#!/bin/env python3

from typing import List, Tuple
import argparse
import requests
import re
import xlrd
import tempfile
import os
import zipfile
import PyPDF2


PLZ_URL = 'https://www.post.at/g/c/postlexikon'
PLZ_BUTTON = re.compile(r'title="PLZ Bestimmungsort" href="(.*?)"')

KGV_URL = 'https://www.bev.gv.at/portal/page?_pageid=713,2601283&_dad=portal&_schema=PORTAL'
KGV_LINK = re.compile(r'<a\s+href="(.*?)"\s*>')

OV_URL = 'https://statistik.gv.at/fileadmin/publications/Ortsverzeichnis_2001__{}.pdf'
OV_NAMES = ['Burgenland', 'Niederoesterreich', 'Wien']

GEM = re.compile(r'^([1-9][0-9]{2} [0-9]{2}) [X0-9]+')
ORT = re.compile(r'^([0-9]{5}) (([A-Za-zÄÖÜäöü(][A-Za-z0-9äöüÄÖÜßẞ():,.-]* ?|[0-9]+\..*?)+) ?'
                 r'([()X0-9.]+ ?|$)')
STRIP_NUM = re.compile(r'[X0-9. -]+$')
STRIP_CODE = re.compile(r' *(Agh|Alm|Alpe|B|Bd|Bwg|Burg|Camp|D|E|Fbk|Fhei|Gh|Hgr|Hot|Indz|Jh|Jhtt|Ki|Kl|Krwk|Ks|M|Mh|'
                        r'Mü|R|Ru|Sa|Sä|Sb|Schh|Schih|Schl|Sdlg|Sgr|St|Stbr|Stt|V|W|We|Ek|Z|Zgl|ZH) *(\([0-9]+\) *)?$')

PlzRow = Tuple[int, str, int, str, int, str]
KgvRow = Tuple[int, str, int, str]
OvRow = Tuple[int, int, str]

WIEN_UMGEBUNG = {
    32401: 30729,
    32402: 30730,
    32403: 31949,
    32404: 31235,
    32405: 30731,
    32406: 30732,
    32407: 30733,
    32408: 32144,
    32409: 30734,
    32410: 30735,
    32411: 30736,
    32412: 31950,
    32413: 30737,
    32415: 31951,
    32416: 31952,
    32417: 30738,
    32418: 30739,
    32419: 30740,
    32421: 31953,
    32423: 31954,
    32424: 30741,
}

GKZ = {}


def get_plz_url() -> str:
    r = requests.get(PLZ_URL, headers={'User-Agent': 'Mozilla/5.0'})
    if r.status_code != 200:
        raise RuntimeError(f'Unexpected response: {r.status_code} {r.reason}')

    matches = PLZ_BUTTON.findall(r.text)
    if len(matches) == 0:
        raise RuntimeError('Unable to find url of file')

    return matches[0]


def download_plz() -> List[PlzRow]:
    f_name = None
    try:
        f = tempfile.NamedTemporaryFile(delete=False)
        with requests.get(get_plz_url(), stream=True) as r:
            if r.status_code != 200:
                raise RuntimeError(f'Unexpected response: {r.status_code} {r.reason}')
            for chunk in r.iter_content(chunk_size=8192):
                f.write(chunk)
        f_name = f.name
        f.close()

        rows = []
        wb = xlrd.open_workbook(f_name)
        sheet = wb.sheet_by_index(0)
        for r in range(1, sheet.nrows):
            row = sheet.row_values(r)
            rows.append((int(row[0]), row[1], int(row[2]), row[3], int(row[4]), row[5]))
        return rows
    finally:
        if f_name is not None:
            os.remove(f_name)


def get_kvg_zip_url() -> str:
    r = requests.get(KGV_URL)
    if r.status_code != 200:
        raise RuntimeError(f'Unexpected response: {r.status_code} {r.reason}')

    matches = KGV_LINK.findall(r.text)
    if len(matches) == 0:
        raise RuntimeError('Unable to find url of zip file')

    return matches[0]


def download_kgv() -> List[KgvRow]:
    with tempfile.NamedTemporaryFile() as f:
        with requests.get(get_kvg_zip_url(), stream=True) as r:
            if r.status_code != 200:
                raise RuntimeError(f'Unexpected response: {r.status_code} {r.reason}')
            for chunk in r.iter_content(chunk_size=8192):
                f.write(chunk)

        rows = []
        with zipfile.ZipFile(f, 'r') as zip_file:
            files = [name for name in zip_file.namelist() if name.endswith('.csv')]
            if len(files) == 0:
                raise RuntimeError('Unable to find csv file in zip')

            with zip_file.open(files[0], 'r') as csv:
                first = True
                for r_line in csv:
                    if first:
                        first = False
                        continue
                    line = r_line.decode('utf8').rstrip()
                    row = [c[1:-1] if c[0] == '"' else int(c) for c in line.split(';')]
                    rows.append((int(row[0]), str(row[1]), int(row[3]), str(row[4])))
        return rows


def download_ov_land(bundesland: str) -> List[OvRow]:
    rows = []
    with tempfile.NamedTemporaryFile() as f:
        r = requests.get(OV_URL.format(bundesland))
        if r.status_code != 200:
            raise RuntimeError(f'Unexpected response: {r.status_code} {r.reason}')
        for chunk in r.iter_content(chunk_size=8192):
            f.write(chunk)

        pdf = PyPDF2.PdfFileReader(f)
        valid = False
        for page in pdf.pages:
            page_num = pdf.getPageNumber(page)
            text = page.extractText()

            if len(text) < 100:
                if text.strip().replace(' ', '') == 'Ortsverzeichnis':
                    valid = True
                    continue
                elif valid and text.strip().replace(' ', '') == 'ALPHABETISCHEVERZEICHNISSE':
                    break
            if not valid:
                continue

            with open(f'out/{bundesland}/{page_num + 1}.txt', 'w+') as o:
                o.write(text)

    return rows


def download_ov() -> None:
    try:
        os.mkdir('out')
        for name in OV_NAMES:
            os.mkdir(f'out/{name}')
            download_ov_land(name)
    except FileExistsError:
        print('Using cache')
        return


def parse_ov() -> List[OvRow]:
    rows = []
    for bundesland in sorted(os.listdir('out')):
        gkz = None
        last = None
        for page_name in sorted(os.listdir(f'out/{bundesland}')):
            with open(f'out/{bundesland}/{page_name}', 'r') as f:
                cont = False
                for line in f:
                    line = line.rstrip()
                    m1 = ORT.match(line)
                    m2 = GEM.match(line)

                    if last is not None:
                        if line == 'Gemeindename':
                            cont = True
                            break
                        elif m1 is None:
                            last = f'{last} {line}'
                            print(last)
                            val = STRIP_NUM.sub('', ORT.match(last).group(0))
                            okz = int(val[:5])
                            name = STRIP_CODE.sub('', val[6:])
                            rows.append((gkz, okz, name))
                            last = None
                            continue
                        else:
                            val = last
                            okz = int(val[:5])
                            name = STRIP_CODE.sub('', val[6:])
                            rows.append((gkz, okz, name))
                            last = None

                    if 'Katastralgemeinden:' in line:
                        p1 = line.find('Katastralgemeinden:')
                        p2 = line.find('Postleitzahl')
                        val = [' '.join(a.split(' ')[:-2]) for a in line[p1 + 20:p2].split(', ')]
                        GKZ[gkz] = val
                        continue

                    if m1:
                        val = STRIP_NUM.sub('', m1.group(0))
                        if m1.group(4) == '':
                            last = val
                        else:
                            okz = int(val[:5])
                            name = STRIP_CODE.sub('', val[6:])
                            rows.append((gkz, okz, name))
                            print(rows[-1])
                    elif m2:
                        if len(line.split(' ')) <= 9:
                            continue
                        gkz = int(m2.group(1).replace(' ', ''))
                        if gkz > 90000:
                            # Gemeinde Wien
                            gkz = 90001
                        elif gkz >= 32400 and gkz <= 32499:
                            # ehem. Bezirk Wien Umgebung
                            gkz = WIEN_UMGEBUNG[gkz]
                if cont:
                    continue
    return rows


def write_sql(plz_rows: List[PlzRow], kgv_rows: List[KgvRow], ov_rows: List[OvRow]) -> None:
    kgv = {kgnr: (kg_name, gkz, gem_name) for kgnr, kg_name, gkz, gem_name in kgv_rows}
    ov = {okz: (o_name, gkz) for plz, _, okz, o_name, gkz, _ in plz_rows}
    ov.update({okz: (name, gkz) for gkz, okz, name in ov_rows})
    gemeinden = {gkz: (gem_name, [], []) for kgnr, kg_name, gkz, gem_name in kgv_rows}

    pr = set()
    with open('plz.sql', 'wb') as f:
        f.write(b"\nINSERT INTO AT_gem VALUES\n")
        for gkz, (name, _, _) in gemeinden.items():
            f.write(f"({gkz:5}, '{name}'),\n".encode('utf8'))
        f.seek(-2, 1)
        f.write(b';\n')

        f.write(b"\nINSERT INTO AT_kg VALUES\n")
        for kgnr, name, gkz, _ in kgv_rows:
            gemeinden[gkz][1].append(kgnr)
            f.write(f"({kgnr:5}, {gkz:5}, '{name}'),\n".encode('utf8'))
        f.seek(-2, 1)
        f.write(b';\n')

        f.write(b"\nINSERT INTO AT_ort VALUES\n")
        for okz, (name, gkz) in ov.items():
            kgnr_o = None

            if name.startswith('Wien '):
                name = name.replace('Wien ', 'Wien, ').replace('.', '. ')
            elif 'Bez.' in name:
                name = name.replace(',', ', ', 1).replace('.Bez.:', '. Bezirk: ').replace('0', '')

            if gkz not in gemeinden:
                print(okz, name, gkz)
                pr.add(gkz)
                continue

            if len(gemeinden[gkz][1]) == 1:
                kgnr_o = gemeinden[gkz][1][0]
            else:
                for kgnr in gemeinden[gkz][1]:
                    n11 = name.lower().replace('-', '').replace('th', 't')
                    n12 = name.lower().replace('-', ' ').replace('th', 't')
                    n21 = kgv[kgnr][0].lower().replace('-', '').replace('th', 't')
                    n22 = kgv[kgnr][0].lower().replace('-', ' ').replace('th', 't')
                    if n11 in n21 or n11 in n22 or n12 in n21 or n12 in n22 or n21 in n11 or n21 in n12 or n22 in n11 or n22 in n12:
                        kgnr_o = kgnr

            f.write(f"({okz:5}, {gkz:5}, {kgnr_o if kgnr_o is not None else 'NULL':>5}, '{name}'),\n".encode('utf8'))
        f.seek(-2, 1)
        f.write(b';\n')

        f.write(b"\nINSERT INTO AT_plz VALUES\n")
        for plz, dest, okz, _, _, _ in plz_rows:
            f.write(f"({plz:4}, {okz:5}, '{dest}'),\n".encode('utf8'))
        f.seek(-2, 1)
        f.write(b';\n')

    p = set()
    for e in pr:
        possible = filter(lambda a: a[2] // 10000 == e // 10000, kgv_rows.copy())
        if e in GKZ:
            for name in GKZ[e]:
                if name == '':
                    continue
                possible = filter(lambda a: a[1] == name, possible)
        possible = list(possible)
        if len(possible) == 1:
            print(f'    {e}: {possible[0][2]},')
            p.add(e)

    u = list(pr - p)
    u.sort()
    print(u)


if __name__ == '__main__':
    parser = argparse.ArgumentParser()
    args = parser.parse_args()

    print('Downloading Ortsverzeichnis from statistik.gv.at')
    download_ov()
    print('Parsing Ortsverzeichnis')
    ov_data = parse_ov()
    print('Downloading PLZ data from www.post.at')
    plz_data = download_plz()
    print('Downloading Katastralgemeindenverzeichnis from www.bev.gv.at')
    kgv_data = download_kgv()
    print('Generating plz.sql')
    write_sql(plz_data, kgv_data, ov_data)
    print('Successfully created plz.sql!')