eic-opticks/ana/datedfolder.py

0001 #!/usr/bin/env python
0002 #
0003 # Copyright (c) 2019 Opticks Team. All Rights Reserved.
0004 #
0005 # This file is part of Opticks
0006 # (see https://bitbucket.org/simoncblyth/opticks).
0007 #
0008 # Licensed under the Apache License, Version 2.0 (the "License");
0009 # you may not use this file except in compliance with the License.
0010 # You may obtain a copy of the License at
0011 #
0012 #   http://www.apache.org/licenses/LICENSE-2.0
0013 #
0014 # Unless required by applicable law or agreed to in writing, software
0015 # distributed under the License is distributed on an "AS IS" BASIS,
0016 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
0017 # See the License for the specific language governing permissions and
0018 # limitations under the License.
0019 #
0020
0021 """
0022 datedfolder.py
0023 ================
0024
0025 Find dated folders beneath a base directory argument or beneath the current directory.
0026
0027 ::
0028
0029     [blyth@localhost issues]$ ~/opticks/ana/datedfolder.py /home/blyth/local/opticks/results/geocache-bench
0030     INFO:__main__:DatedFolder.find searching for date stamped folders beneath : /home/blyth/local/opticks/results/geocache-bench
0031     /home/blyth/local/opticks/results/geocache-bench/OFF_TITAN_RTX/20190422_175618
0032     /home/blyth/local/opticks/results/geocache-bench/ON_TITAN_RTX/20190422_175618
0033     /home/blyth/local/opticks/results/geocache-bench/OFF_TITAN_V/20190422_175618
0034     /home/blyth/local/opticks/results/geocache-bench/ON_TITAN_V/20190422_175618
0035     /home/blyth/local/opticks/results/geocache-bench/OFF_TITAN_V_AND_TITAN_RTX/20190422_175618
0036     [blyth@localhost issues]$
0037
0038 """
0039 from datetime import datetime
0040 import os, re, sys, logging
0041 log = logging.getLogger(__name__)
0042
0043 class DateParser(object):
0044     ptn = re.compile("(\d{4})(\d{2})(\d{2})_(\d{2})(\d{2})(\d{2})")
0045     def __call__(self, name):
0046         """
0047         :param name: basename of a directory
0048         :return dt: datetime if matches the date format 20160819_153245 otherwise return None
0049         """
0050         m = self.ptn.match(name)
0051         if m is None:return None
0052         if len(m.groups()) != 6:return None
0053         dt = datetime(*map(int, m.groups()))
0054         return dt
0055
0056 dateparser = DateParser()
0057
0058 def finddir(base, dirfilter=lambda _:True, relative=True):
0059     """
0060     :param base: directory to walk looking for directories to be dirfiltered
0061     :param dirfilter: function that takes directory path argument and returns true when selected
0062     :param relative: when True returns the base relative path, otherwise absolute paths are returned
0063     :return paths: all directory paths found that satisfy the filter
0064     """
0065     paths = []
0066     for root, dirs, files in os.walk(base):
0067         for name in dirs:
0068             path = os.path.join(root,name)
0069             d = dirfilter(path)
0070             if d is not None:
0071                 paths.append(path[len(base)+1:] if relative else path)
0072             pass
0073         pass
0074     pass
0075     return paths
0076
0077 class DatedFolder(object):
0078     @classmethod
0079     def find(cls, base):
0080         """
0081         :param base: directory
0082         :return dirs, dfolds, dtimes:
0083
0084         Groups of executable runs are grouped together by them using
0085         the same datestamp datedfolder name.
0086         So there can be more dirs that corresponding dfolds and dtimes.
0087         """
0088         while base.endswith("/"):
0089             base = base[:-1]
0090         pass
0091
0092         df = cls()
0093         log.info("DatedFolder.find searching for date stamped folders beneath : %s " % base)
0094
0095         dirs = finddir(base, df)                        # list of dated folders beneath base
0096         dfolds = list(set(map(os.path.basename, dirs))) # list of unique basenames of the dated folders
0097         dtimes = list(map(dateparser, dfolds ))         # list of datetimes
0098
0099         return dirs, dfolds, dtimes
0100
0101     def __call__(self, path):
0102         """
0103         :param path: directory
0104         :return datetime or None:
0105         """
0106         name = os.path.basename(path)
0107         #log.info(name)
0108         return dateparser(name)
0109
0110
0111 if __name__ == '__main__':
0112      logging.basicConfig(level=logging.INFO)
0113      base = sys.argv[1] if len(sys.argv) > 1 else "."
0114      dirs, dfolds, dtimes = DatedFolder.find(base)
0115
0116      print("\n".join(dfolds))
0117      for df in sorted(dfolds):
0118          print("\n".join(filter(lambda _:_.endswith(df),dirs)))
0119      pass
0120
0121
0122