External/wayback-machine-downloader
1
0
Fork 0
mirror of https://github.com/StrawberryMaster/wayback-machine-downloader.git synced 2025-12-29 16:16:06 +00:00
Code Issues Packages Projects Releases Wiki Activity

added new option for regex acceptance you dont have to download entire site when looking for specific path

Browse Source
This commit is contained in:
Gil Cohen
2015-11-06 13:11:26 -05:00
parent ed7948a372
commit 432ca1d5b5
3 changed files with 41 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

14
bin/wayback_machine_downloader
View File

@@ -2,6 +2,7 @@
require_relative '../lib/wayback_machine_downloader'
require 'optparse'
require 'pp'
options = {}
option_parser = OptionParser.new do |opts|
@@ -11,19 +12,26 @@ option_parser = OptionParser.new do |opts|
opts.separator "Download any website from the Wayback Machine."
opts.separator ""
opts.separator "Optional option:"
opts.separator "Optional options:"
opts.on("-t", "--timestamp TIMESTAMP", Integer, "Only files on or before timestamp supplied (ie. 20150806225358)") do |t|
options[:timestamp] = t
end
opts.on("--accept-regex [ACCEPT_REGEX]", String,"Specify a regular expression to download. If a path doesn't meet this regex, it won't get downloaded.") do |accept_regex|
options[:accept_regex] = accept_regex
end
opts.on("-v", "--version", "Display version") do |t|
options[:version] = t
end
end.parse!
if base_url = ARGV[0]
wayback_machine_downloader = WaybackMachineDownloader.new base_url: base_url, timestamp: options[:timestamp]
# this used to be 0. we want to look at the /last/ option.
#
# TODO: this argument needs to be handled better. argument handling is sorta messy.
if base_url = ARGV[-1]
wayback_machine_downloader = WaybackMachineDownloader.new base_url: base_url, timestamp: options[:timestamp], accept_regex: options[:accept_regex]
wayback_machine_downloader.download_files
elsif options[:version]
puts WaybackMachineDownloader::VERSION