archivist-client 0.1.0

checksums.yaml

@@ -0,0 +1,15 @@
+---
+!binary "U0hBMQ==":
+  metadata.gz: !binary |-
+    OTNjMjQ1ZTA3M2RkNzA3NzA0MDg1ZjY2MmMwMzYwODYzYzM1ZTZhMA==
+  data.tar.gz: !binary |-
+    ZGYzZGU5NjBlNzI4NGNkZGM4NmVkNWU1NjY0NTM3OGRhM2ZhYjM4OQ==
+SHA512:
+  metadata.gz: !binary |-
+    MDQ4YzFhNzM2YjE2YjdhYmU1YTU2MzJlZDE4Nzc4NDM0YmJlZWUyYTI1OTdj
+    ZTUwNzk0MmUyOTdmMjRjMjYzZmVmMDM0NjhiZjA5ZDc4NmMyNzQ5YzdiM2Nh
+    NmZhNDVjYTQyNjQ3YjI4MzQ4ZmE0YTY4ZjRjOGNkZTU2MjFjYmU=
+  data.tar.gz: !binary |-
+    NGYzZmVmNTY2ZmEzZDM0NGI0OTJhOTNmZWUyNjViYTk3MGI1NzgxMWYxNTI1
+    NTU0YWE1YjJhYzgxOTY3ZDcyNzFiZWMxZGI5ZDZkNDJjM2Y1YmRiNmY0MTZh
+    M2M5M2EzZWNiZjNiNTlkMzkxNDBlYzU2ZTBmN2EyNDY2NWJlYWQ=

data/README.md

@@ -0,0 +1,107 @@
+Archivist::Client
+=========
+
+an *archive.org* Ruby client
+
+| Project                 |  Archive::Client    |
+|------------------------ | ----------------- |
+| gem name                |  archive-client   |
+| license                 |  MIT              |
+| moldiness               |  [![Maintainer Status](http://stillmaintained.com/wordtreefoundation/archive-client.png)](http://stillmaintained.com/wordtreefoundation/archive-client) |
+| version                 |  [![Gem Version](https://badge.fury.io/rb/archive-client.png)](http://badge.fury.io/rb/archive-client) |
+| dependencies            |  [![Dependency Status](https://gemnasium.com/wordtreefoundation/archive-client.png)](https://gemnasium.com/wordtreefoundation/archive-client) |
+| code quality            |  [![Code Climate](https://codeclimate.com/github/wordtreefoundation/archive-client.png)](https://codeclimate.com/github/wordtreefoundation/archive-client) |
+| continuous integration  |  [![Build Status](https://secure.travis-ci.org/wordtreefoundation/archive-client.png?branch=master)](https://travis-ci.org/wordtreefoundation/archive-client) |
+| test coverage           |  [![Coverage Status](https://coveralls.io/repos/wordtreefoundation/archive-client/badge.png)](https://coveralls.io/r/wordtreefoundation/archive-client) |
+| homepage                |  [https://github.com/wordtreefoundation/archive-client][homepage] |
+| documentation           |  [http://rdoc.info/github/wordtreefoundation/archive-client/frames][documentation] |
+| authors                 |  [Duane Johnson](https://coderbits.com/canadaduane) [![Endorse Duane](https://api.coderwall.com/canadaduane/endorsecount.png)](http://coderwall.com/canadaduane) |
+|                         |  [Peter Boling](https://coderbits.com/pboling) [![Endorse Peter](https://api.coderwall.com/pboling/endorsecount.png)](http://coderwall.com/pboling) |
+
+
+Getting Started
+---------------
+
+```ruby
+require 'archive-client'
+```
+
+Create an Archivist client:
+```ruby
+client = Archivist::Client::Base.new
+```
+
+Search for the books you're interested in:
+```ruby
+books = client.search(:start_year => 1500, :end_year => 1510)
+```
+
+Download them:
+```ruby
+books.each do |book|
+  puts book.download
+end
+```
+
+## Contributors
+
+See the [Network View](https://github.com/wordtreefoundation/archive-client/network) and the [CHANGELOG](https://github.com/wordtreefoundation/archive-client/blob/master/CHANGELOG.md)
+
+## How you can help!
+
+Take a look at the `reek` list which is the file called `REEK` and stat fixing things.  Once you complete a change, run the tests:
+
+```
+bundle exec rspec spec
+```
+
+If the tests pass refresh the `reek` list:
+
+```
+bundle exec rake reek > REEK
+```
+
+Follow the instructions for "Contributing" below.
+
+## Contributing
+
+1. Fork it
+2. Create your feature branch (`git checkout -b my-new-feature`)
+3. Commit your changes (`git commit -am 'Added some feature'`)
+4. Push to the branch (`git push origin my-new-feature`)
+5. Make sure to add tests for it. This is important so I don't break it in a future version unintentionally.
+6. Create new Pull Request
+
+## Versioning
+
+This library aims to adhere to [Semantic Versioning 2.0.0][semver].
+Violations of this scheme should be reported as bugs. Specifically,
+if a minor or patch version is released that breaks backward
+compatibility, a new version should be immediately released that
+restores compatibility. Breaking changes to the public API will
+only be introduced with new major versions.
+
+As a result of this policy, you can (and should) specify a
+dependency on this gem using the [Pessimistic Version Constraint][pvc] with two digits of precision.
+
+For example:
+
+    spec.add_dependency 'archivist-client', '~> 0.1.0'
+
+## References
+
+* [Source Code](http://github.com/wordtreefoundation/archive-client)
+* [Release Announcement](http://www.textual-analysis.org/2013/11/archivist/)
+
+## Legal
+
+* MIT License - See LICENSE file in this project
+* Copyright (c) 2013 Duane Johnson & Word Tree Foundation
+
+[semver]: http://semver.org/
+[pvc]: http://docs.rubygems.org/read/chapter/16#page74
+[documentation]: http://rdoc.info/github/wordtreefoundation/archive-client/frames
+[homepage]: https://github.com/wordtreefoundation/archive-client
+
+[![Bitdeli Badge](https://d2weczhvl823v0.cloudfront.net/wordtreefoundation/archive-client/trend.png)](https://bitdeli.com/free "Bitdeli Badge")
+

data/archivist-client.gemspec

@@ -0,0 +1,37 @@
+# -*- encoding: utf-8 -*-
+lib = File.expand_path('../lib', __FILE__)
+$LOAD_PATH.unshift(lib) unless $LOAD_PATH.include?(lib)
+require 'archivist/client/version'
+
+Gem::Specification.new do |gem|
+  gem.name = "archivist-client"
+  gem.summary = "archivist-client is an archive.org client library"
+  gem.description = "archivist-client queries archive.org for book data and downloads some things."
+  gem.homepage = "http://github.com/canadaduane/archivist-client"
+  gem.authors = ['Duane Johnson','Peter Boling']
+  gem.email = ['duane.johnson@gmail.com', 'peter.boling@gmail.com']
+  gem.licenses = ["MIT"]
+
+  gem.files = %w[archivist-client.gemspec README.md]
+  gem.files += Dir.glob("lib/**/*.rb")
+  gem.files += Dir.glob("spec/**/*")
+
+  gem.test_files = Dir.glob("spec/**/*")
+  gem.require_paths = ["lib"]
+  gem.version = Archivist::Client::VERSION
+  gem.required_ruby_version = '>= 1.9.0'
+
+  gem.add_development_dependency "rake"
+  gem.add_development_dependency "bundler", ">= 1.0.0"
+  gem.add_development_dependency "rspec", "~> 2.6"
+  gem.add_development_dependency "debugger"
+  gem.add_development_dependency "webmock"
+  gem.add_development_dependency "vcr"
+  gem.add_development_dependency "reek", ">= 1.2.8"
+  gem.add_development_dependency "roodi", ">= 2.1.0"
+
+  gem.add_dependency "virtus", "~> 1.0.0"
+  gem.add_dependency "representable"
+  gem.add_dependency "faraday"
+  gem.add_dependency "faraday_middleware"
+end

data/lib/archivist.rb

@@ -0,0 +1,2 @@
+require 'archivist/client/base'
+require 'archivist/client/version'

data/lib/archivist/client/base.rb

@@ -0,0 +1,63 @@
+require 'faraday'
+require 'faraday_middleware'
+require 'archivist/representations'
+
+module Archivist
+  module Client
+    class Base
+      attr_reader :conn
+
+      def initialize(opts = {})
+        @opts = {
+          page: 1,
+          rows: 50
+        }.merge(opts)
+
+        @conn = Faraday.new(url: 'http://archive.org') do |faraday|
+          faraday.use FaradayMiddleware::FollowRedirects
+          faraday.request  :url_encoded             # form-encode POST params
+          faraday.adapter  Faraday.default_adapter  # make requests with Net::HTTP
+        end
+      end
+
+      def query(opts)
+        filters = [
+          'mediatype:texts',
+          '-mediatype:collection'
+        ]
+
+        filters.concat(opts.delete(:filters)) if opts[:filters]
+
+        filters << if opts[:language]
+          "language:#{opts.delete(:language)}"
+        else
+          '(language:eng OR language:English)'
+        end
+
+        if opts[:start_year] && opts[:end_year]
+          start_year = "#{opts.delete(:start_year)}-01-01"
+          end_year = "#{opts.delete(:end_year)}-12-31"
+          filters << "date:[#{start_year} TO #{end_year}]"
+        end
+
+        filters.join(' AND ')
+      end
+
+      def params(opts = {})
+        {
+          q: query(opts),
+          fl: %w(identifier title creator date language mediattype),
+          sort: ['date asc'],
+          output: 'json'
+        }.merge(@opts).merge(opts)
+      end
+
+      def search(opts = {})
+        Model::QueryResponse.new.tap do |qr|
+          response = @conn.get('/advancedsearch.php', params(opts))
+          Representation::QueryResponse.new(qr).from_json(response.body)
+        end
+      end
+    end
+  end
+end

data/lib/archivist/client/version.rb

@@ -0,0 +1,5 @@
+module Archivist
+  module Client
+    VERSION = '0.1.1'
+  end
+end

data/lib/archivist/models.rb

@@ -0,0 +1,7 @@
+require 'archivist/models/body'
+require 'archivist/models/document'
+require 'archivist/models/file_format'
+require 'archivist/models/format_index'
+require 'archivist/models/header'
+require 'archivist/models/parameters'
+require 'archivist/models/query_response'

data/lib/archivist/models/body.rb

@@ -0,0 +1,15 @@
+require 'virtus'
+
+require_relative 'document'
+
+module Archivist
+  module Model
+    class Body
+      include Virtus.model
+
+      attribute :num_found, Integer
+      attribute :start, Integer
+      attribute :docs, Array[Document]
+    end
+  end
+end

data/lib/archivist/models/document.rb

@@ -0,0 +1,48 @@
+require 'virtus'
+require 'date'
+
+module Archivist
+  module Model
+    class Document
+      include Virtus.model
+
+      attribute :identifier, String
+      attribute :title, String
+      attribute :date, Date
+      attribute :languages, Array[String]
+      attribute :creators, Array[String]
+
+      def initialize
+        @conn = Faraday.new(url: "http://archive.org") do |faraday|
+          faraday.use FaradayMiddleware::FollowRedirects
+          faraday.request  :url_encoded             # form-encode POST params
+          # faraday.response :logger                  # log requests to STDOUT
+          faraday.adapter  Faraday.default_adapter  # make requests with Net::HTTP
+        end
+      end
+
+      def format_index
+        response = @conn.get(index_xml_path)
+        Model::FormatIndex.new.tap do |idx|
+          Representation::FormatIndex.new(idx).from_xml(response.body)
+        end
+      end
+
+      def download(format=:text)
+        # e.g. format_index.text_format
+        file_format = format_index.send(:"#{format}_format")
+        # e.g. /download/firstbooknapole00gruagoog/firstbooknapole00gruagoog_djvu.txt
+        @conn.get(download_path(file_format.name)).
+          body.force_encoding('UTF-8')
+      end
+
+      def download_path(file)
+        "/download/#{identifier}/#{file}"
+      end
+
+      def index_xml_path
+        download_path("#{identifier}_files.xml")
+      end
+    end
+  end
+end

data/lib/archivist/models/file_format.rb

@@ -0,0 +1,18 @@
+require 'virtus'
+
+module Archivist
+  module Model
+    class FileFormat
+      include Virtus.model
+
+      attribute :name
+      attribute :source
+      attribute :format
+      attribute :mtime
+      attribute :size
+      attribute :md5
+      attribute :crc32
+      attribute :sha1
+    end
+  end
+end

data/lib/archivist/models/format_index.rb

@@ -0,0 +1,20 @@
+require 'virtus'
+
+module Archivist
+  module Model
+    class FormatIndex
+      include Virtus.model
+
+      attribute :formats
+
+      def find(key)
+        formats.find{ |f| f.format == key }
+      end
+      alias_method :[], :find
+
+      def text_format
+        find("DjVuTXT")
+      end
+    end
+  end
+end

data/lib/archivist/models/header.rb

@@ -0,0 +1,15 @@
+require 'virtus'
+
+require_relative 'parameters'
+
+module Archivist
+  module Model
+    class Header
+      include Virtus.model
+
+      attribute :status, Integer
+      attribute :qtime, Integer
+      attribute :params, Parameters
+    end
+  end
+end

data/lib/archivist/models/parameters.rb

@@ -0,0 +1,17 @@
+require 'virtus'
+
+module Archivist
+  module Model
+    class Parameters
+      include Virtus.model
+
+      attribute :rows, Integer
+      attribute :format, String
+      attribute :start, Integer
+      attribute :q, String
+      attribute :qin, String
+      attribute :fl, String
+      attribute :sort, String
+    end
+  end
+end

data/lib/archivist/models/query_response.rb

@@ -0,0 +1,23 @@
+require 'virtus'
+
+require_relative 'header'
+require_relative 'body'
+
+module Archivist
+  module Model
+    class QueryResponse
+      include Virtus.model
+
+      attribute :header, Header
+      attribute :body, Body
+
+      def each(&block)
+        body && body.docs.each(&block)
+      end
+
+      def empty?
+        body && body.docs.empty?
+      end
+    end
+  end
+end

data/lib/archivist/representations.rb

@@ -0,0 +1,9 @@
+require 'archivist/models'
+
+require 'archivist/representations/body'
+require 'archivist/representations/document'
+require 'archivist/representations/file_format'
+require 'archivist/representations/format_index'
+require 'archivist/representations/header'
+require 'archivist/representations/parameters'
+require 'archivist/representations/query_response'

data/lib/archivist/representations/body.rb

@@ -0,0 +1,19 @@
+require 'archivist/models'
+require 'representable'
+require 'representable/json'
+
+require_relative 'document'
+
+module Archivist
+  module Representation
+    class Body < Representable::Decorator
+      include Representable::JSON
+
+      property :num_found, as: "numFound"
+      property :start
+      collection :docs,
+                 class: Model::Document,
+                 decorator: Document
+    end
+  end
+end

data/lib/archivist/representations/document.rb

@@ -0,0 +1,17 @@
+require 'archivist/models'
+require 'representable'
+require 'representable/json'
+
+module Archivist
+  module Representation
+    class Document < Representable::Decorator
+      include Representable::JSON
+
+      property :identifier
+      property :title
+      property :date
+      collection :languages, as: "language"
+      collection :creators, as: "creator"
+    end
+  end
+end