doc_image_extract 0.0.1 → 0.0.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: f209d30d130de8731ee02197a75cab6f5b3de928
-  data.tar.gz: 7134763b289f13abba156eeb2c514c03abdfa909
+  metadata.gz: 24f47e46b7193367b1bc79efbdffb62f212b13a0
+  data.tar.gz: 35ae2d5a3b4ef9f7592ceabb358f53406f651376
 SHA512:
-  metadata.gz: 1b979d2af4f5ee96ee7f9e6cd0970b7d09cfa6e64d2ad59999c03c24f33cf36a8d4b7e702f3d0ee07a13f60392cc04558f167e0c80a428410f8dec805133daff
-  data.tar.gz: ad499558865070e2f2342499caff1df8b4cb5c656eb90dcbb9da0c884382cd038ae831a5d5d73fcb49926fe8c5698598da0073cc6a9c8ab02f037c13b2718cf2
+  metadata.gz: 84db9d82b9ac6a3fefdcc2fa080c71d9175e8e36788db03279f5bca6b6f77bef65a3ad23d923e7c2849a7a348079ed28381690dc2dd4741af284b32e4ffed7cb
+  data.tar.gz: 8378fbfb6a5e586c9d119e5a107e5c03aa66efff1c0ec8a704239b3731bbc9b32c37ae49b9207b5fa19773a355b71dcf9d3b64531aab2357d40f06dbe85cbabe

data/doc_image_extract.gemspec

@@ -1,11 +1,10 @@
 # coding: utf-8
 lib = File.expand_path('../lib', __FILE__)
 $LOAD_PATH.unshift(lib) unless $LOAD_PATH.include?(lib)
-require 'doc_image_extract/version'
 
 Gem::Specification.new do |spec|
   spec.name          = "doc_image_extract"
-  spec.version       = DocImageExtract::VERSION
+  spec.version       = "0.0.2"
   spec.authors       = ["Eric Barendt"]
   spec.email         = ["ebarendt@gmail.com"]
   spec.summary       = %q{Extracts all images from a given Word document.}

data/lib/doc_image_extract.rb

@@ -1,6 +1,7 @@
-require "doc_image_extract/version"
 require "doc_image_extract/document"
 require "doc_image_extract/exporter"
+require "doc_image_extract/reader"
+require "doc_image_extract/picture"
 
 module DocImageExtract
 end

data/lib/doc_image_extract/document.rb

@@ -1,21 +1,23 @@
-require "poi"
-
 module DocImageExtract
   class Document
-    include Java
-    java_import "org.apache.poi.hwpf.HWPFDocument"
-    java_import "java.io.FileInputStream"
-
     attr_reader :file
 
     def initialize file
       @file = file
     end
 
+    def reader
+      @reader ||= begin
+        if file =~ /docx\Z/
+          Reader::Xwpf.new file
+        else
+          Reader::Hwpf.new file
+        end
+      end
+    end
+
     def pictures
-      document = HWPFDocument.new FileInputStream.new file
-      pictures_table = document.get_pictures_table
-      pictures_table.get_all_pictures
+      reader.pictures
     end
   end
-end
+end

data/lib/doc_image_extract/exporter.rb

@@ -1,7 +1,9 @@
 module DocImageExtract
   class Exporter
     def save picture
-      File.write picture.suggest_full_file_name, picture.get_content
+      if picture.file_name
+        File.write picture.file_name, picture.data
+      end
     end
   end
-end
+end

data/lib/doc_image_extract/picture.rb

@@ -0,0 +1,10 @@
+module DocImageExtract
+  class Picture
+    attr_reader :file_name, :data
+
+    def initialize(file_name, data)
+      @file_name = file_name
+      @data = data
+    end
+  end
+end

data/lib/doc_image_extract/reader.rb

@@ -0,0 +1,2 @@
+require_relative 'reader/hwpf'
+require_relative 'reader/xwpf'

data/lib/doc_image_extract/reader/hwpf.rb

@@ -0,0 +1,26 @@
+require "poi"
+
+module DocImageExtract
+  module Reader
+    class Hwpf
+      java_import java.io.FileInputStream
+      java_import org.apache.poi.hwpf.HWPFDocument
+
+      attr_reader :file
+
+      def initialize(file)
+        @file = file
+      end
+
+      def pictures
+        document = HWPFDocument.new FileInputStream.new file
+        pictures_table = document.get_pictures_table
+        pictures_table.get_all_pictures.map do |picture|
+          filename = picture.suggest_full_file_name
+          data = picture.get_content
+          DocImageExtract::Picture.new filename, data
+        end
+      end
+    end
+  end
+end

data/lib/doc_image_extract/reader/xwpf.rb

@@ -0,0 +1,26 @@
+require "poi"
+
+module DocImageExtract
+  module Reader
+    class Xwpf
+      java_import java.io.FileInputStream
+      java_import org.apache.poi.xwpf.usermodel.XWPFDocument
+
+      attr_reader :file
+
+      def initialize(file)
+        @file = file
+      end
+
+      def pictures
+        document = XWPFDocument.new FileInputStream.new file
+        pictures = document.get_all_pictures
+        pictures.map do |picture|
+          filename = picture.get_file_name
+          data = picture.get_data
+          Picture.new filename, data
+        end
+      end
+    end
+  end
+end

data/lib/poi.rb

@@ -1,2 +1,8 @@
-require "jars/poi-3.9-20121203.jar"
-require "jars/poi-scratchpad-3.9-20121203.jar"
+require "java"
+require "jars/poi-3.10-FINAL-20140208.jar"
+require "jars/poi-scratchpad-3.10-FINAL-20140208.jar"
+require "jars/poi-ooxml-3.10-FINAL-20140208.jar"
+require "jars/poi-ooxml-schemas-3.10-FINAL-20140208.jar"
+require "jars/ooxml-lib/dom4j-1.6.1.jar"
+require "jars/ooxml-lib/stax-api-1.0.1.jar"
+require "jars/ooxml-lib/xmlbeans-2.3.0.jar"

metadata

@@ -1,43 +1,43 @@
 --- !ruby/object:Gem::Specification
 name: doc_image_extract
 version: !ruby/object:Gem::Version
-  version: 0.0.1
+  version: 0.0.2
 platform: ruby
 authors:
 - Eric Barendt
-autorequire: 
+autorequire:
 bindir: bin
 cert_chain: []
-date: 2014-04-27 00:00:00.000000000 Z
+date: 2014-05-02 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler
-  requirement: !ruby/object:Gem::Requirement
+  version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - '>='
       - !ruby/object:Gem::Version
         version: '0'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
+  requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - '>='
       - !ruby/object:Gem::Version
         version: '0'
+  prerelease: false
+  type: :development
 - !ruby/object:Gem::Dependency
   name: rake
-  requirement: !ruby/object:Gem::Requirement
+  version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - '>='
       - !ruby/object:Gem::Version
         version: '0'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
+  requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - '>='
       - !ruby/object:Gem::Version
         version: '0'
+  prerelease: false
+  type: :development
 description: Extracts all images from a given Word document.
 email:
 - ebarendt@gmail.com
@@ -46,7 +46,7 @@ executables:
 extensions: []
 extra_rdoc_files: []
 files:
-- ".gitignore"
+- .gitignore
 - Gemfile
 - LICENSE.txt
 - README.md
@@ -56,32 +56,40 @@ files:
 - lib/doc_image_extract.rb
 - lib/doc_image_extract/document.rb
 - lib/doc_image_extract/exporter.rb
-- lib/doc_image_extract/version.rb
-- lib/jars/poi-3.9-20121203.jar
-- lib/jars/poi-scratchpad-3.9-20121203.jar
+- lib/doc_image_extract/picture.rb
+- lib/doc_image_extract/reader.rb
+- lib/doc_image_extract/reader/hwpf.rb
+- lib/doc_image_extract/reader/xwpf.rb
+- lib/jars/ooxml-lib/dom4j-1.6.1.jar
+- lib/jars/ooxml-lib/stax-api-1.0.1.jar
+- lib/jars/ooxml-lib/xmlbeans-2.3.0.jar
+- lib/jars/poi-3.10-FINAL-20140208.jar
+- lib/jars/poi-ooxml-3.10-FINAL-20140208.jar
+- lib/jars/poi-ooxml-schemas-3.10-FINAL-20140208.jar
+- lib/jars/poi-scratchpad-3.10-FINAL-20140208.jar
 - lib/poi.rb
 homepage: ''
 licenses:
 - MIT
 metadata: {}
-post_install_message: 
+post_install_message:
 rdoc_options: []
 require_paths:
 - lib
 required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
-  - - ">="
+  - - '>='
     - !ruby/object:Gem::Version
       version: '0'
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
-  - - ">="
+  - - '>='
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubyforge_project: 
-rubygems_version: 2.2.2
-signing_key: 
+rubyforge_project:
+rubygems_version: 2.1.9
+signing_key:
 specification_version: 4
 summary: Extracts all images from a given Word document.
 test_files: []

data/lib/doc_image_extract/version.rb

@@ -1,3 +0,0 @@
-module DocImageExtract
-  VERSION = "0.0.1"
-end