Created
May 23, 2013 11:04
-
-
Save zarkzork/5635297 to your computer and use it in GitHub Desktop.
Simple file search in Ruby
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
class String | |
def triads() | |
string = self.clone | |
Enumerator.new do |y| | |
while string.size > 2 | |
y << string.slice(0, 3) | |
string.slice!(0) | |
end | |
end | |
end | |
end | |
class Tokenizer | |
def initialize(content) | |
@content = content | |
end | |
def triads | |
@content.gsub(/[^[:alnum:].:_]+/, ' ').split(' ').map(&:triads).map(&:to_a).flatten | |
end | |
end | |
class FileMatcher | |
def initialize(paths) | |
@db = build_db(paths) | |
end | |
def files_for_query(string) | |
triads = Tokenizer.new(string).triads | |
files = @db[triads.shift].to_a | |
triads.inject(files){ |old_values, triad| | |
old_values.map{ |path, line_num, index| [path, line_num, index + 1] } & @db[triad] | |
}.map{ |o, line_num, _| [o, line_num] }.uniq | |
end | |
private | |
def build_db(paths) | |
storage = Hash.new{ |h,k| h[k] = [] } | |
paths.each do |path| | |
content = IO.readlines(path).map(&:chomp).map{ |s| s.force_encoding('utf-8') } | |
content.each_with_index do |line, line_num| | |
Tokenizer.new(line).triads.each_with_index{ |triad, index| storage[triad] << [path, line_num, index] } | |
end | |
end | |
storage | |
end | |
end | |
class FileIndex | |
def initialize(paths) | |
@index = build_index(paths) | |
end | |
def get_line(path, line) | |
@index[path][line] | |
end | |
private | |
def build_index(paths) | |
paths.inject({}) do |acc, path| | |
lines = IO.readlines(path).map(&:chomp).map{ |s| s.force_encoding('utf-8') } | |
acc[path] = Hash[lines.each_with_index.to_a.map(&:reverse)] | |
acc | |
end | |
end | |
end | |
class Engine | |
def search(string) | |
@file_matcher.files_for_query(string).map do |path_line| | |
line_content = @file_index.get_line(*path_line) | |
if line_content =~ /#{string}/ | |
path_line.push(line_content) | |
else | |
nil | |
end | |
end.compact | |
end | |
def initialize(path, extensions) | |
@paths = get_paths(path, extensions) | |
@file_matcher = FileMatcher.new(@paths) | |
@file_index = FileIndex.new(@paths) | |
end | |
def get_paths(path, extensions) | |
extensions_part = extensions.map{ |o| "-name '*#{o}'" }.join(' -or ') | |
`find #{path} #{extensions_part}`.split("\n") | |
end | |
end | |
class Server | |
def parse_command | |
line = $stdin.readline.chomp | |
command, arguments = line.split(' ', 2) | |
if commands.include?(command) | |
self.send(command, arguments) | |
else | |
puts "ERROR: no such command: #{command}" | |
end | |
end | |
private | |
def commands; %w[init search]; end | |
def init(arguments) | |
path, *extensions_part = arguments.split(' ') | |
@engine = Engine.new('data', ['.rb']) | |
puts 'ok' | |
end | |
def search(arguments) | |
if @engine | |
results = @engine.search(arguments) | |
puts results.map{ |file, line, string| "#{file}:#{line} #{string}"} | |
puts "ok" | |
else | |
puts "ERROR: you need to issue init first." | |
end | |
end | |
end | |
@server = Server.new | |
@server.parse_command while true |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment