Project

General

Profile

Feature #35764 ยป 0001-tokenize-query-strings-for-Issue.like-and-Query-sql_.patch

Go MAEDA, 2021-08-15 07:16

View differences:

app/models/issue.rb
ids.any? ? where(:assigned_to_id => ids) : none
end)
scope :like, (lambda do |q|
q = q.to_s
if q.present?
where("LOWER(#{table_name}.subject) LIKE LOWER(?)", "%#{sanitize_sql_like q}%")
where(*::Query.tokenized_like_conditions("#{table_name}.subject", q))
end
end)
app/models/query.rb
prefix = suffix = nil
prefix = '%' if options[:ends_with]
suffix = '%' if options[:starts_with]
prefix = suffix = '%' if prefix.nil? && suffix.nil?
value = queried_class.sanitize_sql_like value
queried_class.sanitize_sql_for_conditions([
Redmine::Database.like(db_field, '?', :match => options[:match]),
"#{prefix}#{value}#{suffix}"
])
if prefix || suffix
value = queried_class.sanitize_sql_like value
queried_class.sanitize_sql_for_conditions([
Redmine::Database.like(db_field, '?', :match => options[:match]),
"#{prefix}#{value}#{suffix}"
])
else
queried_class.sanitize_sql_for_conditions(
::Query.tokenized_like_conditions(db_field, value, **options)
)
end
end
def self.tokenized_like_conditions(db_field, value, **options)
tokens = Redmine::Search::Tokenizer.new(value).tokens
tokens = [value] unless tokens.present?
sql, values = tokens.map do |token|
[Redmine::Database.like(db_field, '?', options), "%#{sanitize_sql_like token}%"]
end.transpose
[sql.join(" AND "), *values]
end
# Adds a filter for the given custom field
lib/redmine/search.rb
@projects = projects
@cache = options.delete(:cache)
@options = options
# extract tokens from the question
# eg. hello "bye bye" => ["hello", "bye bye"]
@tokens = @question.scan(%r{((\s|^)"[^"]+"(\s|$)|\S+)}).collect {|m| m.first.gsub(%r{(^\s*"\s*|\s*"\s*$)}, '')}
# tokens must be at least 2 characters long
# but for Chinese characters (Chinese HANZI/Japanese KANJI), tokens can be one character
@tokens = @tokens.uniq.select {|w| w.length > 1 || w =~ /\p{Han}/}
# no more than 5 tokens to search for
@tokens.slice! 5..-1
@tokens = Tokenizer.new(@question).tokens
end
# Returns the total result count
......
end
end
class Tokenizer
def initialize(question)
@question = question.to_s
end
def tokens
# extract tokens from the question
# eg. hello "bye bye" => ["hello", "bye bye"]
tokens = @question.scan(%r{((\s|^)"[^"]+"(\s|$)|\S+)}).collect {|m| m.first.gsub(%r{(^\s*"\s*|\s*"\s*$)}, '')}
# tokens must be at least 2 characters long
# but for Chinese characters (Chinese HANZI/Japanese KANJI), tokens can be one character
# no more than 5 tokens to search for
tokens.uniq.select{|w| w.length > 1 || w =~ /\p{Han}/}.first 5
end
end
module Controller
def self.included(base)
base.extend(ClassMethods)
    (1-1/1)