web/ldt/ldt_utils/contentindexer.py
author samuel huron <admin@cybunk.com>
Mon, 14 Jun 2010 09:54:49 +0200
changeset 35 6e5ffd90610c
parent 10 84e31387a741
permissions -rw-r--r--
Update JS form json parsing

import tempfile
import os
import os.path
import shutil
from ldt.utils import zipfileext
import urllib
# import ldt.utils.log
import ldt.utils.xml
from django.conf import settings
from models import Content
import xml
import xml.dom
import xml.dom.minidom
import xml.dom.ext
import xml.xpath
import fnmatch
import Ft
import uuid
import shutil
import lucene
from ldt.ldt_utils import STORE
from ldt.ldt_utils import ANALYZER

def Property(func):
    return property(**func()) 


class ContentIndexer(object):
        
        def __init__(self, contentList, writer, decoupage_blackList = settings.DECOUPAGE_BLACKLIST):
                self.__contentList = contentList
                self.__decoupage_blacklist = decoupage_blackList
                self.__writer = writer
                    
        @Property
        def decoupage_blacklist(): #@NoSelf
            doc = """get blacklist""" #@UnusedVariable
           
            def fget(self):
                if self.__decoupage_blacklist is None:
                    self.__decoupage_blacklist = ()
                return self.__decoupage_blacklist
               
            def fset(self, value):
                self.__decoupage_blacklist = value
               
            def fdel(self):
                del self.__decoupage_blacklist
               
            return locals()
                   
        def index_all(self):
            for content in self.__contentList:
                self.index_content(content)
                
        def index_content(self, content):
            url =content.iri_url()
            filepath = urllib.urlopen(url)
            doc = xml.dom.minidom.parse(filepath)
            doc = Ft.Xml.Domlette.ConvertDocument(doc)
                                   
            self.__writer.deleteDocuments(lucene.Term("iri_id", content.iri_id))
            
            con = xml.xpath.Context.Context(doc, 1, 1, None)
            res = xml.xpath.Evaluate("/iri/body/ensembles/ensemble", context=con)

            for ensemble in res:
                ensembleId = ensemble.getAttributeNS(None,u"id")
                
                for decoupageNode in ensemble.childNodes:
                    if decoupageNode.nodeType != xml.dom.Node.ELEMENT_NODE or decoupageNode.tagName != "decoupage"  or decoupageNode.getAttributeNS(None,u"id") in self.decoupage_blacklist:
                        continue
                    
                    decoupId = decoupageNode.getAttributeNS(None,u"id")
                    res = xml.xpath.Evaluate("elements/element", decoupageNode)
                    for elementNode in res:
                        doc = lucene.Document()
                        elementId = elementNode.getAttributeNS(None,u"id")
                        tags = elementNode.getAttributeNS(None,u"tags")
                        
                        if tags is not None:                            
                            tags.replace(",", ";")
                        
                        if tags is None or len(tags) == 0:
                            tags = ""
                            restagnode = xml.xpath.Evaluate("tag/text()", elementNode)
                            for tagnode in restagnode:
                                tags = tags + " ; " + tagnode.data
                                
                        if tags is None or len(tags) == 0:
                            tags = ""
                            restagnode = xml.xpath.Evaluate("tags/tag/text()", elementNode)
                            for tagnode in restagnode:
                                tags = tags + " ; " + tagnode.data                            
    
                        title = ""
                        for txtRes in xml.xpath.Evaluate("title/text()", elementNode): 
                            title = title + txtRes.data 
                
                        abstract = ""
                        for txtRes in xml.xpath.Evaluate("abstract/text()", elementNode): 
                            abstract = abstract + txtRes.data 
                
                        doc.add(lucene.Field("iri_id", content.iri_id, lucene.Field.Store.YES, lucene.Field.Index.NOT_ANALYZED))
                        doc.add(lucene.Field("ensemble_id", ensembleId, lucene.Field.Store.YES, lucene.Field.Index.NO))
                        doc.add(lucene.Field("decoupage_id", decoupId, lucene.Field.Store.YES, lucene.Field.Index.NO))
                        doc.add(lucene.Field("element_id", elementId, lucene.Field.Store.YES, lucene.Field.Index.NO))                        
                        doc.add(lucene.Field("tags", tags, lucene.Field.Store.NO, lucene.Field.Index.ANALYZED))
                        doc.add(lucene.Field("title", title, lucene.Field.Store.NO, lucene.Field.Index.ANALYZED))
                        doc.add(lucene.Field("abstract", abstract, lucene.Field.Store.NO, lucene.Field.Index.ANALYZED))
                        doc.add(lucene.Field("all", " ".join([tags, title, abstract]), lucene.Field.Store.NO, lucene.Field.Index.ANALYZED))

                        seg = Segment(content=content,
                                      iri_id=content.iri_id,
                                      ensemble_id=ensembleId,
                                      cutting_id=decoupId,
                                      element_id=elementId,
                                      tags=tags,
                                      title=title,
                                      abstract=abstract,
                                      duration=duration,
                                      author=author,
                                      start_ts=start_ts,
                                      date=date_str)
                        seg.save()

            
                        self.__writer.addDocument(doc)
            
            self.__writer.commit()
            
            
class ProjectIndexer(object):
        
        def __init__(self, projectList, writer, decoupage_blackList = settings.DECOUPAGE_BLACKLIST):
                self.__projectList = projectList
                self.__decoupage_blacklist = decoupage_blackList
                self.__writer = writer
                
        @Property
        def decoupage_blacklist(): #@NoSelf
            doc = """get blacklist""" #@UnusedVariable
           
            def fget(self):
                if self.__decoupage_blacklist is None:
                    self.__decoupage_blacklist = ()
                return self.__decoupage_blacklist
               
            def fset(self, value):
                self.__decoupage_blacklist = value
               
            def fdel(self):
                del self.__decoupage_blacklist
               
            return locals()
                   
        def index_all(self):
            for project in self.__projectList:
                self.index_project(project)
 
        def index_project(self, project):
            
            # pocketfilms.utils.log.debug("Indexing project : "+str(project.iri_id))
            doc = xml.dom.minidom.parseString(project.ldt)
            doc = Ft.Xml.Domlette.ConvertDocument(doc) 

            self.__writer.deleteDocuments(lucene.Term("iri_id", project.iri_id))
            
            con = xml.xpath.Context.Context(doc, 1, 1, None)
            res = xml.xpath.Evaluate("/iri/annotations/content", context=con)

            for content in res:
                contentId = content.getAttributeNS(None,u"id")
                
                ensembleId = "ens_perso"
                
                for decoupageNode in content.childNodes:
                    # pocketfilms.utils.log.debug("Indexing content decoupage : "+ repr(decoupageNode.nodeType) + " in " + repr(self.decoupage_blacklist))
                    if decoupageNode.nodeType != xml.dom.Node.ELEMENT_NODE or decoupageNode.tagName != "decoupage"  or decoupageNode.getAttributeNS(None,"id") in self.decoupage_blacklist:
                        continue
                    
                    decoupId = decoupageNode.getAttributeNS(None,u"id")
                    res = xml.xpath.Evaluate("elements/element", decoupageNode)
                    for elementNode in res:
                        doc = lucene.Document()
                        elementId = elementNode.getAttributeNS(None,u"id")
                        tags = elementNode.getAttributeNS(None,u"tags")
                        
                        if tags is not None:                            
                            tags.replace(",", ";")
                        
                        if tags is None or len(tags) == 0:
                            tags = ""
                            restagnode = xml.xpath.Evaluate("tag/text()", elementNode)
                            for tagnode in restagnode:
                                tags = tags + " ; " + tagnode.data
                                
                        if tags is None or len(tags) == 0:
                            tags = ""
                            restagnode = xml.xpath.Evaluate("tags/tag/text()", elementNode)
                            for tagnode in restagnode:
                                tags = tags + " ; " + tagnode.data                            
    
                        title = ""
                        for txtRes in xml.xpath.Evaluate("title/text()", elementNode): 
                            title = title + txtRes.data 
                
                        abstract = ""
                        for txtRes in xml.xpath.Evaluate("abstract/text()", elementNode): 
                            abstract = abstract + txtRes.data 
                
                        doc.add(lucene.Field("project_id", project.iri_id, lucene.Field.Store.YES, lucene.Field.Index.NOT_ANALYZED))              
                        doc.add(lucene.Field("iri_id", contentId, lucene.Field.Store.YES, lucene.Field.Index.NOT_ANALYZED))
                        doc.add(lucene.Field("ensemble_id", ensembleId, lucene.Field.Store.YES, lucene.Field.Index.NO))
                        doc.add(lucene.Field("decoupage_id", decoupId, lucene.Field.Store.YES, lucene.Field.Index.NO))
                        doc.add(lucene.Field("element_id", elementId, lucene.Field.Store.YES, lucene.Field.Index.NO))
                        doc.add(lucene.Field("tags", tags, lucene.Field.Store.NO, lucene.Field.Index.ANALYZED))
                        doc.add(lucene.Field("title", title, lucene.Field.Store.NO, lucene.Field.Index.ANALYZED))
                        doc.add(lucene.Field("abstract", abstract, lucene.Field.Store.NO, lucene.Field.Index.ANALYZED))
                        doc.add(lucene.Field("all", " ".join([tags, title, abstract]), lucene.Field.Store.NO, lucene.Field.Index.ANALYZED))
                        
                        seg = Segment(content=content,
                                      iri_id=content.iri_id,
                                      ensemble_id=ensembleId,
                                      cutting_id=decoupId,
                                      element_id=elementId,
                                      tags=tags,
                                      title=title,
                                      abstract=abstract,
                                      duration=duration,
                                      author=author,
                                      start_ts=start_ts,
                                      date=date_str)
                        seg.save()
                                    
                        self.__writer.addDocument(doc)
            
            self.__writer.commit()