chupa-text-decomposer-pdf 1.0.8 → 1.1.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 206b91fbb36ea3f1df98fbc331c6216d5bcd4463c08f3fd17eae9084fe111fb0
-  data.tar.gz: 321fb688ddfd596037bf93b890869a8c7b1795f38a84c938fd3971889d7e443a
+  metadata.gz: 3d1885ca60dea6d4d7ffb8b4c3aa1f35dda7a1619577bad2b2526bd7d7db6180
+  data.tar.gz: 7d2ccc5f1dd135a5ddf281cf9d2c31215dbcad4943c12f0f03cea32e277a7a2f
 SHA512:
-  metadata.gz: 22b2d8cf3b943b1aa42ee4cab84968979fc3b7e902f2d2b62f803bb537a127e2a37d8c218b6002f17d87039dc366b714b5a0b5d2aa6d0ebfd90fcd98bc7954df
-  data.tar.gz: d5fb58450ac452e7e88e5c3931181c268051df432ddefaa4073eb1cc5baa3af21e1f1615cca2ee94cf3216008d2706856468081c623d6010f7b7e55ff5d473ff
+  metadata.gz: 6ab4e283c338c8f6a9157912781efa345c35c2eb39b5ee08cbdd293e9570deddb4654c138dc704af1093e1714657b7392bf63f0685eadd467ca86d480988af31
+  data.tar.gz: 2745efd44d1e686b53f49f6569a4d9ebcccd4937fb01ea544d1bbc580d8cefa4f958f9cde3e6bd21453a5e29889340f2662655bd09e9fdb36cc1cb6d55651703

data/chupa-text-decomposer-pdf.gemspec

@@ -22,7 +22,7 @@ end
 
 Gem::Specification.new do |spec|
   spec.name = "chupa-text-decomposer-pdf"
-  spec.version = "1.0.8"
+  spec.version = "1.1.0"
   spec.homepage = "https://github.com/ranguba/chupa-text-decomposer-pdf"
   spec.authors = ["Kouhei Sutou"]
   spec.email = ["kou@clear-code.com"]
@@ -39,7 +39,7 @@ Gem::Specification.new do |spec|
   spec.files += Dir.glob("doc/text/*")
   spec.files += Dir.glob("test/**/*")
 
-  spec.add_runtime_dependency("chupa-text", ">= 1.1.8")
+  spec.add_runtime_dependency("chupa-text", ">= 1.1.9")
   spec.add_runtime_dependency("poppler")
 
   spec.add_development_dependency("bundler")

data/doc/text/news.md

@@ -1,5 +1,13 @@
 # News
 
+## 1.0.9: 2019-03-03
+
+### Improvements
+
+  * Changed to use log instead of exception for invalid PDF.
+
+  * Reduced memory usage.
+
 ## 1.0.8: 2019-03-03
 
 ### Improvements

data/lib/chupa-text/decomposers/pdf.rb

@@ -22,6 +22,8 @@ require "poppler"
 module ChupaText
   module Decomposers
     class PDF < Decomposer
+      include Loggable
+
       registry.register("pdf", self)
 
       def target?(data)
@@ -37,6 +39,8 @@ module ChupaText
 
       def decompose(data)
         document = create_document(data)
+        return if document.nil?
+
         text = ""
         document.each do |page|
           page_text = page.get_text
@@ -75,7 +79,9 @@ module ChupaText
         if path.nil?
           file = Tempfile.new(["chupa-text-decomposer-pdf", ".pdf"])
           file.binmode
-          file.write(data.body)
+          data.open do |input|
+            IO.copy_stream(input, file)
+          end
           file.close
           path = file.path
         end
@@ -85,8 +91,14 @@ module ChupaText
           end
         rescue Poppler::Error::Encrypted
           raise ChupaText::EncryptedError.new(data)
-        rescue GLib::Error => error
-          raise ChupaText::InvalidDataError.new(data, error.message)
+        rescue Poppler::Error => poppler_error
+          error do
+            message = "#{log_tag} Failed to process PDF: "
+            message << "#{poppler_error.class}: #{poppler_error.message}\n"
+            message << poppler_error.backtrace.join("\n")
+            message
+          end
+          nil
         end
       end
 
@@ -146,6 +158,10 @@ module ChupaText
 
         Screenshot.new("image/png", [png.string].pack("m*"), "base64")
       end
+
+      def log_tag
+        "[decomposer][pdf]"
+      end
     end
   end
 end

data/test/fixture/empty.pdf

@@ -0,0 +1 @@
+%PDF-1

data/test/test-pdf.rb

@@ -1,4 +1,4 @@
-# Copyright (C) 2013-2017  Kouhei Sutou <kou@clear-code.com>
+# Copyright (C) 2013-2019  Kouhei Sutou <kou@clear-code.com>
 #
 # This library is free software; you can redistribute it and/or
 # modify it under the terms of the GNU Lesser General Public
@@ -32,6 +32,13 @@ class TestPDF < Test::Unit::TestCase
     base_path.join(*components)
   end
 
+  def capture_log(&block)
+    ChupaText::CaptureLogger.capture(&block).collect do |level, message|
+      message = message.split("\n", 2)[0]
+      [level, message]
+    end
+  end
+
   sub_test_case("target?") do
     sub_test_case("extension") do
       def create_data(uri)
@@ -212,5 +219,22 @@ class TestPDF < Test::Unit::TestCase
         end
       end
     end
+
+    sub_test_case("invalid") do
+      def test_empty
+        messages = capture_log do
+          assert_equal([],
+                       decompose(fixture_path("empty.pdf")).collect(&:body))
+        end
+        assert_equal([
+                       [
+                         :error,
+                         "[decomposer][pdf] Failed to process PDF: " +
+                         "Poppler::Error::Damaged: PDF document is damaged",
+                       ],
+                     ],
+                     messages)
+      end
+    end
   end
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: chupa-text-decomposer-pdf
 version: !ruby/object:Gem::Version
-  version: 1.0.8
+  version: 1.1.0
 platform: ruby
 authors:
 - Kouhei Sutou
@@ -16,14 +16,14 @@ dependencies:
     requirements:
     - - ">="
       - !ruby/object:Gem::Version
-        version: 1.1.8
+        version: 1.1.9
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - ">="
       - !ruby/object:Gem::Version
-        version: 1.1.8
+        version: 1.1.9
 - !ruby/object:Gem::Dependency
   name: poppler
   requirement: !ruby/object:Gem::Requirement
@@ -143,6 +143,7 @@ files:
 - lib/chupa-text/decomposers/pdf.rb
 - test/fixture/attributes.odt
 - test/fixture/attributes.pdf
+- test/fixture/empty.pdf
 - test/fixture/encrypted.odt
 - test/fixture/encrypted.pdf
 - test/fixture/multi-pages.odt