lex.xql","document=&language="+str(language)+"&query="+url_quote(str(query)))
         return data
     
+    # WTF: what does this do?
     def getLemma(self, lemma=None, language=None):
         """simular words lemma """
         data = self.getServerData("lt/lemma.xql","language="+str(language)+"&lemma="+urllib.quote(lemma)+"&output=html")
         return data
     
+    # WTF: what does this do?
     def getLemmaQuery(self, query=None, language=None):
         """simular words lemma """
         data = self.getServerData("lt/lemma.xql","language="+str(language)+"&query="+urllib.quote(query)+"&output=html")
         return data
     
+    # WTF: what does this do?
     def getLex(self, query=None, language=None):
         #simular words lemma
         data = self.getServerData("lt/lex.xql","document=&language="+str(language)+"&query="+urllib.quote(query))
         return data
-    
+
+    # WTF: what does this do?
     def getQuery (self,  docinfo=None, pageinfo=None, query=None, queryType=None, pn=1):
          #number of
          docpath = docinfo['textURLPath'] 
@@ -442,9 +470,11 @@ class MpdlXmlTextServer(SimpleItem):
          return tc
     
     def getToc(self, mode="text", docinfo=None):
-        """loads table of contents and stores in docinfo"""
+        """loads table of contents and stores XML in docinfo"""
+        logging.debug("getToc mode=%s"%mode)
         if mode == "none":
-            return docinfo        
+            return docinfo
+              
         if 'tocSize_%s'%mode in docinfo:
             # cached toc
             return docinfo
@@ -460,44 +490,87 @@ class MpdlXmlTextServer(SimpleItem):
         # number of entries in toc
         tocSize = 0
         tocDiv = None
-        
+        # fetch full toc
         pagexml = self.getServerData("doc-query.xql","document=%s&queryType=%s&queryResultPageSize=%s&queryResultPN=%s"%(docpath,queryType, pagesize, pn))
-        
-        # post-processing downloaded xml
-        pagedom = Parse(pagexml)
-        # get number of entries
-        numdivs = pagedom.xpath("//div[@class='queryResultHits']")
-        if len(numdivs) > 0:
-            tocSize = int(getTextFromNode(numdivs[0]))
-        docinfo['tocSize_%s'%mode] = tocSize
+        dom = ET.fromstring(pagexml)
+        # page content is in <div class="queryResultPage">
+        pagediv = None
+        # ElementTree 1.2 in Python 2.6 can't do div[@class='queryResultPage']
+        alldivs = dom.findall("div")
+        for div in alldivs:
+            dc = div.get('class')
+            # page content div
+            if dc == 'queryResultPage':
+                pagediv = div
+                
+            elif dc == 'queryResultHits':
+                docinfo['tocSize_%s'%mode] = intOr0(div.text)
+
+        if pagediv:
+#            # split xml in chunks
+#            tocs = []
+#            tocdivs = pagediv.findall('div')
+#            for p in zip(tocdivs[::2], tocdivs[1::2]):
+#                toc = serialize(p[0])
+#                toc += serialize(p[1])
+#                tocs.append(toc)
+#                logging.debug("pair: %s"%(toc))
+            # store XML in docinfo
+            docinfo['tocXML_%s'%mode] = ET.tostring(pagediv, 'UTF-8')
+
         return docinfo
     
     def getTocPage(self, mode="text", pn=1, pageinfo=None, docinfo=None):
         """returns single page from the table of contents"""
-        # TODO: this should use the cached TOC
+        logging.debug("getTocPage mode=%s, pn=%s"%(mode,pn))
         if mode == "text":
             queryType = "toc"
         else:
             queryType = mode
-        docpath = docinfo['textURLPath']
-        path = docinfo['textURLPath']       
-        pagesize = pageinfo['tocPageSize']
-        pn = pageinfo['tocPN']
+            
+        # check for cached TOC
+        if not docinfo.has_key('tocXML_%s'%mode):
+            self.getToc(mode=mode, docinfo=docinfo)
+            
+        tocxml = docinfo.get('tocXML_%s'%mode, None)
+        if not tocxml:
+            logging.error("getTocPage: unable to find tocXML")
+            return "No ToC"
+        
+        pagesize = int(pageinfo['tocPageSize'])
         url = docinfo['url']
-        selfurl = self.absolute_url()  
+        urlmode = docinfo['mode']
+        selfurl = docinfo['viewerUrl']
         viewMode=  pageinfo['viewMode']
-        characterNormalization = pageinfo ['characterNormalization']
-        #optionToggle =pageinfo ['optionToggle']
         tocMode = pageinfo['tocMode']
-        tocPN = pageinfo['tocPN']  
+        tocPN = int(pageinfo['tocPN'])
+
+        fulltoc = ET.fromstring(tocxml)
         
-        data = self.getServerData("doc-query.xql","document=%s&queryType=%s&queryResultPageSize=%s&queryResultPN=%s&characterNormalization=regPlusNorm"%(docpath,queryType, pagesize, pn))  
-        page = data.replace('page-fragment.xql?document=%s'%str(path),'%s?url=%s&viewMode=%s&tocMode=%s&tocPN=%s'%(selfurl,url, viewMode, tocMode, tocPN))
-        text = page.replace('mode=image','mode=texttool')
-        return text
+        if fulltoc:
+            # paginate
+            #start = (pn - 1) * pagesize * 2
+            #end = start + pagesize * 2
+            #tocdivs = fulltoc[start:end]
+            tocdivs = fulltoc
+            
+            # check all a-tags
+            links = tocdivs.findall(".//a")
+            for l in links:
+                href = l.get('href')
+                if href:
+                    # take pn from href
+                    m = re.match(r'page-fragment\.xql.*pn=(\d+)', href)
+                    if m is not None:
+                        # and create new url
+                        l.set('href', '%s?mode=%s&url=%s&viewMode=%s&pn=%s&tocMode=%s&tocPN=%s'%(selfurl, urlmode, url, viewMode, m.group(1), tocMode, tocPN))
+                    else:
+                        logging.warning("getTocPage: Problem with link=%s"%href)
+                        
+            return serialize(tocdivs)
+    
     
     def manage_changeMpdlXmlTextServer(self,title="",serverUrl="http://mpdl-text.mpiwg-berlin.mpg.de/mpdl/interface/",timeout=40,RESPONSE=None):
-    #def manage_changeMpdlXmlTextServer(self,title="",serverUrl="http://mpdl-text.mpiwg-berlin.mpg.de:30030/mpdl/interface/",timeout=40,RESPONSE=None):
         """change settings"""
         self.title=title
         self.timeout = timeout
@@ -517,4 +590,6 @@ def manage_addMpdlXmlTextServer(self,id,
     newObj = MpdlXmlTextServer(id,title,serverUrl,timeout)
     self.Destination()._setObject(id, newObj)
     if RESPONSE is not None:
-        RESPONSE.redirect('manage_main')
\ No newline at end of file
+        RESPONSE.redirect('manage_main')
+        
+        
\ No newline at end of file