doInitialWarming = int( INCREMENTAL_INDEXER_INITIAL_WARMING)
deleteAfterIndexing = int( DELETE_MARC_RECORDS_FROM_INDEX_QUEUE )
# NOTE: we *cannot* use a job scheduler like Quartz here because we don't want overlapping jobs or jobs building up.
lastRunTime = lastOptimizeTime = lastFacetWarmTime = time.time()
firstTime = 1
keepgoing = 1
count = 0
startDate = int( time.strftime("%d") )
while keepgoing:
now = time.time()
## 1st do incremental indexing if nec.
if firstTime or ( (now - lastRunTime) >= checkInterval ):
if firstTime and doInitialWarming:
#a. warm facets if the very first time it has run
facetWarmer.warmFacets()
lastFacetWarmTime = time.time()
# b. check for PMS/MARC files in indexer-queue directory and handle if necessary.
# NOTE: it only does this check the first time through, so setting deleteAfterIndexing to false
# will not cause the same files to keep getting reindexed.
count += horizonIncrementalIndexer.processFilesInDirectory("./indexer-queue", anselUnicodeConverter = converter, numThreads=numThreads, deleteAfterIndexing=deleteAfterIndexing)
print "[%s] done processing indexer-queue contents" % time.ctime()
# c. do status changes.
if doItemStatusIndexing:
count += horizonIncrementalIndexer.incrementalIndexingJob(commitNonblocking=1)
lastRunTime = time.time()
firstTime = 0
else:
print "not time to run yet, last ran %.4f ago" % (now-lastRunTime)
## 2nd do optimize if nec.
resp = solrConnection.postURL(SOLR_UPDATE_URL, "<add>%s</add>" % data)
if resp.find('<result status="1"') > -1:
print "\nError POSTing documents! Response from Solr was\n\n%s\n\n" % resp
print "committing..."
if nonblocking:
solrConnection.commitNonblocking()
else:
solrConnection.commit()
inStream.close()
return count
if __name__ == "__main__":
anselUnicodeConverter = AnselToUnicode()
# 2 arguments to command line are used to do a ran
if len(sys.argv) == 2:
processFile(sys.argv[1], anselUnicodeConverter)
else:
print "incorrect usage -- specify file to be processed."
sys.exit(1)
print "done indexing, now optimizing"
optimize()
print "done optimizing, now warming facets"
from facetWarmer import *
for i in range(3):
# running warmFacets more than once appears to improve performance.
facetWarmer.warmFacets(server=SOLR_QUERY_URL)
print "all done!"
recordBatch = []
print("+%s+" % bibOn),
# now do last batch
if len(recordBatch) > 0:
data = u''.join(recordBatch)
resp = solrConnection.postURL(SOLR_UPDATE_URL, "<add>%s</add>" % data)
if bibCount > 0:
print "\n[%s] done updating bibs, now committing" % time.ctime()
try:
if commitNonblocking:
solrConnection.commitNonblocking()
else:
solrConnection.commit()
except IOError:
print "Connection reset when talking to Solr, skipping this commit and sleeping 10 sec."
time.sleep(10)
print "[%s] done committing" % time.ctime()
else:
print "[%s] no bibs updated, exiting" % time.ctime()
return bibCount
if __name__ == '__main__':
processFilesInDirectory(HORIZON_BASE_DIR)
# finally, do an optimize here
if DO_OPTIMIZE:
print "starting final optimize"
solrConnection.optimize() # csdebug
facetWarmer.warmFacets()
resp = solrConnection.postURL(SOLR_UPDATE_URL, "<add>%s</add>" % data)
if resp.find('<result status="1"') > -1:
print "\nError POSTing documents! Response from Solr was\n\n%s\n\n" % resp
print "committing..."
if nonblocking:
solrConnection.commitNonblocking()
else:
solrConnection.commit()
inStream.close()
return count
if __name__ == '__main__':
anselUnicodeConverter = AnselToUnicode()
# 2 arguments to command line are used to do a ran
if len(sys.argv) == 2:
processFile(sys.argv[1], anselUnicodeConverter)
else:
print "incorrect usage -- specify file to be processed."
sys.exit(1)
print "done indexing, now optimizing"
optimize()
print "done optimizing, now warming facets"
from facetWarmer import *
for i in range(3):
# running warmFacets more than once appears to improve performance.
facetWarmer.warmFacets(server=SOLR_QUERY_URL)
print "all done!"
