helpers.py

# encoding: utf-8

import logging
import traceback
import ast
import ckan.plugins.toolkit as toolkit
import ckan.logic as logic
import ckan.model as model
import ckan.lib.helpers as helpers
import json
from ckan.common import c
import datetime
from dateutil import parser
from ckan.common import config
import urllib.request, urllib.parse, urllib.error
import hashlib
import re
import csv
import urllib.request, urllib.error, urllib.parse
from ckan.common import request
from urllib.parse import urlsplit, urlunsplit
import ckan.lib.helpers as helpers
from collections import OrderedDict
import pkg_resources


CKAN_TYPES = {'http://dcat-ap.de/def/datasetTypes/collection': 'collection'}

get_action = logic.get_action
log = logging.getLogger(__name__)


def odsh_openness_score_dataset_html(dataset):
    score = -1 
    #dataset = json.loads(dataset)
    resources = dataset.get('resources')
    if resources is None:
        return score 
    for resource in resources:
        r_qa = resource.get('qa')
        if r_qa:
            try:
                qa = None
                # r_qa might be a string of a dictionary when 'dataset' is send from solr
                if isinstance(r_qa, str):
                    qa = ast.literal_eval(r_qa)
                else:
                    qa = r_qa
                resource_score = qa.get('openness_score')
                if resource_score > score:
                    score = resource_score
            except AttributeError as e:
                log.error('Error while calculating openness score %s: %s\nException: %s',
                          e.__class__.__name__,  str(e), traceback.format_exc())
    return score


def odsh_get_resource_details(resource_id):
    resource_details = toolkit.get_action('resource_show')(
        data_dict={'id': resource_id})
    return resource_details


def odsh_get_resource_views(pkg_dict, resource):

    context = {'model': model, 'session': model.Session,
               'user': c.user, 'for_view': True,
               'auth_user_obj': c.userobj}
    return get_action('resource_view_list')(
        context, {'id': resource['id']})


def odsh_get_bounding_box(pkg_dict):
    try:
        extras = pkg_dict.get('extras')
        spatial = None
        for f in extras:
            if 'key' in f and f['key'] == 'spatial':
                spatial = f['value']
                break

        if spatial is not None:
            d = json.loads(spatial)
            if 'coordinates' in d:
                coords = d['coordinates']
                return compute_bounding_box(coords)
    except Exception as e:
        log.error('Error while bounding box %s: %s\nException: %s',
                  e.__class__.__name__,  str(e), traceback.format_exc())
    return None


def compute_bounding_box(coords):
    if len(coords) == 0:
        return None

    if type(coords[0]) != list:
        return [coords[0], coords[0], coords[1], coords[1]]

    coords = [c for sublist in coords for c in sublist]
    if type(coords[0][0]) == list:
        # multipolygon
        coords = [c for sublist in coords for c in sublist]

    minx = min(coords, key=lambda t: t[0])[0]
    maxx = max(coords, key=lambda t: t[0])[0]
    miny = min(coords, key=lambda t: t[1])[1]
    maxy = max(coords, key=lambda t: t[1])[1]

    return [maxx, minx, maxy, miny]


def odsh_get_spatial_text(pkg_dict):
    extras = pkg_dict.get('extras')
    spatial = None
    if extras is None:
        return None
    for f in extras:
        if 'key' in f and f['key'] == 'spatial_text':
            spatial = f['value']
            return spatial
    return None

def extend_search_convert_local_to_utc_timestamp(str_timestamp):
    if not str_timestamp:
        return None 

    if not re.match(r'\d\d\d\d-\d\d-\d\d', str_timestamp):
        raise ValueError('wrong format')
    
    dt = parser.parse(str_timestamp, dayfirst=False).isoformat()

    return dt+"Z"

def odsh_render_datetime(datetime_, fromIso=True):
    date_format='{0.day:02d}.{0.month:02d}.{0.year:04d}'
    if not datetime_:
        return ''
    if not re.match(r'\d\d\d\d-\d\d-\d\d', datetime_):
        return ''
    try:
        if fromIso:
            DATETIME_FORMAT = '%Y-%m-%dT%H:%M:%S'
        else:
            DATETIME_FORMAT = '%Y-%m-%d'

        dt = parser.parse(datetime_, dayfirst=False)
        return date_format.format(dt)

    except:
        return ''


def odsh_resource_formats(as_simple_list = False):
    """
    Retrieves a list of file formats from an RDF file using string manipulation and sorts them alphabetically.

    Returns:
        A list of dictionaries containing key-value pairs for each file format.
    """

    # Path to the RDF file with resource formats
    extension_path = pkg_resources.resource_filename('ckanext.odsh', '')
    filepath = config.get(
        'ckanext.odsh.resource_formats_fallback_filepath', extension_path + '/resources/fileformats.rdf')

    # Read the contents of the RDF file
    with open(filepath, 'r', encoding='utf-8') as file:
        rdf_content = file.read()

    # Regular expression pattern to extract the file format values
    pattern = r'<rdf:Description rdf:about="http:\/\/publications.europa.eu\/resource\/authority\/file-type\/(\w+)">'

    # List to store the key-value pairs
    file_formats = []

    # Find all matches of the pattern in the RDF content
    matches = re.findall(pattern, rdf_content)

    # Iterate over the matches and add key-value pairs to the list
    for match in matches:
        file_formats.append({'key': match, 'value': match})
    
    # Sort the list alphabetically based on the file format values
    file_formats = sorted(file_formats, key=lambda x: x['value'])

    if(as_simple_list):
        file_formats = [obj['value'] for obj in file_formats]
    
    return file_formats


def odsh_encodeurl(url):
    return urllib.parse.quote(url, safe='')


def odsh_create_checksum(in_string):
    hashstring = hashlib.md5(in_string.encode('utf-8')).hexdigest()
    return int(hashstring, base=16)


def odsh_extract_error(key, errors, field='extras'):
    if not errors or not (field in errors):
        return None
    ext = errors[field]
    for item in ext:
        if 'key' in item:
            for error in item['key']:
                if error.startswith(key):
                    return error

def odsh_extract_error_new(key, errors):
    if not errors or not ('__extras' in errors):
        return None
    error = errors['__extras'][0].get(key,None)
    if error:
        return key + ': ' + error

def odsh_extract_value_from_extras(extras, key):
    if not extras:
        return None
    for item in extras:
        if 'key' in item and item['key'].lower() == key.lower():
            if 'value' in item:
                return item['value']
            return None


def presorted_license_options(existing_license_id=None):
    '''Returns [(l.title, l.id), ...] for the licenses configured to be
    offered. Always includes the existing_license_id, if supplied.
    '''
    register = model.Package.get_license_register()
    licenses = list(register.values())
    license_ids = [license.id for license in licenses]
    if existing_license_id and existing_license_id not in license_ids:
        license_ids.insert(0, existing_license_id)
    return [('','')]+[
        (license_id,
         register[license_id].title if license_id in register else license_id)
        for license_id in license_ids]


def odsh_has_more_facets(facet, limit=None, exclude_active=False):
    facets = []
    for facet_item in c.search_facets.get(facet)['items']:
        if not len(facet_item['name'].strip()) or facet_item['count']==0:
            continue
        if not (facet, facet_item['name']) in list(request.args.items()):
            facets.append(dict(active=False, **facet_item))
        elif not exclude_active:
            facets.append(dict(active=True, **facet_item))
    if c.search_facets_limits and limit is None:
        limit = c.search_facets_limits.get(facet)
    if limit is not None and len(facets) > limit:
        return True
    return False


def odsh_public_url():
    return config.get('ckanext.odsh.public_url')

def spatial_extends_available():
    extension_path = pkg_resources.resource_filename('ckanext.odsh', '')
    file_path = config.get('ckanext.odsh.spatial.mapping', extension_path + '/resources/schleswig-holstein_geojson.csv')

    try:
        mapping_file = urllib.request.urlopen(file_path)
    except Exception:
        raise Exception("Could not load spatial mapping file!")

    spatial_text = str()
    spatial = str()
    cr = csv.reader(mapping_file, delimiter="\t")
    result = []
    for row in cr:
        spatial_text = row[1]
        result.append(spatial_text.decode('UTF-8'))
    return result

def odsh_public_resource_url(res):
    home = config.get('ckanext.odsh.public_url')
    if res.get('url_type',None) == 'upload' and 'url' in res:
        f = urlsplit(res['url'])
        return urlunsplit(('', '', f[2], f[3], f[4]))
    else:
        return res['url']


def odsh_show_testbanner():
    return config.get('ckanext.odsh.showtestbanner', False) == True


def odsh_get_facet_items_dict(name, search_facets, limit=None):
    '''
    Gets all facets like 'get_facet_items_dict' but sorted alphabetically
    instead by count.
    '''
    if name == 'groups':
        limit = 20
    facets = helpers.get_facet_items_dict(name, search_facets, limit)
    facets.sort(key=lambda it: (it['display_name'].lower(), -it['count']))
    return facets


def odsh_main_groups():
    '''Return a list of the groups to be shown on the start page.'''

    # Get a list of all the site's groups from CKAN, sorted by number of
    # datasets.
    groups = toolkit.get_action('group_list')(
        data_dict={'all_fields': True})

    return groups


def odsh_now():
    return helpers.render_datetime(datetime.datetime.now(), "%Y-%m-%d")


def odsh_group_id_selected(selected, group_id):
    if type(selected) is not list:
        selected = [selected]
    for g in selected:
        if (isinstance(g, str) and group_id == g) or (type(g) is dict and group_id == g['id']):
            return True

    return False


# def odsh_remove_route(map, routename):
#     route = None
#     for i, r in enumerate(map.matchlist):

#         if r.name == routename:
#             route = r
#             break
#     if route is not None:
#         map.matchlist.remove(route)
#         for key in map.maxkeys:
#             if key == route.maxkeys:
#                 map.maxkeys.pop(key)
#                 map._routenames.pop(route.name)
#                 break


def is_within_last_month(date, date_ref=None):
    '''
    date is a datetime.date object containing the date to be checked
    date_ref is a datetime.date object containing the reference date
    if date_ref is not specified, the date of today is used
    this method is needed by the method OdshPlugin.before_view in plugin.py
    '''
    
    if not date_ref:
        date_ref = datetime.date.today()
    
    [year_ref, month_ref, day_ref] = [date_ref.year, date_ref.month, date_ref.day]

    try:
        if month_ref > 1:
            one_month_ago = datetime.date(year_ref, month_ref-1, day_ref)
        else:
            one_month_ago = datetime.date(year_ref-1, 12, day_ref)
    except ValueError:
        # this happens if month before month_ref has less days than month_ref
        one_month_ago = datetime.date(year_ref, month_ref, 1) - datetime.timedelta(days=1)
    
    if date > one_month_ago:
        return True
    return False

def get_all_datasets_belonging_to_collection(context, collection_name):
    rel_collection_dict = dict({"id": collection_name})
    name_list = list()
    try:
        list_rel_collection = get_action('package_relationships_list')(context, rel_collection_dict)
    except AssertionError:
        #if there does not exist an relationship, returns an empty list
        return name_list 
    for item in list_rel_collection:
        item_object = item.get('object') 
        name_list.append(item_object)
    return name_list

def get_all_datasets_belonging_to_collection_by_dataset(context, dataset_name):
    collection_name = get_collection_name_by_dataset(context, dataset_name)
    if collection_name:
        name_list = get_all_datasets_belonging_to_collection(context, collection_name)
        return name_list
    return list()

def get_collection_name_by_dataset(context, dataset_name):
    rel_dataset_dict = dict({"id" : dataset_name})
    list_rel_dataset = toolkit.get_action('package_relationships_list')(context, rel_dataset_dict)
    if not len(list_rel_dataset):
        return None    
    collection_name = list_rel_dataset[0]['object']
    return collection_name

def get_successor_and_predecessor_dataset(context, pkg_dict):
    dataset_name = pkg_dict.get('name')
    siblings_dicts_with_access = _get_siblings_dicts_with_access(context, pkg_dict)
    if siblings_dicts_with_access:
        n_siblings = len(siblings_dicts_with_access)
        siblings_dicts_sorted_by_date_issued = _sort_siblings_by_name_and_date(siblings_dicts_with_access)
        siblings_names_sorted_by_date_issued = [d['name'] for d in siblings_dicts_sorted_by_date_issued]
        id_current_dataset = siblings_names_sorted_by_date_issued.index(dataset_name)
        predecessor_name = (
            siblings_names_sorted_by_date_issued[id_current_dataset-1] if (id_current_dataset > 0) 
            else None
        )
        successor_name = (
            siblings_names_sorted_by_date_issued[id_current_dataset+1] if (id_current_dataset < n_siblings-1) 
            else None
        )
    else:
        predecessor_name, successor_name = None, None
    return successor_name, predecessor_name

def _get_siblings_dicts_with_access(context, pkg_dict):
    dataset_name = pkg_dict.get('name')
    list_of_siblings = get_all_datasets_belonging_to_collection_by_dataset(context, dataset_name)
    n_siblings = len(list_of_siblings)
    if n_siblings>0:
        siblings_dicts = [get_package_dict(name) for name in list_of_siblings]
        user_has_access = lambda pkg_dict:helpers.check_access('package_show', pkg_dict)
        siblings_dicts_with_access = list(filter(user_has_access, siblings_dicts))
        return siblings_dicts_with_access
    return None

    
def _sort_siblings_by_name_and_date(siblings_dicts):
    '''
    sort by name first and then by date to have a fallback if dates are the same
    '''
    _get_name = lambda pkg_dict:pkg_dict.get('name')
    _get_issued = lambda pkg_dict:odsh_extract_value_from_extras(pkg_dict.get('extras'), 'issued')
    siblings_dicts_sorted_by_name = sorted(siblings_dicts, key=_get_name)
    siblings_dicts_sorted_by_date_issued = sorted(siblings_dicts_sorted_by_name, key=_get_issued)
    return siblings_dicts_sorted_by_date_issued


def get_package_dict(name):
    return model.Package.get(name).as_dict()

def get_successor_and_predecessor_urls(context, pkg_dict):
    successor_name, predecessor_name = get_successor_and_predecessor_dataset(context, pkg_dict)
    successor_url, predecessor_url = (
        helpers.url_for('dataset.read', id=name)
        if name is not None
        else None
        for name in (successor_name, predecessor_name)
    )
    return successor_url, predecessor_url

def short_name_for_category(category_name):
    translations = {
        'soci': 'Bevölkerung',
        'educ': 'Bildung',
        'ener': 'Energie',
        'heal': 'Gesundheit',
        'intr': 'Internationales',
        'just': 'Justiz',
        'agri': 'Landwirtschaft',
        'gove': 'Regierung',
        'regi': 'Regionales',
        'envi': 'Umwelt',
        'tran': 'Verkehr',
        'econ': 'Wirtschaft',
        'tech': 'Wissenschaft',
    }
    return translations.get(category_name)

def odsh_load_mdk_sample_dataset():
    '''
    Load sample dataset (Musterkatalog/Musterdatensatz).
    
    See https://bertelsmannstift.github.io/Musterdatenkatalog/def/musterdatensatz.rdf
    and corresponding mapping in mdk_mapping.json file.
    '''

    extension_path = pkg_resources.resource_filename('ckanext.odsh', '')
    default_sample_data_file_path = extension_path + '/resources/mdk_mapping.json'
    sample_data_file_path = config.get(
        'ckanext.odsh.sample_data_file_path', default_sample_data_file_path)
    
    try:
        with open(sample_data_file_path) as mapping_json:
             MDK_MAPPING = json.loads(mapping_json.read(), object_pairs_hook=OrderedDict)
             default = [{'value': 'Musterdatensatz wählen..', 'key': ''}]
             mdk = [{'key': key, 'value': MDK_MAPPING[key]} for key in MDK_MAPPING]
             result = default+mdk
    except IOError as err:
        log.error(
            'Could not load sample dataset mapping file from {}'
            .format(sample_data_file_path)
        )
        raise
    except ValueError as err:
        log.error(
            'Could not convert sample dataset mapping file from json. \nSample dataset mapping file: {}'
            .format(sample_data_file_path)
        )
        raise
    return result

def odsh_load_raw_mdk_sample_dataset():
    '''
    Load sample dataset (Musterkatalog/Musterdatensatz).
    
    See https://bertelsmannstift.github.io/Musterdatenkatalog/def/musterdatensatz.rdf
    and corresponding mapping in mdk_mapping.json file.
    '''

    extension_path = pkg_resources.resource_filename('ckanext.odsh', '')
    default_sample_data_file_path = extension_path + '/resources/mdk_mapping.json'
    sample_data_file_path = config.get(
        'ckanext.odsh.sample_data_file_path', default_sample_data_file_path)
    
    try:
        with open(sample_data_file_path) as mapping_json:
             result = json.loads(mapping_json.read(), object_pairs_hook=OrderedDict)
    except IOError as err:
        log.error(
            'Could not load sample dataset mapping file from {}'
            .format(sample_data_file_path)
        )
        raise
    except ValueError as err:
        log.error(
            'Could not convert sample dataset mapping file from json. \nSample dataset mapping file: {}'
            .format(sample_data_file_path)
        )
        raise
    return result


def map_dct_type_to_ckan_type(dct_type):
    '''
    matches the field dct:type from a harvested rdf file 
    to the corresponding ckan package type
    '''
    ckan_type = CKAN_TYPES.get(dct_type)
    return ckan_type

def map_ckan_type_to_dct_type(ckan_type):
    DCT_TYPES = _revert_dict(CKAN_TYPES)
    dct_type = DCT_TYPES.get(ckan_type)
    return dct_type

def _revert_dict(d):
    d_inverse = {v: k for k, v in d.items()}
    return d_inverse

def add_pkg_to_collection(id_pkg, id_collection):
    if id_pkg and id_collection:
        relationship_dict = {
            'subject': id_pkg,
            'object': id_collection,
            'type': 'child_of',
        }
        toolkit.get_action('package_relationship_create')(None, relationship_dict)

def correct_missing_relationship(pkg_dict, pkg_relationships_from_model):
    '''
    This function corrects missing relationship in show package.
    Note this fix is only good with one or non relationship. 
    This error is well known but was not fixed. https://github.com/ckan/ckan/issues/3114
    The error causes the deletation of relationships, because package_show is
    used in resource_create to get the package. 
    '''
    if pkg_relationships_from_model:
        relationship_from_model = pkg_relationships_from_model[0]
        relationship_list_from_dict = pkg_dict.get('relationships_as_subject')
        type_pkg = pkg_dict.get('type')
        needs_update = type_pkg == 'dataset' and not relationship_list_from_dict
        if needs_update:
            relationship_for_package = {
                '__extras': {
                    'object_package_id': relationship_from_model.object_package_id,
                    'subject_package_id': relationship_from_model.subject_package_id,
                },
                'comment': relationship_from_model.subject_package_id,
                'id': relationship_from_model.id,
                'type': relationship_from_model.type,
            }
            pkg_dict['relationships_as_subject'].append(relationship_for_package) 
    return pkg_dict

def get_pkg_relationships_from_model(pkg_dict):
    pkg_id = pkg_dict.get('id')
    return model.Package.get(pkg_id).get_relationships()

def load_language_mapping():
    extension_path = pkg_resources.resource_filename('ckanext.odsh', '')
    file_path = config.get('ckanext.odsh.language_mapping', extension_path + '/resources/language_mapping.json')
    with open(file_path) as language_mapping_json:
        LANGUAGE_MAPPING = json.loads(language_mapping_json.read())
    return LANGUAGE_MAPPING

def load_json_to_ordered_dict(json_str):
    return json.loads(json_str, object_pairs_hook=OrderedDict)

def get_language_of_package(pkg_dict):
    LANGUAGE_MAPPING = load_language_mapping()
    language_id = _get_language_id(pkg_dict)
    if not language_id:
        return None
    language = LANGUAGE_MAPPING.get(language_id)
    return language

def get_language_icon(pkg_dict):
    ICONS = {
        "http://publications.europa.eu/resource/authority/language/DAN": '/base/images/icon_lang_danish.png',
        "http://publications.europa.eu/resource/authority/language/ENG": '/base/images/icon_lang_english.png',
    }
    language_id = _get_language_id(pkg_dict)
    if not language_id:
        return None
    return ICONS.get(language_id) 

def _get_language_id(pkg_dict):
    language_id = odsh_extract_value_from_extras(pkg_dict.get('extras'), 'language')
    language_id = pkg_dict.get('language')
    if not language_id:
        language_id = odsh_extract_value_from_extras(
            pkg_dict.get('extras'), 'language'
        )
    if not language_id:
        return None
    language_id_cleaned = re.sub('[\[\]\"]', '', language_id)
    return language_id_cleaned

def get_spatial_for_selection():
    extension_path = pkg_resources.resource_filename('ckanext.odsh', '')
    file_path = config.get('ckanext.odsh.spatial.mapping', extension_path + '/resources/schleswig-holstein_geojson.csv')
    with open(file_path, newline='') as mapping_file:
        cr = csv.reader(mapping_file, delimiter="\t")
        spatial_mapping = list(cr)

    unique_mapping = []
    seen_values = set()
    for key, value, _ in spatial_mapping:
        if value in seen_values:
            continue  # Skip if the value has already been seen

        if "municipalityKey" in key:
            unique_mapping.append({'key': key, 'value': value})
        else:
            # Check if there is a municipality key entry for the value
            municipality_entry = next(
                (entry for entry in spatial_mapping if entry[1] == value and "municipalityKey" in entry[0]),
                None
            )

            if municipality_entry:
                # If a municipality key entry exists, use it instead of the current key
                unique_mapping.append({'key': municipality_entry[0], 'value': value})
            else:
                # Otherwise, use the current key
                unique_mapping.append({'key': key, 'value': value})

        seen_values.add(value)

    unique_mapping.append({'key': '', 'value': ''})
    return unique_mapping


def get_language_for_selection():
    LANGUAGE_MAPPING = load_language_mapping()
    dict_for_select_box = [{'key': key, 'value': LANGUAGE_MAPPING[key]} for key in LANGUAGE_MAPPING]
    return dict_for_select_box

def get_package_dict(name):
    '''
    raises ckan.logic.NotFound if not found
    '''
    package_dict = toolkit.get_action('package_show')(None, {'id': name})
    return package_dict

def size_of_fmt(num, suffix='B'):
    for unit in ['',' k',' M',' G',' T',' P',' E',' Z']:
        if abs(num) < 1000.0:
            return "%3.1f%s%s" % (num, unit, suffix)
        num /= 1000.0
    return "%.1f%s%s" % (num, 'Y', suffix)

def get_resource_size(resource):
    resource_size = resource.get('size')
    if resource_size:
        return size_of_fmt(resource_size)    


def get_address_org(organization):
    list_extras = organization.get('extras')
    address = dict()
    if not list_extras:
        return address
    for extra in list_extras:
            address.update({extra.get('key'):extra.get('value')})
    web = address.get('web')
    if web and not web.startswith('http'):
         web = 'http://' + web
         address.update({'web':web})    
    return address


def get_body_mail(organization, package):
    package_name = package.get('name')
    url = helpers.url_for('dataset.read', id=package_name, qualified = True)
    title = package.get('title')
    anrede = "Sehr geehrte Damen und Herren," + "%0D%0A" +  "%0D%0A" + "zu folgendem Eintrag habe ich eine Anmerkung/Frage:" + "%0D%0A" + "%0D%0A" 
    mail_titel = "Titel: " + title + "%0D%0A"    
    mail_document = "Dokument-ID: " +  package_name + "%0D%0A"
    mail_url = "URL: " +  url + "%0D%0A"  +  "%0D%0A" 
    message =  mail_titel + mail_document  +  mail_url + "Mein Kommentar:" +  "%0D%0A"    +  "%0D%0A"  +  "%0D%0A"  +  "%0D%0A" 
    return anrede + message