Hide keyboard shortcuts

Hot-keys on this page

r m x p   toggle line displays

j k   next/prev highlighted chunk

0   (zero) top of page

1   (one) first highlighted chunk

1# 

2# This file is part of ap_verify. 

3# 

4# Developed for the LSST Data Management System. 

5# This product includes software developed by the LSST Project 

6# (http://www.lsst.org). 

7# See the COPYRIGHT file at the top-level directory of this distribution 

8# for details of code ownership. 

9# 

10# This program is free software: you can redistribute it and/or modify 

11# it under the terms of the GNU General Public License as published by 

12# the Free Software Foundation, either version 3 of the License, or 

13# (at your option) any later version. 

14# 

15# This program is distributed in the hope that it will be useful, 

16# but WITHOUT ANY WARRANTY; without even the implied warranty of 

17# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 

18# GNU General Public License for more details. 

19# 

20# You should have received a copy of the GNU General Public License 

21# along with this program. If not, see <http://www.gnu.org/licenses/>. 

22# 

23 

24"""Command-line program for running and analyzing AP pipeline. 

25 

26In addition to containing ap_verify's main function, this module manages 

27command-line argument parsing. 

28""" 

29 

30__all__ = ["runApVerify", "runIngestion"] 

31 

32import argparse 

33import re 

34 

35import lsst.log 

36 

37from .dataset import Dataset 

38from .ingestion import ingestDataset, ingestDatasetGen3 

39from .metrics import MetricsParser, computeMetrics 

40from .pipeline_driver import ApPipeParser, runApPipeGen2, runApPipeGen3 

41from .workspace import WorkspaceGen2, WorkspaceGen3 

42 

43 

44class _InputOutputParser(argparse.ArgumentParser): 

45 """An argument parser for program-wide input and output. 

46 

47 This parser is not complete, and is designed to be passed to another parser 

48 using the `parent` parameter. 

49 """ 

50 

51 def __init__(self): 

52 # Help and documentation will be handled by main program's parser 

53 argparse.ArgumentParser.__init__(self, add_help=False) 

54 self.add_argument('--dataset', action=_DatasetAction, choices=Dataset.getSupportedDatasets(), 

55 required=True, help='The source of data to pass through the pipeline.') 

56 self.add_argument('--output', required=True, 

57 help='The location of the workspace to use for pipeline repositories.') 

58 

59 gen23 = self.add_mutually_exclusive_group() 

60 # Because store_true and store_false use the same dest, add explicit 

61 # default to avoid ambiguity. 

62 gen23.add_argument('--gen2', dest='useGen3', action='store_false', default=False, 

63 help='Handle the ap_verify dataset using the Gen 2 framework (default).') 

64 gen23.add_argument('--gen3', dest='useGen3', action='store_true', default=False, 

65 help='Handle the ap_verify dataset using the Gen 3 framework.') 

66 

67 

68class _ProcessingParser(argparse.ArgumentParser): 

69 """An argument parser for general run-time characteristics. 

70 

71 This parser is not complete, and is designed to be passed to another parser 

72 using the `parent` parameter. 

73 """ 

74 

75 def __init__(self): 

76 # Help and documentation will be handled by main program's parser 

77 argparse.ArgumentParser.__init__(self, add_help=False) 

78 self.add_argument("-j", "--processes", default=1, type=int, 

79 help="Number of processes to use.") 

80 

81 

82class _ApVerifyParser(argparse.ArgumentParser): 

83 """An argument parser for data needed by the main ap_verify program. 

84 """ 

85 

86 def __init__(self): 

87 argparse.ArgumentParser.__init__( 

88 self, 

89 description='Executes the LSST DM AP pipeline and analyzes its performance using metrics.', 

90 epilog='', 

91 parents=[_InputOutputParser(), _ProcessingParser(), ApPipeParser(), MetricsParser()], 

92 add_help=True) 

93 

94 

95class _IngestOnlyParser(argparse.ArgumentParser): 

96 """An argument parser for data needed by dataset ingestion. 

97 """ 

98 

99 def __init__(self): 

100 argparse.ArgumentParser.__init__( 

101 self, 

102 description='Ingests an ap_verify dataset into a pair of Butler repositories. ' 

103 'The program will create repository(ies) appropriate for --gen2 or --gen3 ' 

104 'in subdirectories of <OUTPUT>. ' 

105 'These repositories may be used directly by ap_verify.py by ' 

106 'passing the same --output argument, or by other programs that accept ' 

107 'Butler repositories as input.', 

108 epilog='', 

109 parents=[_InputOutputParser(), _ProcessingParser()], 

110 add_help=True) 

111 

112 

113class _FormattedType: 

114 """An argparse type converter that requires strings in a particular format. 

115 

116 Leaves the input as a string if it matches, else raises `argparse.ArgumentTypeError`. 

117 

118 Parameters 

119 ---------- 

120 fmt : `str` 

121 A regular expression that values must satisfy to be accepted. The *entire* string must match the 

122 expression in order to pass. 

123 msg : `str` 

124 An error string to display for invalid values. The first "%s" shall be filled with the 

125 invalid argument. 

126 """ 

127 def __init__(self, fmt, msg='"%s" does not have the expected format.'): 

128 fullFormat = fmt 

129 if not fullFormat.startswith('^'): 

130 fullFormat = '^' + fullFormat 

131 if not fullFormat.endswith('$'): 

132 fullFormat += '$' 

133 self._format = re.compile(fullFormat) 

134 self._message = msg 

135 

136 def __call__(self, value): 

137 if self._format.match(value): 

138 return value 

139 else: 

140 raise argparse.ArgumentTypeError(self._message % value) 

141 

142 

143class _DatasetAction(argparse.Action): 

144 """A converter for dataset arguments. 

145 

146 Not an argparse type converter so that the ``choices`` parameter can be 

147 expressed using strings; ``choices`` checks happen after type conversion 

148 but before actions. 

149 """ 

150 def __call__(self, _parser, namespace, values, _option_string=None): 

151 setattr(namespace, self.dest, Dataset(values)) 

152 

153 

154def runApVerify(cmdLine=None): 

155 """Execute the AP pipeline while handling metrics. 

156 

157 This is the main function for ``ap_verify``, and handles logging, 

158 command-line argument parsing, pipeline execution, and metrics 

159 generation. 

160 

161 Parameters 

162 ---------- 

163 cmdLine : `list` of `str` 

164 an optional command line used to execute `runApVerify` from other 

165 Python code. If `None`, `sys.argv` will be used. 

166 

167 Returns 

168 ------- 

169 nFailed : `int` 

170 The number of data IDs that were not successfully processed, up to 127, 

171 or 127 if the task runner framework failed. 

172 """ 

173 lsst.log.configure() 

174 log = lsst.log.Log.getLogger('ap.verify.ap_verify.main') 

175 # TODO: what is LSST's policy on exceptions escaping into main()? 

176 args = _ApVerifyParser().parse_args(args=cmdLine) 

177 log.debug('Command-line arguments: %s', args) 

178 

179 if args.useGen3: 

180 workspace = WorkspaceGen3(args.output) 

181 ingestDatasetGen3(args.dataset, workspace, processes=args.processes) 

182 log.info('Running pipeline...') 

183 # Gen 3 pipeline includes both AP and metrics 

184 return runApPipeGen3(workspace, args, processes=args.processes) 

185 else: 

186 workspace = WorkspaceGen2(args.output) 

187 ingestDataset(args.dataset, workspace) 

188 log.info('Running pipeline...') 

189 apPipeResults = runApPipeGen2(workspace, args, processes=args.processes) 

190 computeMetrics(workspace, apPipeResults.parsedCmd.id, args) 

191 return _getCmdLineExitStatus(apPipeResults.resultList) 

192 

193 

194def _getCmdLineExitStatus(resultList): 

195 """Return the exit status following the conventions of 

196 :ref:`running a CmdLineTask from the command line 

197 <command-line-task-argument-reference>`. 

198 

199 Parameters 

200 ---------- 

201 resultList : `list` [`Struct`] or `None` 

202 A list of `Struct`, as returned by `ApPipeTask.parseAndRun`. Each 

203 element must contain at least an ``exitStatus`` member. 

204 

205 Returns 

206 ------- 

207 exitStatus : `int` 

208 The number of failed runs in ``resultList``, up to 127, or 127 if 

209 ``resultList`` is `None`. 

210 """ 

211 if resultList: 

212 # ApPipeTaskRunner does not override default results handling, exitStatus always defined 

213 return min(127, sum(((res.exitStatus != 0) for res in resultList))) 

214 else: 

215 return 127 

216 

217 

218def runIngestion(cmdLine=None): 

219 """Ingest a dataset, but do not process it. 

220 

221 This is the main function for ``ingest_dataset``, and handles logging, 

222 command-line argument parsing, and ingestion. 

223 

224 Parameters 

225 ---------- 

226 cmdLine : `list` of `str` 

227 an optional command line used to execute `runIngestion` from other 

228 Python code. If `None`, `sys.argv` will be used. 

229 """ 

230 lsst.log.configure() 

231 log = lsst.log.Log.getLogger('ap.verify.ap_verify.ingest') 

232 # TODO: what is LSST's policy on exceptions escaping into main()? 

233 args = _IngestOnlyParser().parse_args(args=cmdLine) 

234 log.debug('Command-line arguments: %s', args) 

235 

236 if args.useGen3: 

237 workspace = WorkspaceGen3(args.output) 

238 ingestDatasetGen3(args.dataset, workspace, processes=args.processes) 

239 else: 

240 workspace = WorkspaceGen2(args.output) 

241 ingestDataset(args.dataset, workspace)