| 1 |
#!/usr/bin/env python
|
| 2 |
import sys, os, string, re
|
| 3 |
import urllib, urllister
|
| 4 |
import urllib2
|
| 5 |
from UnserializePHP import *
|
| 6 |
from orcarcBuilder import *
|
| 7 |
|
| 8 |
class PubDBInfoError:
|
| 9 |
def __init__(self, Collections):
|
| 10 |
print '\nERROR accessing PubDB for Collections: '+Collections+'\n'
|
| 11 |
pass
|
| 12 |
class PubDBInfoNoCollectionError:
|
| 13 |
def __init__(self, Collections):
|
| 14 |
print '\nERROR No Collections found in PubDB : '+Collections+'\n'
|
| 15 |
pass
|
| 16 |
class NoPHPError:
|
| 17 |
def __init__(self, url):
|
| 18 |
#print '\nERROR accessing PHP at '+url+' \n'
|
| 19 |
print 'ERROR accessing PHP: ',url,'isn\'t updated version \n'
|
| 20 |
pass
|
| 21 |
class PubDBInfoResult:
|
| 22 |
def __init__(self,
|
| 23 |
contents):
|
| 24 |
self.contents=contents
|
| 25 |
|
| 26 |
################################################################################
|
| 27 |
# Class to connect to PubDB interface for the analysis and download the data in one shot using the serialized PHP data.
|
| 28 |
################################################################################
|
| 29 |
class PubDBInfo:
|
| 30 |
def __init__(self, pubdburl, Collections):
|
| 31 |
self.Collections= Collections
|
| 32 |
self.PrimaryCollID=string.split(Collections,'-')[0]
|
| 33 |
self.PubDBurl_ = pubdburl
|
| 34 |
self.PubDBInfophp_ = 'pubdb-get-analysisinfo.php'
|
| 35 |
self.protocolPrio_ = ['http', 'rfio', 'mysql' , 'gridftp']
|
| 36 |
|
| 37 |
##########################################################################
|
| 38 |
def GetPubDBInfo(self):
|
| 39 |
"""
|
| 40 |
Get all the catalogues-related info from PubDB and select the best ones if multiple choices are possible
|
| 41 |
"""
|
| 42 |
### extract catalogues related info from pubDB
|
| 43 |
cataloguecoll_map = self.ExtractPubDBInfo()
|
| 44 |
### select the Best catalogues
|
| 45 |
cataloguesinfos=self.SelectBestPubDBInfo(cataloguecoll_map)
|
| 46 |
return cataloguesinfos
|
| 47 |
|
| 48 |
#########################################################################
|
| 49 |
def ExtractPubDBInfo(self):
|
| 50 |
"""
|
| 51 |
Extract all the information from the PubDB analysis interface
|
| 52 |
"""
|
| 53 |
try:
|
| 54 |
#print " contacting PubDb... "+self.PubDBurl_+self.PubDBInfophp_+'?collid='+self.Collections+"\n"
|
| 55 |
f = urllib.urlopen(self.PubDBurl_+self.PubDBInfophp_+'?collid='+self.Collections)
|
| 56 |
except IOError:
|
| 57 |
raise PubDBInfoError(self.Collections)
|
| 58 |
|
| 59 |
data = f.read()
|
| 60 |
#print data
|
| 61 |
if len(data)>0:
|
| 62 |
if data[0]=='<':
|
| 63 |
raise PubDBInfoNoCollectionError(self.Collections)
|
| 64 |
try:
|
| 65 |
catalogues = PHPUnserialize().unserialize(data)
|
| 66 |
except IOError:
|
| 67 |
raise PHPUnserializeError(data)
|
| 68 |
try:
|
| 69 |
catinfos=[]
|
| 70 |
collmap={}
|
| 71 |
for k in catalogues.keys():
|
| 72 |
CollId=catalogues[k]['CollectionId']
|
| 73 |
## get also the collection type
|
| 74 |
CollType=catalogues[k]['CollectionType']
|
| 75 |
## set primary collection flag
|
| 76 |
PrimaryCollFlag=0
|
| 77 |
if ( CollId == self.PrimaryCollID ) : PrimaryCollFlag=1
|
| 78 |
colllist=[]
|
| 79 |
#print ">>> Catalogues for Collection: "+CollId+"\n"
|
| 80 |
|
| 81 |
cat=catalogues[k]['Catalogue']
|
| 82 |
for kcat in cat.keys():
|
| 83 |
##print ("key %s, val %s" %(kcat,cat[kcat]))
|
| 84 |
ContactString=cat[kcat]['ContactString']
|
| 85 |
ContactProtocol=cat[kcat]['ContactProtocol']
|
| 86 |
CatalogueType=cat[kcat]['CatalogueType']
|
| 87 |
ValidationStatus=cat[kcat]['ValidationStatus']
|
| 88 |
#print "CS: "+ContactString
|
| 89 |
#print "CP: "+ContactProtocol
|
| 90 |
#print "CT: "+CatalogueType
|
| 91 |
#print "VS: "+ValidationStatus
|
| 92 |
ce=cat[kcat]['CEs']
|
| 93 |
CElist=[]
|
| 94 |
for kce in ce.keys():
|
| 95 |
##print ("key %s, val %s" %(kce,ce[kce]))
|
| 96 |
CE=ce[kce]
|
| 97 |
CElist.append(ce[kce])
|
| 98 |
#print " CE list :"
|
| 99 |
#for aCE in CElist:
|
| 100 |
# print " CE : "+aCE
|
| 101 |
cc=cat[kcat]['CatalogueContents']
|
| 102 |
for kcc in cc.keys():
|
| 103 |
##print ("key %s, val %s" %(kcc,cc[kcc]))
|
| 104 |
FileType=cc[kcc]['FileType']
|
| 105 |
SE=cc[kcc]['SE']
|
| 106 |
#print "FT: "+FileType
|
| 107 |
#print "SE: "+SE
|
| 108 |
if cc[kcc]['Variables']==None:
|
| 109 |
Variables=''
|
| 110 |
else:
|
| 111 |
for kvar in cc[kcc]['Variables'].keys():
|
| 112 |
Variables=kvar+"="+cc[kcc]['Variables'][kvar]
|
| 113 |
#print "Variables: "+Variables
|
| 114 |
run=cc[kcc]['RunRange']
|
| 115 |
for krun in run.keys():
|
| 116 |
##print ("key %s, val %s" %(krun,run[krun]))
|
| 117 |
reTot = re.compile(r'TotalEvents=(\d*)')
|
| 118 |
TotalEvents=reTot.search(run[krun]).group(1)
|
| 119 |
reFirst= re.compile(r'FirstRun=(\d*)')
|
| 120 |
FirstRun = reFirst.search(run[krun]).group(1)
|
| 121 |
reLast= re.compile(r'LastRun=(\d*)')
|
| 122 |
LastRun = reLast.search(run[krun]).group(1)
|
| 123 |
#print "Nevents: "+TotalEvents
|
| 124 |
#print "First: "+FirstRun
|
| 125 |
#print "Last: "+LastRun
|
| 126 |
#print "----------------------------------"
|
| 127 |
|
| 128 |
## fill a catlogue entry
|
| 129 |
# acatalogue=catalogEntryNew(FileType,ValidationStatus,ContactString,ContactProtocol,CatalogueType,SE,CElist,TotalEvents,FirstRun+'-'+LastRun,Variables)
|
| 130 |
## store collection type and primarycollection flag
|
| 131 |
acatalogue=catalogEntryNew(CollType,PrimaryCollFlag,FileType,ValidationStatus,ContactString,ContactProtocol,CatalogueType,SE,CElist,TotalEvents,FirstRun+'-'+LastRun,Variables)
|
| 132 |
|
| 133 |
## list the catalogues belonging to a given collection
|
| 134 |
colllist.append(acatalogue)
|
| 135 |
|
| 136 |
## dictionary grouping catalogues by CollectionID
|
| 137 |
collmap[CollId]=colllist
|
| 138 |
|
| 139 |
except IOError:
|
| 140 |
raise PHPUnserializeError(data)
|
| 141 |
|
| 142 |
return collmap
|
| 143 |
|
| 144 |
########################################################################
|
| 145 |
def SelectBestPubDBInfo(self,cataloglist):
|
| 146 |
"""
|
| 147 |
Select the lists of needed catalogues (from a set of catalogues refering to the same collection and FileType selects the best on based on protocol)
|
| 148 |
"""
|
| 149 |
|
| 150 |
selectcatalogues=[]
|
| 151 |
|
| 152 |
### for each collection
|
| 153 |
for collid in cataloglist.keys():
|
| 154 |
#print ("key %s, val %s" %(collid,cataloglist[collid]))
|
| 155 |
### get all the possible FileTypes
|
| 156 |
filetypes=[]
|
| 157 |
for catalog in cataloglist[collid]:
|
| 158 |
if catalog.FileType not in filetypes :
|
| 159 |
filetypes.append(catalog.FileType)
|
| 160 |
### dictionary grouping catalogues by FileType
|
| 161 |
ftmap={}
|
| 162 |
for afiletype in filetypes:
|
| 163 |
#print ' filetype is '+afiletype+' for collid='+collid
|
| 164 |
sameFileType=[]
|
| 165 |
for catalog in cataloglist[collid]:
|
| 166 |
if catalog.FileType==afiletype :
|
| 167 |
sameFileType.append(catalog)
|
| 168 |
ftmap[afiletype]=sameFileType
|
| 169 |
### select only one catalogue among the catalouges with the same FileType
|
| 170 |
for ft in ftmap.keys():
|
| 171 |
#print ("key %s, val %s" %(ft,ftmap[ft]))
|
| 172 |
bestcatalog=self.SelectBestCatalog(ftmap[ft])
|
| 173 |
selectcatalogues.append(bestcatalog)
|
| 174 |
|
| 175 |
### return the minimal list of needed catalogues
|
| 176 |
return selectcatalogues
|
| 177 |
|
| 178 |
####################################################
|
| 179 |
def SelectBestCatalog(self,ftcat):
|
| 180 |
"""
|
| 181 |
From a set of catalogues with the same FileTypeand different access protocol, select the one according to access protocols
|
| 182 |
"""
|
| 183 |
sortedProtocols = self.protocolPrio_
|
| 184 |
|
| 185 |
### if just one catalog, just return it!
|
| 186 |
if (len(ftcat)==1):
|
| 187 |
#print '----- Just one catalogue, no selection based on protocol needed'
|
| 188 |
#ftcat[0].dump()
|
| 189 |
#print '---------------------'
|
| 190 |
return ftcat[0]
|
| 191 |
### oterwise select the best catalogue based on protocol
|
| 192 |
for prot in sortedProtocols:
|
| 193 |
for cat in ftcat:
|
| 194 |
if cat.ContactProtocol==prot:
|
| 195 |
#print '----- Catalogue selected based on protocol : '+prot
|
| 196 |
#cat.dump()
|
| 197 |
#print '---------------------'
|
| 198 |
return cat
|
| 199 |
|
