def test_adjust_value(self):
"""Test adjust_value"""
expect = 0
result = adjust_value("0")
self.assertEqual(expect, result)
expect = 1
result = adjust_value("1")
self.assertEqual(expect, result)
expect = 1.1
result = adjust_value("1.1")
self.assertEqual(expect, result)
expect = -1.1
result = adjust_value("-1.1")
self.assertEqual(expect, result)
expect = '2009.05.19 17:41:25'
result = adjust_value("2009.05.19 17:41:25")
self.assertEqual(expect, result)
expect = None
result = adjust_value("null")
self.assertEqual(expect, result)
expect = None
result = adjust_value("(null)")
self.assertEqual(expect, result)
def test_adjust_value(self):
"""Test adjust_value"""
expect = 0
result = adjust_value("0")
self.assertEqual(expect, result)
expect = 1
result = adjust_value("1")
self.assertEqual(expect, result)
expect = 1.1
result = adjust_value("1.1")
self.assertEqual(expect, result)
expect = -1.1
result = adjust_value("-1.1")
self.assertEqual(expect, result)
expect = '2009.05.19 17:41:25'
result = adjust_value("2009.05.19 17:41:25")
self.assertEqual(expect, result)
expect = None
result = adjust_value("null")
self.assertEqual(expect, result)
expect = None
result = adjust_value("(null)")
self.assertEqual(expect, result)
def parser(self, source, api):
"""
RunSummary data-service parser.
"""
row = {}
hold = None
context = ET.iterparse(source, events=("start", "end"))
root = None
for item in context:
event, elem = item
if event == "start" and root is None:
root = elem # the first element is root
if elem.tag == 'cmsdb':
continue
if event == 'start' and elem.tag == 'runInfo':
continue
if event == 'end' and elem.tag == 'runInfo':
yield dict(run=row)
row = {}
elem.clear()
if hold and event == 'end' and elem.tag == hold:
hold = None
continue
if event == 'start':
sub = {}
children = elem.getchildren()
# I don't apply notation conversion to all children
# since those are not likely to overlap
if children:
for child in children:
sub[child.tag] = adjust_value(child.text)
row[elem.tag] = sub
hold = elem.tag
else:
if not hold:
nkey = self.dasmapping.notation2das\
(self.name, elem.tag, api)
row[nkey] = adjust_value(elem.text)
if root:
root.clear()
source.close()
def parser(self, source, api):
"""
RunSummary data-service parser.
"""
row = {}
hold = None
context = ET.iterparse(source, events=("start", "end"))
root = None
for item in context:
event, elem = item
if event == "start" and root is None:
root = elem # the first element is root
if elem.tag == 'cmsdb':
continue
if event == 'start' and elem.tag == 'runInfo':
continue
if event == 'end' and elem.tag == 'runInfo':
yield dict(run=row)
row = {}
elem.clear()
if hold and event == 'end' and elem.tag == hold:
hold = None
continue
if event == 'start':
sub = {}
children = elem.getchildren()
# I don't apply notation conversion to all children
# since those are not likely to overlap
if children:
for child in children:
sub[child.tag] = adjust_value(child.text)
row[elem.tag] = sub
hold = elem.tag
else:
if not hold:
nkey = self.dasmapping.notation2das\
(self.name, elem.tag, api)
row[nkey] = adjust_value(elem.text)
if root:
root.clear()
source.close()
def worker_v2(url, query):
"""
Query RunRegistry service, see documentation at
https://twiki.cern.ch/twiki/bin/viewauth/CMS/DqmRrApi
url=http://pccmsdqm04.cern.ch/runregistry/xmlrpc
"""
server = xmlrpclib.ServerProxy(url)
namespace = 'GLOBAL'
if isinstance(query, str) or isinstance(query, unicode):
try:
data = server.RunLumiSectionRangeTable.exportJson(namespace, query)
except Exception as _err:
data = "{}" # empty response
for row in json.loads(data):
yield row
elif isinstance(query, dict):
iformat = 'tsv_runs' # other formats are xml_all, csv_runs
try:
data = server.RunDatasetTable.export(namespace, iformat, query)
except Exception as _err:
data = "" # empty response
titles = []
for line in data.split('\n'):
if not line:
continue
if not titles:
for title in line.split('\t')[:-1]:
title = title.lower()
if title != 'run_number':
title = title.replace('run_', '')
titles.append(title)
continue
val = line.split('\t')[:-1]
if len(val) != len(titles):
continue
record = {}
for idx in range(0, len(titles)):
key = titles[idx]
record[key] = adjust_value(val[idx])
yield dict(run=record)
def requestquery(self, query, add_to_analytics=True):
"""
Query analyzer which form request query to DAS from a free text-based form.
Return MongoDB request query.
"""
# strip operators while we will match words against them
operators = [o.strip() for o in self.operators]
# find out if input query contains filters/mapreduce functions
mapreduce = []
filters = []
aggregators = []
pat = re.compile(r"^([a-z_]+\.?)+$") # match key.attrib
if query and type(query) is types.StringType:
if query.find("|") != -1:
split_results = query.split("|")
query = split_results[0]
for item in split_results[1:]:
func = item.split("(")[0].strip()
for filter in self.filters:
if item.find(filter) == -1:
continue
for elem in item.replace(filter, '').split(','):
dasfilter = elem.strip()
if not dasfilter:
continue
if not pat.match(dasfilter):
msg = 'Incorrect filter: %s' % dasfilter
raise Exception(msg)
if dasfilter not in filters:
filters.append(dasfilter)
if func in self.aggregators:
aggregators = [agg for agg in get_aggregator(item)]
else:
mapreduce.append(item)
# mapreduce = [i.strip() for i in split_results[1:]]
query = query.strip()
if query[0] == "{" and query[-1] == "}":
mongo_query = json.loads(query)
if mongo_query.keys() != ['fields', 'spec']:
raise Exception("Invalid MongoDB query %s" % query)
if add_to_analytics:
self.analytics.add_query(query, mongo_query)
return mongo_query
# check input query and prepare it for processing
findbracketobj(query) # check brackets in a query
skeys = []
query = query.strip().replace(",", " ")
query = add_spaces(query, operators)
slist = query.split()
idx = 0
# main loop, step over words in query expression and
# findout selection keys and conditions
condlist = []
while True:
if idx >= len(slist):
break
word = slist[idx].strip()
if word in self.daskeys: # look-up for selection keys
try:
next_word = slist[idx+1]
if next_word not in operators and word not in skeys:
skeys.append(word)
except:
pass
if word == slist[-1] and word not in skeys: # last word
skeys.append(word)
elif word in operators: # look-up conditions
oper = word
prev_word = slist[idx-1]
next_word = slist[idx+1]
if word in ['in', 'nin']:
first = next_word
if first.find('[') == -1:
msg = 'No open bracket [ found in query expression'
raise Exception(msg)
arr = []
found_last = False
for item in slist[idx+1:]:
if item.find(']') != -1:
found_last = True
val = item.replace('[', '').replace(']', '')
if val:
arr.append(val)
if not found_last:
msg = 'No closed bracket ] found in query expression'
raise Exception(msg)
value = arr
elif word == 'last':
value = convert2date(next_word)
cdict = dict(key='date', op='in', value=value)
condlist.append(cdict)
value = None
else:
value = next_word
if prev_word == 'date':
if word != 'last': # we already converted date
if type(value) is types.StringType:
value = [das_dateformat(value), time.time()]
elif type(value) is types.ListType:
try:
value1 = das_dateformat(value[0])
value2 = das_dateformat(value[1])
value = [value1, value2]
except:
msg = "Unable to parse %s" % value
raise Exception(msg)
cdict = dict(key='date', op='in', value=value)
condlist.append(cdict)
value = None
idx += 1
if not value:
continue
key = prev_word
value = adjust_value(value)
if key == 'date':
cdict = dict(key=key, op=oper, value=value)
condlist.append(cdict)
continue
for system in self.map.list_systems():
mapkey = self.map.find_mapkey(system, key)
if mapkey:
cdict = dict(key=mapkey, op=oper, value=value)
if cdict not in condlist:
condlist.append(cdict)
else:
if word not in skeys and word in self.daskeys:
skeys.append(word)
idx += 1
if not condlist and skeys: # e.g. --query="dataset"
for key in skeys:
for system, daskeys in self.map.daskeys().items():
if key in daskeys:
mapkey = self.map.find_mapkey(system, key)
cdict = dict(key=mapkey, op="=", value="*")
condlist.append(cdict)
break
# print "\n### condlist", condlist
spec = mongo_exp(condlist)
# print "### spec", spec
if skeys:
fields = skeys
else:
fields = None
mongo_query = dict(fields=fields, spec=spec)
# add mapreduce if it exists
if mapreduce:
mongo_query['mapreduce'] = mapreduce
if filters:
mongo_query['filters'] = filters
if aggregators:
mongo_query['aggregators'] = aggregators
if add_to_analytics:
self.analytics.add_query(query, mongo_query)
return mongo_query
def requestquery(self, query, add_to_analytics=True):
"""
Query analyzer which form request query to DAS from a free text-based form.
Return MongoDB request query.
"""
# strip operators while we will match words against them
operators = [o.strip() for o in self.operators]
# find out if input query contains filters/mapreduce functions
mapreduce = []
filters = []
aggregators = []
pat = re.compile(r"^([a-z_]+\.?)+$") # match key.attrib
if query and type(query) is bytes:
if query.find("|") != -1:
split_results = query.split("|")
query = split_results[0]
for item in split_results[1:]:
func = item.split("(")[0].strip()
for filter in self.filters:
if item.find(filter) == -1:
continue
for elem in item.replace(filter, '').split(','):
dasfilter = elem.strip()
if not dasfilter:
continue
if not pat.match(dasfilter):
msg = 'Incorrect filter: %s' % dasfilter
raise Exception(msg)
if dasfilter not in filters:
filters.append(dasfilter)
if func in self.aggregators:
aggregators = [agg for agg in get_aggregator(item)]
else:
mapreduce.append(item)
# mapreduce = [i.strip() for i in split_results[1:]]
query = query.strip()
if query[0] == "{" and query[-1] == "}":
mongo_query = json.loads(query)
if mongo_query.keys() != ['fields', 'spec']:
raise Exception("Invalid MongoDB query %s" % query)
if add_to_analytics:
self.analytics.add_query(query, mongo_query)
return mongo_query
# check input query and prepare it for processing
findbracketobj(query) # check brackets in a query
skeys = []
query = query.strip().replace(",", " ")
query = add_spaces(query, operators)
slist = query.split()
idx = 0
# main loop, step over words in query expression and
# findout selection keys and conditions
condlist = []
while True:
if idx >= len(slist):
break
word = slist[idx].strip()
if word in self.daskeys: # look-up for selection keys
try:
next_word = slist[idx + 1]
if next_word not in operators and word not in skeys:
skeys.append(word)
except:
pass
if word == slist[-1] and word not in skeys: # last word
skeys.append(word)
elif word in operators: # look-up conditions
oper = word
prev_word = slist[idx - 1]
next_word = slist[idx + 1]
if word in ['in', 'nin']:
first = next_word
if first.find('[') == -1:
msg = 'No open bracket [ found in query expression'
raise Exception(msg)
arr = []
found_last = False
for item in slist[idx + 1:]:
if item.find(']') != -1:
found_last = True
val = item.replace('[', '').replace(']', '')
if val:
arr.append(val)
if not found_last:
msg = 'No closed bracket ] found in query expression'
raise Exception(msg)
value = arr
elif word == 'last':
value = convert2date(next_word)
cdict = dict(key='date', op='in', value=value)
condlist.append(cdict)
value = None
else:
value = next_word
if prev_word == 'date':
if word != 'last': # we already converted date
if type(value) is bytes:
value = [das_dateformat(value), time.time()]
elif type(value) is list:
try:
value1 = das_dateformat(value[0])
value2 = das_dateformat(value[1])
value = [value1, value2]
except:
msg = "Unable to parse %s" % value
raise Exception(msg)
cdict = dict(key='date', op='in', value=value)
condlist.append(cdict)
value = None
idx += 1
if not value:
continue
key = prev_word
value = adjust_value(value)
if key == 'date':
cdict = dict(key=key, op=oper, value=value)
condlist.append(cdict)
continue
for system in self.map.list_systems():
mapkey = self.map.find_mapkey(system, key)
if mapkey:
cdict = dict(key=mapkey, op=oper, value=value)
if cdict not in condlist:
condlist.append(cdict)
else:
if word not in skeys and word in self.daskeys:
skeys.append(word)
idx += 1
if not condlist and skeys: # e.g. --query="dataset"
for key in skeys:
for system, daskeys in self.map.daskeys().items():
if key in daskeys:
mapkey = self.map.find_mapkey(system, key)
cdict = dict(key=mapkey, op="=", value="*")
condlist.append(cdict)
break
# print "\n### condlist", condlist
spec = mongo_exp(condlist)
# print "### spec", spec
if skeys:
fields = skeys
else:
fields = None
mongo_query = dict(fields=fields, spec=spec)
# add mapreduce if it exists
if mapreduce:
mongo_query['mapreduce'] = mapreduce
if filters:
mongo_query['filters'] = filters
if aggregators:
mongo_query['aggregators'] = aggregators
if add_to_analytics:
self.analytics.add_query(query, mongo_query)
return mongo_query
