xupa_emec 1.0.4 → 1.0.5
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- data/.idea/dictionaries/bernardo.xml +3 -0
- data/.idea/encodings.xml +5 -0
- data/.idea/misc.xml +8 -0
- data/.idea/modules.xml +9 -0
- data/.idea/vcs.xml +8 -0
- data/.idea/xupa_emec.iml +63 -0
- data/bin/xupa_emec +3 -2
- data/lib/xupa_emec/crawler.rb +3 -1
- data/lib/xupa_emec/version.rb +1 -1
- data/xupa_emec.gemspec +8 -2
- metadata +10 -4
data/.idea/encodings.xml
ADDED
data/.idea/misc.xml
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
1
|
+
<?xml version="1.0" encoding="UTF-8"?>
|
|
2
|
+
<project version="4">
|
|
3
|
+
<component name="DependencyValidationManager">
|
|
4
|
+
<option name="SKIP_IMPORT_STATEMENTS" value="false" />
|
|
5
|
+
</component>
|
|
6
|
+
<component name="ProjectRootManager" version="2" project-jdk-name="Ruby SDK 1.8.7-p174" project-jdk-type="RUBY_SDK" />
|
|
7
|
+
</project>
|
|
8
|
+
|
data/.idea/modules.xml
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
1
|
+
<?xml version="1.0" encoding="UTF-8"?>
|
|
2
|
+
<project version="4">
|
|
3
|
+
<component name="ProjectModuleManager">
|
|
4
|
+
<modules>
|
|
5
|
+
<module fileurl="file://$PROJECT_DIR$/.idea/xupa_emec.iml" filepath="$PROJECT_DIR$/.idea/xupa_emec.iml" />
|
|
6
|
+
</modules>
|
|
7
|
+
</component>
|
|
8
|
+
</project>
|
|
9
|
+
|
data/.idea/vcs.xml
ADDED
data/.idea/xupa_emec.iml
ADDED
|
@@ -0,0 +1,63 @@
|
|
|
1
|
+
<?xml version="1.0" encoding="UTF-8"?>
|
|
2
|
+
<module type="RUBY_MODULE" version="4">
|
|
3
|
+
<component name="GemRequirementsHolder" version="3">
|
|
4
|
+
<requirement>
|
|
5
|
+
<requirement>
|
|
6
|
+
<dependency name="activesupport" version="0" bound="GREATER_OR_EQUAL" git="false" path="false" doRequire="true" />
|
|
7
|
+
</requirement>
|
|
8
|
+
<source from="Gemfile" />
|
|
9
|
+
</requirement>
|
|
10
|
+
<requirement>
|
|
11
|
+
<requirement>
|
|
12
|
+
<dependency name="bundler" version="1.0.0" bound="LATEST_IN_BRANCH" git="false" path="false" doRequire="true" />
|
|
13
|
+
</requirement>
|
|
14
|
+
<source from="Gemfile" />
|
|
15
|
+
</requirement>
|
|
16
|
+
<requirement>
|
|
17
|
+
<requirement>
|
|
18
|
+
<dependency name="fastercsv" version="0" bound="GREATER_OR_EQUAL" git="false" path="false" doRequire="true" />
|
|
19
|
+
</requirement>
|
|
20
|
+
<source from="Gemfile" />
|
|
21
|
+
</requirement>
|
|
22
|
+
<requirement>
|
|
23
|
+
<requirement>
|
|
24
|
+
<dependency name="jeweler" version="1.5.1" bound="LATEST_IN_BRANCH" git="false" path="false" doRequire="true" />
|
|
25
|
+
</requirement>
|
|
26
|
+
<source from="Gemfile" />
|
|
27
|
+
</requirement>
|
|
28
|
+
<requirement>
|
|
29
|
+
<requirement>
|
|
30
|
+
<dependency name="mechanize" version="0" bound="GREATER_OR_EQUAL" git="false" path="false" doRequire="true" />
|
|
31
|
+
</requirement>
|
|
32
|
+
<source from="Gemfile" />
|
|
33
|
+
</requirement>
|
|
34
|
+
<requirement>
|
|
35
|
+
<requirement>
|
|
36
|
+
<dependency name="nokogiri" version="0" bound="GREATER_OR_EQUAL" git="false" path="false" doRequire="true" />
|
|
37
|
+
</requirement>
|
|
38
|
+
<source from="Gemfile" />
|
|
39
|
+
</requirement>
|
|
40
|
+
<requirement>
|
|
41
|
+
<requirement>
|
|
42
|
+
<dependency name="trollop" version="0" bound="GREATER_OR_EQUAL" git="false" path="false" doRequire="true" />
|
|
43
|
+
</requirement>
|
|
44
|
+
<source from="Gemfile" />
|
|
45
|
+
</requirement>
|
|
46
|
+
</component>
|
|
47
|
+
<component name="NewModuleRootManager">
|
|
48
|
+
<content url="file://$MODULE_DIR$" />
|
|
49
|
+
<orderEntry type="inheritedJdk" />
|
|
50
|
+
<orderEntry type="sourceFolder" forTests="false" />
|
|
51
|
+
<orderEntry type="library" scope="PROVIDED" name="[gem] rake (v0.8.7, /Library/Ruby/Gems/1.8/gems/rake-0.8.7)" level="application" />
|
|
52
|
+
<orderEntry type="library" scope="PROVIDED" name="[gem] bundler (v1.0.0, /Library/Ruby/Gems/1.8/gems/bundler-1.0.0)" level="application" />
|
|
53
|
+
<orderEntry type="library" scope="PROVIDED" name="[gem] activesupport (v3.0.0, /Library/Ruby/Gems/1.8/gems/activesupport-3.0.0)" level="application" />
|
|
54
|
+
<orderEntry type="library" scope="PROVIDED" name="[gem] fastercsv (v1.5.3, /Library/Ruby/Gems/1.8/gems/fastercsv-1.5.3)" level="application" />
|
|
55
|
+
<orderEntry type="library" scope="PROVIDED" name="[gem] nokogiri (v1.4.3.1, /Users/bernardo/.bundle/ruby/1.8/gems/nokogiri-1.4.3.1)" level="application" />
|
|
56
|
+
<orderEntry type="library" scope="PROVIDED" name="[gem] trollop (v1.16.2, /Library/Ruby/Gems/1.8/gems/trollop-1.16.2)" level="application" />
|
|
57
|
+
<orderEntry type="library" scope="PROVIDED" name="[gem] jeweler (v1.5.1, /Library/Ruby/Gems/1.8/gems/jeweler-1.5.1)" level="application" />
|
|
58
|
+
<orderEntry type="library" scope="PROVIDED" name="[gem] mechanize (v1.0.0, /Library/Ruby/Gems/1.8/gems/mechanize-1.0.0)" level="application" />
|
|
59
|
+
<orderEntry type="library" scope="PROVIDED" name="[gem] nokogiri (v1.4.3.1, /Users/bernardo/.bundle/ruby/1.8/gems/nokogiri-1.4.3.1)" level="application" />
|
|
60
|
+
<orderEntry type="library" scope="PROVIDED" name="[gem] git (v1.2.5, /Library/Ruby/Gems/1.8/gems/git-1.2.5)" level="application" />
|
|
61
|
+
</component>
|
|
62
|
+
</module>
|
|
63
|
+
|
data/bin/xupa_emec
CHANGED
|
@@ -37,7 +37,7 @@ File.open(opts[:entrada], "r") do |input|
|
|
|
37
37
|
iess_to_search.each_with_index do |line, index|
|
|
38
38
|
|
|
39
39
|
raw_name = line.css('td:nth-child(2)').text.strip
|
|
40
|
-
ies_search_name = raw_name.split('-').max{|a,b| a.length <=> b.length } #pega o nome maior
|
|
40
|
+
ies_search_name = raw_name.split('-').max{|a,b| a.length <=> b.length }.strip #pega o nome maior
|
|
41
41
|
|
|
42
42
|
puts
|
|
43
43
|
puts "#{index+1} - Buscando nome da instituição '#{ies_search_name}'..."
|
|
@@ -51,7 +51,8 @@ File.open(opts[:entrada], "r") do |input|
|
|
|
51
51
|
out_csv << new_hash
|
|
52
52
|
end
|
|
53
53
|
else
|
|
54
|
-
|
|
54
|
+
result = crawler.crawl(ies_search_name)
|
|
55
|
+
out_csv << result if result
|
|
55
56
|
end
|
|
56
57
|
|
|
57
58
|
end
|
data/lib/xupa_emec/crawler.rb
CHANGED
|
@@ -18,7 +18,9 @@ module XupaEmec
|
|
|
18
18
|
})
|
|
19
19
|
|
|
20
20
|
|
|
21
|
-
|
|
21
|
+
ies_name_cell = r.search('#lista_resultado > table > tbody > tr > td')[2]
|
|
22
|
+
return unless ies_name_cell
|
|
23
|
+
ies_url = ies_name_cell.inner_html.match( /detalhamento\/(.*)\' \)/ )[1]
|
|
22
24
|
|
|
23
25
|
puts "Buscando dados de '#{ies_search_name}' em #{ies_url}..."
|
|
24
26
|
ies_data = agent.get("http://emec.mec.gov.br/emec/consulta-ies/index/#{ies_url}")
|
data/lib/xupa_emec/version.rb
CHANGED
data/xupa_emec.gemspec
CHANGED
|
@@ -5,11 +5,11 @@
|
|
|
5
5
|
|
|
6
6
|
Gem::Specification.new do |s|
|
|
7
7
|
s.name = %q{xupa_emec}
|
|
8
|
-
s.version = "1.0.
|
|
8
|
+
s.version = "1.0.5"
|
|
9
9
|
|
|
10
10
|
s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
|
|
11
11
|
s.authors = ["Bernardo de P\303\241dua"]
|
|
12
|
-
s.date = %q{2010-12-
|
|
12
|
+
s.date = %q{2010-12-10}
|
|
13
13
|
s.default_executable = %q{xupa_emec}
|
|
14
14
|
s.description = %q{Puxa dados sobre instituições de ensino superior do site do MEC para o formato CSV.}
|
|
15
15
|
s.email = %q{berpasan@gmail.com}
|
|
@@ -20,6 +20,12 @@ Gem::Specification.new do |s|
|
|
|
20
20
|
]
|
|
21
21
|
s.files = [
|
|
22
22
|
".document",
|
|
23
|
+
".idea/dictionaries/bernardo.xml",
|
|
24
|
+
".idea/encodings.xml",
|
|
25
|
+
".idea/misc.xml",
|
|
26
|
+
".idea/modules.xml",
|
|
27
|
+
".idea/vcs.xml",
|
|
28
|
+
".idea/xupa_emec.iml",
|
|
23
29
|
"Gemfile",
|
|
24
30
|
"Gemfile.lock",
|
|
25
31
|
"LICENSE.txt",
|
metadata
CHANGED
|
@@ -1,13 +1,13 @@
|
|
|
1
1
|
--- !ruby/object:Gem::Specification
|
|
2
2
|
name: xupa_emec
|
|
3
3
|
version: !ruby/object:Gem::Version
|
|
4
|
-
hash:
|
|
4
|
+
hash: 29
|
|
5
5
|
prerelease: false
|
|
6
6
|
segments:
|
|
7
7
|
- 1
|
|
8
8
|
- 0
|
|
9
|
-
-
|
|
10
|
-
version: 1.0.
|
|
9
|
+
- 5
|
|
10
|
+
version: 1.0.5
|
|
11
11
|
platform: ruby
|
|
12
12
|
authors:
|
|
13
13
|
- "Bernardo de P\xC3\xA1dua"
|
|
@@ -15,7 +15,7 @@ autorequire:
|
|
|
15
15
|
bindir: bin
|
|
16
16
|
cert_chain: []
|
|
17
17
|
|
|
18
|
-
date: 2010-12-
|
|
18
|
+
date: 2010-12-10 00:00:00 -02:00
|
|
19
19
|
default_executable: xupa_emec
|
|
20
20
|
dependencies:
|
|
21
21
|
- !ruby/object:Gem::Dependency
|
|
@@ -206,6 +206,12 @@ extra_rdoc_files:
|
|
|
206
206
|
- README.rdoc
|
|
207
207
|
files:
|
|
208
208
|
- .document
|
|
209
|
+
- .idea/dictionaries/bernardo.xml
|
|
210
|
+
- .idea/encodings.xml
|
|
211
|
+
- .idea/misc.xml
|
|
212
|
+
- .idea/modules.xml
|
|
213
|
+
- .idea/vcs.xml
|
|
214
|
+
- .idea/xupa_emec.iml
|
|
209
215
|
- Gemfile
|
|
210
216
|
- Gemfile.lock
|
|
211
217
|
- LICENSE.txt
|