def _makeRegex(pattern, startp):
pattern = learning.safeRegexLiteral(pattern)
if startp:
if pattern.startswith('\n'):
pattern = pattern[1:]
return '([\\r\\n]+)%s # BREAK BEFORE SEEING "%s" AT THE START OF A LINE' % (pattern, pattern)
else:
return '([\\r\\n]+).{0,%s}%s # BREAK BEFORE SEEING "%s" NEAR THE BEGINNING OF A LINE' % (MAX_FORWARD_CHARS, pattern, pattern)
else:
if pattern.endswith('\n'):
pattern = pattern[:-1]
return '%s([\\r\\n]+) # BREAK AFTER SEEING "%s" AT THE END OF A LINE' % (pattern, pattern)
else:
return '%s.{0,%s}([\\r\\n]+) # BREAK AFTER SEEING "%s" NEAR THE END OF A LINE' % (pattern, MAX_BACKWARDS_CHARS, pattern)
def learnSTRPTime(text, timevalues):
timevalues = timevalues.split(',')
positions = getPositions(text, timevalues)
if positions == None:
return None, None
if positions[0] == -1 or positions[1] == -1 or positions[
3] == -1 or positions[4] == -1:
print "Warning: month, day, hour, and minute are required."
return None, None
posDict = {}
valueDict = {}
count = 0
first = -1
last = -1
for position in positions:
fieldname = FIELDNAMES[count]
posDict[fieldname] = position
value = timevalues[count]
valueDict[fieldname] = value
count += 1
if first < 0 or -1 < position < first:
first = position
if position > last:
last = position + len(value)
# sort by position
fieldAndPos = posDict.items()
fieldAndPos.sort(lambda x, y: x[1] - y[1])
#print fieldAndPos
prefix = text[:first]
suffix = text[last:]
#print "TEXT:", text
#print "PREFIX:", prefix
#print "SUFFIX:", suffix
#print "VALUES:", valueDict
prefixRegex = learning.generateSearchRegex(prefix)
strpformat = buildSTRPTime(text, fieldAndPos, valueDict)
print "TIME_PREFIX: '%s'" % prefixRegex
print "TIME_FORMAT:", strpformat
try:
verify(text, prefixRegex, strpformat, suffix)
return (prefixRegex, strpformat, suffix)
except:
print "Error determining timeformat"
return None
def learnSTRPTime(text, timevalues):
timevalues = timevalues.split(',')
positions = getPositions(text, timevalues)
if positions == None:
return None, None
if positions[0] == -1 or positions[1] == -1 or positions[3] == -1 or positions[4] == -1:
print "Warning: month, day, hour, and minute are required."
return None, None
posDict = {}
valueDict = {}
count = 0
first = -1
last = -1
for position in positions:
fieldname = FIELDNAMES[count]
posDict[fieldname] = position
value = timevalues[count]
valueDict[fieldname] = value
count += 1
if first < 0 or -1 < position < first:
first = position
if position > last:
last = position + len(value)
# sort by position
fieldAndPos = posDict.items()
fieldAndPos.sort( lambda x, y: x[1] - y[1] )
#print fieldAndPos
prefix = text[:first]
suffix = text[last:]
#print "TEXT:", text
#print "PREFIX:", prefix
#print "SUFFIX:", suffix
#print "VALUES:", valueDict
prefixRegex = learning.generateSearchRegex(prefix)
strpformat = buildSTRPTime(text, fieldAndPos, valueDict)
print "TIME_PREFIX: '%s'" % prefixRegex
print "TIME_FORMAT:", strpformat
try:
verify(text, prefixRegex, strpformat, suffix)
return (prefixRegex, strpformat, suffix)
except:
print "Error determining timeformat"
return None
