libis-format 1.3.8.1 → 2.0.0

metadata

@@ -1,114 +1,87 @@
 --- !ruby/object:Gem::Specification
 name: libis-format
 version: !ruby/object:Gem::Version
-  version: 1.3.8.1
+  version: 2.0.0
 platform: ruby
 authors:
 - Kris Dekeyser
+autorequire: 
 bindir: bin
 cert_chain: []
-date: 1980-01-02 00:00:00.000000000 Z
+date: 2020-11-09 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
-  name: chromaprint
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: 0.0.2
-  type: :runtime
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: 0.0.2
-- !ruby/object:Gem::Dependency
-  name: deep_dive
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '0.3'
-  type: :runtime
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - "~>"
-      - !ruby/object:Gem::Version
-        version: '0.3'
-- !ruby/object:Gem::Dependency
-  name: libis-tools
+  name: rake
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.1'
-  type: :runtime
+        version: '10.3'
+  type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.1'
+        version: '10.3'
 - !ruby/object:Gem::Dependency
-  name: mail
+  name: rspec
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '2.9'
-  type: :runtime
+        version: '3.1'
+  type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '2.9'
+        version: '3.1'
 - !ruby/object:Gem::Dependency
-  name: mini_magick
+  name: awesome_print
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ">="
       - !ruby/object:Gem::Version
-        version: 5.0.1
-  type: :runtime
+        version: '0'
+  type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ">="
       - !ruby/object:Gem::Version
-        version: 5.0.1
+        version: '0'
 - !ruby/object:Gem::Dependency
-  name: msg_extractor
+  name: equivalent-xml
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '0.1'
-  type: :runtime
+        version: '0.5'
+  type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '0.1'
+        version: '0.5'
 - !ruby/object:Gem::Dependency
-  name: naturally
+  name: nokogiri
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ">="
       - !ruby/object:Gem::Version
-        version: '2.2'
-  type: :runtime
+        version: '0'
+  type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - ">="
       - !ruby/object:Gem::Version
-        version: '2.2'
+        version: '0'
 - !ruby/object:Gem::Dependency
-  name: new_rfc_2047
+  name: libis-tools
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
@@ -125,156 +98,109 @@ dependencies:
   name: os
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - '='
       - !ruby/object:Gem::Version
-        version: '1.1'
+        version: 0.9.6
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - "~>"
+    - - '='
       - !ruby/object:Gem::Version
-        version: '1.1'
+        version: 0.9.6
 - !ruby/object:Gem::Dependency
-  name: pdfinfo
+  name: mini_magick
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.4'
+        version: '4.3'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.4'
+        version: '4.3'
 - !ruby/object:Gem::Dependency
-  name: pdfkit
+  name: deep_dive
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '0.8'
+        version: '0.3'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '0.8'
+        version: '0.3'
 - !ruby/object:Gem::Dependency
-  name: word_wrap
+  name: chromaprint
   requirement: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.0'
+        version: 0.0.2
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: '1.0'
-- !ruby/object:Gem::Dependency
-  name: awesome_print
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-- !ruby/object:Gem::Dependency
-  name: equivalent-xml
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-- !ruby/object:Gem::Dependency
-  name: rake
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-- !ruby/object:Gem::Dependency
-  name: rspec
-  requirement: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
-  type: :development
-  prerelease: false
-  version_requirements: !ruby/object:Gem::Requirement
-    requirements:
-    - - ">="
-      - !ruby/object:Gem::Version
-        version: '0'
+        version: 0.0.2
 - !ruby/object:Gem::Dependency
-  name: nokogiri
+  name: naturally
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
-  type: :development
+        version: '2.1'
+  type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
-description: Collection of tools and classes that help to identify file formats and
-  create derivative copies.
+        version: '2.1'
+description: Collection of tools and classes that help to identify formats of binary
+  files and create derivative copies (e.g. PDF from Word).
 email:
 - kris.dekeyser@libis.be
 executables:
-- droid_tool
-- fido_tool
+- droid
+- fido
 - formatinfo
 - libis_format
-- pdf_tool
+- pdf_copy
 extensions: []
 extra_rdoc_files: []
 files:
+- ".coveralls.yml"
+- ".gitignore"
+- ".travis.yml"
 - Gemfile
 - LICENSE.txt
 - README.md
-- bin/droid_tool
-- bin/fido_tool
+- Rakefile
+- base/Dockerfile
+- base/Dockerfile.alpine
+- base/Dockerfile.rvm
+- base/rework_path
+- bin/droid
+- bin/fido
 - bin/formatinfo
 - bin/libis_format
-- bin/pdf_tool
-- data/AdobeRGB1998.icc
+- bin/pdf_copy
 - data/ISOcoated_v2_eci.icc
 - data/PDFA_def.ps
 - data/ead.xsd
+- data/eciRGB_v2.icc
 - data/lias_formats.xml
 - data/types.yml
 - data/xlink.xsd
+- docker_cfg.yml
 - lib/libis-format.rb
 - lib/libis/format.rb
 - lib/libis/format/cli/convert.rb
@@ -287,39 +213,43 @@ files:
 - lib/libis/format/converter/audio_converter.rb
 - lib/libis/format/converter/base.rb
 - lib/libis/format/converter/chain.rb
-- lib/libis/format/converter/email_converter.rb
 - lib/libis/format/converter/fop_pdf_converter.rb
+- lib/libis/format/converter/image_assembler.rb
 - lib/libis/format/converter/image_converter.rb
+- lib/libis/format/converter/image_splitter.rb
+- lib/libis/format/converter/image_watermarker.rb
 - lib/libis/format/converter/jp2_converter.rb
 - lib/libis/format/converter/office_converter.rb
+- lib/libis/format/converter/pdf_assembler.rb
 - lib/libis/format/converter/pdf_converter.rb
+- lib/libis/format/converter/pdf_optimizer.rb
+- lib/libis/format/converter/pdf_splitter.rb
+- lib/libis/format/converter/pdf_watermarker.rb
 - lib/libis/format/converter/repository.rb
 - lib/libis/format/converter/spreadsheet_converter.rb
 - lib/libis/format/converter/video_converter.rb
 - lib/libis/format/converter/xslt_converter.rb
 - lib/libis/format/identifier.rb
+- lib/libis/format/info.rb
+- lib/libis/format/library.rb
 - lib/libis/format/tool.rb
 - lib/libis/format/tool/droid.rb
-- lib/libis/format/tool/eml_to_pdf.rb
 - lib/libis/format/tool/extension_identification.rb
-- lib/libis/format/tool/ff_mpeg.rb
+- lib/libis/format/tool/ffmpeg.rb
 - lib/libis/format/tool/fido.rb
 - lib/libis/format/tool/file_tool.rb
 - lib/libis/format/tool/fop_pdf.rb
 - lib/libis/format/tool/identification_tool.rb
-- lib/libis/format/tool/mail_to_pdf.rb
-- lib/libis/format/tool/msg_to_pdf.rb
 - lib/libis/format/tool/office_to_pdf.rb
+- lib/libis/format/tool/pdf_copy.rb
 - lib/libis/format/tool/pdf_merge.rb
 - lib/libis/format/tool/pdf_optimizer.rb
 - lib/libis/format/tool/pdf_split.rb
 - lib/libis/format/tool/pdf_to_pdfa.rb
-- lib/libis/format/tool/pdf_tool.rb
 - lib/libis/format/tool/pdfa_validator.rb
 - lib/libis/format/tool/spreadsheet_to_ods.rb
-- lib/libis/format/type_database.rb
-- lib/libis/format/type_database_impl.rb
 - lib/libis/format/version.rb
+- lib/libis/format/yaml_loader.rb
 - libis-format.gemspec
 - tools/PdfTool.jar
 - tools/bcpkix-jdk15on-1.49.jar
@@ -374,13 +304,13 @@ files:
 - tools/fop/lib/xmlgraphics-commons-2.3.jar
 - tools/fop/lib/xmlgraphics-commons.LICENSE.txt
 - tools/fop/lib/xmlgraphics-commons.NOTICE.txt
-- tools/pdf2pdfa
-- tools/pdfbox/pdfbox-app-3.0.3.jar
-- tools/pdfbox/preflight-app-3.0.3.jar
+- tools/pdfbox/pdfbox-app-2.0.13.jar
+- tools/pdfbox/preflight-app-2.0.13.jar
 homepage: ''
 licenses:
 - MIT
 metadata: {}
+post_install_message: 
 rdoc_options: []
 require_paths:
 - lib
@@ -388,14 +318,15 @@ required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
-      version: '3.2'
+      version: '0'
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 4.0.8
+rubygems_version: 3.0.8
+signing_key: 
 specification_version: 4
 summary: LIBIS File format format services.
 test_files: []

data/lib/libis/format/converter/email_converter.rb

@@ -1,51 +0,0 @@
-# frozen_string_literal: true
-
-require_relative 'base'
-
-require 'libis/format/tool/msg_to_pdf'
-require 'libis/format/tool/eml_to_pdf'
-require 'libis/format/type_database'
-require 'rexml/document'
-
-module Libis
-  module Format
-    module Converter
-      class EmailConverter < Libis::Format::Converter::Base
-        def self.input_types
-          %i[MSG EML]
-        end
-
-        def self.output_types(format = nil)
-          return [] unless input_types.include?(format)
-
-          %i[PDF]
-        end
-
-        def email_convert(_)
-          # force usage of this converter
-        end
-
-        def convert(source, target, format, opts = {})
-          super
-
-          tool = tool_for_format(opts, source)
-          tool.run(source, target)
-        rescue StandardError => e
-          { command: { status: -1 }, errors: [{ error: e.message, error_class: e.class, error_trace: e.backtrace }] }
-        end
-
-        private
-
-        def tool_for_format(opts, source)
-          if opts[:source_format] == :MSG || File.extname(source).casecmp('.msg').zero?
-            Format::Tool::MsgToPdf
-          elsif opts[:source_format] == :EML || File.extname(source).casecmp('.eml').zero?
-            Format::Tool::EmlToPdf
-          else
-            raise "Unsupported file extension #{File.extname(source)}"
-          end
-        end
-      end
-    end
-  end
-end

data/lib/libis/format/tool/eml_to_pdf.rb

@@ -1,168 +0,0 @@
-# rubocop:disable Style/Documentation, Metrics/*
-# frozen_string_literal: true
-
-require_relative 'mail_to_pdf'
-
-require 'mail'
-require 'word_wrap'
-
-module Libis
-  module Format
-    module Tool
-      class EmlToPdf < MailToPdf
-        protected
-
-        def open_email(source)
-          eml = File.read(source)
-          begin
-            msg = Mail.new(eml)
-          rescue StandardError => e
-            eml.force_encoding('ASCII-8BIT')
-            msg = Mail.new(eml)
-            @warnings << "Failed to parse message with default encoding, forced ASCII-8BIT: #{e.message}"
-          end
-          msg
-        rescue StandardError => e
-          raise "Failed to open message: #{e.message}"
-        end
-
-        def close_email(_msg)
-          true
-        end
-
-        def get_body_html(msg)
-          # Get the encoding
-
-          if !msg.multipart?
-            HTML_WRAPPER_TEMPLATE % WordWrap.ww(msg.decoded, 120, false)
-
-          elsif msg.html_part
-            body = msg.html_part.body.decoded
-
-            # if the encoding is not UTF-8, then the HTML may contain metadata that specifies the encoding
-            # the browser will use that metadata to render the HTML correctly, so we should not force it to UTF-8
-            encoding = body.scan(/<\?xml\s[^?>]*encoding="([^"]*)"[^?>]*\?>/).flatten.first
-            encoding ||= body.scan(/<meta\s+[^>]*charset=["']?([^"'>\s]+)["']?[^>]*>/).flatten.first
-            encoding ||= msg.html_part.charset
-            encoding ||= msg.content_type_parameters['charset'] || msg.charset || 'UTF-8'
-
-            body.force_encoding(encoding) unless encoding.casecmp(body.encoding.name).zero?
-
-            body
-          elsif msg.text_part
-            HTML_WRAPPER_TEMPLATE % msg.text_part.decoded
-
-          else
-            HTML_WRAPPER_TEMPLATE % ''
-
-          end
-        end
-
-        def get_subject(msg)
-          msg.subject || 'No Subject'
-        end
-
-        def get_headers(msg)
-          headers = {}
-          html = ''
-
-          field_list = msg.header_fields
-
-          %w[From To Cc Subject Date].each do |key|
-            value = find_hdr(field_list, key)
-            next unless value
-
-            if value.is_a? Time
-              begin
-                headers[key.downcase.to_sym] = value.iso8601
-                html += hdr_html(key, value.rfc2822)
-              rescue StandardError => e
-                logger.warn "Failed to parse date header '#{value}': #{e.message}"
-              end
-            else
-              headers[key.downcase.to_sym] = value
-              html += hdr_html(key, value)
-            end
-          end
-
-          [headers, html]
-        end
-
-        def get_inline_attachment_data(attachments, cid)
-          attachments.each do |attachment|
-            next unless attachment.has_content_id?
-            next unless attachment.cid == cid
-
-            begin
-              attachment.data.rewind
-            rescue NoMethodError
-              # do nothing, attachment.data is not a stream
-            end
-            return {
-              mime_type: attachment.mime_type,
-              base64: Base64.strict_encode64(attachment.read)
-            }
-          end
-          nil
-        end
-
-        def get_file_attachments(attachments, _used_files)
-          attachments.select do |attachment|
-            !attachment.inline? &&
-              !attachment.has_content_id? &&
-              attachment.mime_type != 'message/rfc822'
-          end
-        end
-
-        def get_mail_attachments(attachments)
-          attachments.select do |attachment|
-            !attachment.inline? &&
-              !attachment.has_content_id? &&
-              attachment.mime_type == 'message/rfc822'
-          end
-        end
-
-        def get_attachment_info(attachment)
-          if attachment.mime_type == 'message/rfc822'
-            sub_msg = Mail.new(attachment.body.to_s)
-            subject = sub_msg.subject.to_s.strip
-
-            {
-              embedded_msg: sub_msg,
-              filename: subject
-            }
-
-          elsif attachment.filename
-
-            {
-              data: attachment.decoded,
-              filename: attachment.filename
-            }
-
-          else
-            {
-              filename: attachment.content_id.to_s
-            }
-          end
-        end
-
-        private
-
-        def find_hdr(list, key)
-          hdr = list.find { |x| x.name.to_s =~ /^#{key}$/i }
-          return nil unless hdr
-
-          field = hdr.field
-          return hdr.value unless field
-
-          return field.decoded unless field.is_a? Mail::CommonDateField
-
-          return field.date_time.to_time.localtime if field.respond_to?(:date_time)
-
-          DateTime.parse(field.decoded).to_time.localtime
-        end
-      end
-    end
-  end
-end
-# rubocop:enable Style/Documentation, Metrics/*