filerary 0.0.1 → 0.0.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: e0e03b2bc1bcc04674c5c134c54fbcd6385cda9d
-  data.tar.gz: ca7cb1b5ebfb1ee1b3cafbf77dbb6f11ebff18c1
+  metadata.gz: 6cb82ae7cf6e47684a98d9fd942ed055f624b9cf
+  data.tar.gz: 827a84c3b79c5e50300cc814edcd849b927b047f
 SHA512:
-  metadata.gz: 3860577c8ba2591d8b506dd22e5aa668f255f006c7e7e99541e916a4b0bb4fcc68d8768646da7ee4c90a980b85f8c6db74700fb29eccab40c256bb3ae71c494b
-  data.tar.gz: a5a6bc5c8b8fbbfbb274e62887a29f7ca825a505533d4bf3917ec8a4c2427d914aaa4f71c9685d6ccdd9209fcdd613ab55d9120a8a2b5bdecc5cc0a780df4e9c
+  metadata.gz: 0822bb86cbe996ef71a856bd54a3ca6fb504527b4fa3e818ccaf4bc79edc04a235da33cb959246ed9961bd341040ccfea7be3c3922f90e62c7061f2fc256d4c1
+  data.tar.gz: a7ae84144ba9abda1bfd1ce4c8826aa5f44bd1da9ed8914d7692d7349b46253fbf55b844d889ae1f7fba00ec808dd3db8ec067bd3f125ab2c842390296eb3151

data/lib/filerary/librarian.rb

@@ -1,6 +1,9 @@
 require "fileutils"
+require "uri"
 require "grn_mini"
 require "chupa-text"
+gem "chupa-text-decomposer-pdf"
+gem "chupa-text-decomposer-libreoffice"
 
 module Filerary
   class Librarian
@@ -60,14 +63,20 @@ module Filerary
       extractor = ChupaText::Extractor.new
       extractor.apply_configuration(ChupaText::Configuration.default)
 
-      extractor.extract(path) do |text_data|
-        text = text_data.body
+      begin
+        extractor.extract(URI.encode(path)) do |text_data|
+          text = text_data.body
+        end
+      rescue URI::InvalidURIError
+        return path
       end
 
-      # TODO: I want to specify encoding in ChupaText side.
-      text.force_encoding(Encoding.default_external || "UTF-8")
+      return path unless text
 
-      text
+      # TODO: I want to specify encoding in ChupaText side.
+      text.force_encoding("UTF-8")
+      return text if text.valid_encoding?
+      text.force_encoding(Encoding.default_external)
     end
   end
 end

data/lib/filerary/version.rb

@@ -1,3 +1,3 @@
 module Filerary
-  VERSION = "0.0.1"
+  VERSION = "0.0.2"
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: filerary
 version: !ruby/object:Gem::Version
-  version: 0.0.1
+  version: 0.0.2
 platform: ruby
 authors:
 - Masafumi Yokoyama