rea 0.0.1
Sign up to get free protection for your applications and to get access to all the features.
- data/CHANGELOG.md +5 -0
- data/Gemfile +16 -0
- data/Gemfile.lock +48 -0
- data/LICENCE.md +22 -0
- data/Manifest.txt +15 -0
- data/README.md +7 -0
- data/Rakefile +23 -0
- data/bin/rea +8 -0
- data/lib/rea-dblp/rea/dblp.rb +59 -0
- data/lib/rea-dblp/rea/dblp/entry_parser.rb +48 -0
- data/lib/rea-dblp/rea/dblp/grab.rb +14 -0
- data/lib/rea-dblp/rea/dblp/parser.rb +36 -0
- data/lib/rea-dblp/rea/dblp/query_result_parser.rb +38 -0
- data/lib/rea-dblp/rea/dblp/search.rb +18 -0
- data/lib/rea.rb +28 -0
- data/lib/rea/command.rb +62 -0
- data/lib/rea/command/search.rb +52 -0
- data/lib/rea/command/utils.rb +31 -0
- data/lib/rea/errors.rb +62 -0
- data/lib/rea/loader.rb +8 -0
- data/lib/rea/version.rb +14 -0
- data/lib/rea/work_key.rb +33 -0
- data/rea.gemspec +192 -0
- data/rea.noespec +33 -0
- data/spec/rea-dblp/fixtures/entries/Damas2005.xml +17 -0
- data/spec/rea-dblp/fixtures/entries/Damas2009.xml +16 -0
- data/spec/rea-dblp/fixtures/entries/Erroneous.xml +1 -0
- data/spec/rea-dblp/fixtures/entries/QueryEntry.xml +15 -0
- data/spec/rea-dblp/fixtures/entries/Unrecognized.xml +3 -0
- data/spec/rea-dblp/fixtures/query_results/Erroneous.xml +1 -0
- data/spec/rea-dblp/fixtures/query_results/QueryResult.xml +25 -0
- data/spec/rea-dblp/fixtures/query_results/Unrecognized.xml +3 -0
- data/spec/rea-dblp/test_entry_parser.rb +72 -0
- data/spec/rea-dblp/test_grab.rb +18 -0
- data/spec/rea-dblp/test_query_result_parser.rb +38 -0
- data/spec/rea-dblp/test_search.rb +18 -0
- data/spec/spec_helper.rb +12 -0
- data/spec/test_rea.rb +12 -0
- data/spec/work_key/test_parse.rb +28 -0
- data/spec/work_key/test_to_ruby_literal.rb +20 -0
- data/spec/work_key/test_to_s.rb +12 -0
- data/tasks/debug_mail.rake +75 -0
- data/tasks/debug_mail.txt +13 -0
- data/tasks/gem.rake +68 -0
- data/tasks/spec_test.rake +71 -0
- data/tasks/unit_test.rake +76 -0
- data/tasks/yard.rake +51 -0
- metadata +215 -0
data/CHANGELOG.md
ADDED
data/Gemfile
ADDED
@@ -0,0 +1,16 @@
|
|
1
|
+
source 'http://rubygems.org'
|
2
|
+
|
3
|
+
group :runtime do
|
4
|
+
gem "http", "~> 0.1.0"
|
5
|
+
gem "nokogiri", "~> 1.5"
|
6
|
+
gem "quickl", "~> 0.4.2"
|
7
|
+
gem "pdf-toolkit", "= 1.0.0.rc1"
|
8
|
+
gem "epath", "~> 0.0.1"
|
9
|
+
gem "alf", "~> 0.12.0"
|
10
|
+
gem "bibtex-ruby", "~> 2.0"
|
11
|
+
end
|
12
|
+
|
13
|
+
group :development do
|
14
|
+
gem "rake", "~> 0.9.2"
|
15
|
+
gem "rspec", "~> 2.8.0"
|
16
|
+
end
|
data/Gemfile.lock
ADDED
@@ -0,0 +1,48 @@
|
|
1
|
+
GEM
|
2
|
+
remote: http://rubygems.org/
|
3
|
+
specs:
|
4
|
+
alf (0.12.0)
|
5
|
+
backports (~> 2.3.0)
|
6
|
+
myrrha (~> 1.2.1)
|
7
|
+
quickl (~> 0.4.2)
|
8
|
+
backports (2.3.0)
|
9
|
+
bibtex-ruby (2.0.4)
|
10
|
+
latex-decode (>= 0.0.6)
|
11
|
+
multi_json (~> 1.0)
|
12
|
+
diff-lcs (1.1.3)
|
13
|
+
epath (0.0.1)
|
14
|
+
http (0.1.0)
|
15
|
+
latex-decode (0.0.12)
|
16
|
+
unicode (~> 0.4)
|
17
|
+
latex-decode (0.0.12-java)
|
18
|
+
multi_json (1.0.4)
|
19
|
+
myrrha (1.2.2)
|
20
|
+
nokogiri (1.5.0)
|
21
|
+
nokogiri (1.5.0-java)
|
22
|
+
pdf-toolkit (1.0.0.rc1)
|
23
|
+
quickl (0.4.2)
|
24
|
+
rake (0.9.2.2)
|
25
|
+
rspec (2.8.0)
|
26
|
+
rspec-core (~> 2.8.0)
|
27
|
+
rspec-expectations (~> 2.8.0)
|
28
|
+
rspec-mocks (~> 2.8.0)
|
29
|
+
rspec-core (2.8.0)
|
30
|
+
rspec-expectations (2.8.0)
|
31
|
+
diff-lcs (~> 1.1.2)
|
32
|
+
rspec-mocks (2.8.0)
|
33
|
+
unicode (0.4.1)
|
34
|
+
|
35
|
+
PLATFORMS
|
36
|
+
java
|
37
|
+
ruby
|
38
|
+
|
39
|
+
DEPENDENCIES
|
40
|
+
alf (~> 0.12.0)
|
41
|
+
bibtex-ruby (~> 2.0)
|
42
|
+
epath (~> 0.0.1)
|
43
|
+
http (~> 0.1.0)
|
44
|
+
nokogiri (~> 1.5)
|
45
|
+
pdf-toolkit (= 1.0.0.rc1)
|
46
|
+
quickl (~> 0.4.2)
|
47
|
+
rake (~> 0.9.2)
|
48
|
+
rspec (~> 2.8.0)
|
data/LICENCE.md
ADDED
@@ -0,0 +1,22 @@
|
|
1
|
+
# The MIT Licence
|
2
|
+
|
3
|
+
Copyright (c) 2012 - Bernard Lambeau
|
4
|
+
|
5
|
+
Permission is hereby granted, free of charge, to any person obtaining
|
6
|
+
a copy of this software and associated documentation files (the
|
7
|
+
"Software"), to deal in the Software without restriction, including
|
8
|
+
without limitation the rights to use, copy, modify, merge, publish,
|
9
|
+
distribute, sublicense, and/or sell copies of the Software, and to
|
10
|
+
permit persons to whom the Software is furnished to do so, subject to
|
11
|
+
the following conditions:
|
12
|
+
|
13
|
+
The above copyright notice and this permission notice shall be
|
14
|
+
included in all copies or substantial portions of the Software.
|
15
|
+
|
16
|
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
|
17
|
+
EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
|
18
|
+
MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
|
19
|
+
NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE
|
20
|
+
LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION
|
21
|
+
OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION
|
22
|
+
WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
|
data/Manifest.txt
ADDED
data/README.md
ADDED
data/Rakefile
ADDED
@@ -0,0 +1,23 @@
|
|
1
|
+
begin
|
2
|
+
gem "bundler", "~> 1.0"
|
3
|
+
require "bundler/setup"
|
4
|
+
rescue LoadError => ex
|
5
|
+
puts ex.message
|
6
|
+
abort "Bundler failed to load, (did you run 'gem install bundler' ?)"
|
7
|
+
end
|
8
|
+
|
9
|
+
# Dynamically load the gem spec
|
10
|
+
$gemspec_file = File.expand_path('../rea.gemspec', __FILE__)
|
11
|
+
$gemspec = Kernel.eval(File.read($gemspec_file))
|
12
|
+
|
13
|
+
# We run tests by default
|
14
|
+
task :default => :test
|
15
|
+
|
16
|
+
#
|
17
|
+
# Install all tasks found in tasks folder
|
18
|
+
#
|
19
|
+
# See .rake files there for complete documentation.
|
20
|
+
#
|
21
|
+
Dir["tasks/*.rake"].each do |taskfile|
|
22
|
+
load taskfile
|
23
|
+
end
|
data/bin/rea
ADDED
@@ -0,0 +1,59 @@
|
|
1
|
+
module Rea
|
2
|
+
module DBLP
|
3
|
+
|
4
|
+
def grab(identifier)
|
5
|
+
Grab.new.call(identifier)
|
6
|
+
end
|
7
|
+
module_function :grab
|
8
|
+
|
9
|
+
def search(query)
|
10
|
+
Search.new.call(query)
|
11
|
+
end
|
12
|
+
module_function :search
|
13
|
+
|
14
|
+
def url
|
15
|
+
"http://www.dblp.org"
|
16
|
+
end
|
17
|
+
module_function :url
|
18
|
+
|
19
|
+
def shortcut
|
20
|
+
"dblp"
|
21
|
+
end
|
22
|
+
module_function :shortcut
|
23
|
+
|
24
|
+
private
|
25
|
+
|
26
|
+
include Rea::ErrorUtils
|
27
|
+
|
28
|
+
def entry_url(identifier)
|
29
|
+
"#{url}/rec/bibtex/#{identifier}.xml"
|
30
|
+
end
|
31
|
+
|
32
|
+
def search_url
|
33
|
+
"#{url}/search/api"
|
34
|
+
end
|
35
|
+
|
36
|
+
def get(url)
|
37
|
+
got = Http.get(url, :response => :object)
|
38
|
+
case got.status
|
39
|
+
when 200...300
|
40
|
+
not_found_error!(url) if got.body.empty?
|
41
|
+
got.body
|
42
|
+
when 400...500
|
43
|
+
not_found_error!(url)
|
44
|
+
when 500...600
|
45
|
+
third_party_error!(url)
|
46
|
+
else
|
47
|
+
unexpected_error!(url)
|
48
|
+
end
|
49
|
+
end
|
50
|
+
|
51
|
+
Rea::register_provider(self)
|
52
|
+
end # module DBLP
|
53
|
+
end # module Rea
|
54
|
+
require_relative 'dblp/parser'
|
55
|
+
require_relative 'dblp/entry_parser'
|
56
|
+
require_relative 'dblp/query_result_parser'
|
57
|
+
require_relative 'dblp/grab'
|
58
|
+
require_relative 'dblp/search'
|
59
|
+
|
@@ -0,0 +1,48 @@
|
|
1
|
+
module Rea
|
2
|
+
module DBLP
|
3
|
+
class EntryParser < Parser
|
4
|
+
|
5
|
+
def parse_text(entry, source = nil)
|
6
|
+
entries = []
|
7
|
+
parse_xml(entry, source) do |doc|
|
8
|
+
doc.xpath("/dblp/*").each do |entry|
|
9
|
+
entries << send(:"parse_#{entry.name}", entry)
|
10
|
+
end
|
11
|
+
end
|
12
|
+
entries
|
13
|
+
end
|
14
|
+
|
15
|
+
private
|
16
|
+
|
17
|
+
###
|
18
|
+
|
19
|
+
def parse_inproceedings(entry)
|
20
|
+
{ :kind => "inproceedings",
|
21
|
+
:key => work_key(entry.attribute("key").text),
|
22
|
+
:title => entry.xpath("title").text,
|
23
|
+
:source => entry.xpath("booktitle").text,
|
24
|
+
:year => entry.xpath("year").text,
|
25
|
+
:authors => entry.xpath("author").map(&:text) }
|
26
|
+
end
|
27
|
+
|
28
|
+
def parse_article(entry)
|
29
|
+
{ :kind => "article",
|
30
|
+
:key => work_key(entry.attribute("key").text),
|
31
|
+
:title => entry.xpath("title").text,
|
32
|
+
:source => entry.xpath("journal").text,
|
33
|
+
:year => entry.xpath("year").text,
|
34
|
+
:authors => entry.xpath("author").map(&:text) }
|
35
|
+
end
|
36
|
+
|
37
|
+
def parse_queryentry(entry)
|
38
|
+
{ :kind => entry.xpath('type').text,
|
39
|
+
:key => work_key(entry.attribute("key").text),
|
40
|
+
:title => entry.xpath("title").text,
|
41
|
+
:source => entry.xpath("venue").text,
|
42
|
+
:year => entry.xpath("year").text,
|
43
|
+
:authors => entry.xpath("authors/author").map(&:text) }
|
44
|
+
end
|
45
|
+
|
46
|
+
end # class EntryParser
|
47
|
+
end # module DBLP
|
48
|
+
end # module Rea
|
@@ -0,0 +1,36 @@
|
|
1
|
+
module Rea
|
2
|
+
module DBLP
|
3
|
+
class Parser
|
4
|
+
include Rea::ErrorUtils
|
5
|
+
|
6
|
+
def self.parse(*args, &bl)
|
7
|
+
new.parse(*args, &bl)
|
8
|
+
end
|
9
|
+
|
10
|
+
def parse(path)
|
11
|
+
if path.respond_to?(:to_path)
|
12
|
+
parse_text(Path(path).read, path)
|
13
|
+
else
|
14
|
+
parse_text(path, nil)
|
15
|
+
end
|
16
|
+
end
|
17
|
+
|
18
|
+
private
|
19
|
+
|
20
|
+
def work_key(identifier)
|
21
|
+
WorkKey.new("dblp", identifier)
|
22
|
+
end
|
23
|
+
|
24
|
+
def parse_xml(xml, source)
|
25
|
+
options = Nokogiri::XML::ParseOptions::STRICT
|
26
|
+
parsed = Nokogiri::XML(xml.to_s, source.to_s, nil, options)
|
27
|
+
block_given? ? yield(parsed) : parsed
|
28
|
+
rescue NoMethodError
|
29
|
+
parse_format_error!(source)
|
30
|
+
rescue Nokogiri::XML::SyntaxError
|
31
|
+
parse_error!(source)
|
32
|
+
end
|
33
|
+
|
34
|
+
end # class Parser
|
35
|
+
end # module DBLP
|
36
|
+
end # module Rea
|
@@ -0,0 +1,38 @@
|
|
1
|
+
module Rea
|
2
|
+
module DBLP
|
3
|
+
class QueryResultParser < Parser
|
4
|
+
|
5
|
+
def parse_text(result, source = nil)
|
6
|
+
entries = []
|
7
|
+
parse_xml(result, source) do |doc|
|
8
|
+
unrecognized!(source) if doc.xpath('/result').empty?
|
9
|
+
doc.xpath("/result/hits/hit").each do |entry|
|
10
|
+
entries << send(:"xml_parse_hit", entry)
|
11
|
+
end
|
12
|
+
end
|
13
|
+
entries
|
14
|
+
end
|
15
|
+
|
16
|
+
private
|
17
|
+
|
18
|
+
def fix_xmlns(xml)
|
19
|
+
xml.gsub %r{<result>},
|
20
|
+
%q{<result xmlns:dblp="http://www.dblp.org/xmlns/dblp">}
|
21
|
+
end
|
22
|
+
|
23
|
+
def xml_parse_hit(hit)
|
24
|
+
key = hit.xpath('url').text[%r{http://www.dblp.org/rec/bibtex/(.*)}, 1]
|
25
|
+
content = <<-XML.gsub(/\n\s*/, "")
|
26
|
+
<dblp>
|
27
|
+
<queryentry key="#{key}">
|
28
|
+
#{hit.xpath('title').inner_text}
|
29
|
+
</queryentry>
|
30
|
+
</dblp>
|
31
|
+
XML
|
32
|
+
EntryParser.parse(content).first
|
33
|
+
end
|
34
|
+
|
35
|
+
end # class QueryResultParser
|
36
|
+
end # module DBLP
|
37
|
+
end # module Rea
|
38
|
+
|
@@ -0,0 +1,18 @@
|
|
1
|
+
module Rea
|
2
|
+
module DBLP
|
3
|
+
class Search
|
4
|
+
include DBLP
|
5
|
+
|
6
|
+
def call(query)
|
7
|
+
unless query.size == 1 and query[:title]
|
8
|
+
not_implemented_error!("Unable to search with #{query.inspect}")
|
9
|
+
end
|
10
|
+
title = query[:title].gsub(/\s+/, "%20")
|
11
|
+
xml = get("#{search_url}?q=#{title}")
|
12
|
+
tuples = QueryResultParser.parse(xml)
|
13
|
+
Relation(tuples)
|
14
|
+
end
|
15
|
+
|
16
|
+
end # class Search
|
17
|
+
end # module DBLP
|
18
|
+
end # module Rea
|
data/lib/rea.rb
ADDED
@@ -0,0 +1,28 @@
|
|
1
|
+
require_relative "rea/version"
|
2
|
+
require_relative "rea/loader"
|
3
|
+
require_relative "rea/errors"
|
4
|
+
|
5
|
+
#
|
6
|
+
# Research Exploring Assistant
|
7
|
+
#
|
8
|
+
module Rea
|
9
|
+
|
10
|
+
def WorkKey(str)
|
11
|
+
WorkKey.parse(str)
|
12
|
+
end
|
13
|
+
module_function :WorkKey
|
14
|
+
|
15
|
+
def register_provider(provider)
|
16
|
+
providers << provider
|
17
|
+
end
|
18
|
+
module_function :register_provider
|
19
|
+
|
20
|
+
def providers
|
21
|
+
@providers ||= []
|
22
|
+
end
|
23
|
+
module_function :providers
|
24
|
+
|
25
|
+
end # module Rea
|
26
|
+
|
27
|
+
require_relative 'rea/work_key'
|
28
|
+
require_relative 'rea-dblp/rea/dblp'
|
data/lib/rea/command.rb
ADDED
@@ -0,0 +1,62 @@
|
|
1
|
+
require 'rea'
|
2
|
+
module Rea
|
3
|
+
#
|
4
|
+
# rea - The Research Exploring Assistant
|
5
|
+
#
|
6
|
+
# SYNOPSIS
|
7
|
+
# #{program_name} [--version] [--help] COMMAND [cmd opts] ARGS...
|
8
|
+
#
|
9
|
+
# OPTIONS
|
10
|
+
# #{summarized_options}
|
11
|
+
#
|
12
|
+
# COMMANDS
|
13
|
+
# #{summarized_subcommands}
|
14
|
+
#
|
15
|
+
# See '#{program_name} help COMMAND' for more information on a specific command.
|
16
|
+
#
|
17
|
+
class Command < Quickl::Delegator(__FILE__, __LINE__)
|
18
|
+
|
19
|
+
# Install options
|
20
|
+
options do |opt|
|
21
|
+
@trace = false
|
22
|
+
opt.on('--trace') do
|
23
|
+
@trace = true
|
24
|
+
end
|
25
|
+
opt.on_tail("--help", "Show help") do
|
26
|
+
raise Quickl::Help
|
27
|
+
end
|
28
|
+
opt.on_tail("--version", "Show version") do
|
29
|
+
raise Quickl::Exit,
|
30
|
+
"rea #{Rea::VERSION} (c) 2012, Bernard Lambeau"
|
31
|
+
end
|
32
|
+
end
|
33
|
+
|
34
|
+
def execute(*args)
|
35
|
+
super
|
36
|
+
rescue Quickl::Error
|
37
|
+
raise
|
38
|
+
rescue Rea::Error => ex
|
39
|
+
if @trace
|
40
|
+
dump_exception(ex)
|
41
|
+
else
|
42
|
+
$stderr.puts ex.message
|
43
|
+
end
|
44
|
+
rescue Exception => ex
|
45
|
+
puts "Sorry, an internal error occured. Please report --trace to developers."
|
46
|
+
dump_exception(ex) if @trace
|
47
|
+
end
|
48
|
+
|
49
|
+
private
|
50
|
+
|
51
|
+
def dump_exception(ex, indent = 0)
|
52
|
+
$stderr.puts ex.message.gsub(/^/m," "*indent)
|
53
|
+
$stderr.puts ex.backtrace.map{|s|
|
54
|
+
s.gsub(/^/m," "*(indent+1))
|
55
|
+
}.join("\n")
|
56
|
+
dump_exception(ex.cause, indent + 1) if ex.respond_to?(:cause)
|
57
|
+
end
|
58
|
+
|
59
|
+
end # class Command
|
60
|
+
end # module Rea
|
61
|
+
require_relative "command/utils"
|
62
|
+
require_relative "command/search"
|