def __init__(self):
PrintAbsorptionCountsByBranching2.__bases__[0].__init__(self)
self.e0 = CountDict()
self.e1 = CountDict()
self.e2 = CountDict()
self.e3 = CountDict()
self.e0_examples = defaultdict(FixedSizeList)
self.e1_examples = defaultdict(FixedSizeList)
self.e2_examples = defaultdict(FixedSizeList)
self.e3_examples = defaultdict(FixedSizeList)
class PrintCommaCountsByBranching(FixedTgrep(r'{, $ /.+/} > /.+/')):
'''Prints comma occurrence counts by descending frequency together with the rule used.'''
def __init__(self):
PrintCommaCountsByBranching.__bases__[0].__init__(self)
self.envs = CountDict()
def output(self):
as_left = {}
as_right = {}
for (l, r, p), f in self.envs.iteritems():
if l == ',':
as_left[(l, r, p)] = f
else:
as_right[(l, r, p)] = f
print ", _ -> _"
print "--------"
for (l, r, p), f in sorted_by_value_desc(as_left):
print "% 10d [%28s] %s %20s -> %s" % (f, analyse(C(l), C(r), C(p)), l, r, p)
print "_ , -> _"
print "--------"
for (l, r, p), f in sorted_by_value_desc(as_right):
print "% 10d [%28s] %20s %s -> %s" % (f, analyse(C(l), C(r), C(p)), l, r, p)
def match_generator(self, deriv, expr):
return find_all(deriv, expr)
def match_callback(self, match_node, bundle):
is_left_child = match_node.parent.lch is match_node
if is_left_child:
self.envs[ (str(match_node.cat), str(match_node.parent.rch.cat), str(match_node.parent.cat)) ] += 1
else:
self.envs[ (str(match_node.parent.lch.cat), str(match_node.cat), str(match_node.parent.cat)) ] += 1
class PrintCommaCountsByBranching(FixedTgrep(r"{, $ /.+/} > /.+/")):
"""Prints comma occurrence counts by descending frequency together with the rule used."""
def __init__(self):
PrintCommaCountsByBranching.__bases__[0].__init__(self)
self.envs = CountDict()
def output(self):
as_left = {}
as_right = {}
for (l, r, p), f in self.envs.iteritems():
if l == ",":
as_left[(l, r, p)] = f
else:
as_right[(l, r, p)] = f
print ", _ -> _"
print "--------"
for (l, r, p), f in sorted_by_value_desc(as_left):
print "% 10d [%28s] %s %20s -> %s" % (f, analyse(C(l), C(r), C(p)), l, r, p)
print "_ , -> _"
print "--------"
for (l, r, p), f in sorted_by_value_desc(as_right):
print "% 10d [%28s] %20s %s -> %s" % (f, analyse(C(l), C(r), C(p)), l, r, p)
def match_generator(self, deriv, expr):
return find_all(deriv, expr)
def match_callback(self, match_node, bundle):
is_left_child = match_node.parent.lch is match_node
if is_left_child:
self.envs[(str(match_node.cat), str(match_node.parent.rch.cat), str(match_node.parent.cat))] += 1
else:
self.envs[(str(match_node.parent.lch.cat), str(match_node.cat), str(match_node.parent.cat))] += 1
def __init__(self):
Filter.__init__(self)
self.freqs = CountDict()
def __init__(self):
Filter.__init__(self)
self.counter = CountDict()
def __init__(self):
DumpRules.__bases__[0].__init__(self)
self.rules = CountDict()
def __init__(self):
self.counts = CountDict()
self.total = 0
def __init__(self):
CountRuleOccurrencesByType.__bases__[0].__init__(self)
self.counts = CountDict()
def __init__(self):
PrintCommaCountsByBranching.__bases__[0].__init__(self)
self.envs = CountDict()
def __init__(self):
Filter.__init__(self)
self.envs = CountDict()
def __init__(self):
PrintCommaCountsByBranching.__bases__[0].__init__(self)
self.envs = CountDict()