#########
# This script ...
# Runtime on my local machine ~seconds
# This script generates the file geopolar_filepaths_{start_date}_{end_date}.txt
# Run TWICE: `python 00a_generate_satellite_urls.py` from the environment
# First time set DATASET = 'geopolar', second time DATASET = 'mur'
#########
import pathlib

import pandas as pd

# Define the root directory of the repository
REPO_ROOT = pathlib.Path(__file__).resolve().parents[2]
# Define global variables. To be set by user (or if time made into CLI args)
# Define the dates of the images based on the start and end dates of the analysis

# REPO_ROOT = '/Users/rwegener/repos/chesapeake_mhw/'
START_DATE, END_DATE = "20020901", "20230831"
DATASET = "geopolar"
if DATASET not in ["geopolar", "mur"]:
    raise KeyError("Valid options for DATASET are `geopolar` or `mur`")


# Define a function which creates a Geopolar filepath, given a date
def make_geopolar_url(time):
    yyyy = time.strftime("%Y")
    yyyymmdd = time.strftime("%Y%m%d")

    # Organization swithes from STAR to OSPO in 2017
    if time.year >= 2017:
        org = "OSPO"
    else:
        org = "STAR"

    return (
        "https://coastwatch.noaa.gov/pub/socd2/coastwatch/sst_blended/sst5km/"
        f"night/ghrsst/{yyyy}/{yyyymmdd}000000-{org}-L4_GHRSST-SSTfnd-Geo_Polar"
        "_Blended_Night-GLOB-v02.0-fv01.0.nc"
    )


# def make_mur_url_OLD(time):
#     '''link used for full dataset download'''
#     yyyymmdd = time.strftime('%Y%m%d')
#     return (
#         'https://archive.podaac.earthdata.nasa.gov/podaac-ops-cumulus-protected/'
#         f'MUR-JPL-L4-GLOB-v4.1/{yyyymmdd}090000-JPL-L4_GHRSST-SSTfnd-MUR-GLOB-v02.0'
#         '-fv04.1.nc'
#     )

# def make_mur_url_opendap(time):
#     yyyymmdd = time.strftime('%Y%m%d')
#     yyyy = time.strftime('%Y')
#     jjj = time.strftime('%j')
#     return (
#         'https://opendap.jpl.nasa.gov/opendap/OceanTemperature/ghrsst/data/GDS2/L4/'
#         f'GLOB/JPL/MUR/v4.1/{yyyy}/{jjj}/{yyyymmdd}090000-JPL-L4_GHRSST-SSTfnd-MUR-GLOB-'
#         'v02.0-fv04.1.nc'
#     )


def make_mur_url_s3(time):
    yyyymmdd = time.strftime("%Y%m%d")
    return (
        "s3://podaac-ops-cumulus-protected/MUR-JPL-L4-GLOB-v4.1/"
        f"{yyyymmdd}090000-JPL-L4_GHRSST-SSTfnd-MUR-GLOB-v02.0-fv04.1.nc"
    )


# Manually define known dates where geopolar data was removed for quality reasons
geopolar_missing_dates = [
    pd.Timestamp(2017, 2, 1),
    pd.Timestamp(2018, 2, 8),
    pd.Timestamp(2018, 3, 6),
    pd.Timestamp(2018, 3, 12),
    pd.Timestamp(2018, 3, 17),
    pd.Timestamp(2019, 5, 25),
    pd.Timestamp(2021, 5, 21),
]

mur_missing_dates = [
    pd.Timestamp(2021, 2, 20),
    pd.Timestamp(2021, 2, 21),
    pd.Timestamp(2022, 11, 9),
]

# Choose the url function and missing dates list corresponding to the dataset
# being processed
if DATASET == "geopolar":
    make_url = make_geopolar_url
    missing_dates = geopolar_missing_dates
elif DATASET == "mur":
    make_url = make_mur_url_s3
    missing_dates = mur_missing_dates
else:
    raise Exception("Invalid value for DATASET")

# Create a list of dates with a daily frequency
dates = pd.date_range(START_DATE, end=END_DATE)
try:
    dates = dates.drop(missing_dates)
except KeyError:
    pass

# Create a list that contains all the filepaths
all_filepaths = [make_url(d) for d in dates]

# Write the names of the filepaths to a text file, each on a new line
output_filename = f"filepaths_{DATASET}_{START_DATE}_{END_DATE}.txt"


output_path = REPO_ROOT / "data" / "interim" / output_filename
# Create directory if it does not exist
if not pathlib.Path.exists(REPO_ROOT / "data" / "interim"):
    pathlib.Path.mkdir(REPO_ROOT / "data" / "interim")

with open(output_path, "w") as f:
    f.writelines(filepath + "\n" for filepath in all_filepaths)

print("Generated file ", output_path)