Created
January 23, 2013 07:37
-
-
Save majioa/4602879 to your computer and use it in GitHub Desktop.
Sample downloader tool in Ruby. it reads HTTP or FTP sites from /local/misc/misc/sites and stores results into /local/misc/mirrors.
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
#!/usr/bin/ruby | |
require 'net/ftp' | |
require 'net/http' | |
require 'fileutils' | |
class String | |
def to_l(srcl,dstl) | |
val = gsub(/(["'\&\(\)])/) { "\\" + $1 } | |
`echo #{val} |iconv -f #{srcl} -t #{dstl}`.sub(/\n/,"") | |
end | |
end | |
class Net::FTP | |
def mirror(folder) | |
printf("Curdir %s\n", FileUtils.pwd) if $verbose > 1 | |
printf("Remdir %s\n", pwd) if $verbose > 1 | |
name = folder.to_l("cp1251","utf-8") | |
unless $simulate > 0 then | |
begin | |
FileUtils.cd(name) | |
rescue Errno::ENOENT | |
FileUtils.makedirs(name) | |
retry | |
end | |
end | |
printf("Folder %s\n", name) if $verbose > 0 | |
chdir(folder) | |
list.each {|line| | |
line =~ /^([d-])[rwx-]{1,9} +\d+ \w+ +\w+ +(\d+) (\w{3} +\d{1,2} +[\d:]+) (.*)$/ | |
next if $4 == "." or $4 == ".." | |
if $1 == "d" then | |
printf("Down to folder %s\n", $4) if $verbose > 1 | |
mirror($4) | |
else | |
file = $4.to_l("cp1251","utf-8") | |
if File.size?(file) == $2.to_i then | |
printf("File %s is exist and the same as remote: not downloading\n", file) if $verbose > 0 | |
next; | |
end | |
printf("Downloading file %s\n", file) if $verbose > 0 | |
begin | |
getbinaryfile($4, $4.to_l("cp1251","utf-8"), 1024) unless $simulate > 0 | |
# rescue Errno::EEXIST | |
end | |
end | |
} | |
FileUtils.cd("..") | |
chdir("..") | |
end | |
end | |
def log(msg) | |
puts msg | |
end | |
def error(msg) | |
log msg | |
exit | |
end | |
#puts "Initialization" | |
p = `ps aux`#|grep downloader.rb` | |
c = 0 | |
val = $0.gsub(/.*\//,"") | |
while p =~ /ruby.*#{val}/o | |
c = c + 1 | |
p = p.sub(/#{val}/o, "") | |
end | |
exit if c > 1 | |
#init vars | |
$verbose = 1 | |
$simulate = 0 | |
root_folder = "/local/misc/mirrors" | |
sitef = IO.read('/local/misc/misc/sites'); | |
#change to specified path | |
begin | |
FileUtils.cd(root_folder) | |
rescue Errno::ENOENT | |
FileUtils.mkdir(root_folder) | |
retry | |
end | |
sitef.each { |line| | |
FileUtils.cd(root_folder) | |
next if line == "\n" | |
puts "Reading the new line" if $verbose > 0 | |
if line =~ /^(.*) +(.*?)$/ then | |
site = $1 | |
folder_path = $2 | |
else | |
site = line | |
end | |
if (site.gsub(/%(..)/) {|s| $1.hex.chr } !~ /^ *(ftp|http):\/\/(?:(.*?)\/)?(?:(.*)\/)*(?:(.*)\/?)?$/) then | |
puts "Wrong site name, please check" | |
exit 1 | |
end | |
ftp_addr = $2 | |
folder_path = $2 + '/' + $3.to_s.to_l("cp1251","utf-8") if !folder_path | |
remote_dir = $3 | |
folder = $4 | |
puts ftp_addr, folder_path, remote_dir, folder if $verbose > 2 | |
puts "Connection to site #{ftp_addr}" if $verbose > 0 | |
printf("Site = %s, folder = %s, folder = %s\n", ftp_addr.to_l("cp1251","utf-8"), folder_path, folder.to_l("cp1251","utf-8")) if $verbose > 1 | |
begin | |
ftp = Net::FTP.new(ftp_addr.gsub(/\//,"")) | |
rescue Errno::ECONNREFUSED | |
printf("Connection to server %s refused\n", ftp_addr.gsub(/\//,"")) | |
exit 1 | |
end | |
begin | |
ftp.login | |
rescue Net::FTPTempError | |
next | |
end | |
begin | |
FileUtils.cd(folder_path) | |
rescue Errno::ENOENT | |
FileUtils.makedirs(folder_path) | |
retry | |
end | |
ftp.chdir(remote_dir) if remote_dir | |
ftp.mirror(folder) | |
ftp.close | |
} |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment