anemoi-utils 0.3.13__py3-none-any.whl → 0.3.15__py3-none-any.whl

anemoi/utils/_version.py

@@ -12,5 +12,5 @@ __version__: str
 __version_tuple__: VERSION_TUPLE
 version_tuple: VERSION_TUPLE
 
-__version__ = version = '0.3.13'
-__version_tuple__ = version_tuple = (0, 3, 13)
+__version__ = version = '0.3.15'
+__version_tuple__ = version_tuple = (0, 3, 15)

anemoi/utils/cli.py

@@ -12,6 +12,11 @@ import os
 import sys
 import traceback
 
+try:
+    import argcomplete
+except ImportError:
+    argcomplete = None
+
 LOG = logging.getLogger(__name__)
 
 
@@ -100,6 +105,8 @@ def register_commands(here, package, select, fail=None):
 def cli_main(version, description, commands):
     parser = make_parser(description, commands)
     args, unknown = parser.parse_known_args()
+    if argcomplete:
+        argcomplete.autocomplete(parser)
 
     if args.version:
         print(version)

anemoi/utils/dates.py

@@ -8,21 +8,14 @@
 
 import calendar
 import datetime
+import re
 
-from .hindcasts import HindcastDatesTimes
-
-
-def normalise_frequency(frequency):
-    if isinstance(frequency, int):
-        return frequency
-    assert isinstance(frequency, str), (type(frequency), frequency)
+import isodate
 
-    unit = frequency[-1].lower()
-    v = int(frequency[:-1])
-    return {"h": v, "d": v * 24}[unit]
+from .hindcasts import HindcastDatesTimes
 
 
-def no_time_zone(date):
+def _no_time_zone(date):
     """Remove time zone information from a date.
 
     Parameters
@@ -40,13 +33,15 @@ def no_time_zone(date):
 
 
 # this function is use in anemoi-datasets
-def as_datetime(date):
+def as_datetime(date, keep_time_zone=False):
     """Convert a date to a datetime object, removing any time zone information.
 
     Parameters
     ----------
     date : datetime.date or datetime.datetime or str
         The date to convert.
+    keep_time_zone : bool, optional
+        If True, the time zone information is kept, by default False.
 
     Returns
     -------
@@ -54,18 +49,150 @@ def as_datetime(date):
         The datetime object.
     """
 
+    tidy = _no_time_zone if not keep_time_zone else lambda x: x
+
     if isinstance(date, datetime.datetime):
-        return no_time_zone(date)
+        return tidy(date)
 
     if isinstance(date, datetime.date):
-        return no_time_zone(datetime.datetime(date.year, date.month, date.day))
+        return tidy(datetime.datetime(date.year, date.month, date.day))
 
     if isinstance(date, str):
-        return no_time_zone(datetime.datetime.fromisoformat(date))
+        return tidy(datetime.datetime.fromisoformat(date))
 
     raise ValueError(f"Invalid date type: {type(date)}")
 
 
+def frequency_to_timedelta(frequency):
+    """Convert a frequency to a timedelta object.
+
+    Parameters
+    ----------
+    frequency : int or str or datetime.timedelta
+        The frequency to convert. If an integer, it is assumed to be in hours. If a string, it can be in the format:
+
+        - "1h" for 1 hour
+        - "1d" for 1 day
+        - "1m" for 1 minute
+        - "1s" for 1 second
+        - "1:30" for 1 hour and 30 minutes
+        - "1:30:10" for 1 hour, 30 minutes and 10 seconds
+        - "PT10M" for 10 minutes (ISO8601)
+
+        If a timedelta object is provided, it is returned as is.
+
+    Returns
+    -------
+    datetime.timedelta
+        The timedelta object.
+
+    Raises
+    ------
+    ValueError
+        Exception raised if the frequency cannot be converted to a timedelta.
+    """
+
+    if isinstance(frequency, datetime.timedelta):
+        return frequency
+
+    if isinstance(frequency, int):
+        return datetime.timedelta(hours=frequency)
+
+    assert isinstance(frequency, str), (type(frequency), frequency)
+
+    try:
+        return frequency_to_timedelta(int(frequency))
+    except ValueError:
+        pass
+
+    if re.match(r"^\d+[hdms]$", frequency, re.IGNORECASE):
+        unit = frequency[-1].lower()
+        v = int(frequency[:-1])
+        unit = {"h": "hours", "d": "days", "s": "seconds", "m": "minutes"}[unit]
+        return datetime.timedelta(**{unit: v})
+
+    m = frequency.split(":")
+    if len(m) == 2:
+        return datetime.timedelta(hours=int(m[0]), minutes=int(m[1]))
+
+    if len(m) == 3:
+        return datetime.timedelta(hours=int(m[0]), minutes=int(m[1]), seconds=int(m[2]))
+
+    # ISO8601
+    try:
+        return isodate.parse_duration(frequency)
+    except isodate.isoerror.ISO8601Error:
+        pass
+
+    raise ValueError(f"Cannot convert frequency {frequency} to timedelta")
+
+
+def frequency_to_string(frequency):
+    """Convert a frequency (i.e. a datetime.timedelta) to a string.
+
+    Parameters
+    ----------
+    frequency : datetime.timedelta
+        The frequency to convert.
+
+    Returns
+    -------
+    str
+        A string representation of the frequency.
+    """
+
+    frequency = frequency_to_timedelta(frequency)
+
+    total_seconds = frequency.total_seconds()
+    assert int(total_seconds) == total_seconds, total_seconds
+    total_seconds = int(total_seconds)
+
+    seconds = total_seconds
+
+    days = seconds // (24 * 3600)
+    seconds %= 24 * 3600
+    hours = seconds // 3600
+    seconds %= 3600
+    minutes = seconds // 60
+    seconds %= 60
+
+    if days > 0 and hours == 0 and minutes == 0 and seconds == 0:
+        return f"{days}d"
+
+    if days == 0 and hours > 0 and minutes == 0 and seconds == 0:
+        return f"{hours}h"
+
+    if days == 0 and hours == 0 and minutes > 0 and seconds == 0:
+        return f"{minutes}m"
+
+    if days == 0 and hours == 0 and minutes == 0 and seconds > 0:
+        return f"{seconds}s"
+
+    if days > 0:
+        return f"{total_seconds}s"
+
+    return str(frequency)
+
+
+def frequency_to_seconds(frequency):
+    """Convert a frequency to seconds.
+
+    Parameters
+    ----------
+    frequency : _type_
+        _description_
+
+    Returns
+    -------
+    int
+        Number of seconds.
+    """
+
+    result = frequency_to_timedelta(frequency).total_seconds()
+    assert int(result) == result, result
+    return int(result)
+
+
 DOW = {
     "monday": 0,
     "tuesday": 1,
@@ -142,7 +269,7 @@ class DateTimes:
         """
         self.start = as_datetime(start)
         self.end = as_datetime(end)
-        self.increment = datetime.timedelta(hours=increment)
+        self.increment = frequency_to_timedelta(increment)
         self.day_of_month = _make_day(day_of_month)
         self.day_of_week = _make_week(day_of_week)
         self.calendar_months = _make_months(calendar_months)
@@ -273,8 +400,7 @@ def datetimes_factory(*args, **kwargs):
 
         kwargs = kwargs.copy()
         if "frequency" in kwargs:
-            freq = kwargs.pop("frequency")
-            kwargs["increment"] = normalise_frequency(freq)
+            kwargs["increment"] = kwargs.pop("frequency")
         return DateTimes(**kwargs)
 
     if not any((isinstance(x, dict) or isinstance(x, list)) for x in args):

anemoi/utils/humanize.py

@@ -15,6 +15,8 @@ import re
 import warnings
 from collections import defaultdict
 
+from anemoi.utils.dates import as_datetime
+
 
 def bytes_to_human(n: float) -> str:
     """Convert a number of bytes to a human readable string
@@ -625,3 +627,60 @@ def shorten_list(lst, max_length=5):
         if isinstance(lst, tuple):
             return tuple(result)
         return result
+
+
+def _compress_dates(dates):
+    dates = sorted(dates)
+    if len(dates) < 3:
+        yield dates
+        return
+
+    prev = first = dates.pop(0)
+    curr = dates.pop(0)
+    delta = curr - prev
+    while curr - prev == delta:
+        prev = curr
+        if not dates:
+            break
+        curr = dates.pop(0)
+
+    yield (first, prev, delta)
+    if dates:
+        yield from _compress_dates([curr] + dates)
+
+
+def compress_dates(dates):
+    """Compress a list of dates into a human-readable format.
+
+    Parameters
+    ----------
+    dates : list
+        A list of dates, as datetime objects or strings.
+
+    Returns
+    -------
+    str
+        A human-readable string representing the compressed dates.
+    """
+
+    dates = [as_datetime(_) for _ in dates]
+    result = []
+
+    for n in _compress_dates(dates):
+        if isinstance(n, list):
+            result.extend([str(_) for _ in n])
+        else:
+            result.append(" ".join([str(n[0]), "to", str(n[1]), "by", str(n[2])]))
+
+    return result
+
+
+def print_dates(dates):
+    """Print a list of dates in a human-readable format.
+
+    Parameters
+    ----------
+    dates : list
+        A list of dates, as datetime objects or strings.
+    """
+    print(compress_dates(dates))

anemoi/utils/s3.py

@@ -343,6 +343,7 @@ def upload(source, target, *, overwrite=False, resume=False, verbosity=1, progre
     """
 
     uploader = Upload()
+
     if os.path.isdir(source):
         uploader.transfer_folder(
             source=source,

anemoi/utils/text.py

@@ -7,6 +7,7 @@
 
 """Text utilities"""
 
+import re
 from collections import defaultdict
 
 # https://en.wikipedia.org/wiki/Box-drawing_character
@@ -32,6 +33,62 @@ def dotted_line(width=84) -> str:
     return "┈" * width
 
 
+# Regular expression to match ANSI escape codes
+_ansi_escape = re.compile(r"\x1b\[([0-9;]*[mGKH])")
+
+
+def _has_ansi_escape(s):
+    return _ansi_escape.search(s) is not None
+
+
+def _split_tokens(s):
+    """Split a string into a list of visual characters with their lenghts."""
+    from wcwidth import wcswidth
+
+    initial = s
+    out = []
+
+    # Function to probe the number of bytes needed to encode the first character
+    def probe_utf8(s):
+        for i in range(1, 5):
+            try:
+                s[:i].encode("utf-8")
+            except UnicodeEncodeError:
+                return i - 1
+        return 1
+
+    while s:
+        match = _ansi_escape.match(s)
+        if match:
+            token = match.group(0)
+            s = s[len(token) :]
+            out.append((token, 0))
+        else:
+            i = probe_utf8(s)
+            token = s[:i]
+            s = s[i:]
+            out.append((token, wcswidth(token)))
+
+    assert "".join(token for (token, _) in out) == initial, (out, initial)
+    return out
+
+
+def visual_len(s):
+    """Compute the length of a string as it appears on the terminal."""
+    if isinstance(s, str):
+        s = _split_tokens(s)
+    assert isinstance(s, (tuple, list)), (type(s), s)
+    if len(s) == 0:
+        return 0
+    for _ in s:
+        assert isinstance(_, tuple), s
+        assert len(_) == 2, s
+    n = 0
+    for _, width in s:
+        n += width
+    return n
+
+
 def boxed(text, min_width=80, max_width=80) -> str:
     """Put a box around a text
 
@@ -57,27 +114,44 @@ def boxed(text, min_width=80, max_width=80) -> str:
 
     """
 
-    lines = text.split("\n")
-    width = max(len(_) for _ in lines)
+    lines = []
+    for line in text.split("\n"):
+        line = line.strip()
+        line = _split_tokens(line)
+        lines.append(line)
+
+    width = max(visual_len(_) for _ in lines)
 
     if min_width is not None:
         width = max(width, min_width)
 
     if max_width is not None:
+
+        def shorten_line(line, max_width):
+            if visual_len(line) > max_width:
+                while visual_len(line) >= max_width:
+                    line = line[:-1]
+                line.append(("…", 1))
+            return line
+
         width = min(width, max_width)
-        lines = []
-        for line in text.split("\n"):
-            if len(line) > max_width:
-                line = line[: max_width - 1] + "…"
-            lines.append(line)
-        text = "\n".join(lines)
+        lines = [shorten_line(line, max_width) for line in lines]
+
+    def pad_line(line, width):
+        line = line + [" "] * (width - visual_len(line))
+        return line
+
+    lines = [pad_line(line, width) for line in lines]
 
     box = []
     box.append("┌" + "─" * (width + 2) + "┐")
     for line in lines:
-        box.append(f"│ {line:{width}} │")
-
+        s = "".join(_[0] for _ in line)
+        if _has_ansi_escape(s):
+            s += "\x1b[0m"
+        box.append(f"│ {s} │")
     box.append("└" + "─" * (width + 2) + "┘")
+
     return "\n".join(box)
 
 
@@ -241,7 +315,7 @@ def table(rows, header, align, margin=0):
                ['B', 120, 1],
                ['C', 9, 123]],
                ['C1', 'C2', 'C3'],
-               ['<', '>', '>']))
+               ['<', '>', '>'])
         C1 │  C2 │  C3
         ───┼─────┼────
         Aa │  12 │   5

{anemoi_utils-0.3.13.dist-info → anemoi_utils-0.3.15.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: anemoi-utils
-Version: 0.3.13
+Version: 0.3.15
 Summary: A package to hold various functions to support training of ML models on ECMWF data.
 Author-email: "European Centre for Medium-Range Weather Forecasts (ECMWF)" <software.support@ecmwf.int>
 License: Apache License
@@ -223,6 +223,7 @@ Classifier: Programming Language :: Python :: Implementation :: CPython
 Classifier: Programming Language :: Python :: Implementation :: PyPy
 Requires-Python: >=3.9
 License-File: LICENSE
+Requires-Dist: isodate
 Requires-Dist: pyyaml
 Requires-Dist: tomli
 Requires-Dist: tqdm

{anemoi_utils-0.3.13.dist-info → anemoi_utils-0.3.15.dist-info}/RECORD

@@ -1,25 +1,25 @@
 anemoi/utils/__init__.py,sha256=zZZpbKIoGWwdCOuo6YSruLR7C0GzvzI1Wzhyqaa0K7M,456
 anemoi/utils/__main__.py,sha256=cLA2PidDTOUHaDGzd0_E5iioKYNe-PSTv567Y2fuwQk,723
-anemoi/utils/_version.py,sha256=drg-3xPikZkJrT_ckut_6xmzJvTagr6bFEW0FoqfxwQ,413
+anemoi/utils/_version.py,sha256=9bARtG-NIsnxdzdFmq2OMzx2-WMRy4fvkJadsRpAZKk,413
 anemoi/utils/caching.py,sha256=EZ4bRG72aTvTxzrbYCgjFpdIn8OtA1rzoRmGg8caWsI,1919
 anemoi/utils/checkpoints.py,sha256=1_3mg4B-ykTVfIvIUEv7IxGyREx_ZcilVbB3U-V6O6I,5165
-anemoi/utils/cli.py,sha256=4uMy6LyL9VPg9ZVNKsimahT5BBVPQgWWk7LUel5dQZE,3545
+anemoi/utils/cli.py,sha256=SWb5_itARlDCq6yEf-VvagTioSW2phKXXFMW2ihXu18,3678
 anemoi/utils/config.py,sha256=s8eqlHsuak058_NJXGMOoT2HenwiZJKcZ9plUWvO7tw,8865
-anemoi/utils/dates.py,sha256=P-AW-dViogxn4t8Q10sO1HldJ2-lez6ILrx6z59tA10,7454
+anemoi/utils/dates.py,sha256=6jLdWkPGmdCKyMJonB5MNPe_RuCBURm1Q0frjHqLk54,10655
 anemoi/utils/grib.py,sha256=gVfo4KYQv31iRyoqRDwk5tiqZDUgOIvhag_kO0qjYD0,3067
 anemoi/utils/hindcasts.py,sha256=X8k-81ltmkTDHdviY0SJgvMg7XDu07xoc5ALlUxyPoo,1453
-anemoi/utils/humanize.py,sha256=-MRGgGG_STTWzIagL4RQnOAooqAIigCcuN-RZ6CK75M,14982
+anemoi/utils/humanize.py,sha256=w1D17HqTT89xvNgeZlLDw_GNkWngEpv23JHk5A9tGtU,16265
 anemoi/utils/provenance.py,sha256=NL36lM_aCw3fG6VIAouZCRBAJv8a6M3x9cScrFxCMcA,9579
-anemoi/utils/s3.py,sha256=thSweHcsbp86CzGh_EMG98iGzJ4rRf2FeuQkXYv1xN8,18505
-anemoi/utils/text.py,sha256=4Zlc4r9dzRjkKL9xqp2vuQsoJY15bJ3y_Xv3YW_XsmU,8510
+anemoi/utils/s3.py,sha256=kDzbs4nVD2lQuppSe88NVSNpy0wSZpuzkmcAgN2irkU,18506
+anemoi/utils/text.py,sha256=5HBqNwhifus4d3OUnod5q1VgCBdEpzE7o0IR0S85knw,10397
 anemoi/utils/timer.py,sha256=EQcucuwUaGeSpt2S1APJlwSOu6kC47MK9f4h-r8c_AY,990
 anemoi/utils/commands/__init__.py,sha256=qAybFZPBBQs0dyx7dZ3X5JsLpE90pwrqt1vSV7cqEIw,706
 anemoi/utils/commands/config.py,sha256=KEffXZh0ZQfn8t6LXresfd94kDY0gEyulx9Wto5ttW0,824
 anemoi/utils/mars/__init__.py,sha256=RAeY8gJ7ZvsPlcIvrQ4fy9xVHs3SphTAPw_XJDtNIKo,1750
 anemoi/utils/mars/mars.yaml,sha256=R0dujp75lLA4wCWhPeOQnzJ45WZAYLT8gpx509cBFlc,66
-anemoi_utils-0.3.13.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-anemoi_utils-0.3.13.dist-info/METADATA,sha256=SBTqiIl_wfFDAOH_6Bu5Tol4LJV6ZE8AtuATA0n2Pyg,15447
-anemoi_utils-0.3.13.dist-info/WHEEL,sha256=-oYQCr74JF3a37z2nRlQays_SX2MqOANoqVjBBAP2yE,91
-anemoi_utils-0.3.13.dist-info/entry_points.txt,sha256=LENOkn88xzFQo-V59AKoA_F_cfYQTJYtrNTtf37YgHY,60
-anemoi_utils-0.3.13.dist-info/top_level.txt,sha256=DYn8VPs-fNwr7fNH9XIBqeXIwiYYd2E2k5-dUFFqUz0,7
-anemoi_utils-0.3.13.dist-info/RECORD,,
+anemoi_utils-0.3.15.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+anemoi_utils-0.3.15.dist-info/METADATA,sha256=embyjlpZISW0daGqXYuZAJ9t4zHRe1IxfiDA213Mcis,15470
+anemoi_utils-0.3.15.dist-info/WHEEL,sha256=Mdi9PDNwEZptOjTlUcAth7XJDFtKrHYaQMPulZeBCiQ,91
+anemoi_utils-0.3.15.dist-info/entry_points.txt,sha256=LENOkn88xzFQo-V59AKoA_F_cfYQTJYtrNTtf37YgHY,60
+anemoi_utils-0.3.15.dist-info/top_level.txt,sha256=DYn8VPs-fNwr7fNH9XIBqeXIwiYYd2E2k5-dUFFqUz0,7
+anemoi_utils-0.3.15.dist-info/RECORD,,

{anemoi_utils-0.3.13.dist-info → anemoi_utils-0.3.15.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (71.0.3)
+Generator: setuptools (73.0.1)
 Root-Is-Purelib: true
 Tag: py3-none-any