"""
    ulmo.twc.kbdi.core
    ~~~~~~~~~~~~~~~~~~~~~

    This module provides direct access to `Texas Weather Connection`_ `Daily
    Keetch-Byram Drought Index (KBDI)`_ dataset.

    .. _Texas Weather Connection: http://twc.tamu.edu/
    .. _Daily Keetch-Byram Drought Index (KBDI): http://twc.tamu.edu/drought/kbdi
"""

import datetime
import os

import numpy as np
import pandas

from ulmo import util

CSV_SWITCHOVER = pandas.Timestamp('2016-10-01')

def get_data(county=None, start=None, end=None, as_dataframe=False, data_dir=None):
    """Retreives data.

    Parameters
    ----------
    county : ``None`` or str
        If specified, results will be limited to the county corresponding to the
        given 5-character Texas county fips code i.e. 48???.
    end : ``None`` or date (see :ref:`dates-and-times`)
        Results will be limited to data on or before this date. Default is the
        current date.
    start : ``None`` or date (see :ref:`dates-and-times`)
        Results will be limited to data on or after this date. Default is the
        start of the calendar year for the end date.
    as_dataframe: bool
        If ``False`` (default), a dict with a nested set of dicts will be
        returned with data indexed by 5-character Texas county FIPS code. If ``True``
        then a pandas.DataFrame object will be returned.  The pandas dataframe
        is used internally, so setting this to ``True`` is a little bit faster
        as it skips a serialization step.
    data_dir : ``None`` or directory path
        Directory for holding downloaded data files. If no path is provided
        (default), then a user-specific directory for holding application data
        will be used (the directory will depend on the platform/operating
        system).


    Returns
    -------
    data : dict or pandas.Dataframe
        A dict or pandas.DataFrame representing the data. See the
        ``as_dataframe`` parameter for more.
    """
    if end is None:
        end_date = datetime.date.today()
    else:
        end_date = util.convert_date(end)
    if start is None:
        start_date = datetime.date(end_date.year, 1, 1)
    else:
        start_date = util.convert_date(start)
    if data_dir is None:
        data_dir = os.path.join(util.get_ulmo_dir(), 'twc/kbdi')

    df = pandas.concat([
        _date_dataframe(date, data_dir)
        for date in pandas.period_range(start_date, end_date, freq='D')
    ], ignore_index=True)
    fips_df = _fips_dataframe()
    df = pandas.merge(df, fips_df, left_on='county', right_on='name')
    del df['name']

    if county:
        df = df[df['fips'] == county]

    if as_dataframe:
        return df
    else:
        return _as_data_dict(df)


def _as_data_dict(df):
    df['date'] = df['date'].map(str)
    county_dict = {}
    for county in df['fips'].unique():
        county_df = df[df['fips'] == county]
        county_data = county_df.T.drop(['fips'])
        values = [v.to_dict() for k, v in county_data.iteritems()]
        county_dict[county] = values

    return county_dict


def _date_dataframe(date, data_dir):

    if date.to_timestamp() < CSV_SWITCHOVER:
        url = _get_text_url(date)
        with _open_data_file(url, data_dir) as data_file:
            date_df = _parse_text_file(data_file)
    else:
        url = _get_csv_url(date)
        with _open_data_file(url, data_dir) as data_file:
            date_df = _parse_csv_file(data_file)

    date_df['date'] = pandas.Period(date, freq='D')

    return date_df


def _fips_dataframe():
    # fips codes from http://www.census.gov/geo/www/ansi/national.txt
    # with names adjusted to match twc kbdi: DEWITT --> DE WITT
    codes = (
        ('ANDERSON', 48001),
        ('ANDREWS', 48003),
        ('ANGELINA', 48005),
        ('ARANSAS', 48007),
        ('ARCHER', 48009),
        ('ARMSTRONG', 48011),
        ('ATASCOSA', 48013),
        ('AUSTIN', 48015),
        ('BAILEY', 48017),
        ('BANDERA', 48019),
        ('BASTROP', 48021),
        ('BAYLOR', 48023),
        ('BEE', 48025),
        ('BELL', 48027),
        ('BEXAR', 48029),
        ('BLANCO', 48031),
        ('BORDEN', 48033),
        ('BOSQUE', 48035),
        ('BOWIE', 48037),
        ('BRAZORIA', 48039),
        ('BRAZOS', 48041),
        ('BREWSTER', 48043),
        ('BRISCOE', 48045),
        ('BROOKS', 48047),
        ('BROWN', 48049),
        ('BURLESON', 48051),
        ('BURNET', 48053),
        ('CALDWELL', 48055),
        ('CALHOUN', 48057),
        ('CALLAHAN', 48059),
        ('CAMERON', 48061),
        ('CAMP', 48063),
        ('CARSON', 48065),
        ('CASS', 48067),
        ('CASTRO', 48069),
        ('CHAMBERS', 48071),
        ('CHEROKEE', 48073),
        ('CHILDRESS', 48075),
        ('CLAY', 48077),
        ('COCHRAN', 48079),
        ('COKE', 48081),
        ('COLEMAN', 48083),
        ('COLLIN', 48085),
        ('COLLINGSWORTH', 48087),
        ('COLORADO', 48089),
        ('COMAL', 48091),
        ('COMANCHE', 48093),
        ('CONCHO', 48095),
        ('COOKE', 48097),
        ('CORYELL', 48099),
        ('COTTLE', 48101),
        ('CRANE', 48103),
        ('CROCKETT', 48105),
        ('CROSBY', 48107),
        ('CULBERSON', 48109),
        ('DALLAM', 48111),
        ('DALLAS', 48113),
        ('DAWSON', 48115),
        ('DE WITT', 48123),
        ('DEAF SMITH', 48117),
        ('DELTA', 48119),
        ('DENTON', 48121),
        ('DEWITT', 48123),
        ('DICKENS', 48125),
        ('DIMMIT', 48127),
        ('DONLEY', 48129),
        ('DUVAL', 48131),
        ('EASTLAND', 48133),
        ('ECTOR', 48135),
        ('EDWARDS', 48137),
        ('EL PASO', 48141),
        ('ELLIS', 48139),
        ('ERATH', 48143),
        ('FALLS', 48145),
        ('FANNIN', 48147),
        ('FAYETTE', 48149),
        ('FISHER', 48151),
        ('FLOYD', 48153),
        ('FOARD', 48155),
        ('FORT BEND', 48157),
        ('FRANKLIN', 48159),
        ('FREESTONE', 48161),
        ('FRIO', 48163),
        ('GAINES', 48165),
        ('GALVESTON', 48167),
        ('GARZA', 48169),
        ('GILLESPIE', 48171),
        ('GLASSCOCK', 48173),
        ('GOLIAD', 48175),
        ('GONZALES', 48177),
        ('GRAY', 48179),
        ('GRAYSON', 48181),
        ('GREGG', 48183),
        ('GRIMES', 48185),
        ('GUADALUPE', 48187),
        ('HALE', 48189),
        ('HALL', 48191),
        ('HAMILTON', 48193),
        ('HANSFORD', 48195),
        ('HARDEMAN', 48197),
        ('HARDIN', 48199),
        ('HARRIS', 48201),
        ('HARRISON', 48203),
        ('HARTLEY', 48205),
        ('HASKELL', 48207),
        ('HAYS', 48209),
        ('HEMPHILL', 48211),
        ('HENDERSON', 48213),
        ('HIDALGO', 48215),
        ('HILL', 48217),
        ('HOCKLEY', 48219),
        ('HOOD', 48221),
        ('HOPKINS', 48223),
        ('HOUSTON', 48225),
        ('HOWARD', 48227),
        ('HUDSPETH', 48229),
        ('HUNT', 48231),
        ('HUTCHINSON', 48233),
        ('IRION', 48235),
        ('JACK', 48237),
        ('JACKSON', 48239),
        ('JASPER', 48241),
        ('JEFF DAVIS', 48243),
        ('JEFFERSON', 48245),
        ('JIM HOGG', 48247),
        ('JIM WELLS', 48249),
        ('JOHNSON', 48251),
        ('JONES', 48253),
        ('KARNES', 48255),
        ('KAUFMAN', 48257),
        ('KENDALL', 48259),
        ('KENEDY', 48261),
        ('KENT', 48263),
        ('KERR', 48265),
        ('KIMBLE', 48267),
        ('KING', 48269),
        ('KINNEY', 48271),
        ('KLEBERG', 48273),
        ('KNOX', 48275),
        ('LA SALLE', 48283),
        ('LAMAR', 48277),
        ('LAMB', 48279),
        ('LAMPASAS', 48281),
        ('LAVACA', 48285),
        ('LEE', 48287),
        ('LEON', 48289),
        ('LIBERTY', 48291),
        ('LIMESTONE', 48293),
        ('LIPSCOMB', 48295),
        ('LIVE OAK', 48297),
        ('LLANO', 48299),
        ('LOVING', 48301),
        ('LUBBOCK', 48303),
        ('LYNN', 48305),
        ('MADISON', 48313),
        ('MARION', 48315),
        ('MARTIN', 48317),
        ('MASON', 48319),
        ('MATAGORDA', 48321),
        ('MAVERICK', 48323),
        ('MCCULLOCH', 48307),
        ('MCLENNAN', 48309),
        ('MCMULLEN', 48311),
        ('MEDINA', 48325),
        ('MENARD', 48327),
        ('MIDLAND', 48329),
        ('MILAM', 48331),
        ('MILLS', 48333),
        ('MITCHELL', 48335),
        ('MONTAGUE', 48337),
        ('MONTGOMERY', 48339),
        ('MOORE', 48341),
        ('MORRIS', 48343),
        ('MOTLEY', 48345),
        ('NACOGDOCHES', 48347),
        ('NAVARRO', 48349),
        ('NEWTON', 48351),
        ('NOLAN', 48353),
        ('NUECES', 48355),
        ('OCHILTREE', 48357),
        ('OLDHAM', 48359),
        ('ORANGE', 48361),
        ('PALO PINTO', 48363),
        ('PANOLA', 48365),
        ('PARKER', 48367),
        ('PARMER', 48369),
        ('PECOS', 48371),
        ('POLK', 48373),
        ('POTTER', 48375),
        ('PRESIDIO', 48377),
        ('RAINS', 48379),
        ('RANDALL', 48381),
        ('REAGAN', 48383),
        ('REAL', 48385),
        ('RED RIVER', 48387),
        ('REEVES', 48389),
        ('REFUGIO', 48391),
        ('ROBERTS', 48393),
        ('ROBERTSON', 48395),
        ('ROCKWALL', 48397),
        ('RUNNELS', 48399),
        ('RUSK', 48401),
        ('SABINE', 48403),
        ('SAN AUGUSTINE', 48405),
        ('SAN JACINTO', 48407),
        ('SAN PATRICIO', 48409),
        ('SAN SABA', 48411),
        ('SCHLEICHER', 48413),
        ('SCURRY', 48415),
        ('SHACKELFORD', 48417),
        ('SHELBY', 48419),
        ('SHERMAN', 48421),
        ('SMITH', 48423),
        ('SOMERVELL', 48425),
        ('STARR', 48427),
        ('STEPHENS', 48429),
        ('STERLING', 48431),
        ('STONEWALL', 48433),
        ('SUTTON', 48435),
        ('SWISHER', 48437),
        ('TARRANT', 48439),
        ('TAYLOR', 48441),
        ('TERRELL', 48443),
        ('TERRY', 48445),
        ('THROCKMORTON', 48447),
        ('TITUS', 48449),
        ('TOM GREEN', 48451),
        ('TRAVIS', 48453),
        ('TRINITY', 48455),
        ('TYLER', 48457),
        ('UPSHUR', 48459),
        ('UPTON', 48461),
        ('UVALDE', 48463),
        ('VAL VERDE', 48465),
        ('VAN ZANDT', 48467),
        ('VICTORIA', 48469),
        ('WALKER', 48471),
        ('WALLER', 48473),
        ('WARD', 48475),
        ('WASHINGTON', 48477),
        ('WEBB', 48479),
        ('WHARTON', 48481),
        ('WHEELER', 48483),
        ('WICHITA', 48485),
        ('WILBARGER', 48487),
        ('WILLACY', 48489),
        ('WILLIAMSON', 48491),
        ('WILSON', 48493),
        ('WINKLER', 48495),
        ('WISE', 48497),
        ('WOOD', 48499),
        ('YOAKUM', 48501),
        ('YOUNG', 48503),
        ('ZAPATA', 48505),
        ('ZAVALA', 48507),
    )

    df = pandas.DataFrame(np.array(codes))
    df = df.rename(columns={0: 'name', 1: 'fips'})
    df['fips'] = df['fips'].astype(int)
    return df


def _get_text_url(date):
    return 'http://twc.tamu.edu/weather_images/summ/summ%s.txt' % date.strftime('%Y%m%d')

def _get_csv_url(date):
    return 'http://twc.tamu.edu/weather_images/summ/summ%s.csv' % date.strftime('%Y%m%d')

def _parse_text_file(data_file):
    """
    example:
        COUNTY                        KBDI_AVG   KBDI_MAX    KBDI_MIN
                ----------------------------------------------------------------
                ANDERSON                         262       485        47
                ANDREWS                          485       614       357
                ...
    """

    dtype = [
        ('county', '|U15'),
        ('avg', 'i4'),
        ('max', 'i4'),
        ('min', 'i4'),
    ]

    if not data_file.readline().lower().startswith(b'county'):
        return pandas.DataFrame()
    data_file.seek(0)

    data_array = np.genfromtxt(
        data_file, delimiter=[31, 11, 11, 11], dtype=dtype, skip_header=2,
        skip_footer=1, autostrip=True)
    dataframe = pandas.DataFrame(data_array)
    return dataframe

def _parse_csv_file(data_file):
    """
    example:
        County,Min,Max,Average,Change
        Anderson,429,684,559,+5
        Andrews,92,356,168,+7
    """

    if not data_file.readline().lower().startswith(b'county'):
        return pandas.DataFrame()
    data_file.seek(0)

    dataframe = pandas.read_csv(data_file)
    dataframe.columns = dataframe.columns.str.lower()
    dataframe = dataframe.rename(columns={'average':'avg'})
    dataframe.county = dataframe.county.str.upper()
    dataframe = dataframe[['county','avg','max','min']]

    return dataframe

def _open_data_file(url, data_dir):
    """returns an open file handle for a data file; downloading if necessary or
    otherwise using a previously downloaded file
    """
    file_name = url.rsplit('/', 1)[-1]
    file_path = os.path.join(data_dir, file_name)
    return util.open_file_for_url(url, file_path, check_modified=True, use_bytes=True)