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 

39from .metrics import MetricsParser, computeMetrics 

40from .pipeline_driver import ApPipeParser, runApPipe 

41from .workspace import Workspace 

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 

60class _ApVerifyParser(argparse.ArgumentParser): 

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

62 """ 

63 

64 def __init__(self): 

65 argparse.ArgumentParser.__init__( 

66 self, 

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

68 epilog='', 

69 parents=[_InputOutputParser(), ApPipeParser(), MetricsParser()], 

70 add_help=True) 

71 

72 

73class _IngestOnlyParser(argparse.ArgumentParser): 

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

75 """ 

76 

77 def __init__(self): 

78 argparse.ArgumentParser.__init__( 

79 self, 

80 description='Ingests a dataset into a pair of Butler repositories.' 

81 'The program will create a data repository in <OUTPUT>/ingested and a calib repository ' 

82 'in <OUTPUT>/calibingested. ' 

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

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

85 'Butler repositories as input.', 

86 epilog='', 

87 parents=[_InputOutputParser()], 

88 add_help=True) 

89 

90 

91class _FormattedType: 

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

93 

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

95 

96 Parameters 

97 ---------- 

98 fmt : `str` 

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

100 expression in order to pass. 

101 msg : `str` 

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

103 invalid argument. 

104 """ 

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

106 fullFormat = fmt 

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

108 fullFormat = '^' + fullFormat 

109 if not fullFormat.endswith('$'): 

110 fullFormat += '$' 

111 self._format = re.compile(fullFormat) 

112 self._message = msg 

113 

114 def __call__(self, value): 

115 if self._format.match(value): 

116 return value 

117 else: 

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

119 

120 

121class _DatasetAction(argparse.Action): 

122 """A converter for dataset arguments. 

123 

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

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

126 but before actions. 

127 """ 

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

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

130 

131 

132def runApVerify(cmdLine=None): 

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

134 

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

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

137 generation. 

138 

139 Parameters 

140 ---------- 

141 cmdLine : `list` of `str` 

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

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

144 

145 Returns 

146 ------- 

147 nFailed : `int` 

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

149 or 127 if the task runner framework failed. 

150 """ 

151 lsst.log.configure() 

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

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

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

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

156 

157 workspace = Workspace(args.output) 

158 ingestDataset(args.dataset, workspace) 

159 

160 log.info('Running pipeline...') 

161 apPipeResults = runApPipe(workspace, args) 

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

163 

164 return _getCmdLineExitStatus(apPipeResults.resultList) 

165 

166 

167def _getCmdLineExitStatus(resultList): 

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

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

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

171 

172 Parameters 

173 ---------- 

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

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

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

177 

178 Returns 

179 ------- 

180 exitStatus : `int` 

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

182 ``resultList`` is `None`. 

183 """ 

184 if resultList: 

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

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

187 else: 

188 return 127 

189 

190 

191def runIngestion(cmdLine=None): 

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

193 

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

195 command-line argument parsing, and ingestion. 

196 

197 Parameters 

198 ---------- 

199 cmdLine : `list` of `str` 

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

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

202 """ 

203 lsst.log.configure() 

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

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

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

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

208 

209 workspace = Workspace(args.output) 

210 ingestDataset(args.dataset, workspace)