datahunter 0.2.0 → 0.2.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 16e4a869538392e0dc2618a3dcd32862c3d015ae
-  data.tar.gz: a282c2f92377da9c5ae53d3c7bf58da488326bed
+  metadata.gz: a69c30628131f2921984bddf97d7285b72df1fec
+  data.tar.gz: db641793fbb347d570595ee76cf47e5e5534e4f1
 SHA512:
-  metadata.gz: 56691ba5f70b12183dc7aecf14f28a595f46336c32709f46b9955bbec7f1df7fffb2b9221ca3fce65b34373e686ff6b53e6489ee0b0ecbd8dfd2d5cf9b5ae252
-  data.tar.gz: 55db3407764898e9b357aae8b2962015bfe3dd7027314ec19a59aec05cf04347a15bc8da986977a4fc9e430b68119bffe89c2ea09651bc3e1eb9a8f2da2079ee
+  metadata.gz: 962d15c0e12844ec7c99fa69ed8f1e566b629c5f5b2f4d6e239d36907a54b9d761fbbf3166b02f58cbe17ab1bad5e7e5f9a6e2a84dce1d4b76639a1bb6ff28cb
+  data.tar.gz: ed0e562e83f233bb4491ace62c157c0dca9a417d413d4a319fef8bbc6e7cc1d1a9fb2d5abe63cbdc88af73a74deaf2b2cc5d56c17c596d83d92bed8e2257a65d

data/bin/hunter

@@ -53,7 +53,7 @@ command :find do |c|
             
             case ask "### get the data? (y/n)".colorize(:yellow)
             when 'y'
-              if datasets[index].has_key?("resources")
+              if datasets[index].has_key?("resources") and datasets[index]["resources"].any?
                 Datahunter.download_the_data datasets[index]
               else
                 Datahunter.open_in_browser datasets[index]["uri"]

data/lib/datahunter/base.rb

@@ -2,14 +2,16 @@ require 'colorize'
 require 'launchy'
 require 'downloadr'
 require 'addressable/uri'
-# require 'Pathname'
 
 module Datahunter
 
 #  DATASETS_URL = "http://localhost:3000/api/datasets/"
   DATASETS_URL = "http://shrouded-harbor-5877.herokuapp.com/api/datasets/"
   FEEDBACK_URL = "https://docs.google.com/forms/d/1yNzZjCCXvWHQCbWz4sx-nui3LafeeLcT7FF9T-vbKvw/viewform"
-  REQUEST_URL = "https://docs.google.com/forms/d/1NRKWmb_mcpKJmrutXvZSZnysM_v0rfLhjD897H3Myrw/viewform?usp=send_form"
+  REQUEST_URL =
+    "https://docs.google.com/forms/d/1NRKWmb_mcpKJmrutXvZSZnysM_v0rfLhjD897H3Myrw/viewform?usp=send_form"
+
+  @extensions = ["json", "csv", "xml", "zip", "gz", "xls", "xlsx", "tsv"]
 
   def self.datasets_url tag, geo=nil, temp=nil
     tag = tag.downcase.split.first if tag
@@ -41,7 +43,44 @@ module Datahunter
     puts ("score: ".colorize(:green) + "#{dataset["huntscore"]}")
     puts
   end
- 
+  
+  def self.print_downloadable_links resources
+    resources.each_with_index do |dl, i|
+      puts ("#{i}. ".colorize(:yellow) +
+            "#{dl["title"]} - ".colorize(:blue) + 
+            "#{dl["format"]}".colorize(:green))
+    end
+  end
+
+  def self.parse_extension_from_uri uri
+    uri.basename.split(".").last
+  end
+  
+  def self.download_file url, format="", alt_url=""
+    uri = Addressable::URI.parse(url)
+    extension = Datahunter.parse_extension_from_uri uri
+
+    if !@extensions.include? extension
+      Launchy.open(url, options = {})
+    else
+      location = Dir.pwd
+      uri = Addressable::URI.parse(url)
+      file_name = uri.basename
+      loc = location + "/" + file_name
+
+      case ask ("Create/overwrite #{loc}?(y/rename/n)".colorize(:yellow))
+      when 'rename'
+        loc = ask "Path/to/filename: ".colorize(:yellow)
+      when 'n'
+        abort("Ok then")
+      end
+      puts "Start downloading..."
+      Downloadr::HTTP.download(url, loc)
+      puts "Your file has been downloaded ;)".colorize(:green)
+      Datahunter.print_excuse_and_alternative_url_message alt_url
+    end
+  end
+
   def self.download_the_data dataset
     resources = dataset["resources"]
     number_of_downloadable_links = resources.size
@@ -52,11 +91,11 @@ module Datahunter
       Datahunter.print_downloadable_links resources
       dl = ask("### which one? (0/1/...)".colorize(:yellow), Integer) {|i| i.in = 0..(number_of_downloadable_links - 1)}
     end
-
+    
     dl = dl.to_i
     Datahunter.download_file(resources[dl]["url"], resources[dl]["format"], dataset["uri"])
   end
-
+  
   def self.open_in_browser url
     if url =~ /\A#{URI::regexp}\z/
       puts "You can't download this dataset directly, but you should from there"
@@ -90,45 +129,13 @@ module Datahunter
          "if you just want to give us a feedback, don't hesitate!".colorize(:red)
   end
 
-  def self.print_bad_uri_message
-    puts "The URL given by the publisher is not valid. We'll try to find out why "\
-         "as soon as we can!".colorize(:red)
-  end
-
-  private
-  
-  def self.print_downloadable_links resources
-    resources.each_with_index do |dl, i|
-      puts ("#{i}. ".colorize(:yellow) +
-            "#{dl["title"]} - ".colorize(:blue) + 
-            "#{dl["format"]}".colorize(:green))
-    end
-  end
-  
-  def self.download_file url, format="", alt_url=""
-    if format == "HTML"
-      Launchy.open(url, options = {})
-    else
-      location = Dir.pwd
-      uri = Addressable::URI.parse(url)
-      file_name = uri.basename
-      loc = location + "/" + file_name
-
-      case ask ("Create/overwrite #{loc}?(y/rename/n)".colorize(:yellow))
-      when 'rename'
-        loc = ask "Path/to/filename: ".colorize(:yellow)
-      when 'n'
-        abort("Ok then")
-      end
-      puts "Start downloading..."
-      Downloadr::HTTP.download(url, loc)
-      puts "Your file has been downloaded ;)".colorize(:green)
-      Datahunter.print_excuse_and_alternative_url_message alt_url
-    end
-  end
-
   def self.print_excuse_and_alternative_url_message alt_url=""
     puts "If this is not the file you expected, it's maybe because publisher don't always keep the metadata up-to-date. We try to clean most of uri's and check the url. Anyway you may be able to download your file by hand here:"
     puts "#{alt_url}".colorize(:blue)
   end
+
+  def self.print_bad_uri_message
+    puts "The URL given by the publisher is not valid. We'll try to find out why "\
+         "as soon as we can!".colorize(:red)
+  end
 end

data/lib/datahunter/version.rb

@@ -1,3 +1,3 @@
 module Datahunter
-  VERSION = "0.2.0"
+  VERSION = "0.2.1"
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: datahunter
 version: !ruby/object:Gem::Version
-  version: 0.2.0
+  version: 0.2.1
 platform: ruby
 authors:
 - Terpo
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-01-30 00:00:00.000000000 Z
+date: 2015-02-07 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler