rika 0.9.4-java → 0.9.5-java
Sign up to get free protection for your applications and to get access to all the features.
- data/lib/rika.rb +24 -22
- data/lib/rika/version.rb +1 -1
- data/spec/rika_spec.rb +0 -4
- metadata +2 -2
data/lib/rika.rb
CHANGED
@@ -1,3 +1,5 @@
|
|
1
|
+
# encoding: utf-8
|
2
|
+
|
1
3
|
raise "You need to run JRuby to use Rika" unless RUBY_PLATFORM =~ /java/
|
2
4
|
|
3
5
|
require "rika/version"
|
@@ -13,30 +15,32 @@ end
|
|
13
15
|
module Rika
|
14
16
|
import org.apache.tika.metadata.Metadata
|
15
17
|
import org.apache.tika.Tika
|
18
|
+
import java.io.FileInputStream
|
19
|
+
import java.net.URL
|
16
20
|
|
17
21
|
class Parser
|
18
22
|
|
19
|
-
def initialize(
|
20
|
-
|
21
|
-
@uri = uri
|
23
|
+
def initialize(file_location, max_content_length = -1)
|
24
|
+
@uri = file_location
|
22
25
|
@tika = Tika.new
|
23
26
|
@tika.set_max_string_length(max_content_length)
|
24
27
|
@metadata = Metadata.new
|
28
|
+
|
29
|
+
@is_file = File.exists?(@uri)
|
30
|
+
is_http = URI(@uri).scheme == "http" && Net::HTTP.get_response(URI(@uri)).is_a?(Net::HTTPSuccess) if !@is_file
|
25
31
|
|
26
|
-
if
|
27
|
-
self.parse_file
|
28
|
-
elsif p.parse(@uri).scheme == 'http' || p.parse(@uri).scheme == 'https'
|
29
|
-
self.parse_url
|
30
|
-
else
|
32
|
+
if !@is_file && !is_http
|
31
33
|
raise IOError, "File does not exist or can't be reached."
|
32
34
|
end
|
33
35
|
end
|
34
36
|
|
35
37
|
def content
|
36
|
-
|
38
|
+
self.parse
|
39
|
+
@content
|
37
40
|
end
|
38
41
|
|
39
42
|
def metadata
|
43
|
+
self.parse
|
40
44
|
metadata_hash = {}
|
41
45
|
|
42
46
|
@metadata.names.each do |name|
|
@@ -47,33 +51,31 @@ module Rika
|
|
47
51
|
end
|
48
52
|
|
49
53
|
def media_type
|
50
|
-
@media_type
|
54
|
+
@media_type ||= @tika.detect(input_stream)
|
51
55
|
end
|
52
56
|
|
53
57
|
def available_metadata
|
58
|
+
self.parse
|
54
59
|
@metadata.names.to_a
|
55
60
|
end
|
56
61
|
|
57
62
|
def metadata_exists?(name)
|
63
|
+
self.parse
|
58
64
|
@metadata.get(name) != nil
|
59
65
|
end
|
60
66
|
|
61
67
|
protected
|
62
68
|
|
63
|
-
def
|
64
|
-
|
65
|
-
@metadata.set("filename", File.basename(@uri))
|
66
|
-
@media_type = @tika.detect(java.io.FileInputStream.new(java.io.File.new(@uri)))
|
67
|
-
@content = @tika.parse_to_string(input_stream, @metadata)
|
69
|
+
def parse
|
70
|
+
@content ||= @tika.parse_to_string(input_stream, @metadata).to_s.strip
|
68
71
|
end
|
69
72
|
|
70
|
-
def
|
71
|
-
|
72
|
-
|
73
|
-
|
74
|
-
|
75
|
-
|
76
|
-
@content = @tika.parse_to_string(input_stream, @metadata)
|
73
|
+
def input_stream
|
74
|
+
if @is_file
|
75
|
+
FileInputStream.new(java.io.File.new(@uri))
|
76
|
+
else
|
77
|
+
URL.new(@uri).open_stream
|
78
|
+
end
|
77
79
|
end
|
78
80
|
end
|
79
81
|
end
|
data/lib/rika/version.rb
CHANGED
data/spec/rika_spec.rb
CHANGED
@@ -90,10 +90,6 @@ describe Rika::Parser do
|
|
90
90
|
@txt_parser.metadata["nonsense"].should be_nil
|
91
91
|
end
|
92
92
|
|
93
|
-
it "should return metadata from a text file" do
|
94
|
-
@txt_parser.metadata["filename"].should == "text_file.txt"
|
95
|
-
end
|
96
|
-
|
97
93
|
it "should return metadata from a docx file" do
|
98
94
|
@docx_parser.metadata["Page-Count"].should == "1"
|
99
95
|
end
|
metadata
CHANGED
@@ -1,7 +1,7 @@
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
2
2
|
name: rika
|
3
3
|
version: !ruby/object:Gem::Version
|
4
|
-
version: 0.9.
|
4
|
+
version: 0.9.5
|
5
5
|
prerelease:
|
6
6
|
platform: java
|
7
7
|
authors:
|
@@ -9,7 +9,7 @@ authors:
|
|
9
9
|
autorequire:
|
10
10
|
bindir: bin
|
11
11
|
cert_chain: []
|
12
|
-
date: 2012-09-
|
12
|
+
date: 2012-09-29 00:00:00.000000000 Z
|
13
13
|
dependencies:
|
14
14
|
- !ruby/object:Gem::Dependency
|
15
15
|
name: rspec
|