#!/usr/bin/python3.10
# -*- coding: utf-8 -*-

"""************************
AnkIdentification - Supplementary code 3-1 : Creating the list of all plant species.
Aurélien VALENTIN - from 06/11/2022 to 29/12/2022
************************"""

# Importation and initialisation
import json, sys
dict_common_names = {}

# Definition of the progressbar function, from https://gist.github.com/ChesterChowWOV/2b35c551b339adbf459363322aac5b4b
def progressbar(it, prefix="", size=60, file=sys.stdout):
    count = len(it)
    def show(j):
        x = int(size*j/count)
        file.write("{}[{}{}] {}/{} {}%\r".format(prefix, "█"*x, "."*(size-x), j, count, round(j/count * 100, 2)))
        file.flush()
    show(0)
    for i, item in enumerate(it):
        yield item
        show(i+1)
    file.write("\n")
    file.flush()

# Creation of the list
with open("0229588-220831081235567.csv", encoding="utf-8") as fpi:
    fpi.readline()
    for line in progressbar(fpi.readlines(), "Processing species", 40):
        try:
            specie = line.split("\t")[1].split(" ")[0] + " " + line.split("\t")[1].split(" ")[1]
            dict_common_names[specie] = ""
            # Looking for another scientific name accepted
            specie2 = line.split("\t")[3].split(" ")[0] + " " + line.split("\t")[3].split(" ")[1]
            if specie2 != specie:
                dict_common_names[specie2] = ""
        except: pass

# Generation of the json file
with open("_dict_species_raw.json", "w") as fpo : json.dump(dict_common_names, fpo)
print(dict_common_names)