def flatten_strings(self):
return (utf8decode(k)[0] for k in self.flatten())
def prefix_string(self):
"""Returns the labels of the path from the root to the current arc as
a decoded unicode string.
"""
return utf8decode(self.prefix_bytes())[0]
def peek_key_string(self):
"""Returns the next closest key in the graph as a decoded unicode
string.
"""
return utf8decode(self.peek_key_bytes())[0]
def flatten_strings(self):
return (utf8decode(k)[0] for k in self.flatten())
def within(graph, text, k=1, prefix=0, address=None):
"""Yields a series of keys in the given graph within ``k`` edit distance of
``text``. If ``prefix`` is greater than 0, all keys must match the first
``prefix`` characters of ``text``.
"""
text = to_labels(text)
if address is None:
address = graph._root
sofar = emptybytes
accept = False
if prefix:
prefixchars = text[:prefix]
arc = graph.find_path(prefixchars, address=address)
if arc is None:
return
sofar = emptybytes.join(prefixchars)
address = arc.target
accept = arc.accept
stack = [(address, k, prefix, sofar, accept)]
seen = set()
while stack:
state = stack.pop()
# Have we already tried this state?
if state in seen:
continue
seen.add(state)
address, k, i, sofar, accept = state
# If we're at the end of the text (or deleting enough chars would get
# us to the end and still within K), and we're in the accept state,
# yield the current result
if (len(text) - i <= k) and accept:
yield utf8decode(sofar)[0]
# If we're in the stop state, give up
if address is None:
continue
# Exact match
if i < len(text):
arc = graph.find_arc(address, text[i])
if arc:
stack.append((arc.target, k, i + 1, sofar + text[i],
arc.accept))
# If K is already 0, can't do any more edits
if k < 1:
continue
k -= 1
arcs = graph.arc_dict(address)
# Insertions
stack.extend((arc.target, k, i, sofar + char, arc.accept)
for char, arc in iteritems(arcs))
# Deletion, replacement, and transpo only work before the end
if i >= len(text):
continue
char = text[i]
# Deletion
stack.append((address, k, i + 1, sofar, False))
# Replacement
for char2, arc in iteritems(arcs):
if char2 != char:
stack.append((arc.target, k, i + 1, sofar + char2, arc.accept))
# Transposition
if i < len(text) - 1:
char2 = text[i + 1]
if char != char2 and char2 in arcs:
# Find arc from next char to this char
target = arcs[char2].target
if target:
arc = graph.find_arc(target, char)
if arc:
stack.append((arc.target, k, i + 2,
sofar + char2 + char, arc.accept))
def peek_key_string(self):
"""Returns the next closest key in the graph as a decoded unicode
string.
"""
return utf8decode(self.peek_key_bytes())[0]
def prefix_string(self):
"""Returns the labels of the path from the root to the current arc as
a decoded unicode string.
"""
return utf8decode(self.prefix_bytes())[0]
def within(graph, text, k=1, prefix=0, address=None):
"""Yields a series of keys in the given graph within ``k`` edit distance of
``text``. If ``prefix`` is greater than 0, all keys must match the first
``prefix`` characters of ``text``.
"""
text = to_labels(text)
if address is None:
address = graph._root
sofar = emptybytes
accept = False
if prefix:
prefixchars = text[:prefix]
arc = graph.find_path(prefixchars, address=address)
if arc is None:
return
sofar = emptybytes.join(prefixchars)
address = arc.target
accept = arc.accept
stack = [(address, k, prefix, sofar, accept)]
seen = set()
while stack:
state = stack.pop()
# Have we already tried this state?
if state in seen:
continue
seen.add(state)
address, k, i, sofar, accept = state
# If we're at the end of the text (or deleting enough chars would get
# us to the end and still within K), and we're in the accept state,
# yield the current result
if (len(text) - i <= k) and accept:
yield utf8decode(sofar)[0]
# If we're in the stop state, give up
if address is None:
continue
# Exact match
if i < len(text):
arc = graph.find_arc(address, text[i])
if arc:
stack.append((arc.target, k, i + 1, sofar + text[i],
arc.accept))
# If K is already 0, can't do any more edits
if k < 1:
continue
k -= 1
arcs = graph.arc_dict(address)
# Insertions
stack.extend((arc.target, k, i, sofar + char, arc.accept)
for char, arc in iteritems(arcs))
# Deletion, replacement, and transpo only work before the end
if i >= len(text):
continue
char = text[i]
# Deletion
stack.append((address, k, i + 1, sofar, False))
# Replacement
for char2, arc in iteritems(arcs):
if char2 != char:
stack.append((arc.target, k, i + 1, sofar + char2, arc.accept))
# Transposition
if i < len(text) - 1:
char2 = text[i + 1]
if char != char2 and char2 in arcs:
# Find arc from next char to this char
target = arcs[char2].target
if target:
arc = graph.find_arc(target, char)
if arc:
stack.append((arc.target, k, i + 2,
sofar + char2 + char, arc.accept))
def from_bytes(self, bs):
return utf8decode(bs)[0]
def from_bytes(self, bs):
return utf8decode(bs)[0]