I'd like to add a single line to the top a of file with Ruby like this:
# initial file contents
something
else
# file contents after prepending "hello" on its own line
hello
something
else
The following code just replaces the contents of the entire file:
f = File.new('myfile', 'w')
f.write "test string"
This is a pretty common task:
original_file = './original_file'
new_file = original_file + '.new'
Set up the test:
File.open(original_file, 'w') do |fo|
%w[something else].each { |w| fo.puts w }
end
This is the actual code:
File.open(new_file, 'w') do |fo|
fo.puts 'hello'
File.foreach(original_file) do |li|
fo.puts li
end
end
Rename the old file to something safe:
File.rename(original_file, original_file + '.old')
File.rename(new_file, original_file)
Show that it works:
puts `cat #{original_file}`
puts '---'
puts `cat #{original_file}.old`
Which outputs:
hello
something
else
---
something
else
You don't want to try to load the file completely into memory. That'll work until you get a file that is bigger than your RAM allocation, and the machine goes to a crawl, or worse, crashes.
Instead, read it line by line. Reading individual lines is still extremely fast, and is scalable. You'll have to have enough room on your drive to store the original and the temporary file.
fwiw this seems to work:
#!usr/bin/ruby
f = File.open("myfile", "r+")
lines = f.readlines
f.close
lines = ["something\n"] + lines
output = File.new("myfile", "w")
lines.each { |line| output.write line }
output.close
I have came up with something like this, it is a little bit more descriptive and less cryptic than other solutions I've seen:
def file_prepend(file, str)
new_contents = ""
File.open(file, 'r') do |fd|
contents = fd.read
new_contents = str << contents
end
# Overwrite file but now with prepended string on it
File.open(file, 'w') do |fd|
fd.write(new_contents)
end
end
And you can use it like this:
file_prepend("target_file.txt", "hello world!\n")
As some have said, probably don’t use this for larger files, but it is a simple start.
rd = IO.read 'myfile'
IO.write 'myfile', "hello\n" + rd
No mechanism exists to do what you want to do easily.
Instead, you will need to open the file, delete the file, open a new file by the old name for writing, write your content, and then write the new file from the old file's contents. That's pretty convoluted sounding but is straightforward in code:
$ cat prepend.rb
#!/usr/bin/ruby
File.open("passwd", "r") do |orig|
File.unlink("passwd")
File.open("passwd", "w") do |new|
new.write "test string"
new.write(orig.read())
end
end
Note that the mechanism I've used doesn't bother checking errors -- you should probably handle errors on each File.open() request and the File.unlink() request -- and it assumes that the entire contents of the file will fit in memory. A short example:
$ rm passwd
$ cp /etc/passwd .
$ ./prepend.rb
$ head passwd
test stringroot:x:0:0:root:/root:/bin/bash
daemon:x:1:1:daemon:/usr/sbin:/bin/sh
bin:x:2:2:bin:/bin:/bin/sh
sys:x:3:3:sys:/dev:/bin/sh
sync:x:4:65534:sync:/bin:/bin/sync
games:x:5:60:games:/usr/games:/bin/sh
man:x:6:12:man:/var/cache/man:/bin/sh
lp:x:7:7:lp:/var/spool/lpd:/bin/sh
mail:x:8:8:mail:/var/mail:/bin/sh
news:x:9:9:news:/var/spool/news:/bin/sh
$
If you want to handle files that might not fit entirely in memory, you should code a loop sort of like this (untested -- better consider it pseudo-code):
while (data=orig.read(4096)) {
new.write(data)
}
As an alternative, you could write to a temporary file, and if the writing process succeeds, then delete the file and rename the temporary file in place. Whichever approach makes the most sense to you.
You can try this:
File.copy_stream(myfile,tempfile)
f = File.open(myfile,'w')
f.write("Hello\n#{File.open(tempfile,'r').read}")
f.close
File.delete(tempfile)
From the command line, you can do:
ruby -i -pe '$_= "prepended text\n"+$_ if $. == 1' myfile
or more efficiently
ruby -i -pe 'BEGIN { gets; print "prepended text\n" + $_ }; ' myfile
Sadly, it turns out, the -i (in-place) options isn't truly in-place, though (and nor is seds in-place option for that matter) -- my file will have a different inode after the op.
That made me sad, because effectively, I can't filter through (prepend to) a huge file if I don't have enough diskspace for two copies of it.
It's not really difficult to do it in-place, though (filtering in general, not just for the first line). All you need to do is:
1) make sure, you have separate read and write pointers (or separate File objects for reading and writing) 2) make sure you have buffered the unread parts that you are going to rewrite 3) truncate to file at the end if your filtering operation should end up with a shorter file
I wrote wrapper class for that at https://github.org/pjump/i_rewriter .
With it, you can do
Rewriter.new("myfile") do |line, index|
index == 0 ? "prepended text\n" + line : line
end
or with the executable:
$ i_rewriter 'index == 0 ? "prepended text\n" + line : line' myfile
Use carefully (it can corrupt your file if interrupted).
Pure, but works. Minimize file-related operations.
`#!/bin/ruby
inv_file = File.open("./1.txt","r+") { |f|
#Read file string by-string
until f.eof?
#Searching for some
if f.readline[/condition here/]
#remember position
offset = f.pos
#Find it? - stop reading
break
end
end
#Remember what contains rest of file
tail = f.readlines
#Again to offset
f.pos = offset
#insert what necessary
f.puts "INSERTED"
#reconstruct tail
f.write tail.join
}`
