from nltk.chunk import ChunkParserI 
from nltk.chunk.util import conlltags2tree 
from nltk.corpus import gazetteers 

class LocationChunker(ChunkParserI): 
    def __init__(self): 
        self.locations = set(gazetteers.words()) 
        self.lookahead = 0
        for loc in self.locations: 
            nwords = loc.count(' ') 
        if nwords > self.lookahead: 
            self.lookahead = nwords 

nltk.chunk 中的 ChunkParserI 是什么？它究竟是为了什么？另外，请解释代码。分块和解析有什么区别？