src/ldt/ldt/ldt_utils/utils.py
author ymh <ymh.work@gmail.com>
Wed, 13 Jun 2012 15:04:18 +0200
changeset 665 f4fed46efbde
parent 654 69b3c4d7a7c1
child 713 202fb1255d8b
child 718 5e27a39d3742
permissions -rw-r--r--
correct warnings about timezones, remove dependance on pytz

from copy import deepcopy
from django.conf import settings
from ldt.indexation import get_searcher, get_results_list
from django.utils.translation import ugettext as _
from StringIO import StringIO
import datetime
import django.core.urlresolvers
import lxml.etree
import urllib
import uuid

__BOOLEAN_DICT = {
    'false':False,
    'true':True,
    '0':False,
    '1':True,
    't': True,
    'f':False
}

def reduce_text_node(element_node, xpath_str=None):
    node_list = []
    if xpath_str is not None:
        node_list = element_node.xpath(xpath_str, smart_strings=False)
    else:
        node_list = [element_node.text]
    return reduce(lambda t, s: t + s, node_list , "")

def boolean_convert(bool):
    if bool is None:
        return False
    if bool is True or bool is False:
        return bool
    key = str(bool).lower()
    return __BOOLEAN_DICT.get(key, False)

def generate_uuid():
    return unicode(uuid.uuid1())

class LdtSearch(object):

    def query(self, field, query):
        indexSearcher = get_searcher()
        hits = get_results_list(field, query)
    
        res = []
        for hit in hits:
            doc = indexSearcher.doc(hit.doc)
            res.append({"iri_id":doc.get("iri_id"), "ensemble_id":doc.get("ensemble_id"), "decoupage_id":doc.get("decoupage_id"), "element_id":doc.get("element_id"), "project_id":doc.get("project_id")})
        indexSearcher.close()
        return res

    def query_all(self, query):        
        return self.query("all", query)
        

class LdtUtils(object):
    
    def generate_ldt(self, contentList, title=u"", author=u"IRI Web", web_url=u"", startSegment=None, projects=None):

        iri = lxml.etree.Element(u'iri')
        doc = lxml.etree.ElementTree(iri)
    
        project = lxml.etree.SubElement(iri, u'project')
        project.set(u"id", generate_uuid())
        project.set(u"title", unicode(title))
        project.set(u"user", author)
        project.set(u"abstract", u"")
    
        medias = lxml.etree.SubElement(iri, u"medias")
        for content in contentList:
            videopath = unicode(settings.STREAM_URL)
            if content.videopath != None :
                videopath = unicode(content.videopath)
            media = lxml.etree.SubElement(medias, "media")
            media.set(u"id", content.iri_id)
            media.set(u"src", content.iri_url(web_url))
            media.set(u"video", videopath)
            media.set(u"pict", u"")
            media.set(u"extra", u"")
    
        if projects is None:
            projects = []
        annotations_nodes = {}
        for project in projects:
            ldtdoc = lxml.etree.fromstring(project.ldt_encoded)
            res = ldtdoc.xpath("/iri/annotations/content")

            for content in res:
                contentid = content.get("id")
                if annotations_nodes.has_key(contentid):
                    contentnode = annotations_nodes[contentid]
                else:
                    contentnode = {"id":contentid, "ensembles":[]}
                    annotations_nodes[contentid] = contentnode
                for ens in content:
                    if ens.tag.endswith("ensemble"):
                        contentnode["ensembles"].append(deepcopy(ens))

        annotations = lxml.etree.SubElement(iri, "annotations")
        if len(annotations_nodes) > 0:
            for content in contentList:
                if content.iri_id in annotations_nodes:
                    contentnode = annotations_nodes[content.iri_id]
                    if contentnode is not None:
                        content_node = lxml.etree.SubElement(annotations, "content")
                        content_node.set("id", contentnode["id"])
                        content_node.text = u""
                        for ens in contentnode["ensembles"]:
                            content_node.append(ens)
    
    
        displays = lxml.etree.SubElement(iri, "displays")
        if len(contentList) > 0:
            display = lxml.etree.SubElement(displays, "display")
            display.set(u"id", u"0")
            display.set(u"title", u"generated")
            display.set(u"idsel", contentList[0].iri_id)
            display.set(u"tc", u"0")
            for content in contentList:
                contentd = lxml.etree.SubElement(display, "content")
                contentd.set(u"id", content.iri_id)
                filepath = urllib.urlopen(content.iri_url())
            
                udoc = lxml.etree.parse(filepath)
                res = udoc.xpath("/iri/body/ensembles/ensemble/decoupage")
                for decoupagenode in res:
                    decoupage_id = decoupagenode.get(u"id")
                    ensemble_id = decoupagenode.getparent().get(u"id")
                    decoupage_id = decoupagenode.get(u"id")
                    ensemble_id = decoupagenode.getparent().get(u"id")
                    decoupage = lxml.etree.SubElement(contentd, "decoupage")
                    decoupage.set(u"id", decoupage_id)
                    decoupage.set(u"idens", ensemble_id)
            if startSegment is not None:
                activeSegment = lxml.etree.SubElement(display, "activeSegment")
                idas = lxml.etree.SubElement(activeSegment, "id")
                idas.set(u"idctt", startSegment["idcontent"])
                idas.set(u"idens" , startSegment["idgroup"])
                idas.set(u"idcut", startSegment["idcutting"])
                idas.set(u"idseg", startSegment["idsegment"])
    
        lxml.etree.SubElement(iri, "edits")
        return doc
        
    def generate_init(self, url, method, search=None):
                
        iri = lxml.etree.Element('iri')
 
        elementFiles = lxml.etree.SubElement(iri, 'files')    
        elementInit = lxml.etree.SubElement(elementFiles, 'init')
        elementfile = lxml.etree.SubElement(elementInit, 'file')
            
        elementfile.set('src', settings.WEB_URL + django.core.urlresolvers.reverse(method, args=url))
        elementfile.set('src', django.core.urlresolvers.reverse(method, args=url))
        if(search):
            elementfile.set("segsel", settings.WEB_URL + django.core.urlresolvers.reverse(search, args=url))

        lxml.etree.SubElement(elementFiles, 'recent')
        lxml.etree.SubElement(elementFiles, 'library')
        
        return iri 


class LdtAnnotation:
    
    def __init__(self, project, force_save=False):
        self.project = project
        self.parser = lxml.etree.XMLParser(remove_blank_text=True)
        self.ldtdoc = lxml.etree.parse(StringIO(project.ldt_encoded), self.parser)
        self.to_add = True
        
    #   add( a['media'], a['type'],  a['type_title, a[data], '', a['tags'], begin, dur, author, date)
    def add(self, media, cutting_id, cutting_title, title, text, tags_list, begin, dur, author, date, view_id="0", color="2194379", audio_scr="", audio_href=""):
        """
        Add an annotation to a project. begin and dur must be strings. Default color is yellow.
        """
        
        if dur < 0:
            self.to_add = False
            return False
            
        # We check if the project references the media.
        path_media = self.ldtdoc.xpath('/iri/medias/media[@id="%s"]' % media)
        if len(path_media) == 0:
            self.to_add = False
            return False
        
        # We get the content node
        path_annotations = self.ldtdoc.xpath('/iri/annotations')[0]
        path_content = path_annotations.xpath('content[@id="%s"]' % media)
        # If the content node does not exist, we create it
        if len(path_content) == 0:
            path_content = lxml.etree.SubElement(path_annotations, 'content')
            path_content.set('id', media)
            path_content = [path_content]
        
        # We generate the cutting id if necessary
        if cutting_id is None or cutting_id=="" :
            cutting_id = 'c_' + generate_uuid()
        
        # We get the ensemble node
        path_ensemble = path_content[0].xpath('ensemble[decoupage[@id="%s"]]' % cutting_id)
        if len(path_ensemble) == 0:
            # If the ensemble node does not exist, we create it
            path_ensemble = lxml.etree.SubElement(path_content[0], 'ensemble')
            path_ensemble.set('id', 'g_' + generate_uuid())
            path_ensemble.set('title', _('Personal cutting'))
            path_ensemble.set('author', 'IRI')
            path_ensemble.set('abstract', 'IRI')
            path_ensemble = [path_ensemble]
        #else:
        #    path_ensemble = path_content[0].xpath('ensemble')
        
            
        # We get the elements node in the good decoupage node
        ensemble_id = path_ensemble[0].get('id')
        decoupage_elements = path_ensemble[0].xpath('decoupage[@id="%s"]/elements' % cutting_id)
        if len(decoupage_elements) == 0:
            # If the decoupage node does not exist, we create it and its elements node 
            decoupage = lxml.etree.SubElement(path_ensemble[0], 'decoupage')
            #cutting_id = "c_" + generate_uuid()
            decoupage.set('id', cutting_id)
            decoupage.set('author', author)
            decoupage_title = lxml.etree.SubElement(decoupage, 'title')
            decoupage_title.text = cutting_title
            lxml.etree.SubElement(decoupage, 'abstract')
            decoupage_elements = lxml.etree.SubElement(decoupage, 'elements')
            decoupage_elements = [decoupage_elements]
        #else:
        #    cutting_id = path_ensemble[0].xpath('decoupage[title="%s"]' % cutting_title)[0].get('id')     
        
        # We add the cutting to the view
        if view_id is not None and view_id!="" :
            path_view = self.ldtdoc.xpath('/iri/displays/display[@id="%s"]' % view_id)
            if len(path_view) == 0:
                path_view = self.ldtdoc.xpath('/iri/displays/display[@title="Init view"]')
            else :
                content_display = path_view[0].xpath('content[@id="%s"]' % media)
                if len(content_display) == 0:
                    content_display = lxml.etree.SubElement(path_view[0], 'content')
                    content_display.set('id', media)
                    content_display = [content_display]
                    # We add the decoupage node to the content node
                    dec = lxml.etree.SubElement(content_display[0], 'decoupage')
                    dec.set('idens', ensemble_id)
                    dec.set('id', cutting_id)
                    dec.set('tagsSelect', '')
        
        # We add the annotation/element node
        element = lxml.etree.SubElement(decoupage_elements[0], 'element')
        id_annotation = 's_' + generate_uuid()
        element.set('id', id_annotation)
        element.set('begin', begin)
        element.set('dur', dur)
        element.set('author', author)
        element.set('date', date)
        element.set('color', color)
        element.set('src', "")
        abstract = lxml.etree.SubElement(element, 'abstract')
        abstract.text = text
        title_node = lxml.etree.SubElement(element, 'title')
        title_node.text = title
        audio = lxml.etree.SubElement(element, 'audio')
        audio.set('source', audio_scr)
        audio.text = audio_href
        tags = lxml.etree.SubElement(element, 'tags')
                                
        polemics = self.get_polemic_syntax(title)
        
        if polemics:
            meta = lxml.etree.SubElement(element, 'meta')
            polemics_node = lxml.etree.SubElement(meta, 'polemics')
            
            for polemic in polemics:
                polemic_node = lxml.etree.SubElement(polemics_node, 'polemic')
                polemic_node.text = polemic
        
        for tag in tags_list:
            tag_node = lxml.etree.SubElement(tags, 'tag')
            tag_node.text = tag
        
        return cutting_id, id_annotation
    
    def save(self):
        if self.to_add:
            self.project.ldt = lxml.etree.tostring(self.ldtdoc, pretty_print=True)
            self.project.save()
       
    def get_polemic_syntax(self, text):
        polemics = []
        for key in settings.SYNTAX.keys():
            if key in text:
                polemics.append(settings.SYNTAX[key])
                
        return polemics
    
    def __del__(self):
        self.save()

def create_ldt(project, user, cuttings=[]):
    """create xml"""

    
    contentList = project.contents.all()
    
    # create a dom
    iri = lxml.etree.Element('iri')
    doc = lxml.etree.ElementTree(iri)
    
    #node project
    elementProject = lxml.etree.SubElement(iri, 'project')
    
    elementProject.set('abstract', project.description)
    elementProject.set('title', project.title)
    elementProject.set('user', user.username)
    elementProject.set('id', project.ldt_id)

    #node medias
    elementMedias = lxml.etree.SubElement(iri, 'medias')
    
    idsel = None      
    for content in contentList:
        if not idsel:
            idsel = content.iri_id
        elementMedia = lxml.etree.SubElement(elementMedias, 'media')
        elementMedia.set('id', content.iri_id)
        elementMedia.set('src', content.iri_url())
        
        if content.videopath != None :
            elementMedia.set('video', content.videopath)
        else:
            elementMedia.set('video', settings.STREAM_URL)
        elementMedia.set('pict', "")
        elementMedia.set('extra', "")
        
    if not idsel:
        idsel = ""

    #node annotations
    annotations = lxml.etree.SubElement(iri, 'annotations')
    
    #node displays
    elementDisplays = lxml.etree.SubElement(iri, 'displays')
    elementDisplay = lxml.etree.SubElement(elementDisplays, 'display')
    elementDisplay.set('id', '0')
    elementDisplay.set('title', 'Init view')
    elementDisplay.set('idsel', idsel)
    elementDisplay.set('tc', '0')
    elementDisplay.set('zoom', '0')
    elementDisplay.set('scroll', '0')
    elementDisplay.set('infoBAB', '')

    
    #node content
    for content in contentList:
        elementContent = lxml.etree.SubElement(elementDisplay, 'content')
        elementContent.set('id', content.iri_id)

        if not 'http' in content.iriurl:
        #eg: "iiiielizabethrosse/ENMI08-III_elizabethrosse.iri"
            url = content.iri_url()
        else:
            url = content.iriurl
        file = urllib.urlopen(url)
        doc = lxml.etree.parse(file)
        res = doc.xpath("/iri/body/ensembles/ensemble/decoupage")        

        #node decoupage
        for decoupagenode in res:
            decoupage_id = decoupagenode.get(u"id")
            parent = decoupagenode.getparent()
            ensemble_id = parent.get(u"id")
            elementDecoupage = lxml.etree.SubElement(elementContent, 'decoupage')
            elementDecoupage.set('idens', ensemble_id)
            elementDecoupage.set('id', decoupage_id)
    
    #node edits
    lxml.etree.SubElement(iri, 'edits')
    
    # add cuttings to the first content if necessary
    if cuttings and contentList:
        content = contentList[0]
        
        ensemble_id = "g_%s" % generate_uuid()
        cuttings_id = []
        
        content_node = lxml.etree.SubElement(annotations, 'content')
        content_node.set('id', content.iri_id)
        ensemble = lxml.etree.SubElement(content_node, 'ensemble')
        ensemble.set('title', _('Personal cutting'))
        ensemble.set('id', ensemble_id)
        ensemble.set('author', 'undefined')
        ensemble.set('abstract', '')
        ensemble.set('idProject', project.ldt_id)
        
        #Add cuttings to tag annotations
        for cutting in cuttings:
            cutting_node = lxml.etree.SubElement(ensemble, 'decoupage')
            cutting_node.set('author', 'perso')
            cutting_id = "c_%s" % generate_uuid()
            cuttings_id.append(cutting_id)
            cutting_node.set('id', cutting_id)
            title = lxml.etree.SubElement(cutting_node, 'title')
            title.text = cutting
            lxml.etree.SubElement(cutting_node, 'abstract')
            lxml.etree.SubElement(cutting_node, 'elements')
            
        #Add cuttings to tag displays
        content_node = elementDisplays.xpath('display/content[@id=\'%s\']' % content.iri_id)[0]
        for cutting_id in cuttings_id:
            cutting_node = lxml.etree.SubElement(content_node, 'decoupage')
            cutting_node.set('idens', ensemble_id)
            cutting_node.set('id', cutting_id)
            cutting_node.set('tagSelect', '')            
        
    
    #write dom in Project.ldt 
    project.ldt = lxml.etree.tostring(iri, pretty_print=True)
    
    #save Project
    project.save()
    return project        


def copy_ldt(project, new_project, user):
    new_project.ldt_id = str(uuid.uuid1())
    new_project.created_by = user.username
    new_project.changed_by = user.username
    new_project.state = 1
    
    
    """create xml"""
    
    ldt = lxml.etree.fromstring(project.ldt_encoded)
    res = ldt.xpath("/iri/project")
    for elementProject in res:
        elementProject.set('abstract', project.get_description())
        elementProject.set('title', new_project.title)
        elementProject.set('user', user.username)
        elementProject.set('id', new_project.ldt_id)
        
    new_project.ldt = lxml.etree.tostring(ldt, pretty_print=True)

    #save Project
    new_project.save()
    return new_project

def create_empty_iri(file, content, username):
    
    iri = lxml.etree.Element('iri')
    doc = lxml.etree.ElementTree(iri)

    head = lxml.etree.SubElement(iri, 'head')
    meta_id = lxml.etree.SubElement(head, 'meta')
    meta_id.set(u'name', u'id')
    meta_id.set(u'content', unicode(content.iri_id))
    meta_title = lxml.etree.SubElement(head, 'meta')
    meta_title.set(u'name', u'title')
    meta_title.set(u'content', unicode(content.title))
    meta_abstract = lxml.etree.SubElement(head, 'meta')
    meta_abstract.set(u'name', u'abstract')
    meta_abstract.set(u'content', unicode(content.description))
    meta_author = lxml.etree.SubElement(head, 'meta')
    meta_author.set(u'name', u'author')
    meta_author.set(u'content', unicode(username))
    meta_contributor = lxml.etree.SubElement(head, 'meta')
    meta_contributor.set(u'name', u'contributor')
    meta_contributor.set(u'content', unicode(username))
    meta_date = lxml.etree.SubElement(head, 'meta')
    meta_date.set(u'name', u'date')
    meta_date.set(u'content', unicode(datetime.date.today().isoformat()))
    meta_copyright = lxml.etree.SubElement(head, 'meta')
    meta_copyright.set(u'name', u'copyright')
    meta_copyright.set(u'content', u'IRI')
    meta_type = lxml.etree.SubElement(head, 'meta')
    meta_type.set(u'name', u'type')
    meta_type.set(u'content', u'video')

    body = lxml.etree.SubElement(iri, 'body')
    lxml.etree.SubElement(body, 'ensembles')
    lxml.etree.SubElement(body, 'links')

    medias = lxml.etree.SubElement(body, 'medias')

    media_video = lxml.etree.SubElement(medias, 'media')
    media_video.set(u'id', u'video')
    video = lxml.etree.SubElement(media_video, 'video')
    video.set(u'src', unicode(content.stream_src))
    video.set(u'id', unicode(content.iri_id))
    video.set(u'dur', unicode(content.duration))
    video.set(u'begin', u'0')

    media_tool = lxml.etree.SubElement(medias, 'media')
    media_tool.set(u'id', u'tool')
    lxml.etree.SubElement(media_tool, 'tool')

    lxml.etree.SubElement(body, 'display')

    doc.write(file, pretty_print=True)


def update_iri(filepath, content, username):
    
    # open xml
    doc = lxml.etree.parse(filepath)
    
    res = doc.xpath("/iri/head/meta")
    # update meta
    
    for meta_node in res:
        meta_name = meta_node.get("name")
        content_attr = None
        if meta_name == u'id':
            content_attr = unicode(content.iri_id)
        elif meta_name == u'title':
            content_attr = unicode(content.title)
        elif meta_name == u'abstract':
            content_attr = unicode(content.description)
        elif meta_name == u'contributor':
            content_attr = unicode(username)
        elif meta_name == u"date":
            content_attr = unicode(datetime.date.today().isoformat())
        if content_attr is not None:
            meta_node.set(u"content", content_attr)

    res = doc.xpath("/iri/body/medias/media[@id='video']/video")
    
    if len(res) > 0:
        video_node = res[0]
        video_node.set(u'src', unicode(content.stream_src))
        video_node.set(u'dur', unicode(content.duration))
        video_node.set(u'id', unicode(content.iri_id))
    # update video
    
    f = open(filepath, "w")
    try:
        doc.write(f, encoding="UTF-8", pretty_print=True, xml_declaration=True)
    finally:
        f.close()
        
def clean_description(description):
    """ Remove html tags added by flash if necessary """
    new_desc = u''
    begin_str = "KERNING=\"0\">"  
                
    for chunk in description.split("<TEXTFORMAT"):
        begin = chunk.find(begin_str) + len(begin_str)
        end = chunk.find("</FONT")
    
        if begin > 0 and end > 0:
            new_desc = new_desc +  chunk[begin:end] + "<br />"
        
    if new_desc:
        return new_desc
    return None