def upgrade_downgrade(action):
    """Upgrade or downgrade index holdings.

    Correct items_count and public_items_count for holdings of type serial.
    :param str action: upgrade or downgrade.
    """
    index = HoldingsSearch.Meta.index
    query = HoldingsSearch()\
        .filter('term', holdings_type='serial') \
        .source(['pid'])

    ids = [(h.meta.id, h.pid) for h in query.scan()]
    count = 0

    LOGGER.info(f'Indexing {len(ids)} records ....')
    for (_id, pid) in ids:
        document = Document.get(_id, index=index, using=current_search_client)
        items_count, public_items_count = get_counts(pid, action)

        document.update(items_count=items_count,
                        public_items_count=public_items_count,
                        index=index,
                        using=current_search_client,
                        refresh=True)
        count += 1
        LOGGER.info(f'{count} records indexed.')
Beispiel #2
0
    def update(cls, _id, date, data):
        """Update all data for a record.

        :param str _id: Elasticsearch document ID.
        :param str date: Log date, useful for getting the right index.
        :param dict data: New record data.
        """
        index = cls.get_index({'date': date})

        document = Document.get(_id, index=index, using=current_search_client)

        # Assign each properties to the document
        for key, item in data.items():
            document[key] = item

        result = document.save(
            index=index,
            using=current_search_client,
            refresh=True,
        )

        if result != 'updated':
            raise Exception('Operation log cannot be updated.')
Beispiel #3
0
 def search_by_id(self, id):
     return Document.get(id, self.__es, self.__index)
Beispiel #4
0
    def get(self, request):
        id = request.GET.get("id", "")
        doc = Document.get(id=id, index='policynew', using=client).to_dict()
        link = doc['link']
        title = doc['title']

        policy_object_list = []
        # 从数据库中取出推荐的政策
        policy_recommend = PolicyRecommend.objects.filter(title=title)
        if len(policy_recommend) > 0:
            policy_object = policy_recommend[0]
            policy_str = policy_object.recommend
            policy_items = policy_str.split(",")[:3]
            policy_names = [item.split(' ')[0] for item in policy_items]
            policy_scores = [
                item.split(' ')[1] for item in policy_items
                if len(item.split(' ')[1]) == 7
            ]
            print(policy_scores)
            policy_recommend_urls = []
            for policy_name in policy_names:
                resp = Search(using=client,
                              index='policynew').query("match",
                                                       title=policy_name)
                response = resp.execute()
                link_str = response['hits']['hits'][0]["_id"]
                if (policy_name == response[0].title):
                    policy_recommend_url = "http://127.0.0.1:8000/detail/?id=" + link_str
                else:
                    policy_recommend_url = ""
                policy_recommend_urls.append(policy_recommend_url)
            policy_object_list = list(
                zip(policy_names, policy_scores, policy_recommend_urls))
            # relation_policies_dict[relation_policy] = relation_policy_url

        # 从数据库中取出相关的政策
        relation_policies_dict = {}
        relations = RelationPolicies.objects.filter(title=title)
        if len(relations) == 1:
            # print(relations[0].relation_policies)
            relation_policies = set(relations[0].relation_policies.split(','))
            # print(relation_policies)
            for relation_policy in relation_policies:
                if (relation_policy != ''):
                    resp = Search(using=client, index='policynew').query(
                        "match", title=relation_policy)
                    response = resp.execute()
                    link_str = response['hits']['hits'][0]["_id"]
                    if (relation_policy == response[0].title):
                        relation_policy_url = "http://127.0.0.1:8000/detail/?id=" + link_str
                    else:
                        relation_policy_url = ""
                    relation_policies_dict[
                        relation_policy] = relation_policy_url

        # 生成结果返回对象
        result_dict = {}
        result_dict['link'] = link
        if policy_object_list != []:
            print(policy_object_list)
            result_dict['policy_object_list'] = policy_object_list
        if relation_policies_dict != {}:
            result_dict['relation_policies_dict'] = relation_policies_dict
        return render(
            request, 'detail.html', {
                'link': link,
                'relation_policies_dict': relation_policies_dict,
                'policy_object_list': policy_object_list
            })
Beispiel #5
0
def merge_generated_parameters(params, idx, hash):
    """

    :param params:
    :param paramsfile:
    :param idx:
    :return:
    """

    layer_id = "%s_%s" % (hash, socket.gethostname())
    es = Elasticsearch(
        current_app.config.get("ELASTIC").split(","),
        verify_certs=False,
        timeout=120
    )

    #See if the hash exists
    try:
        doc = Document.get(id=layer_id, using=es, index=".datashader_layers")
    except NotFoundError:
        doc = None

    if not doc:
        #if not, create the hash in the db but only if it does not already exist
        try:
            doc = Document(_id=layer_id,
                            creating_host=socket.gethostname(),
                            creating_pid=os.getpid(),
                            creating_timestamp=datetime.now(),
                            generated_params=None,
                            params=params)
            doc.save(using=es, index=".datashader_layers", op_type="create", skip_empty=False)
            current_app.logger.debug("Created Hash document")
        except ConflictError:
            current_app.logger.debug("Hash document now exists, continuing")

        #re-fetch to get sequence number correct
        doc = Document.get(id=layer_id, using=es, index=".datashader_layers")

    #Check for generator timeouts:
    if doc.to_dict().get("generated_params", {}).get("generation_start_time") and \
                datetime.now() > datetime.strptime(doc.to_dict().get("generated_params", {}).get("generation_start_time"),"%Y-%m-%dT%H:%M:%S.%f")+timedelta(seconds=5*60):
        #Something caused the worker generating the params to time out so clear that entry
        try:
            doc.update(using=es, index=".datashader_layers", retry_on_conflict=0, refresh=True, \
                generated_params=None)
        except ConflictError:
            current_app.logger.debug("Abandoned resetting parameters due to conflict, other process has completed.")

    #Loop-check if the generated params are in missing/in-process/complete
    timeout_at = datetime.now()+timedelta(seconds=45)
    while doc.to_dict().get("generated_params", {}).get("complete", False) == False:
        if datetime.now() > timeout_at:
            current_app.logger.info("Hit timeout waiting for generated parameters to be placed into database")
            break
        #If missing, mark them as in generation
        if not doc.to_dict().get("generated_params", None):
            #Mark them as being generated but do so with concurrenty control
            #https://www.elastic.co/guide/en/elasticsearch/reference/current/optimistic-concurrency-control.html
            current_app.logger.info("Discovering generated parameters")
            generated_params = dict()
            generated_params["complete"] = False
            generated_params["generation_start_time"] = datetime.now()
            generated_params["generating_host"] = socket.gethostname()
            generated_params["generating_pid"] = os.getpid()
            try:
                doc.update(using=es, index=".datashader_layers", retry_on_conflict=0, refresh=True, \
                    generated_params=generated_params)
            except ConflictError:
                current_app.logger.debug("Abandoned generating parameters due to conflict, will wait for other process to complete.")
                break
            #Generate and save off parameters
            current_app.logger.warn("Discovering generated params")
            generated_params.update(generate_global_params(params, idx))
            generated_params["generation_complete_time"] = datetime.now()
            generated_params["complete"] = True
            #Store off generated params
            doc.update(using=es, index=".datashader_layers", retry_on_conflict=0, refresh=True, \
                    generated_params=generated_params)
            break
        else:
            time.sleep(1)
            doc = Document.get(id=layer_id, using=es, index=".datashader_layers")

    #We now have params so use them
    params["generated_params"] = doc.to_dict().get("generated_params")
    return params