def free_text_parser(sentence, daskeys, default_key="dataset"):
"""Parse sentence and construct DAS QL expresion"""
found = 0
for word in sentence.split():
if word.find('=') != -1 or word in daskeys:
found = 1
dasquery = ''
if not found:
for word in sentence.split():
if PAT_RUN.match(word):
dasquery += 'run=%s ' % word
elif PAT_BLOCK.match(word):
dasquery += 'block=%s ' % word
elif PAT_FILE.match(word):
dasquery += 'file=%s ' % word
elif PAT_RELEASE.match(word):
if word.find('CMSSW_') == -1:
word = 'CMSSW_' + word
if word.find('*') != -1:
dasquery += 'release=%s ' % word
else:
dasquery += 'release=%s* ' % word
elif PAT_SITE.match(word):
if word.find('*') != -1:
dasquery += 'site=%s ' % word
else:
dasquery += 'site=%s* ' % word
elif PAT_SE.match(word):
dasquery += 'site=%s ' % word
elif PAT_TIERS.match(word):
if word.find('*') != -1:
dasquery += 'tier=%s ' % word
else:
dasquery += 'tier=*%s* ' % word
elif PAT_DATATYPE.match(word):
dasquery += 'datatype=%s ' % word
else:
if word.find('*') != -1:
dasquery += 'dataset=%s ' % word
elif len(word) > 0 and word[0] == '/':
dasquery += 'dataset=%s* ' % word
else:
dasquery += 'dataset=*%s* ' % word
if dasquery:
if dasquery.find(default_key) != -1:
select_key = default_key
else:
select_key = dasquery.split('=')[0]
dasquery = select_key + ' ' + dasquery
return dasquery.strip()
return sentence
def free_text_parser(sentence, daskeys, default_key="dataset"):
"""Parse sentence and construct DAS QL expresion"""
found = 0
for word in sentence.split():
if word.find('=') != -1 or word in daskeys:
found = 1
dasquery = ''
if not found:
for word in sentence.split():
if PAT_RUN.match(word):
dasquery += 'run=%s ' % word
elif PAT_BLOCK.match(word):
dasquery += 'block=%s ' % word
elif PAT_FILE.match(word):
dasquery += 'file=%s ' % word
elif PAT_RELEASE.match(word):
if word.find('CMSSW_') == -1:
word = 'CMSSW_' + word
if word.find('*') != -1:
dasquery += 'release=%s ' % word
else:
dasquery += 'release=%s* ' % word
elif PAT_SITE.match(word):
if word.find('*') != -1:
dasquery += 'site=%s ' % word
else:
dasquery += 'site=%s* ' % word
elif PAT_SE.match(word):
dasquery += 'site=%s ' % word
elif PAT_TIERS.match(word):
if word.find('*') != -1:
dasquery += 'tier=%s ' % word
else:
dasquery += 'tier=*%s* ' % word
elif PAT_DATATYPE.match(word):
dasquery += 'datatype=%s ' % word
else:
if word.find('*') != -1:
dasquery += 'dataset=%s ' % word
elif len(word) > 0 and word[0] == '/':
dasquery += 'dataset=%s* ' % word
else:
dasquery += 'dataset=*%s* ' % word
if dasquery:
if dasquery.find(default_key) != -1:
select_key = default_key
else:
select_key = dasquery.split('=')[0]
dasquery = select_key + ' ' + dasquery
return dasquery.strip()
return sentence
def autocomplete_helper(query, dasmgr, daskeys):
"""
Interface to the DAS keylearning system, for a
as-you-type suggestion system. This is a call for AJAX
in the page rather than a user-visible one.
This returns a list of JS objects, formatted like::
{'css': '<ul> css class', 'value': 'autocompleted text', 'info': '<html> text'}
Some of the work done here could be moved client side, and
only calls that actually require keylearning look-ups
forwarded. Given the number of REs used, this may be necessary
if load increases.
"""
uinput = str(query)
qsplit = uinput.split()
last_word = qsplit[-1] # operate on last word in a query
if last_word.find(',') != -1:
last_word = last_word.strip().replace(',', '').split()[-1]
result = []
prev = ""
if len(qsplit) != 1:
prev = ' '.join(qsplit[:-1])
query = last_word
if RE_DBSQL_0.match(query):
#find...
match1 = RE_DBSQL_1.match(query)
match2 = RE_DBSQL_2.match(query)
if match1:
daskey = match1.group(1)
if daskey in daskeys:
if match2:
operator = match2.group(3)
value = match2.group(4)
if operator == '=' or operator == 'like':
result.append({'css': 'ac-warning sign', 'value':'%s=%s' % (daskey, value),
'info': "This appears to be a DBS-QL query, but the key (<b>%s</b>) is a valid DAS key, and the condition should <b>probably</b> be expressed like this." % (daskey)})
else:
result.append({'css': 'ac-warning sign', 'value':daskey,
'info': "This appears to be a DBS-QL query, but the key (<b>%s</b>) is a valid DAS key. However, I'm not sure how to interpret the condition (<b>%s %s<b>)." % (daskey, operator, value)})
else:
result.append({'css': 'ac-warning sign', 'value': daskey,
'info': 'This appears to be a DBS-QL query, but the key (<b>%s</b>) is a valid DAS key.' % daskey})
else:
result.append({'css': 'ac-error sign', 'value': '',
'info': "This appears to be a DBS-QL query, and the key (<b>%s</b>) isn't known to DAS." % daskey})
key_search = dasmgr.keylearning.key_search(daskey)
#do a key search, and add info elements for them here
for keys, members in key_search.items():
result.append({'css': 'ac-info', 'value': ' '.join(keys),
'info': 'Possible keys <b>%s</b> (matching %s).' % (', '.join(keys), ', '.join(members))})
if not key_search:
result.append({'css': 'ac-error sign', 'value': '',
'info': 'No matches found for <b>%s</b>.' % daskey})
else:
result.append({'css': 'ac-error sign', 'value': '',
'info': 'This appears to be a DBS-QL query. DAS queries are of the form <b>key</b><span class="faint">[ operator value]</span>'})
elif RE_HASPIPE.match(uinput) and RE_SUBKEY.match(query):
subkey = RE_SUBKEY.match(query).group(1)
daskey = subkey.split('.')[0]
if daskey in daskeys and dasmgr.keylearning.col and\
dasmgr.keylearning.col.count():
if dasmgr.keylearning.has_member(subkey):
result.append({'css': 'ac-info', 'value': subkey,
'info': 'Correct DAS query'})
else:
result.append({'css': 'ac-warning sign', 'value': subkey,
'info': "Correct DAS query, but <b>%s</b> is not known in DAS keylearning system" % subkey})
key_search = dasmgr.keylearning.key_search(subkey, daskey)
for keys, members in key_search.items():
for member in members:
result.append({'css': 'ac-info', 'value':'%s' % member,
'info': 'Possible member match <b>%s</b> (for daskey <b>%s</b>)' % (member, daskey)})
elif RE_HASPIPE.match(uinput):
keystr = uinput.split('|')[0]
keys = set()
for keymatch in RE_KEYS.findall(keystr):
if keymatch[0]:
keys.add(keymatch[0])
else:
keys.add(keymatch[2])
keys = list(keys)
if not keys:
result.append({'css':'ac-error sign', 'value': '',
'info': "You seem to be trying to write a pipe command without any keys."})
agg_pat = AGG_PAT.match(query)
flt_pat = FLT_PAT.match(query)
daskey = query.split('.')[0]
if agg_pat:
matches = filter(lambda x: x.startswith(query), das_aggregators())
if matches:
for match in matches:
result.append({'css': 'ac-info', 'value': '%s' % match,
'info': 'Aggregated function <b>%s</b>' % (match)})
elif flt_pat:
matches = filter(lambda x: x.startswith(query), das_filters())
if matches:
for match in matches:
result.append({'css': 'ac-info', 'value': '%s' % match,
'info': 'Filter function <b>%s</b>' % (match)})
elif daskey.strip() == '|':
result.append({'css': 'ac-warning sign', 'value': query,
'info': 'DAS pipe must follow either by filter or aggregator function'})
elif daskey not in daskeys and daskey.find('(') == -1:
result.append({'css': 'ac-warning sign', 'value': query,
'info': '<b>%s</b> is neither aggregator, filter or DAS key' % query})
elif PAT_RELEASE.match(query):
if query[0] == 'C': # CMS releases all starts with CMSSW
release = '%s*' % query
else:
release = 'CMSSW_%s*' % query
result.append({'css': 'ac-info', 'value': 'release=%s' % release, 'info': 'Seems like CMSSW release'})
result.append({'css': 'ac-info', 'value': 'dataset=*%s*' % query, 'info': 'Seems like dataset pattern'})
elif PAT_TIERS.match(query):
result.append({'css': 'ac-info', 'value': 'tier=*%s*' % query, 'info': 'Seems like data tier'})
result.append({'css': 'ac-info', 'value': 'dataset=*%s*' % query, 'info': 'Seems like dataset pattern'})
elif PAT_SLASH.match(query):
if PAT_FILE.match(query):
result.append({'css': 'ac-info', 'value': 'file=%s' % query, 'info': 'Seems like file pattern'})
elif PAT_BLOCK.match(query):
result.append({'css': 'ac-info', 'value': 'block=%s' % query, 'info': 'Seems like block name'})
else:
result.append({'css': 'ac-info', 'value': 'block=%s*' % query, 'info': 'Seems like block name'})
result.append({'css': 'ac-info', 'value': 'file=%s*' % query, 'info': 'Seems like file pattern'})
result.append({'css': 'ac-info', 'value': 'dataset=%s*' % query, 'info': 'Seems like dataset pattern'})
elif PAT_RUN.match(query):
result.append({'css': 'ac-info', 'value': 'run=%s' % query, 'info': 'Seems like run number'})
elif PAT_DATATYPE.match(query):
result.append({'css': 'ac-info', 'value': 'datatype=%s*' % query, 'info': 'Seems like data type'})
result.append({'css': 'ac-info', 'value': 'dataset=%s*' % query, 'info': 'Seems like dataset pattern'})
elif PAT_SITE.match(query):
result.append({'css': 'ac-info', 'value': 'site=%s*' % query, 'info': 'Seems like site name'})
result.append({'css': 'ac-info', 'value': 'dataset=%s*' % query, 'info': 'Seems like dataset pattern'})
elif PAT_SE.match(query):
result.append({'css': 'ac-info', 'value': 'site=%s' % query, 'info': 'Seems like SE'})
elif RE_K_SITE.match(query):
result.append({'css': 'ac-info', 'value': query, 'info': 'Valid DAS key: site'})
if query.find('=') == -1:
result.append({'css': 'ac-info', 'value': 'dataset=*%s*' % query, 'info': 'Seems like dataset pattern'})
elif RE_K_FILE.match(query):
result.append({'css': 'ac-info', 'value': query, 'info': 'Valid DAS key: file'})
if query.find('=') == -1:
result.append({'css': 'ac-info', 'value': 'dataset=*%s*' % query, 'info': 'Seems like dataset pattern'})
elif RE_K_PR_DATASET.match(query):
result.append({'css': 'ac-info', 'value': query, 'info': 'Valid DAS key: primary_dataset'})
if query.find('=') == -1:
result.append({'css': 'ac-info', 'value': 'dataset=*%s*' % query, 'info': 'Seems like dataset pattern'})
elif RE_K_JOBSUMMARY.match(query):
result.append({'css': 'ac-info', 'value': query, 'info': 'Valid DAS key: jobsummary'})
if query.find('=') == -1:
result.append({'css': 'ac-info', 'value': 'dataset=*%s*' % query, 'info': 'Seems like dataset pattern'})
elif RE_K_MONITOR.match(query):
result.append({'css': 'ac-info', 'value': query, 'info': 'Valid DAS key: monitor'})
if query.find('=') == -1:
result.append({'css': 'ac-info', 'value': 'dataset=*%s*' % query, 'info': 'Seems like dataset pattern'})
elif RE_K_TIER.match(query):
result.append({'css': 'ac-info', 'value': query, 'info': 'Valid DAS key: tier'})
if query.find('=') == -1:
result.append({'css': 'ac-info', 'value': 'dataset=*%s*' % query, 'info': 'Seems like dataset pattern'})
elif RE_K_RELEASE.match(query):
result.append({'css': 'ac-info', 'value': query, 'info': 'Valid DAS key: release'})
if query.find('=') == -1:
result.append({'css': 'ac-info', 'value': 'dataset=*%s*' % query, 'info': 'Seems like dataset pattern'})
elif RE_K_CONFIG.match(query):
result.append({'css': 'ac-info', 'value': query, 'info': 'Valid DAS key: config'})
if query.find('=') == -1:
result.append({'css': 'ac-info', 'value': 'dataset=*%s*' % query, 'info': 'Seems like dataset pattern'})
elif RE_K_GROUP.match(query):
result.append({'css': 'ac-info', 'value': query, 'info': 'Valid DAS key: group'})
if query.find('=') == -1:
result.append({'css': 'ac-info', 'value': 'dataset=*%s*' % query, 'info': 'Seems like dataset pattern'})
elif RE_K_CHILD.match(query):
result.append({'css': 'ac-info', 'value': query, 'info': 'Valid DAS key: child'})
if query.find('=') == -1:
result.append({'css': 'ac-info', 'value': 'dataset=*%s*' % query, 'info': 'Seems like dataset pattern'})
elif RE_K_PARENT.match(query):
result.append({'css': 'ac-info', 'value': query, 'info': 'Valid DAS key: parent'})
if query.find('=') == -1:
result.append({'css': 'ac-info', 'value': 'dataset=*%s*' % query, 'info': 'Seems like dataset pattern'})
elif RE_K_DATASET.match(query):
result.append({'css': 'ac-info', 'value': query, 'info': 'Valid DAS key: dataset'})
if query.find('=') == -1:
result.append({'css': 'ac-info', 'value': 'dataset=*%s*' % query, 'info': 'Seems like dataset pattern'})
elif RE_K_RUN.match(query):
result.append({'css': 'ac-info', 'value': query, 'info': 'Valid DAS key: run'})
if query.find('=') == -1:
result.append({'css': 'ac-info', 'value': 'dataset=*%s*' % query, 'info': 'Seems like dataset pattern'})
elif RE_K_BLOCK.match(query):
result.append({'css': 'ac-info', 'value': query, 'info': 'Valid DAS key: block'})
if query.find('=') == -1:
result.append({'css': 'ac-info', 'value': 'dataset=*%s*' % query, 'info': 'Seems like dataset pattern'})
elif RE_K_STATUS.match(query):
result.append({'css': 'ac-info', 'value': query, 'info': 'Valid DAS key: status'})
if query.find('=') == -1:
result.append({'css': 'ac-info', 'value': 'dataset=*%s*' % query, 'info': 'Seems like dataset pattern'})
elif RE_K_DATASET.match(query):
#/something...
result.append({'css': 'ac-warning sign', 'value': query,
'info':'''Seems like dataset query'''})
elif RE_SITE.match(query):
#T{0123}_...
result.append({'css': 'ac-warning sign', 'value':'site=%s' % query,
'info':'''Seems like site query. The correct syntax is <b>site=TX_YY_ZZZ</b>'''})
result.append({'css': 'ac-info', 'value': 'dataset=*%s*' % query, 'info': 'Seems like dataset pattern'})
else:
#we've no idea what you're trying to accomplish, do a search
result.append({'css': 'ac-info', 'value': 'dataset=*%s*' % query,
'info': 'Seems like dataset pattern'})
if prev:
new_result = []
for idict in result:
newval = prev + ' ' + idict['value']
new_result.append({'css':idict['css'], 'value':newval, 'info':idict['info']})
return new_result
return result
