Sha256: 56734ce489299b4eca0569174a8494ffc0fbf5aa9defedb07f0e6f825932a171
Contents?: true
Size: 1.43 KB
Versions: 5
Compression:
Stored size: 1.43 KB
Contents
module Internals
module Tokenizers
# The base indexing tokenizer.
#
# Override in indexing subclasses and define in configuration.
#
class Index < Base
def self.default= new_default
@default = new_default
end
def self.default
@default ||= new
end
# Default indexing preprocessing hook.
#
# Does:
# 1. Character substitution.
# 2. Downcasing.
# 3. Remove illegal expressions.
# 4. Remove non-single stopwords. (Stopwords that occur with other words)
#
def preprocess text
text = substitute_characters text
text.downcase!
remove_illegals text
# we do not remove single stopwords for an entirely different
# reason than in the query tokenizer.
# An indexed thing with just name "UND" (a possible stopword) should not lose its name.
#
remove_non_single_stopwords text
text
end
# Default indexing pretokenizing hook.
#
# Does:
# 1. Split the text into words.
# 2. Normalize each word.
#
def pretokenize text
words = split text
words.collect! do |word|
normalize_with_patterns word
word
end
end
# Does not actually return a token, but a
# symbol "token".
#
def token_for text
symbolize text
end
end
end
end
Version data entries
5 entries across 5 versions & 1 rubygems