#!/usr/bin/python
# -*- coding: utf-8 -*-

import cgitb
cgitb.enable()
import sys, os, cgi, json
from HitEx.matching_set2 import MatchingSet
from partial_tree_kernel import calculate_similarity

#activate_this = os.path.join(os.path.dirname(__file__), 'venv/bin/activate_this.py')
#execfile(activate_this, dict(__file__=activate_this))

def create_mset_json(parameters, criteria):
    ms = MatchingSet()
    ms.initialize(parameters, criteria)
    ms.create_set()
    url = ms.get_url()
    j = ms.to_json()
  
    return (ms, j, url)

def main():
    """Main CGI handler."""
    # Print JSON header
    print_header()
    
    # Convert form fields to regular dictionary
    form = dict((field, cgi.FieldStorage().getvalue(field)) for field in cgi.FieldStorage().keys())
    default_parameters = { 
        #"query_type": "lemma", # wordform or cqp or lemma
        #"query_w" : u'bröd',   # u"huset", [deprel = "SS" & lemma contains "språk"].decode("utf-8")
        #"query_w" : '[ref = "01" & pos = "NN" & lemma contains "dörr"]'.decode("utf-8"),
        #"query_w" :'[msd = "VB.INF.AKT" &]'.decode("utf-8"),
        "query_pos" : "any",
        #"corpus_list":["ROM99","GP2012","LASBART"], #randomly pick one of them? "BLOGGMIX2012"
        "corpus_list":["rom99","gp2010","gp2011","gp2012","gp2013","gp2d","attasidor","lasbart","suc3","talbanken"], #"wikipedia-sv" "sweachum","sweacsam"
        "max_kwics": 300,      # nr KWICs to process (limited for efficiency reasons)
        "maxhit": 10,          # maximum number of matches to return
        "target_edge" : "end", # to which edge the keyword should be close to
        "proportion" : 50,     # within which percentage of the sent the keyword should appear
        "target_cefr" : "B1",
        "voc_thr" : 0,         # percentage of words above the target CEFR level
        "min_len" : 6,
        "max_len" : 20, 
        "non_alpha_thr": 30,
        "non_lemmatized_thr": 30,
        "lex_to_func_thr": "", #0.8,
        "sensitive_voc_cats": ["all"], # ["sex", "violence", "other", "religion", "secretion"],
        "preserve_bad":True}
 
    default_criteria = {
        "well_formedness":{"root":"filter", "sent_tokenization":"filter", 
                        "elliptic":"filter", "non_alpha":"filter", 
                        "non_lemmatized":"filter"}, 
        "isolability":{"struct_conn":"filter", "yn_answer":"filter", 
                     "anaphora-PN":"filter", "anaphora-AB":"filter"}, 
        "readability":"filter", 
        "typicality": "ranker",
        "sensitive_voc": "filter",
        "other_criteria":{"length":"filter",
                         "proper_name":"filter",
                         "repkw":"filter",
                         "kw_position":"",
                         "modal_verb":"",
                         "participle":"",  # same as korp
                         "sverb":"",       # same as korp
                         "interrogative":"filter",
                         "neg_form":"filter",
                         "abbrev":"filter",
                         "direct_speech":"filter",
                         "diff_voc_kelly":"rank",
                         "svalex_fr":"filter",
                         "out_of_svalex":"filter"}}

    # Get input values + analyze text
   		 
    if form.get("command", "") == "hitex":
        if "query_type" in form:
            query_type = form["query_type"]
        else:
            json.dumps({"Error":"Missing 'query_type'"})
        if "query_w" in form:
            word = form["query_w"]
        else:
            json.dumps({"Error":"Missing 'query_w'"})
        #default_parameters["target_cefr"] = form.get("target_cefr", "any")
        parameters = {}
        criteria = {}
        print "started"
        if form["use_defaults"]:
            parameters = default_parameters
            parameters["query_type"] = query_type
            parameters["query_w"] = word
            criteria = default_criteria
        else:
            for p in default_parameters:
                if p in form:
                    parameters[p] = form[p]
            for cr in default_criteria:
                if type(default_criteria[cr]) == dict:
                    for sub_cr in default_criteria[cr]:
                        if sub_cr in form:
                            parameters[sub_cr] = form[sub_cr]
                elif cr in form:
                    parameters[cr] = form[cr]
        mset, json_ms, url = create_mset_json(parameters, criteria)
        result = json_ms

    else:
        result = json.dumps({"Error":"MMX Unknown or missing command. Command option(s): complexity, log","code":501})
    print_object(json.loads(result), form)

def print_header():
    """Prints the JSON header."""
    print "Content-Type: application/json"
    print "Access-Control-Allow-Origin: *"
    print "Access-Control-Allow-Methods: POST, GET, OPTIONS"
    print "Access-Control-Allow-Headers: Content-Type"
    print


def print_object(obj, form):
    """Prints an object in JSON format.
    The CGI form can contain optional parameters 'callback' and 'indent'
    which change the output format.
    """
    callback = form.get("callback")
    if callback: print callback + "(",
    try:
        indent = int(form.get("indent"))
        print json.dumps(obj, sort_keys=True, indent=indent),
    except:
        print json.dumps(obj, separators=(",",":"))
    if callback: print ")",
    print

if __name__ == "__main__":
    main()