def dfs(self, node, esp=0, maxDist=0.3):
"""
Navega em profundidade na árvore buscando nós adjacentes com distancia
de edição menor que maxDist e os agrupa na mesma componente.
@param Node node: No atual da dfs
@param int esp: Nível da árvore
@param float maxDist: Proporção máxima de diferença para ser agrupado
na mesma componente
"""
global s1
global s2
s1 = ""
s2 = ""
result = [False] * len(node.childNodes)
match = False
for x in xrange(0,len(node.childNodes)):
c = node.childNodes[x]
if x == 0 or c.height < 3:
s1 = c.str
match = False
self._comp += 1
continue
s2 = c.str
# print 'str:', s1, s2
if len(s1) <= (len(s2) * (1.0+maxDist)) and\
len(s2) <= (len(s1) * (1.0 +maxDist)):
if float(stringDistance(s1,s2))/max(len(s1), len(s2)) > maxDist :
s1 = c.str
match = False
self._comp += 1
continue
else:
s1 = c.str
result[x] = self._comp
if not match:
result[x-1] = self._comp
match = True
else:
s1 = c.str
match = False
self._comp += 1
#for
node.result = result
for x in xrange(0,len(node.childNodes)):
if not result[x]:
self.dfs(node.childNodes[x], esp+1)
def test_stringDistance(self):
"""
Test if costs and distance value are correct
"""
print '\n - Test: utils.distances.stringDistance()'
add = 1
change = 2
delete = 4
cost = { 'a': add, 'd': delete, 'c': change }
dist = distances.stringDistance('aa','aba', cost)
self.assertEqual(dist, add, '1: %d!=%d' % (dist, change))
dist = distances.stringDistance('a', 'ab', cost)
self.assertEqual(dist, add, '2: %d!=%d' % (dist, change))
dist = distances.stringDistance('a', 'ba', cost)
self.assertEqual(dist, add, '3: %d!=%d' % (dist, change))
dist = distances.stringDistance('a', 'b', cost)
self.assertEqual(dist, change, '4: %d!=%d' % (dist, change))
dist = distances.stringDistance('ba', 'a', cost)
self.assertEqual(dist, delete, '5: %d!=%d' % (dist, change))
dist = distances.stringDistance('aba', 'aa', cost)
self.assertEqual(dist, delete, '6: %d!=%d' % (dist, change))
def match(node, maxDist=0, height=3, tags=False, printtag=False):
global __debug
if __debug:
print "Debug mode: eri.utils.match.match()"
s1 = ""
s2 = ""
result = [False] * len(node.childNodes)
match = False
_comp = 0
for x in xrange(0,len(node.childNodes)):
c = node.childNodes[x]
#primary test to not match low height
if x == 0 or c.height < height:
s1 = c.tags if tags else c.str
match = False
_comp += 1
continue
s2 = c.tags if tags else c.str
if printtag or __debug:
print 'str:', s1, s2
# match test
if len(s1) <= (len(s2) * (1.0+maxDist)) or\
len(s2) <= (len(s1) * (1.0+maxDist)):
d = float(stringDistance(s1,s2))/max(len(s1), len(s2))
if __debug:
print 'distance:', d
#save match component
if d <= maxDist :
s1 = c.tags if tags else c.str
result[x] = _comp
result[x-1] = _comp if not match else result[x-1]
match = True
continue
# not match
s1 = c.tags if tags else c.str
match = False
_comp += 1
#for
if __debug:
print "return: eri.utils.match.match()"
return result
from eri.utils.distances import stringDiff, stringDistance
a = "eturn one that starts earliest in"
b = "starts earliest in a, and of all those maximal matching"
for i in xrange(1, 10000):
stringDiff(a,b)
stringDistance(a,b)
