4 |
|
from DBSAPI.dbsApiException import * |
5 |
|
import common |
6 |
|
from crab_util import * |
7 |
+ |
import os |
8 |
|
|
9 |
|
|
10 |
|
# ####################################### |
71 |
|
# #################################### |
72 |
|
# class to find and extact info from published data |
73 |
|
class DataDiscovery: |
74 |
< |
def __init__(self, datasetPath, cfg_params): |
74 |
> |
def __init__(self, datasetPath, cfg_params, skipAnBlocks): |
75 |
|
|
76 |
|
# Attributes |
77 |
|
self.datasetPath = datasetPath |
78 |
|
self.cfg_params = cfg_params |
79 |
+ |
self.skipBlocks = skipAnBlocks |
80 |
|
|
81 |
|
self.eventsPerBlock = {} # DBS output: map fileblocks-events for collection |
82 |
|
self.eventsPerFile = {} # DBS output: map files-events |
83 |
|
self.blocksinfo = {} # DBS output: map fileblocks-files |
84 |
|
self.maxEvents = 0 # DBS output: max events |
85 |
+ |
self.parent = {} # DBS output: max events |
86 |
|
|
87 |
|
# #################################### |
88 |
|
def fetchDBSInfo(self): |
102 |
|
if (self.cfg_params.has_key('CMSSW.runselection')): |
103 |
|
runselection = parseRange2(self.cfg_params['CMSSW.runselection']) |
104 |
|
|
105 |
+ |
common.logger.debug(6,"runselection is: %s"%runselection) |
106 |
|
## service API |
107 |
|
args = {} |
108 |
|
args['url'] = dbs_url |
109 |
|
args['level'] = 'CRITICAL' |
110 |
|
|
111 |
+ |
## check if has been requested to use the parent info |
112 |
+ |
useParent = self.cfg_params.get('CMSSW.use_parent',False) |
113 |
+ |
|
114 |
+ |
## check if has been asked for a non default file to store/read analyzed fileBlocks |
115 |
+ |
defaultName = common.work_space.shareDir()+'AnalyzedBlocks.txt' |
116 |
+ |
fileBlocks_FileName = os.path.abspath(self.cfg_params.get('CMSSW.fileblocks_file',defaultName)) |
117 |
+ |
|
118 |
|
api = DBSAPI.dbsApi.DbsApi(args) |
119 |
|
try: |
120 |
|
if len(runselection) <= 0 : |
121 |
< |
files = api.listDatasetFiles(self.datasetPath) |
121 |
> |
if useParent: |
122 |
> |
allowedRetriveValue = ['retrive_parent', |
123 |
> |
'retrive_block', |
124 |
> |
'retrive_lumi', |
125 |
> |
'retrive_run' |
126 |
> |
] |
127 |
> |
files = api.listFiles(path=self.datasetPath, retriveList=allowedRetriveValue) |
128 |
> |
common.logger.debug(5,"Set of input parameters used for DBS query : \n"+str(allowedRetriveValue)) |
129 |
> |
common.logger.write("Set of input parameters used for DBS query : \n"+str(allowedRetriveValue)) |
130 |
> |
else: |
131 |
> |
files = api.listDatasetFiles(self.datasetPath) |
132 |
|
else : |
133 |
< |
files = api.listFiles(path=self.datasetPath, details=True) |
133 |
> |
files=[] |
134 |
> |
for arun in runselection: |
135 |
> |
try: |
136 |
> |
filesinrun = api.listFiles(path=self.datasetPath,retriveList=allowedRetriveValue,runNumber=arun) |
137 |
> |
files.extend(filesinrun) |
138 |
> |
except: |
139 |
> |
msg="WARNING: problem extracting info from DBS for run %s "%arun |
140 |
> |
common.logger.message(msg) |
141 |
> |
pass |
142 |
> |
|
143 |
|
except DbsBadRequest, msg: |
144 |
|
raise DataDiscoveryError(msg) |
145 |
|
except DBSError, msg: |
146 |
|
raise DataDiscoveryError(msg) |
147 |
|
|
148 |
+ |
anFileBlocks = [] |
149 |
+ |
if self.skipBlocks: anFileBlocks = readTXTfile(self, fileBlocks_FileName) |
150 |
+ |
|
151 |
|
# parse files and fill arrays |
152 |
|
for file in files : |
153 |
< |
filename = file['LogicalFileName'] |
154 |
< |
if filename.find('.dat') < 0 : |
155 |
< |
fileblock = file['Block']['Name'] |
156 |
< |
events = file['NumberOfEvents'] |
157 |
< |
continue_flag = 0 |
158 |
< |
if len(runselection) > 0 : |
159 |
< |
runslist = file['RunsList'] |
160 |
< |
for run in runslist : |
161 |
< |
runnumber = run['RunNumber'] |
162 |
< |
for selected_run in runselection : |
130 |
< |
if runnumber == selected_run : |
131 |
< |
continue_flag = 1 |
132 |
< |
else : |
133 |
< |
continue_flag = 1 |
134 |
< |
|
135 |
< |
if continue_flag == 1 : |
153 |
> |
parList = [] |
154 |
> |
# skip already analyzed blocks |
155 |
> |
fileblock = file['Block']['Name'] |
156 |
> |
if fileblock not in anFileBlocks : |
157 |
> |
filename = file['LogicalFileName'] |
158 |
> |
# asked retry the list of parent for the given child |
159 |
> |
if useParent: parList = [x['LogicalFileName'] for x in file['ParentList']] |
160 |
> |
self.parent[filename] = parList |
161 |
> |
if filename.find('.dat') < 0 : |
162 |
> |
events = file['NumberOfEvents'] |
163 |
|
# number of events per block |
164 |
|
if fileblock in self.eventsPerBlock.keys() : |
165 |
|
self.eventsPerBlock[fileblock] += events |
166 |
|
else : |
167 |
|
self.eventsPerBlock[fileblock] = events |
141 |
– |
|
168 |
|
# number of events per file |
169 |
|
self.eventsPerFile[filename] = events |
170 |
< |
|
170 |
> |
|
171 |
|
# number of events per block |
172 |
|
if fileblock in self.blocksinfo.keys() : |
173 |
|
self.blocksinfo[fileblock].append(filename) |
174 |
|
else : |
175 |
|
self.blocksinfo[fileblock] = [filename] |
176 |
< |
|
176 |
> |
|
177 |
|
# total number of events |
178 |
|
self.maxEvents += events |
179 |
+ |
if self.skipBlocks and len(self.eventsPerBlock.keys()) == 0: |
180 |
+ |
msg = "No new fileblocks available for dataset: "+str(self.datasetPath) |
181 |
+ |
raise CrabException(msg) |
182 |
|
|
183 |
+ |
saveFblocks='' |
184 |
|
for block in self.eventsPerBlock.keys() : |
185 |
+ |
saveFblocks += str(block)+'\n' |
186 |
|
common.logger.debug(6,"DBSInfo: total nevts %i in block %s "%(self.eventsPerBlock[block],block)) |
187 |
< |
|
187 |
> |
writeTXTfile(self, fileBlocks_FileName , saveFblocks) |
188 |
> |
|
189 |
|
if len(self.eventsPerBlock) <= 0: |
190 |
|
raise NotExistingDatasetError(("\nNo data for %s in DBS\nPlease check" |
191 |
|
+ " dataset path variables in crab.cfg") |
220 |
|
""" |
221 |
|
return self.blocksinfo |
222 |
|
|
223 |
+ |
# ################################################# |
224 |
+ |
def getParent(self): |
225 |
+ |
""" |
226 |
+ |
return parent grouped by file |
227 |
+ |
""" |
228 |
+ |
return self.parent |
229 |
+ |
|
230 |
|
######################################################################## |