实施Norvig拼写检查器时出错

Question

I've tried implementing Peter Norvig's spell checker in cython, and I get the following errors when I try to compile:

Error compiling Cython file:
------------------------------------------------------------
...
cimport cython
cimport re
       ^
------------------------------------------------------------

spell.pyx:2:8: 're.pxd' not found

Error compiling Cython file:
------------------------------------------------------------
...
cimport cython
cimport re
cimport collections
   ^
------------------------------------------------------------

spell.pyx:3:8: 'collections.pxd' not found

Error compiling Cython file:
------------------------------------------------------------
...
cdef char* words(char* text): return re.findall('[a-z]+', text.lower()) 

cdef char* train(char* features):
    cdef char f
    cdef char* model    
    model = collections.defaultdict(lambda: 1)
                              ^
------------------------------------------------------------

spell.pyx:10:35: Storing unsafe C derivative of temporary Python reference

Error compiling Cython file:
------------------------------------------------------------
...
   cdef char* splits
   cdef char* deletes
   cdef char* transposes
   cdef char* replaces
   cdef char* inserts
   splits     = [(word[:i], word[i:]) for i in range(len(word) + 1)]
               ^
------------------------------------------------------------

spell.pyx:30:16: Storing unsafe C derivative of temporary Python reference

Error compiling Cython file:
------------------------------------------------------------
...
   cdef char* deletes
   cdef char* transposes
   cdef char* replaces
   cdef char* inserts
   splits     = [(word[:i], word[i:]) for i in range(len(word) + 1)]
   deletes    = [a + b[1:] for a, b in splits if b]
               ^
------------------------------------------------------------

spell.pyx:31:16: Storing unsafe C derivative of temporary Python reference

Error compiling Cython file:
------------------------------------------------------------
...
   cdef char* transposes
   cdef char* replaces
   cdef char* inserts
   splits     = [(word[:i], word[i:]) for i in range(len(word) + 1)]
   deletes    = [a + b[1:] for a, b in splits if b]
   transposes = [a + b[1] + b[0] + b[2:] for a, b in splits if len(b)>1]
               ^
------------------------------------------------------------

spell.pyx:32:16: Storing unsafe C derivative of temporary Python reference

Error compiling Cython file:
------------------------------------------------------------
...
   cdef char* replaces
   cdef char* inserts
   splits     = [(word[:i], word[i:]) for i in range(len(word) + 1)]
   deletes    = [a + b[1:] for a, b in splits if b]
   transposes = [a + b[1] + b[0] + b[2:] for a, b in splits if len(b)>1]
   replaces   = [a + c + b[1:] for a, b in splits for c in alphabet if b]
               ^
------------------------------------------------------------

spell.pyx:33:16: Storing unsafe C derivative of temporary Python reference

Error compiling Cython file:
------------------------------------------------------------
...
   cdef char* inserts
   splits     = [(word[:i], word[i:]) for i in range(len(word) + 1)]
   deletes    = [a + b[1:] for a, b in splits if b]
   transposes = [a + b[1] + b[0] + b[2:] for a, b in splits if len(b)>1]
   replaces   = [a + c + b[1:] for a, b in splits for c in alphabet if b]
   inserts    = [a + c + b     for a, b in splits for c in alphabet]
           ^
------------------------------------------------------------

spell.pyx:34:16: Storing unsafe C derivative of temporary Python reference

Error compiling Cython file:
------------------------------------------------------------
...
   return set(deletes + transposes + replaces + inserts)

cdef char* known_edits2(char* word):
    cdef char* e2 
    cdef char* e1
    return set(e2 for e1 in edits1(word) for e2 in edits1(e1) if e2 in NWORDS)
                                                       ^
------------------------------------------------------------

spell.pyx:40:60: Cannot assign type 'char' to 'char *'

Error compiling Cython file:
------------------------------------------------------------
...

cdef char* correct(char* word):

    cdef char*  candidates;
    cdef char* key;
    candidates = known([word]) or known(edits1(word)) or known_edits2(word) or [word]
                       ^
------------------------------------------------------------

spell.pyx:50:28: Cannot assign type 'char *' to 'char'

My code is:

cimport cython
cimport re
cimport collections

cdef char* words(char* text): return re.findall('[a-z]+', text.lower()) 

cdef char* train(char* features):
    cdef char f
    cdef char* model    
    model = collections.defaultdict(lambda: 1)
    for f in features:
        model[f] += 1
    return model

cdef char* NWORDS
NWORDS = train(words(file('big.txt')).read())

cdef char* alphabet = 'abcdefghijklmnopqrstuvwxyz'

cdef char* edits1(char* word):
   cdef int i
   cdef char* a
   cdef char* b
   cdef char* c
   cdef char* splits
   cdef char* deletes
   cdef char* transposes
   cdef char* replaces
   cdef char* inserts
   splits     = [(word[:i], word[i:]) for i in range(len(word) + 1)]
   deletes    = [a + b[1:] for a, b in splits if b]
   transposes = [a + b[1] + b[0] + b[2:] for a, b in splits if len(b)>1]
   replaces   = [a + c + b[1:] for a, b in splits for c in alphabet if b]
   inserts    = [a + c + b     for a, b in splits for c in alphabet]
   return set(deletes + transposes + replaces + inserts)

cdef char* known_edits2(char* word):
    cdef char* e2 
    cdef char* e1
    return set(e2 for e1 in edits1(word) for e2 in edits1(e1) if e2 in NWORDS)

cdef char* known(char* words): 
    cdef char* w
    return set(w for w in words if w in NWORDS)

cdef char* correct(char* word):

    cdef char*  candidates;
    cdef char* key;
    candidates = known([word]) or known(edits1(word)) or known_edits2(word) or [word]
    return max(candidates, key=NWORDS.get)

This is the first time I've worked in cython or anything to do with the C language. By my understanding, it seems like the regular python libraries cannot be imported in cython, so how would I get the collections or regex library?

Answer 1

To import a Python module or a Cython module (.pyx) from within Cython you use the "import" statement - the same one as in Python. To import Cython's definition file (.pxd) you use "cimport".

See:

• http://docs.cython.org/src/userguide/sharing_declarations.html
• http://docs.cython.org/src/reference/language_basics.html