Source code for pycif.plugins.datastreams.fields.lmdz_outfields_nc.fetch

import os
import pandas as pd
import datetime
from .....utils import path




def fetch(
    ref_dir, ref_file, date_interval, target_dir, tracer=None, component=None
):
    """Link monthly LMDz ``trajq`` files and build 3-hourly sub-intervals.

    Iterates over dates spaced by ``tracer.file_freq`` across
    ``date_interval``, links each existing monthly file into ``target_dir``,
    and builds 3-hourly sub-intervals covering the whole month (8 per day).

    Args:
        ref_dir: Directory where the original files are found.
        ref_file: Date-format pattern for the original file names.
        date_interval: List of two dates, the beginning and end of the
            period to fetch.
        target_dir: Directory where links to the original files are
            created.
        tracer: Tracer/component configuration; ``tracer.file_freq`` gives
            the frequency at which monthly files are available.
        component: Unused.

    Returns:
        A tuple ``(list_files, list_dates)`` of dictionaries keyed by
        monthly file date. ``list_dates`` maps each key to one 3-hourly
        ``[start, end]`` interval per 3-hour period of the month;
        ``list_files`` maps each key to the (repeated) path of that
        month's file.
    """
    # Reshape input interval to include full months
    datei, datef = date_interval

    list_period_dates = \
        pd.date_range(datei, datef, freq=tracer.file_freq)
    list_dates = {}
    list_files = {}
    for dd in list_period_dates:
        file = dd.strftime(f"{ref_dir}/{ref_file}")
        if os.path.isfile(file):
            # Fetching
            target_file = f"{target_dir}/{os.path.basename(file)}"
            path.link(file, target_file)

            # Time stamps
            file_hours = pd.date_range(
                dd, periods=8 * pd.DatetimeIndex([dd]).days_in_month[0],
                freq="3h")
            list_dates[dd] = [
                [hh0, hh0 + datetime.timedelta(hours=3)]
                for hh0 in file_hours]

            list_files[dd] = (len(list_dates[dd]) * [target_file])

    return list_files, list_dates