import pandas as pd
import datetime
import string
import math
import re
from constants import knr_consts
from utils.validation import valid_str_duration, is_empty_value

def format_duration(duration):
    if valid_str_duration(duration): return duration
    h_m_s = str(datetime.timedelta(seconds=round(float(duration))))
    if len(h_m_s.split(":")[0]) == 1: return "0" + h_m_s
    else: return h_m_s

def format_date(col):
    return pd.to_datetime(col, dayfirst=True, format="mixed").dt.strftime('%d-%b-%Y')

# sales start date can be empty for video content
def format_video_date(date):
    return pd.to_datetime(date, dayfirst=True, errors="coerce", format="mixed").dt.strftime('%d-%b-%Y')

def strip_all_strings(df):
    return df.applymap(lambda x: x.strip() if isinstance(x, str) else x)

def format_contributor_types(types):
    if is_empty_value(types): return ''
    types_array = str(types).strip().split('|')
    formatted_types = []
    for item in types_array:
        # Only valid items will be appended to formatted_types, invalid string values, extra pipes, and empty strings are ignored
        x = item.upper().translate(str.maketrans('', '', f'{string.punctuation} ')).strip()
        if x in knr_consts.MAIN_PERFORMER_CATEGORY_MAPS:
            formatted_types.append('Main Performer')
        elif x in knr_consts.FEATURING_PERFORMER_CATEGORY_MAPS:
            formatted_types.append('Featuring Performer')
        elif x in knr_consts.SESSION_MUSICIAN_CATEGORY_MAPS:
            formatted_types.append('Session Musician')
    piped_types = '|'.join(formatted_types)
    return piped_types

def format_contributor_names(names):
    if is_empty_value(names): return ''
    names_array = list(filter(lambda x: x != "", [x.strip() for x in names.split('|')]))
    formatted_names = []
    for name in names_array:
        if len(name.strip().split(' ')) == 1:
            formatted_names.append(name.strip())
        else:
            [firstname, *middlename, lastname] = name.strip().split(' ')
            if not middlename:
                formatted_names.append(f'{lastname}, {firstname}')
            else:
                formatted_names.append(f'{lastname}, {firstname} {" ".join(middlename)}')
    piped_names = '|'.join(formatted_names)
    return piped_names

def format_piped_values(values):
    if is_empty_value(values): return ''
    # The below ensures any extra pipes/whitespace are removed from piped values (e.g. "x || y " becomes "x|y")
    return "|".join(list(filter(lambda x: x != "", [x.strip() for x in values.split('|')])))

def format_participant_names(names):
    # Any blank primary artists would have been excluded in the earlier validation steps
    # Featuring artists can legitimately be blank
    if is_empty_value(names): return ''
    piped_names = re.sub("(, &)|(,)", '|', str(names))
    formatted_names = format_piped_values(piped_names)
    return formatted_names

def format_explicit_value(value):
    if value == '' or type(value) == float and math.isnan(float(value)): return value
    elif value.lower().strip() == "explicit": return "Yes"
    elif value.lower().strip() == "clean": return "No"
    else: return math.nan

def format_dataframe(ingest_template_df):
    ingest_template_df = ingest_template_df.assign(**{
        'Product Primary Artists': ingest_template_df.loc[:,'Product Primary Artists'].apply(lambda x: format_participant_names(x)),
        'Product Featuring Artists': ingest_template_df.loc[:,'Product Featuring Artists'].apply(lambda x: format_participant_names(x)),
        'Primary Artist': ingest_template_df.loc[:,'Primary Artist'].apply(lambda x: format_participant_names(x)),
        'Featuring Artist': ingest_template_df.loc[:,'Featuring Artist'].apply(lambda x: format_participant_names(x)),
        'Product Type': 'Audio',
        'Sales Start Date': format_date(ingest_template_df.loc[:,'Sales Start Date']),
        'Collection Ownership Start Date': format_date(ingest_template_df.loc[:,'Collection Ownership Start Date']),
        'Collection Ownership End Date': format_date(ingest_template_df.loc[:,'Collection Ownership End Date']),
        'Duration': ingest_template_df.loc[:,'Duration'].apply(lambda x: format_duration(x)),
        'Disc Number': ingest_template_df.loc[:,'Disc Number'].fillna('1'),
        'Track Number': ingest_template_df.loc[:,'Track Number'].fillna('1'),
        'Contributor Legal Name': ingest_template_df.loc[:,'Contributor Legal Name'].apply(lambda x: format_contributor_names(x)),
        'Contributor Type': ingest_template_df.loc[:,'Contributor Type'].apply(lambda x: format_contributor_types(x)),
        'Contributor Role(s)': ingest_template_df.loc[:,'Contributor Role(s)'].apply(lambda x: format_piped_values(x)),
        'Collection Ownership Percentage of Rights': ingest_template_df.loc[:,'Collection Ownership Percentage of Rights'].fillna(100),
        'Collection Ownership Country': ingest_template_df.loc[:,'Collection Ownership Country'].apply(lambda x: format_piped_values(x)),
        'Collection Ownership Country - Exclusion': ingest_template_df.loc[:, 'Collection Ownership Country - Exclusion'].apply(lambda x: format_piped_values(x)).replace(knr_consts.NO_TERRITORY, ''),
        'Country of Label/Individual/Party Which Funded Recording': ingest_template_df.loc[:,'Country of Label/Individual/Party Which Funded Recording'].apply(lambda x: format_piped_values(x)),
        'Country of Recording': ingest_template_df.loc[:,'Country of Recording'].apply(lambda x: format_piped_values(x)),
        'Explicit': ingest_template_df.loc[:,'Explicit'].apply(lambda x: format_explicit_value(x))
    })

    invalid_rights_start_dates = ingest_template_df.loc[pd.to_datetime(ingest_template_df.loc[:,'Collection Ownership Start Date'], errors="coerce").isna()]
    for i, row in invalid_rights_start_dates.iterrows():
        release_yr = ingest_template_df.at[i, 'First Year of Release']
        ingest_template_df.at[i, 'Collection Ownership Start Date'] = f'01-Jan-{release_yr}'

    ingest_template_df = strip_all_strings(ingest_template_df)

    return ingest_template_df

def format_video_dataframe(ingest_template_df):
    ingest_template_df = ingest_template_df.assign(**{
        'Product Primary Artists': ingest_template_df.loc[:,'Product Primary Artists'].apply(lambda x: format_participant_names(x)),
        'Product Featuring Artists': ingest_template_df.loc[:,'Product Featuring Artists'].apply(lambda x: format_participant_names(x)),
        'Primary Artist': ingest_template_df.loc[:,'Primary Artist'].apply(lambda x: format_participant_names(x)),
        'Featuring Artist': ingest_template_df.loc[:,'Featuring Artist'].apply(lambda x: format_participant_names(x)),
        'Product Type': 'Video',
        'Sales Start Date': format_video_date(ingest_template_df.loc[:,'Sales Start Date']),
        'Collection Ownership Start Date': format_date(ingest_template_df.loc[:,'Collection Ownership Start Date']),
        'Collection Ownership End Date': format_date(ingest_template_df.loc[:,'Collection Ownership End Date']),
        'Duration': ingest_template_df.loc[:,'Duration'].apply(lambda x: format_duration(x)),
        'Contributor Legal Name': ingest_template_df.loc[:,'Contributor Legal Name'].apply(lambda x: format_contributor_names(x)),
        'Contributor Type': ingest_template_df.loc[:,'Contributor Type'].apply(lambda x: format_contributor_types(x)),
        'Contributor Role(s)': ingest_template_df.loc[:,'Contributor Role(s)'].apply(lambda x: format_piped_values(x)),
        'Collection Ownership Percentage of Rights': ingest_template_df.loc[:,'Collection Ownership Percentage of Rights'].fillna(100),
        'Collection Ownership Country': ingest_template_df.loc[:,'Collection Ownership Country'].apply(lambda x: format_piped_values(x)),
        'Collection Ownership Country - Exclusion': ingest_template_df.loc[:, 'Collection Ownership Country - Exclusion'].apply(lambda x: format_piped_values(x)).replace(knr_consts.NO_TERRITORY, ''),
        'Country of Label/Individual/Party Which Funded Recording': ingest_template_df.loc[:,'Country of Label/Individual/Party Which Funded Recording'].apply(lambda x: format_piped_values(x)),
        'Country of Recording': ingest_template_df.loc[:,'Country of Recording'].apply(lambda x: format_piped_values(x)),
        'Explicit': ingest_template_df.loc[:,'Explicit'].apply(lambda x: format_explicit_value(x))
    })

    invalid_rights_start_dates = ingest_template_df.loc[pd.to_datetime(ingest_template_df.loc[:,'Collection Ownership Start Date'], errors="coerce").isna()]
    for i, row in invalid_rights_start_dates.iterrows():
        release_yr = ingest_template_df.at[i, 'First Year of Release']
        ingest_template_df.at[i, 'Collection Ownership Start Date'] = f'01-Jan-{release_yr}'

    ingest_template_df = strip_all_strings(ingest_template_df)

    return ingest_template_df

def format_date_for_error_file(col):
    return pd.to_datetime(col, dayfirst=True, errors="ignore").dt.date