shared_count-cli 0.0.2 → 0.0.3

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: dbac3484620be6939c0dd0941d049ae62c693901
-  data.tar.gz: af1587d6aef436efdd9aa2d07eb4b3a9073fd27b
+  metadata.gz: 1fe6bc3a8a598572968a5b70e68f8a7fd534d272
+  data.tar.gz: 27bf9e7bac99f698e0b8b453232615b5cfd747c4
 SHA512:
-  metadata.gz: 08ea24413ded85fa2ab27f187769e178459dcaf64ee67690c60e6670c090b1fda30de1d7422abab6635ab45ad4b4f0eaa704b02075f99987b1a8f6eb9c3c4fe8
-  data.tar.gz: 35fd4f6093cab45615bde109e1ae42fda27614a798c950375acdd9bba7ee8668f9773ac52ff30169f5ff7c16283f37962f5981415aeb91fbe2009ebb05e8f029
+  metadata.gz: facacde3a1c30afee7008f08dd4760f0b04cd7d4778a4b74433e46190fab6b6b591d01fff8d367ddbc53e7d69e5109a70d18a82ee8ee4fb105c2e18ad87deda2
+  data.tar.gz: 61adc0fac22c6c50c90a0b6f48453810684b362233ee1ee889de6f80f525da26b05283af0df9f1b5ec76f19b4473ae52ee49289035a4312cc8a99a7e5ce3c39d

data/README.md

@@ -22,13 +22,19 @@ $ shared_count-cli ~/Desktop/file?.txt # => ~/Desktop/file1.txt ~/Desktop/file2.
 $ shared_count-cli ~/Desktop/file?.txt > ~/Desktop/output.csv
 ```
 
-* Launch it in debug mode
+* Run it in debug mode
 
 ```ruby
 $ DEBUG=true shared_count-cli ~/Desktop/file?.txt > ~/Desktop/output.csv
 $ tail -f shared_count-cli.log
 ```
 
+* For more fine control see the available command line options
+
+```ruby
+$ shared_count-cli --help
+```
+
 ## Contributing
 
 1. Fork it ( http://github.com/<my-github-username>/shared_count-cli/fork )

data/bin/shared_count-cli

@@ -1,7 +1,45 @@
 #!/usr/bin/env ruby
 
+require "getoptlong"
+
+opts = GetoptLong.new(
+  ['--help', '-h', GetoptLong::NO_ARGUMENT],
+  ['--concurrency', '-c', GetoptLong::REQUIRED_ARGUMENT],
+  ['--iteration-size', GetoptLong::REQUIRED_ARGUMENT]
+)
+
+concurrency, iteration_size = nil
+opts.each do |opt, arg|
+  case opt
+  when '--help'
+    puts <<-EOF
+shared_count-cli [OPTION] ... file1 [file2]
+
+-h, --help:
+  show help
+
+--concurrency n, -c n:
+  use at most n threads, defaults to 50
+
+--iteration-size n:
+  process at most n URLs per iteration (group of x threads), defaults to 1000
+
+EOF
+    exit
+  when '--concurrency'
+    concurrency = arg.to_i
+  when '--iteration-size'
+    iteration_size = arg.to_i
+  end
+end
+
 require_relative "../lib/shared_count/cli"
 
+SharedCount::Cli.configure do |config|
+  config.concurrency = concurrency
+  config.iteration_size = iteration_size
+end
+
 urls = ARGF.readlines
 urls.reject! { |url| url.empty? }
 urls.map! { |url| url.start_with?("http") ? url : "http://#{url}" }

data/lib/shared_count/cli.rb

@@ -18,18 +18,22 @@ module SharedCount
     MAX_CONCURRENCY = 50
 
     class << self
+      attr_writer :concurrency, :iteration_size
+
       def run(lines)
         configure_shared_count_client
+        logger.info "Using #{concurrency} threads"
+        logger.info "The iteration size is #{iteration_size} URLs"
 
-        iterations, mod =  lines.length.divmod(LINES_PER_ITERATION)
+        iterations, mod =  lines.length.divmod(iteration_size)
         iterations += 1 if mod > 0
         results = Queue.new
 
         iterations.times do |iteration|
           logger.error "Iteration ##{iteration + 1}"
           queue = Queue.new
-          from = LINES_PER_ITERATION * iteration
-          lines[from, LINES_PER_ITERATION].each { |url| queue.push(url) }
+          from = iteration_size * iteration
+          lines[from, iteration_size].each { |url| queue.push(url) }
           thread_count = [MAX_CONCURRENCY, lines.length].min
 
           threads = (0...thread_count).map do |thread|
@@ -108,6 +112,18 @@ module SharedCount
         end
       end
 
+      def configure
+        yield self
+      end
+
+      def concurrency
+        @concurrency ||= MAX_CONCURRENCY
+      end
+
+      def iteration_size
+        @iteration_size ||= LINES_PER_ITERATION
+      end
+
     private
 
       def logger

data/lib/shared_count/cli/version.rb

@@ -1,5 +1,5 @@
 module SharedCount
   module Cli
-    VERSION = "0.0.2"
+    VERSION = "0.0.3"
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: shared_count-cli
 version: !ruby/object:Gem::Version
-  version: 0.0.2
+  version: 0.0.3
 platform: ruby
 authors:
 - Cristian Rasch
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-01-06 00:00:00.000000000 Z
+date: 2014-01-07 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: shared_count_api