我跳过检查你看看这行是否已经存在，通常是像这样（这里我想用'BAR'替换'FOO'）： / p>

  full_path_to_read = File.expand_path（'〜/ test1.txt'）
 full_path_to_write = File.expand_path（'〜/ test2.txt'）

 File.open（full_path_to_read）do | source_file | 
 contents = source_file.read 
 contents.gsub！（/ FOO /，'BAR'）
 File.open（full_path_to_write，w +）{| f | f.write（contents）} 
 end 
 

使用 expand_path 在这里也可能有点儿迂腐，但是我喜欢它，所以我不会不小心打开一些我并不想要的文件。

I have this little program I write in ruby. I found a nice piece of code here, on SO, to find and replace something in a file, but it doesn't seems to work. Here's the code:

#!/usr/bin/env ruby

DOC = "test.txt"
FIND = /,,^M/
SEP = "\n"

#make substitution
File.read(DOC).gsub(FIND, SEP)

#Check if the line already exist
unique_lines = File.readlines(DOC).uniq

#Save the result in a new file
File.open('test2.txt', 'w') { |f| f.puts(unique_lines) }

Thanks everybody !

I skip the check you make to see if the line already exists and usually go with something like this (here I want to replace 'FOO' with 'BAR'):

full_path_to_read = File.expand_path('~/test1.txt')
full_path_to_write = File.expand_path('~/test2.txt')

File.open(full_path_to_read) do |source_file|
  contents = source_file.read
  contents.gsub!(/FOO/, 'BAR')
  File.open(full_path_to_write, "w+") { |f| f.write(contents) }
end

The use of expand_path is also probably a bit pedantic here, but I like it just so that I don't accidentally clobber some file I didn't mean to.