Skip to content

Option to get date from filename (works with diag files) #18

Description

@mefrediani

My diag files didn't have xtime and there was no way to get the plot date from attributes.

import re

Replace the if block in line 436 with:
logger.debug(f"Plotting time step {i}") times.append( resolve_timestring(uxds, config_d["dataset"]["files"], i) )

And add these functions:

def parse_time_from_filename(filepath: str) -> str:
    """
    Parse a timestamp from filename patterns like:
      diag.2025-06-06_21.00.00.nc
    and return a string in YYYY-MM-DD_HH:MM:SS format.
    """
    basename = os.path.basename(filepath)
    match = re.search(r"(\d{4}-\d{2}-\d{2})_(\d{2}\.\d{2}\.\d{2})", basename)
    if not match:
        raise ValueError(
            f"Could not parse timestamp from filename '{basename}'. "
            "Expected pattern like 'diag.YYYY-MM-DD_HH.MM.SS.nc'."
        )

    date_part, time_part = match.groups()
    timestring = f"{date_part}_{time_part.replace('.', ':')}"
    try:
        datetime.strptime(timestring, "%Y-%m-%d_%H:%M:%S")
    except ValueError as e:
        raise ValueError(
            f"Parsed timestamp '{timestring}' from '{basename}' is invalid."
        ) from e
    return timestring
def resolve_timestring(
    uxds: ux.UxDataset,
    datafiles: list[str],
    time_index: int,
) -> str:
    """
    Resolve plot timestamp from xtime when present, otherwise from input filename.
    time_index == -1 is used for variables without a Time dimension.
    """
    if "xtime" in uxds:
        idx = 0 if time_index == -1 else time_index
        return "".join(uxds["xtime"].isel(Time=idx).values.astype(str))

    if not datafiles:
        raise ValueError("No input files available to derive timestamp")

    if time_index == -1:
        filepath = datafiles[0]
    else:
        n_time = uxds.sizes.get("Time", 0)
        if n_time != len(datafiles):
            raise ValueError(
                "'xtime' is missing and timestamp mapping is ambiguous: "
                f"dataset Time dimension has {n_time} steps, but only "
                f"{len(datafiles)} input file(s). Use one file per timestep "
                "or provide xtime in input data."
            )
        if time_index >= len(datafiles):
            raise ValueError(
                f"Time index {time_index} is out of range for {len(datafiles)} input files"
            )
        filepath = datafiles[time_index]

    timestring = parse_time_from_filename(filepath)
    logger.warning(
        f"'xtime' variable not found in input data; using timestamp parsed from "
        f"filename '{os.path.basename(filepath)}' -> {timestring}"
    )
    return timestring

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions