Python EventsPortal Beispiele

Beispiel #1

def addAllDataToSolrFromUrl(sourceUrl, patternUrl, solrUrl):
    """
    add data to a Solr index crawling events from a URL
    """

    data = getCoordinatesForAllData(sourceUrl, patternUrl)

    EventsPortal.addDataToSolr(data, solrUrl)

Beispiel #2

Datei: GetCoordinatesForAllData.py Projekt: elixirhub/events-portal-scraping-scripts

def addAllDataToSolrFromUrl(sourceUrl,patternUrl,solrUrl):
    """
    add data to a Solr index crawling events from a URL
    """

    data = getCoordinatesForAllData(sourceUrl,patternUrl)

    EventsPortal.addDataToSolr(data, solrUrl)

Beispiel #3

def getCoordinatesForAllData(sourceUrl, patternUrl):

    data = EventsPortal.getAllEventsData(sourceUrl, patternUrl)

    for docs in data:
        latitudeField = docs['latitude']

        if latitudeField == None:

            geocity = docs['locationCity']

            geolocator = Nominatim()
            locatorCity = geocity

            location = geolocator.geocode(locatorCity)

            latitude = location.latitude

            longitude = location.longitude

            docs['latitudetest'] = latitude

            docs['longitudetest'] = longitude

        else:
            pass

    return data

Beispiel #4

Datei: GetCoordinatesForAllData.py Projekt: elixirhub/events-portal-scraping-scripts

def getCoordinatesForAllData(sourceUrl,patternUrl):

    data = EventsPortal.getAllEventsData(sourceUrl,patternUrl)


    for docs in data:
        latitudeField = docs['latitude']

        if latitudeField == None:

            geocity = docs['locationCity']

            geolocator = Nominatim()
            locatorCity = geocity

            location = geolocator.geocode(locatorCity)

            latitude = location.latitude

            longitude = location.longitude

            docs['latitudetest'] = latitude

            docs['longitudetest'] = longitude

        else:
            pass

    return  data

Beispiel #5

Datei: DeleteDataFromUrlTest.py Projekt: elixirhub/events-portal-scraping-scripts

__author__ = 'chuqiao'
import EventsPortal
# DELETE ALL DATA
# EventsPortal.deleteDataInSolr()
# ADD DATA FROM 2 SORCES

# EventsPortal.addDataToSolrFromUrl("http://www.elixir-europe.org:8080/events", "http://www.elixir-europe.org:8080/events")
# EventsPortal.addDataToSolrFromUrl("http://bioevents-portal.org/eventsfull/test?state=published&field_type_tid=All", "http://bioevents-portal.org/events","localhost:8983/solr/event_portal")

# DELETE DATA FROM 1 SOURCE

EventsPortal.deleteDataInSolrFromUrl("http://bioevents-portal.org/eventsfull/test?state=published&field_type_tid=All","139.162.217.53:8983/solr/eventsportal")



# EventsPortal.deleteDataInSolrByQuery('source:("http://www.elixir-europe.org:8080/events")')
# EventsPortal.deleteDataInSolrByQuery('source:("http://localhost/ep/events" AND "state=published" AND "field_type_tid=All")')

Beispiel #6

Datei: AddDataTest.py Projekt: elixirhub/events-portal-scraping-scripts

    # add the handlers to the logger
    logger.addHandler(ch)
    logger.addHandler(fh)



# EventsPortal.addDataToSolrFromUrl("http://www.elixir-europe.org:8080/events", "http://www.elixir-europe.org:8080/events");
logger()

# EventsPortal.addDataToSolrFromUrl("http://bioevents-portal.org/eventsfull/test?state=published&field_type_tid=All", "http://bioevents-portal.org/events","http://139.162.217.53:8983/solr/eventsportal");
# EventsPortal.addDataToSolrFromUrl("http://bioevents-portal.org/eventsfull", "http://bioevents-portal.org/events","http://139.162.217.53:8983/solr/eventsportal");
# EventsPortal.addDataToSolrFromUrl("http://bioevents-portal.org/eventsfull/upcoming?state=published&field_type_tid=All", "http://bioevents-portal.org/events","http://139.162.217.53:8983/solr/eventsportal");



# EventsPortal.addDataToSolrFromUrl("http://bioevents-portal.org/eventsfull", "http://bioevents-portal.org/events","http://localhost:8983/solr/event_portal");

# EventsPortal.addDataToSolrFromUrl("http://bioevents-portal.org/eventsfull/test?state=published&field_type_tid=All", "http://bioevents-portal.org/events","139.162.217.53:8983/solr/eventsportal/")



if  __name__ == '__main__':

    logger.info('start at %s' % datetime.now())

    EventsPortal.addDataToSolrFromUrl("http://bioevents-portal.org/eventsfull",
                                      "http://bioevents-portal.org/events",
                                      "139.162.217.53:8983/solr/eventsportal/")


    logger.info('finish at %s' % datetime.now())

Beispiel #7

    # create formatter and add it to the handlers
    formatter = logging.Formatter(
        '%(asctime)s - %(name)s - %(levelname)s - %(message)s')
    fh.setFormatter(formatter)
    ch.setFormatter(formatter)
    # add the handlers to the logger
    logger.addHandler(ch)
    logger.addHandler(fh)


# EventsPortal.addDataToSolrFromUrl("http://www.elixir-europe.org:8080/events", "http://www.elixir-europe.org:8080/events");
logger()

# EventsPortal.addDataToSolrFromUrl("http://bioevents-portal.org/eventsfull/test?state=published&field_type_tid=All", "http://bioevents-portal.org/events","http://139.162.217.53:8983/solr/eventsportal");
# EventsPortal.addDataToSolrFromUrl("http://bioevents-portal.org/eventsfull", "http://bioevents-portal.org/events","http://139.162.217.53:8983/solr/eventsportal");
# EventsPortal.addDataToSolrFromUrl("http://bioevents-portal.org/eventsfull/upcoming?state=published&field_type_tid=All", "http://bioevents-portal.org/events","http://139.162.217.53:8983/solr/eventsportal");

# EventsPortal.addDataToSolrFromUrl("http://bioevents-portal.org/eventsfull", "http://bioevents-portal.org/events","http://localhost:8983/solr/event_portal");

# EventsPortal.addDataToSolrFromUrl("http://bioevents-portal.org/eventsfull/test?state=published&field_type_tid=All", "http://bioevents-portal.org/events","139.162.217.53:8983/solr/eventsportal/")

if __name__ == '__main__':

    logger.info('start at %s' % datetime.now())

    EventsPortal.addDataToSolrFromUrl(
        "http://bioevents-portal.org/eventsfull",
        "http://bioevents-portal.org/events",
        "139.162.217.53:8983/solr/eventsportal/")

    logger.info('finish at %s' % datetime.now())

Beispiel #8

Datei: DeleteDataFromUrlTest.py Projekt: elixirhub/events-portal-scraping-scripts

__author__ = 'chuqiao'
import EventsPortal
# DELETE ALL DATA
# EventsPortal.deleteDataInSolr()
# ADD DATA FROM 2 SORCES

# EventsPortal.addDataToSolrFromUrl("http://www.elixir-europe.org:8080/events", "http://www.elixir-europe.org:8080/events")
# EventsPortal.addDataToSolrFromUrl("http://bioevents-portal.org/eventsfull/test?state=published&field_type_tid=All", "http://bioevents-portal.org/events","localhost:8983/solr/event_portal")

# DELETE DATA FROM 1 SOURCE

EventsPortal.deleteDataInSolrFromUrl(
    "http://bioevents-portal.org/eventsfull/test?state=published&field_type_tid=All",
    "139.162.217.53:8983/solr/eventsportal")

# EventsPortal.deleteDataInSolrByQuery('source:("http://www.elixir-europe.org:8080/events")')
# EventsPortal.deleteDataInSolrByQuery('source:("http://localhost/ep/events" AND "state=published" AND "field_type_tid=All")')

Beispiel #9

__author__ = 'chuqiao'
import EventsPortal
EventsPortal.deleteDataInSolr("http://139.162.217.53:8983/solr/eventsportal")