#!/usr/bin/env python

# gquery.py: functions that deal with / transform SPARQL queries in grlc

import yaml
import json
from rdflib.plugins.sparql.parser import Query, UpdateUnit
from rdflib.plugins.sparql.processor import translateQuery
from flask import request, has_request_context
from pyparsing import ParseException
from pprint import pformat
import traceback
import re
import requests

# grlc modules
import grlc.static as static
import grlc.glogging as glogging

glogger = glogging.getGrlcLogger(__name__)

XSD_PREFIX = 'PREFIX xsd: <http://www.w3.org/2001/XMLSchema#>'

import SPARQLTransformer

def guess_endpoint_uri(rq, gh_repo):
    Guesses the endpoint URI from (in this order):
    - An endpoint parameter in URL
    - An #+endpoint decorator
    - A endpoint.txt file in the repo
    Otherwise assigns a default one
    if auth == ('none', 'none'):
        auth = None

    if has_request_context() and "endpoint" in request.args:
        endpoint = request.args['endpoint']
        glogger.info("Endpoint provided in request: " + endpoint)
        return endpoint, auth

    # Decorator
        decorators = get_yaml_decorators(rq)
        endpoint = decorators['endpoint']
        auth = None
        glogger.info("Decorator guessed endpoint: " + endpoint)
    except (TypeError, KeyError):
        # File
            endpoint_content = gh_repo.getTextFor({'download_url': 'endpoint.txt'})
            endpoint = endpoint_content.strip().splitlines()[0]
            auth = None
            glogger.info("File guessed endpoint: " + endpoint)
        # TODO: except all is really ugly
            # Default
            endpoint = static.DEFAULT_ENDPOINT
            if auth == ('none', 'none'):
                auth = None
            glogger.warning("No endpoint specified, using default ({})".format(endpoint))

    return endpoint, auth

def count_query_results(query, endpoint):
    Returns the total number of results that query 'query' will generate
    WARNING: This is too expensive just for providing a number of result pages
             Providing a dummy count for now

    # number_results_query, repl = re.subn("SELECT.*FROM", "SELECT COUNT (*) FROM", query)
    # if not repl:
    #     number_results_query = re.sub("SELECT.*{", "SELECT COUNT(*) {", query)
    # number_results_query = re.sub("GROUP\s+BY\s+[\?\_\(\)a-zA-Z0-9]+", "", number_results_query)
    # number_results_query = re.sub("ORDER\s+BY\s+[\?\_\(\)a-zA-Z0-9]+", "", number_results_query)
    # number_results_query = re.sub("LIMIT\s+[0-9]+", "", number_results_query)
    # number_results_query = re.sub("OFFSET\s+[0-9]+", "", number_results_query)
    # glogger.debug("Query for result count: " + number_results_query)
    # # Preapre HTTP request
    # headers = { 'Accept' : 'application/json' }
    # data = { 'query' : number_results_query }
    # count_json = requests.get(endpoint, params=data, headers=headers).json()
    # count = int(count_json['results']['bindings'][0]['callret-0']['value'])
    # glogger.info("Paginated query has {} results in total".format(count))
    # return count

    return 1000

def _getDictWithKey(key, dict_list):
    """ Returns the first dictionary in dict_list which contains the given key"""
    for d in dict_list:
        if key in d:
            return d
    return None

def get_parameters(rq, variables, endpoint, query_metadata, auth=None):
        ?_name The variable specifies the API mandatory parameter name. The value is incorporated in the query as plain literal.
        ?__name The parameter name is optional.
        ?_name_iri The variable is substituted with the parameter value as a IRI (also: number or literal).
        ?_name_en The parameter value is considered as literal with the language 'en' (e.g., en,it,es, etc.).
        ?_name_integer The parameter value is considered as literal and the XSD datatype 'integer' is added during substitution.
        ?_name_prefix_datatype The parameter value is considered as literal and the datatype 'prefix:datatype' is added during substitution. The prefix must be specified according to the SPARQL syntax.

    # variables = translateQuery(Query.parseString(rq, parseAll=True)).algebra['_vars']

    ## Aggregates
    internal_matcher = re.compile("__agg_\d+__")
    ## Basil-style variables
    variable_matcher = re.compile(

    parameters = {}
    for v in variables:
        if internal_matcher.match(v):

        match = variable_matcher.match(v)
        # TODO: currently only one parameter per triple pattern is supported
        if match:
            vname = match.group('name')
            vrequired = True if match.group('required') == '_' else False
            vtype = 'string'
            # All these can be None
            vcodes = get_enumeration(rq, vname, endpoint, query_metadata, auth)
            vdefault = get_defaults(rq, vname, query_metadata)
            vlang = None
            vdatatype = None
            vformat = None

            mtype = match.group('type')
            muserdefined = match.group('userdefined')

            if mtype in ['number', 'literal', 'string']:
                vtype = mtype
            elif mtype in ['iri']:  # TODO: proper form validation of input parameter uris
                vtype = 'string'
                vformat = 'iri'
            elif mtype:
                vtype = 'string'

                if mtype in static.XSD_DATATYPES:
                    vdatatype = 'xsd:{}'.format(mtype)
                elif len(mtype) == 2:
                    vlang = mtype
                elif muserdefined:
                    vdatatype = '{}:{}'.format(mtype, muserdefined)

            parameters[vname] = {
                'original': '?{}'.format(v),
                'required': vrequired,
                'name': vname,
                'type': vtype

            # Possibly None parameter attributes
            if vcodes is not None:
                parameters[vname]['enum'] = sorted(vcodes)
            if vlang is not None:
                parameters[vname]['lang'] = vlang
            if vdatatype is not None:
                parameters[vname]['datatype'] = vdatatype
            if vformat is not None:
                parameters[vname]['format'] = vformat
            if vdefault is not None:
                parameters[vname]['default'] = vdefault

            glogger.info('Finished parsing the following parameters: {}'.format(parameters))

    return parameters

def get_defaults(rq, v, metadata):
    Returns the default value for a parameter or None
    glogger.debug("Metadata with defaults: {}".format(metadata))
    if 'defaults' not in metadata:
        return None
    defaultsDict = _getDictWithKey(v, metadata['defaults'])
    if defaultsDict:
        return defaultsDict[v]
    return None

def get_enumeration(rq, v, endpoint, metadata={}, auth=None):
    Returns a list of enumerated values for variable 'v' in query 'rq'
    # glogger.debug("Metadata before processing enums: {}".format(metadata))
    # We only fire the enum filling queries if indicated by the query metadata
    if 'enumerate' not in metadata:
        return None
    enumDict = _getDictWithKey(v, metadata['enumerate'])
    if enumDict:
        return enumDict[v]
    if v in metadata['enumerate']:
        return get_enumeration_sparql(rq, v, endpoint, auth)
    return None

def get_enumeration_sparql(rq, v, endpoint, auth=None):
    Returns a list of enumerated values for variable 'v' in query 'rq'
    glogger.info('Retrieving enumeration for variable {}'.format(v))
    vcodes = []
    # tpattern_matcher = re.compile(".*(FROM\s+)?(?P<gnames>.*)\s+WHERE.*[\.\{][\n\t\s]*(?P<tpattern>.*\?" + re.escape(v) + ".*?\.).*", flags=re.DOTALL)
    # tpattern_matcher = re.compile(".*?((FROM\s*)(?P<gnames>(\<.*\>)+))?\s*WHERE\s*\{(?P<tpattern>.*)\}.*", flags=re.DOTALL)

    # WHERE is optional too!!
    tpattern_matcher = re.compile(".*?(FROM\s*(?P<gnames>\<.*\>+))?\s*(WHERE\s*)?\{(?P<tpattern>.*)\}.*",

    tp_match = tpattern_matcher.match(rq)
    if tp_match:
        vtpattern = tp_match.group('tpattern')
        gnames = tp_match.group('gnames')
        glogger.debug("Detected graph names: {}".format(gnames))
        glogger.debug("Detected BGP: {}".format(vtpattern))
        glogger.debug("Matched triple pattern with parameter")
        if gnames:
            codes_subquery = re.sub("SELECT.*\{.*\}.*",
                                    "SELECT DISTINCT ?" + v + " FROM " + gnames + " WHERE { " + vtpattern + " }", rq,
            codes_subquery = re.sub("SELECT.*\{.*\}.*",
                                    "SELECT DISTINCT ?" + v + " WHERE { " + vtpattern + " }", rq,
        glogger.debug("Codes subquery: {}".format(codes_subquery))
        codes_json = requests.get(endpoint, params={'query': codes_subquery},
                                  headers={'Accept': static.mimetypes['json'],
                                           'Authorization': 'token {}'.format(static.ACCESS_TOKEN)}, auth=auth).json()
        for code in codes_json['results']['bindings']:
        glogger.debug("No match between variable name and query.")

    return vcodes

def get_yaml_decorators(rq):
    Returns the yaml decorator metadata only (this is needed by triple pattern fragments)
    # glogger.debug('Guessing decorators for query {}'.format(rq))
    if not rq:
        return None

    yaml_string = ''
    query_string = ''
    if isinstance(rq, dict):  # json query (sparql transformer)
        if 'grlc' in rq:
            yaml_string = rq['grlc']
            query_string = rq
    else:  # classic query
        yaml_string = "\n".join([row.lstrip('#+') for row in rq.split('\n') if row.startswith('#+')])
        query_string = "\n".join([row for row in rq.split('\n') if not row.startswith('#+')])

    query_metadata = None
    if type(yaml_string) == dict:
        query_metadata = yaml_string
    elif type(yaml_string) == str:
        try:  # Invalid YAMLs will produce empty metadata
            query_metadata = yaml.load(yaml_string)
        except (yaml.parser.ParserError, yaml.scanner.ScannerError) as e:
                query_metadata = json.loads(yaml_string)
            except json.JSONDecodeError:
                glogger.warning("Query decorators could not be parsed; check your YAML syntax")

    # If there is no YAML string
    if query_metadata is None:
        query_metadata = {}
    query_metadata['query'] = query_string

    # glogger.debug("Parsed query decorators: {}".format(query_metadata))

    return query_metadata

def enable_custom_function_prefix(rq, prefix):
    """Add SPARQL prefixe header if the prefix is used in the given query."""
    if ' %s:' % prefix in rq or '(%s:' % prefix in rq and not 'PREFIX %s:' % prefix in rq:
        rq = 'PREFIX %s: <:%s>\n' % (prefix, prefix) + rq
    return rq

def get_metadata(rq, endpoint):
    Returns the metadata 'exp' parsed from the raw query file 'rq'
    'exp' is one of: 'endpoint', 'tags', 'summary', 'request', 'pagination', 'enumerate'
    query_metadata = get_yaml_decorators(rq)
    query_metadata['type'] = 'UNKNOWN'
    query_metadata['original_query'] = rq

    if isinstance(rq, dict):  # json query (sparql transformer)
        rq, proto, opt = SPARQLTransformer.pre_process(rq)
        rq = rq.strip()
        query_metadata['proto'] = proto
        query_metadata['opt'] = opt
        query_metadata['query'] = rq

    rq = enable_custom_function_prefix(rq, 'bif')
    rq = enable_custom_function_prefix(rq, 'sql')

        # THE PARSING
        # select, describe, construct, ask
        parsed_query = translateQuery(Query.parseString(rq, parseAll=True))
        query_metadata['type'] = parsed_query.algebra.name
        if query_metadata['type'] == 'SelectQuery':
            # Projection variables
            query_metadata['variables'] = parsed_query.algebra['PV']
            # Parameters
            query_metadata['parameters'] = get_parameters(rq, parsed_query.algebra['_vars'], endpoint, query_metadata)
        elif query_metadata['type'] == 'ConstructQuery':
            # Parameters
            query_metadata['parameters'] = get_parameters(rq, parsed_query.algebra['_vars'], endpoint, query_metadata)
                "Query type {} is currently unsupported and no metadata was parsed!".format(query_metadata['type']))
    except ParseException as pe:
        glogger.warning("Could not parse regular SELECT, CONSTRUCT, DESCRIBE or ASK query")
        # glogger.warning(traceback.print_exc())

        # insert queries won't parse, so we regex
        # glogger.info("Trying to parse INSERT query")
        # if static.INSERT_PATTERN in rq:
        #     query_metadata['type'] = 'InsertQuery'
        #     query_metadata['parameters'] = [u'_g_iri']

            # update query
            glogger.info("Trying to parse UPDATE query")
            parsed_query = UpdateUnit.parseString(rq, parseAll=True)
            query_metadata['type'] = parsed_query[0]['request'][0].name
            if query_metadata['type'] == 'InsertData':
                query_metadata['parameters'] = {
                    'g': {'datatype': None, 'enum': [], 'lang': None, 'name': 'g', 'original': '?_g_iri',
                          'required': True, 'type': 'iri'},
                    'data': {'datatype': None, 'enum': [], 'lang': None, 'name': 'data', 'original': '?_data',
                             'required': True, 'type': 'literal'}}

            glogger.info("Update query parsed with {}".format(query_metadata['type']))
            # if query_metadata['type'] == 'InsertData':
            #     query_metadata['variables'] = parsed_query.algebra['PV']
            glogger.error("Could not parse query")

    glogger.debug("Finished parsing query of type {}".format(query_metadata['type']))
    glogger.debug("All parsed query metadata (from decorators and content): ")
    glogger.debug(pformat(query_metadata, indent=32))

    return query_metadata

def paginate_query(query, results_per_page, get_args):
    """Modify the given query so that it can be paginated. The paginated query will 
    split display a maximum of `results_per_page`."""
    page = get_args.get('page', 1)

    glogger.info("Paginating query for page {}, {} results per page".format(page, results_per_page))

    # If contains LIMIT or OFFSET, remove them
    glogger.debug("Original query: " + query)
    no_limit_query = re.sub("((LIMIT|OFFSET)\s+[0-9]+)*", "", query)
    glogger.debug("No limit query: " + no_limit_query)

    # Append LIMIT results_per_page OFFSET (page-1)*results_per_page
    paginated_query = no_limit_query + " LIMIT {} OFFSET {}".format(results_per_page,
                                                                    (int(page) - 1) * results_per_page)
    glogger.debug("Paginated query: " + paginated_query)

    return paginated_query

def rewrite_query(query, parameters, get_args):
    """Rewrite query to replace query parameters for given values."""
    glogger.debug("Query parameters")
    requireXSD = False

    required_params = {}
    for k, v in parameters.items():
        if parameters[k]['required']:
            required_params[k] = v
    requiredParams = set(required_params.keys())
    providedParams = set(get_args.keys())
    glogger.debug("Required parameters: {} Request args: {}".format(requiredParams, providedParams))
    assert requiredParams.issubset(providedParams), 'Provided parameters do not cover the required parameters!'

    for pname, p in list(parameters.items()):
        # Get the parameter value from the GET request
        v = get_args.get(pname, None)
        # If the parameter has a value
        if not v:

        if isinstance(query, dict):  # json query (sparql transformer)
            if '$values' not in query:
                query['$values'] = {}
            values = query['$values']

            if not p['original'] in values:
                values[p['original']] = v
            elif isinstance(values[p['original']], list):
                values[p['original']] = [values[p['original']], v]


        # IRI
        if p['type'] == 'iri':  # TODO: never reached anymore, since iris are now type=string with format=iri
            query = query.replace(p['original'], "{}{}{}".format('<', v, '>'))
        # A number (without a datatype)
        elif p['type'] == 'number':
            query = query.replace(p['original'], v)
        # Literals
        elif p['type'] == 'literal' or p['type'] == 'string':
            # If it's a iri
            if 'format' in p and p['format'] == 'iri':
                query = query.replace(p['original'], "{}{}{}".format('<', v, '>'))
            # If there is a language tag
            if 'lang' in p and p['lang']:
                query = query.replace(p['original'], "\"{}\"@{}".format(v, p['lang']))
            elif 'datatype' in p and p['datatype']:
                query = query.replace(p['original'], "\"{}\"^^{}".format(v, p['datatype']))
                if 'xsd' in p['datatype']:
                    requireXSD = True
                query = query.replace(p['original'], "\"{}\"".format(v))

    if isinstance(query, dict):  # json query (sparql transformer)
        rq, proto, opt = SPARQLTransformer.pre_process(query)
        query = rq.strip()

    if requireXSD and XSD_PREFIX not in query:
        query = query.replace('SELECT', XSD_PREFIX + '\n\nSELECT')

    glogger.debug("Query rewritten as: " + query)

    return query