from __future__ import print_function, unicode_literals, absolute_import, division
import sys
import re
import json

import requests
from bs4 import BeautifulSoup
from Levenshtein import distance

from tsammalexdata.util import DataProvider


def search_fuzzy(name):
    """
    http://eol.org/search?q=Ammodaucus+leucothricus&search=Go
<div id='main'>
Did you mean: <a href="/search?q=ammodaucus+leucotrichus">Ammodaucus leucotrichus</a>, <a href="/search?q=ammodaucus+leucotrichus+coss.">Ammodaucus leucotrichus coss.</a>, <a href="/search?q=oreocereus+leucotrichus">Oreocereus leucotrichus</a>, <a href="/search?q=acrolophus+leucotricha">Acrolophus leucotricha</a>, <a href="/search?q=ammodiscus+catinus">Ammodiscus catinus</a>, <a href="/search?q=ammodiscus+excertus">Ammodiscus excertus</a> ?
<div class='filtered_search'>
    """
    res = requests.get('http://eol.org/search', params=dict(q=name, search='Go'))
    maindiv = BeautifulSoup(res.text).find('div', id='main')
    candidate = None
    if maindiv:
        for a in maindiv.find_all('a'):
            if a['href'].startswith('/search?q='):
                if distance(name.decode('utf8'), a.text.strip()) < 3:
                    candidate = a.text
                #print(a.text)
    if candidate:
        print(name.decode('utf8'), '->', candidate)
    return candidate


class EOL(DataProvider):
    host = 'eol.org'

    def _path(self, name, id=None):
        path = '1.0'
        if id:
            path += '/%s' % id
        return 'api/%s/%s.json' % (name, path)

    def _api(self, name, id=None, **kw):
        try:
            return self.get(self._path(name, id), **kw)
        except ValueError:
            return {}

    def get_id(self, name):
        res = self._api('search', q=name, page=1, exact='false')
        if res.get('results'):
            for result in res['results']:
                if result['title'] == name:
                    return result['id']
            return res['results'][0]['id']

    def get_taxon_concept(self, data):
        if data.get('taxonConcepts'):
            # augment the data with complete classification info for one taxonomy:
            for tc in data['taxonConcepts']:
                if tc['nameAccordingTo'].startswith('Species 2000'):
                    # this is our preferred taxonomy ...
                    return tc
            # ... but any will do :)
            return data['taxonConcepts'][0]

    def get_info(self, id):
        """Extract classification and synonym/common name data from EOL for a given taxon.

        :param id: EOL id of a taxon.
        :return: dict with information.
        """
        kw = dict(
            images=0,
            videos=0,
            sounds=0,
            maps=0,
            text=0,
            iucn='true',
            subjects='overview',
            licenses='all',
            details='true',
            common_names='true',
            synonyms='true',
            references='false',
            vetted=0)
        data = self._api('pages', id, **kw)
        if isinstance(data, list):
            return {}
        tc = self.get_taxon_concept(data)
        if tc:
            taxonomy = self._api('hierarchy_entries', tc['identifier'])
            data.update(ancestors=taxonomy['ancestors'])
        return data

    def update(self, taxon, data):
        for ancestor in data.get('ancestors', []):
            if 'taxonRank' not in ancestor:
                continue
            for k in 'kingdom phylum class order family genus'.split():
                if ancestor['taxonRank'] == k:
                    taxon[k] = ancestor['scientificName'].split()[0]
                    break
        tc = self.get_taxon_concept(data)
        if tc and 'taxonRank' in tc:
            taxon['taxonRank'] = tc['taxonRank'].lower()
        for vn in data.get('vernacularNames', []):
            if vn.get('language') == 'en' and vn.get('eol_preferred'):
                taxon['english_name'] = vn['vernacularName']
                break


if __name__ == '__main__':
    api = EOL()
    #args = sys.argv[1:]
    #if args:
    #    search_fuzzy(args[0])
    if sys.argv[1:]:
        if len(sys.argv[1:]) > 1:
            api.refresh(*sys.argv[1:])
        else:
            print(json.dumps(api.cli(sys.argv[1]), indent=4))