kajian 0.1.2
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.
- checksums.yaml +7 -0
- data/.gitignore +9 -0
- data/Gemfile +15 -0
- data/LICENSE.txt +21 -0
- data/README.md +96 -0
- data/Rakefile +2 -0
- data/bin/console +14 -0
- data/bin/setup +8 -0
- data/kajian.gemspec +21 -0
- data/lib/kajian.rb +31 -0
- data/lib/kajian/adapter.rb +29 -0
- data/lib/kajian/adapter/jadwal_kajian.json +55 -0
- data/lib/kajian/adapter/jadwal_kajian.rb +50 -0
- data/lib/kajian/fungsi_bantu.rb +26 -0
- data/lib/kajian/koleksi_adapter.rb +29 -0
- data/lib/kajian/makro.rb +189 -0
- data/lib/kajian/version.rb +3 -0
- metadata +63 -0
checksums.yaml
ADDED
@@ -0,0 +1,7 @@
|
|
1
|
+
---
|
2
|
+
SHA1:
|
3
|
+
metadata.gz: f1ebed3e1cb139a9c2029bf4e9fc664c5ff482cc
|
4
|
+
data.tar.gz: fc6f259d249a41e3d6fe018cad10fa8b1dddfb8c
|
5
|
+
SHA512:
|
6
|
+
metadata.gz: feeb22ce0e1643bf02814aa530290215d66ad912a469ea44e696d70f4f36ca345aba1b603422a5019d3ef72c54d2b804d75fcce06a89261e5545744c7e51ccf5
|
7
|
+
data.tar.gz: f0e82ad4036f2be42cc3907e73440fa557cfcccb7dbd608ef91d9185f07b39fab100b798e343366b03c358d39f58493046956a0ca6a18cc01364d449284a9dfb
|
data/.gitignore
ADDED
data/Gemfile
ADDED
@@ -0,0 +1,15 @@
|
|
1
|
+
# source 'https://rubygems.org'
|
2
|
+
|
3
|
+
# Specify your gem's dependencies in kajian.gemspec
|
4
|
+
gemspec
|
5
|
+
|
6
|
+
gem 'nokogiri', '~> 1.8'
|
7
|
+
|
8
|
+
group :development do
|
9
|
+
gem 'bundler', '~> 1.14'
|
10
|
+
gem 'rake', '~> 12.1'
|
11
|
+
end
|
12
|
+
|
13
|
+
group :development, :test do
|
14
|
+
gem 'pry', '~> 0.10'
|
15
|
+
end
|
data/LICENSE.txt
ADDED
@@ -0,0 +1,21 @@
|
|
1
|
+
The MIT License (MIT)
|
2
|
+
|
3
|
+
Copyright (c) 2017 Adrian Setyadi
|
4
|
+
|
5
|
+
Permission is hereby granted, free of charge, to any person obtaining a copy
|
6
|
+
of this software and associated documentation files (the "Software"), to deal
|
7
|
+
in the Software without restriction, including without limitation the rights
|
8
|
+
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
9
|
+
copies of the Software, and to permit persons to whom the Software is
|
10
|
+
furnished to do so, subject to the following conditions:
|
11
|
+
|
12
|
+
The above copyright notice and this permission notice shall be included in
|
13
|
+
all copies or substantial portions of the Software.
|
14
|
+
|
15
|
+
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
16
|
+
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
17
|
+
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
18
|
+
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
19
|
+
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
20
|
+
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
|
21
|
+
THE SOFTWARE.
|
data/README.md
ADDED
@@ -0,0 +1,96 @@
|
|
1
|
+
# Kajian
|
2
|
+
|
3
|
+
[](https://rubygems.org/gems/kajian)
|
4
|
+
|
5
|
+
Pustaka ruby untuk mengekstrak data acara kajian Islam dari berbagai situs di
|
6
|
+
Indonesia. Dengan menggunakan DSL (_Domain Specific Language_), mudah untuk
|
7
|
+
membuat adapter untuk situs-situs lain yang akan diekstrak data acara kajiannya.
|
8
|
+
|
9
|
+
Gem ini mengikutsertakan sebuah adapter untuk situs
|
10
|
+
[Jadwal Kajian](http://jadwalkajian.com) dengan simbol adapter `:jadwal_kajian`.
|
11
|
+
|
12
|
+
|
13
|
+
## Pemasangan
|
14
|
+
|
15
|
+
Tambahkan baris berikut ini pada Gemfile Anda:
|
16
|
+
|
17
|
+
```ruby
|
18
|
+
gem 'kajian'
|
19
|
+
```
|
20
|
+
|
21
|
+
Lalu jalankan:
|
22
|
+
|
23
|
+
$ bundle
|
24
|
+
|
25
|
+
Atau Anda pasang sendiri seperti ini:
|
26
|
+
|
27
|
+
$ gem install kajian
|
28
|
+
|
29
|
+
|
30
|
+
## Penggunaan
|
31
|
+
|
32
|
+
Jangan lupa `require` sebelum melihat kajian.
|
33
|
+
|
34
|
+
```ruby
|
35
|
+
require 'kajian'
|
36
|
+
```
|
37
|
+
|
38
|
+
|
39
|
+
Lihat semua kajian
|
40
|
+
|
41
|
+
```ruby
|
42
|
+
Kajian.lihat(:semua).semua
|
43
|
+
```
|
44
|
+
|
45
|
+
|
46
|
+
Lihat semua kajian untuk daerah/kota "Bekasi"
|
47
|
+
|
48
|
+
```ruby
|
49
|
+
Kajian.lihat(:semua).bekasi
|
50
|
+
```
|
51
|
+
|
52
|
+
|
53
|
+
Lihat semua kajian yang dipublikasikan melalui situs jadwal kajian
|
54
|
+
|
55
|
+
```ruby
|
56
|
+
Kajian.lihat(:jadwal_kajian).semua
|
57
|
+
```
|
58
|
+
|
59
|
+
|
60
|
+
Lihat semua kajian yang dipublikasikan melalui situs jadwal kajian
|
61
|
+
daerah/kota "Tangerang"
|
62
|
+
|
63
|
+
```ruby
|
64
|
+
Kajian.lihat(:jadwal_kajian).tangerang
|
65
|
+
```
|
66
|
+
|
67
|
+
|
68
|
+
Tambah adapter lain, contoh [Kajian Muslim](https://kajianmuslim.com)
|
69
|
+
|
70
|
+
```ruby
|
71
|
+
require 'kajian/adapter/kajian_muslim'
|
72
|
+
|
73
|
+
Kajian.lihat(:kajian_muslim).malang
|
74
|
+
|
75
|
+
Kajian.lihat(:jadwal_kajian, :kajian_muslim).jakarta
|
76
|
+
|
77
|
+
Kajian.lihat(:semua).bandung
|
78
|
+
```
|
79
|
+
|
80
|
+
|
81
|
+
## TODO
|
82
|
+
|
83
|
+
* Dokumentasi cara membuat adapter kajian dengan menggunakan DSL Kajian
|
84
|
+
* Pencarian acara kajian berdasarkan isi kolom (termasuk tanggal) yang efisien.
|
85
|
+
* Sortir berdasarkan kolom.
|
86
|
+
* Pengecualian adapter.
|
87
|
+
|
88
|
+
|
89
|
+
## Kontribusi
|
90
|
+
|
91
|
+
Laporan _bug_ dan _pull request_ dapat diajukan melalui Github https://github.com/styd/kajian.
|
92
|
+
|
93
|
+
|
94
|
+
## Lisensi
|
95
|
+
|
96
|
+
Gem ini tersedia sebagai _open source_ sesuai ketentuan [MIT License](http://opensource.org/licenses/MIT).
|
data/Rakefile
ADDED
data/bin/console
ADDED
@@ -0,0 +1,14 @@
|
|
1
|
+
#!/usr/bin/env ruby
|
2
|
+
|
3
|
+
require "bundler/setup"
|
4
|
+
require "kajian"
|
5
|
+
|
6
|
+
# You can add fixtures and/or initialization code here to make experimenting
|
7
|
+
# with your gem easier. You can also use a different console, if you like.
|
8
|
+
|
9
|
+
# (If you use this, don't forget to add pry to your Gemfile!)
|
10
|
+
# require "pry"
|
11
|
+
# Pry.start
|
12
|
+
|
13
|
+
require "irb"
|
14
|
+
IRB.start(__FILE__)
|
data/bin/setup
ADDED
data/kajian.gemspec
ADDED
@@ -0,0 +1,21 @@
|
|
1
|
+
# coding: utf-8
|
2
|
+
lib = File.expand_path('../lib', __FILE__)
|
3
|
+
$LOAD_PATH.unshift(lib) unless $LOAD_PATH.include?(lib)
|
4
|
+
require 'kajian/version'
|
5
|
+
|
6
|
+
Gem::Specification.new do |spec|
|
7
|
+
spec.name = "kajian"
|
8
|
+
spec.version = Kajian::VERSION
|
9
|
+
spec.authors = ["Adrian Setyadi"]
|
10
|
+
spec.email = ["adrian@setyadi.pro"]
|
11
|
+
spec.summary = %q{Pustaka ruby untuk mengekstrak data acara kajian Islam dari berbagai situs di Indonesia.}
|
12
|
+
spec.description = %q{Pustaka ruby untuk mengekstrak data acara kajian Islam dari berbagai situs di Indonesia. Dengan menggunakan DSL (Domain Specific Language), mudah untuk menambahkan situs-situs lain untuk diekstrak data acara kajiannya.}
|
13
|
+
spec.homepage = "https://github.com/styd/kajian"
|
14
|
+
spec.license = "MIT"
|
15
|
+
spec.files = `git ls-files -z`.split("\x0").reject do |f|
|
16
|
+
f.match(%r{^(test|spec|features)/})
|
17
|
+
end
|
18
|
+
spec.bindir = "exe"
|
19
|
+
spec.executables = spec.files.grep(%r{^exe/}) { |f| File.basename(f) }
|
20
|
+
spec.require_paths = ["lib"]
|
21
|
+
end
|
data/lib/kajian.rb
ADDED
@@ -0,0 +1,31 @@
|
|
1
|
+
require "kajian/version"
|
2
|
+
require "kajian/fungsi_bantu"
|
3
|
+
require "kajian/makro"
|
4
|
+
require "kajian/koleksi_adapter"
|
5
|
+
require "kajian/adapter"
|
6
|
+
require "kajian/adapter/jadwal_kajian"
|
7
|
+
|
8
|
+
module Kajian
|
9
|
+
TidakAdaAdapter = Class.new(StandardError)
|
10
|
+
|
11
|
+
class << self
|
12
|
+
def adapter
|
13
|
+
@@adapter ||= Kajian::Makro.adapter
|
14
|
+
end
|
15
|
+
|
16
|
+
def lihat *sumber_sumber
|
17
|
+
sumber_sumber = [:semua] if sumber_sumber.empty?
|
18
|
+
if sumber_sumber.length == 1
|
19
|
+
if sumber_sumber.first == :semua
|
20
|
+
raise TidakAdaAdapter unless Kajian.adapter
|
21
|
+
Kajian::KoleksiAdapter.new(Kajian.adapter)
|
22
|
+
else
|
23
|
+
s = sumber_sumber.first
|
24
|
+
Kajian::Adapter.new(s)
|
25
|
+
end
|
26
|
+
else
|
27
|
+
Kajian::KoleksiAdapter.new(sumber_sumber)
|
28
|
+
end
|
29
|
+
end
|
30
|
+
end
|
31
|
+
end
|
@@ -0,0 +1,29 @@
|
|
1
|
+
require 'kajian/fungsi_bantu'
|
2
|
+
|
3
|
+
module Kajian
|
4
|
+
class Adapter
|
5
|
+
include FungsiBantu
|
6
|
+
|
7
|
+
def initialize(sumber)
|
8
|
+
if sumber.kind_of?(Class)
|
9
|
+
@kelas_sumber = sumber
|
10
|
+
@simbol_sumber = adapter_ke_simbol(sumber)
|
11
|
+
elsif sumber.kind_of?(String) or sumber.kind_of?(Symbol)
|
12
|
+
@kelas_sumber = simbol_ke_adapter(sumber)
|
13
|
+
@simbol_sumber = sumber.to_sym
|
14
|
+
end
|
15
|
+
end
|
16
|
+
|
17
|
+
def semua
|
18
|
+
{@simbol_sumber => @kelas_sumber.dokumen}
|
19
|
+
end
|
20
|
+
|
21
|
+
def [](*daerah_daerah)
|
22
|
+
{@simbol_sumber => @kelas_sumber.dokumen(*daerah_daerah)}
|
23
|
+
end
|
24
|
+
|
25
|
+
def method_missing(m)
|
26
|
+
self[m]
|
27
|
+
end
|
28
|
+
end
|
29
|
+
end
|
@@ -0,0 +1,55 @@
|
|
1
|
+
{
|
2
|
+
"aceh": "/aceh",
|
3
|
+
"bali": "/bali",
|
4
|
+
"balikpapan": "/balikpapan",
|
5
|
+
"barabai": "/barabai",
|
6
|
+
"bandung": "/bandung",
|
7
|
+
"banten": "/banten",
|
8
|
+
"batam": "/batam",
|
9
|
+
"bekasi": "/bekasi",
|
10
|
+
"bogor": "/bogor",
|
11
|
+
"bojonegoro": "/bojonegoro",
|
12
|
+
"boyolali": "/boyolali",
|
13
|
+
"cilacap": "/cilacap",
|
14
|
+
"cilegon": "/cilegon",
|
15
|
+
"denpasar": "/denpasar",
|
16
|
+
"depok": "/depok",
|
17
|
+
"garut": "/garut",
|
18
|
+
"gorontalo": "/gorontalo",
|
19
|
+
"indramayu": "/indramayu",
|
20
|
+
"jakarta": "/jakarta",
|
21
|
+
"jambi": "/jambi",
|
22
|
+
"jayapura": "/jayapura",
|
23
|
+
"kampar": "/kampar",
|
24
|
+
"karawang": "/karawang",
|
25
|
+
"kebumen": "/kebumen",
|
26
|
+
"kediri": "/kediri",
|
27
|
+
"kutai_kartanegara": "/kutai-kartanegara",
|
28
|
+
"madiun": "/madiun",
|
29
|
+
"magelang": "/magelang",
|
30
|
+
"makassar": "/makassar",
|
31
|
+
"malang": "/malang",
|
32
|
+
"medan": "/medan",
|
33
|
+
"manokwari": "/manokwari",
|
34
|
+
"palangkaraya": "/palangkaraya",
|
35
|
+
"pamulang": "/pamulang",
|
36
|
+
"pasuruan": "/pasuruan",
|
37
|
+
"pekalongan": "/pekalongan",
|
38
|
+
"pekanbaru": "/pekanbaru",
|
39
|
+
"pemalang": "/pemalang",
|
40
|
+
"ponorogo": "/ponorogo",
|
41
|
+
"salatiga": "/salatiga",
|
42
|
+
"samarinda": "/samarinda",
|
43
|
+
"sangatta": "/sangatta",
|
44
|
+
"semarang": "/semarang",
|
45
|
+
"sidoarjo": "/sidoarjo",
|
46
|
+
"solo": "/solo",
|
47
|
+
"sukoharjo": "/sukoharjo",
|
48
|
+
"sumedang": "/sumedang",
|
49
|
+
"surabaya": "/surabaya",
|
50
|
+
"surakarta": "/surakarta",
|
51
|
+
"tangerang": "/tangerang",
|
52
|
+
"tarakan": "/tarakan",
|
53
|
+
"tangerang_selatan": "/tangerang-selatan",
|
54
|
+
"yogyakarta": "/yogyakarta"
|
55
|
+
}
|
@@ -0,0 +1,50 @@
|
|
1
|
+
require 'kajian/makro'
|
2
|
+
|
3
|
+
module Kajian
|
4
|
+
class Adapter
|
5
|
+
class JadwalKajian
|
6
|
+
extend Kajian::Makro
|
7
|
+
|
8
|
+
@url = 'http://jadwalkajian.com'
|
9
|
+
|
10
|
+
@direktori_master = json_ke_hash("#{simbol}.json")
|
11
|
+
|
12
|
+
# buang_direktori(
|
13
|
+
# :aceh, :bali, :balikpapan, :barabai, :bandung, :banten, :batam, :bojonegoro,
|
14
|
+
# :boyolali, :cilacap, :cilegon, :denpasar, :garut, :gorontalo, :indramayu,
|
15
|
+
# :jayapura, :kampar, :karawang, :kebumen, :kediri, :kutai_kartanegara,
|
16
|
+
# :madiun, :magelang, :makassar, :malang, :medan, :manokwari, :palangkaraya,
|
17
|
+
# :pamulang, :pasuruan, :pekalongan, :pemalang, :ponorogo, :salatiga,
|
18
|
+
# :samarinda, :sangatta, :semarang, :sidoarjo, :solo, :sukoharjo, :sumedang,
|
19
|
+
# :surabaya, :surakarta, :tarakan, :yogyakarta
|
20
|
+
# )
|
21
|
+
|
22
|
+
# pakai_direktori(:bekasi, :jakarta)
|
23
|
+
|
24
|
+
blok(css: 'div.eventon_list_event.event')
|
25
|
+
|
26
|
+
tema(css: 'span.evcal_event_title')
|
27
|
+
|
28
|
+
penceramah(css: 'span.evcal_event_subtitle', substitusi: {/dr\.?/ => 'Dr.'})
|
29
|
+
|
30
|
+
tempat do |blok|
|
31
|
+
blok.css('div.evo_location').map do |lokasi|
|
32
|
+
lokasi.css('p').map(&:text).join("\n")
|
33
|
+
end
|
34
|
+
end
|
35
|
+
|
36
|
+
tanggal do |blok|
|
37
|
+
"#{blok.css('span.evcal_cblock .evo_date span.start').first.text.to_i} " +
|
38
|
+
"#{blok.css('span.evcal_cblock').first["data-smon"]}"
|
39
|
+
end
|
40
|
+
|
41
|
+
waktu(css: 'div.evo_time p', hilangkan: /^\(.*\)\s+/)
|
42
|
+
|
43
|
+
gambar do |blok|
|
44
|
+
blok.css('div.evo_metarow_locImg').map do |img|
|
45
|
+
img.attributes.values[1].value[/(?<=url\().*(?=\))/]
|
46
|
+
end
|
47
|
+
end
|
48
|
+
end
|
49
|
+
end
|
50
|
+
end
|
@@ -0,0 +1,26 @@
|
|
1
|
+
module Kajian
|
2
|
+
module FungsiBantu
|
3
|
+
# Konversi dari kelas adapter ke simbol.
|
4
|
+
# Misal:
|
5
|
+
# Kajian::Adapter::JadwalKajian ke :jadwal_kajian
|
6
|
+
def adapter_ke_simbol(kelas_adapter)
|
7
|
+
kelas_adapter.name
|
8
|
+
.sub(/(\w+::)+/, '')
|
9
|
+
.gsub(/([A-Z]+)([A-Z][a-z])/, '\1_\2')
|
10
|
+
.gsub(/([a-z\d])([A-Z])/, '\1_\2')
|
11
|
+
.tr("-", "_")
|
12
|
+
.downcase
|
13
|
+
.to_sym
|
14
|
+
end
|
15
|
+
|
16
|
+
# Konversi dari simbol ke kelas adapter.
|
17
|
+
# Misal:
|
18
|
+
# :jadwal_kajian ke Kajian::Adapter::JadwalKajian
|
19
|
+
def simbol_ke_adapter(simbol)
|
20
|
+
nama_adapter = simbol.to_s
|
21
|
+
.sub(/^[a-z\d]*/) { |match| match.capitalize }
|
22
|
+
.gsub!(/(?:_)([a-z\d]*)/i) {"#{$1.capitalize}"}
|
23
|
+
Kernel.const_get("Kajian::Adapter::#{nama_adapter}")
|
24
|
+
end
|
25
|
+
end
|
26
|
+
end
|
@@ -0,0 +1,29 @@
|
|
1
|
+
require 'kajian/adapter'
|
2
|
+
|
3
|
+
module Kajian
|
4
|
+
class KoleksiAdapter
|
5
|
+
def initialize(sumber_sumber)
|
6
|
+
@adapter_adapter = sumber_sumber
|
7
|
+
end
|
8
|
+
|
9
|
+
def semua
|
10
|
+
koleksi = {}
|
11
|
+
@adapter_adapter.map do |adapter|
|
12
|
+
koleksi = koleksi.merge(Kajian::Adapter.new(adapter).semua)
|
13
|
+
end
|
14
|
+
koleksi
|
15
|
+
end
|
16
|
+
|
17
|
+
def [](*daerah_daerah)
|
18
|
+
koleksi = {}
|
19
|
+
@adapter_adapter.map do |adapter|
|
20
|
+
koleksi = koleksi.merge(Kajian::Adapter.new(adapter)[*daerah_daerah])
|
21
|
+
end
|
22
|
+
koleksi
|
23
|
+
end
|
24
|
+
|
25
|
+
def method_missing(m)
|
26
|
+
self[m]
|
27
|
+
end
|
28
|
+
end
|
29
|
+
end
|
data/lib/kajian/makro.rb
ADDED
@@ -0,0 +1,189 @@
|
|
1
|
+
module Kajian
|
2
|
+
KesalahanDaerah = Class.new(StandardError)
|
3
|
+
|
4
|
+
module Makro
|
5
|
+
require 'kajian/fungsi_bantu'
|
6
|
+
require 'open-uri'
|
7
|
+
require 'nokogiri'
|
8
|
+
require 'json'
|
9
|
+
require 'thread'
|
10
|
+
|
11
|
+
KOLOM = %i[tema penceramah tempat tanggal waktu gambar]
|
12
|
+
BULAN = {
|
13
|
+
"januari" => "january",
|
14
|
+
"februari" => "february",
|
15
|
+
"pebruari" => "february",
|
16
|
+
"maret" => "march",
|
17
|
+
"mei" => "may",
|
18
|
+
"juni" => "june",
|
19
|
+
"juli" => "july",
|
20
|
+
"agustus" => "august",
|
21
|
+
"oktober" => "october",
|
22
|
+
"desember" => "december",
|
23
|
+
}
|
24
|
+
|
25
|
+
extend FungsiBantu
|
26
|
+
|
27
|
+
@@adapter = []
|
28
|
+
|
29
|
+
class << self
|
30
|
+
def extended(adapter)
|
31
|
+
@@adapter << adapter
|
32
|
+
adapter.simbol = adapter_ke_simbol(adapter)
|
33
|
+
end
|
34
|
+
|
35
|
+
def adapter
|
36
|
+
@@adapter
|
37
|
+
end
|
38
|
+
end
|
39
|
+
|
40
|
+
attr_accessor :simbol, *KOLOM.map {|k| "proc_#{k}"}
|
41
|
+
attr_reader :url, :direktori_master, :direktori_salinan, *KOLOM
|
42
|
+
attr_writer :proc_blok,
|
43
|
+
|
44
|
+
def dokumen *daerah_daerah
|
45
|
+
set_direktori!
|
46
|
+
@buang_direktori ||= []
|
47
|
+
@buang_direktori.each {|daerah| direktori_salinan.delete(daerah)}
|
48
|
+
unless daerah_daerah.empty?
|
49
|
+
daerah_daerah = daerah_daerah.map(&:to_sym)
|
50
|
+
daerah_daerah.each do |daerah|
|
51
|
+
raise Kajian::KesalahanDaerah, "daerah '#{daerah}' tidak ditemukan" \
|
52
|
+
unless direktori_salinan.has_key?(daerah)
|
53
|
+
end
|
54
|
+
direktori_salinan.keep_if{|k,_| daerah_daerah.include?(k)}
|
55
|
+
end
|
56
|
+
|
57
|
+
jadwal = []
|
58
|
+
direktori_salinan.to_a.each_slice(10) do |slice|
|
59
|
+
slice.map do |daerah, dir|
|
60
|
+
Thread.new do
|
61
|
+
jadwal << [daerah, @proc_blok.(daerah, dir)] rescue [daerah, []]
|
62
|
+
end
|
63
|
+
end.map(&:join)
|
64
|
+
end
|
65
|
+
jadwal.to_h
|
66
|
+
end
|
67
|
+
|
68
|
+
def json_ke_hash(file_name = "#{simbol}.json")
|
69
|
+
JSON(
|
70
|
+
File.read(File.join(File.dirname(caller[0][/^.*\.rb/]), file_name)),
|
71
|
+
symbolize_names: true
|
72
|
+
)
|
73
|
+
end
|
74
|
+
|
75
|
+
def set_direktori!
|
76
|
+
@direktori_salinan = direktori_master.dup
|
77
|
+
end
|
78
|
+
|
79
|
+
def buang_direktori(*args)
|
80
|
+
set_direktori!
|
81
|
+
@pakai_direktori = direktori_master.keys - args.map(&:to_sym)
|
82
|
+
@buang_direktori = args
|
83
|
+
end
|
84
|
+
|
85
|
+
def pakai_direktori(*args)
|
86
|
+
set_direktori!
|
87
|
+
@buang_direktori = direktori_master.keys - args.map(&:to_sym)
|
88
|
+
@pakai_direktori = args.map(&:to_sym)
|
89
|
+
end
|
90
|
+
|
91
|
+
def blok(filter_blok)
|
92
|
+
@blok = if block_given?
|
93
|
+
proc {|laman| yield(laman)}
|
94
|
+
elsif filter_blok
|
95
|
+
filter_blok
|
96
|
+
end
|
97
|
+
|
98
|
+
@proc_blok = proc do |daerah, dir|
|
99
|
+
if instance_variable_get("@laman_#{daerah}").nil?
|
100
|
+
laman = Nokogiri::HTML(open(File.join(self.url, dir)))
|
101
|
+
instance_variable_set("@laman_#{daerah}", laman)
|
102
|
+
end
|
103
|
+
|
104
|
+
laman ||= instance_variable_get("@laman_#{daerah}")
|
105
|
+
|
106
|
+
if @blok.kind_of? Proc
|
107
|
+
@blok.(laman)
|
108
|
+
elsif filter_blok
|
109
|
+
laman.public_send(*filter_blok.to_a.first)
|
110
|
+
end.map do |b|
|
111
|
+
kajian = {}
|
112
|
+
KOLOM.each do |k|
|
113
|
+
konten = self.public_send("proc_#{k}").(b) rescue nil
|
114
|
+
kajian[k] = konten.kind_of?(Array) ? (k == :penceramah ? konten : konten.first) : konten
|
115
|
+
end
|
116
|
+
kajian
|
117
|
+
end
|
118
|
+
end
|
119
|
+
end
|
120
|
+
|
121
|
+
KOLOM.each do |k|
|
122
|
+
define_method k do |filter = {}, &block|
|
123
|
+
nilai_k = if block
|
124
|
+
proc {|blok| block.(blok)}
|
125
|
+
elsif filter[:css]
|
126
|
+
Hash[*filter.assoc(:css)]
|
127
|
+
elsif filter[:xpath]
|
128
|
+
Hash[*filter.assoc(:xpath)]
|
129
|
+
end
|
130
|
+
|
131
|
+
instance_variable_set("@#{k}", nilai_k)
|
132
|
+
|
133
|
+
proc_k = proc do |blok|
|
134
|
+
teks_awal = if instance_variable_get("@#{k}").kind_of? Proc
|
135
|
+
instance_variable_get("@#{k}").(blok)
|
136
|
+
elsif nilai_k
|
137
|
+
konten = blok.public_send(*nilai_k.to_a.first)
|
138
|
+
konten = if k == :gambar
|
139
|
+
konten.map {|img| img["src"]}
|
140
|
+
else
|
141
|
+
konten.map(&:text)
|
142
|
+
end
|
143
|
+
k == :penceramah ? konten : konten.first
|
144
|
+
end
|
145
|
+
|
146
|
+
if k == :penceramah
|
147
|
+
teks_awal.map do |teks|
|
148
|
+
teks_bersih = bersihkan(teks, filter[:hilangkan], k)
|
149
|
+
teks_sub = substitusi(teks_bersih, filter[:substitusi])
|
150
|
+
konversi(teks_sub, filter[:parser])
|
151
|
+
end
|
152
|
+
else
|
153
|
+
teks_bersih = bersihkan(teks_awal, filter[:hilangkan], k)
|
154
|
+
teks_sub = substitusi(teks_bersih, filter[:substitusi])
|
155
|
+
filter[:parser] ||= Date if k == :tanggal
|
156
|
+
konversi(teks_sub, filter[:parser])
|
157
|
+
end
|
158
|
+
end
|
159
|
+
|
160
|
+
self.public_send("proc_#{k}=", proc_k)
|
161
|
+
end
|
162
|
+
end
|
163
|
+
|
164
|
+
private
|
165
|
+
def bersihkan(teks, harus_hilang, nama_kolom)
|
166
|
+
teks.gsub!(harus_hilang, '') if harus_hilang
|
167
|
+
if nama_kolom == :tema
|
168
|
+
teks.sub(/\s+((B|b)ersama (Usta(d)?(z)?(ah)?)? .*)?$/, '')
|
169
|
+
elsif nama_kolom == :penceramah
|
170
|
+
teks.sub(/^(((A|a)l(\-?|\s+))?(U|u)stad?z?(ah)?)\s+/, '')
|
171
|
+
elsif nama_kolom == :tanggal
|
172
|
+
BULAN.each do |asal_kata, terjemahan|
|
173
|
+
teks.sub!(asal_kata, terjemahan)
|
174
|
+
end
|
175
|
+
teks
|
176
|
+
else
|
177
|
+
teks
|
178
|
+
end
|
179
|
+
end
|
180
|
+
|
181
|
+
def substitusi(teks, sub_hash)
|
182
|
+
sub_hash ? teks.gsub(*sub_hash.to_a.first) : teks
|
183
|
+
end
|
184
|
+
|
185
|
+
def konversi(teks, parser)
|
186
|
+
parser ? parser.parse(teks) : teks
|
187
|
+
end
|
188
|
+
end
|
189
|
+
end
|
metadata
ADDED
@@ -0,0 +1,63 @@
|
|
1
|
+
--- !ruby/object:Gem::Specification
|
2
|
+
name: kajian
|
3
|
+
version: !ruby/object:Gem::Version
|
4
|
+
version: 0.1.2
|
5
|
+
platform: ruby
|
6
|
+
authors:
|
7
|
+
- Adrian Setyadi
|
8
|
+
autorequire:
|
9
|
+
bindir: exe
|
10
|
+
cert_chain: []
|
11
|
+
date: 2017-09-28 00:00:00.000000000 Z
|
12
|
+
dependencies: []
|
13
|
+
description: Pustaka ruby untuk mengekstrak data acara kajian Islam dari berbagai
|
14
|
+
situs di Indonesia. Dengan menggunakan DSL (Domain Specific Language), mudah untuk
|
15
|
+
menambahkan situs-situs lain untuk diekstrak data acara kajiannya.
|
16
|
+
email:
|
17
|
+
- adrian@setyadi.pro
|
18
|
+
executables: []
|
19
|
+
extensions: []
|
20
|
+
extra_rdoc_files: []
|
21
|
+
files:
|
22
|
+
- ".gitignore"
|
23
|
+
- Gemfile
|
24
|
+
- LICENSE.txt
|
25
|
+
- README.md
|
26
|
+
- Rakefile
|
27
|
+
- bin/console
|
28
|
+
- bin/setup
|
29
|
+
- kajian.gemspec
|
30
|
+
- lib/kajian.rb
|
31
|
+
- lib/kajian/adapter.rb
|
32
|
+
- lib/kajian/adapter/jadwal_kajian.json
|
33
|
+
- lib/kajian/adapter/jadwal_kajian.rb
|
34
|
+
- lib/kajian/fungsi_bantu.rb
|
35
|
+
- lib/kajian/koleksi_adapter.rb
|
36
|
+
- lib/kajian/makro.rb
|
37
|
+
- lib/kajian/version.rb
|
38
|
+
homepage: https://github.com/styd/kajian
|
39
|
+
licenses:
|
40
|
+
- MIT
|
41
|
+
metadata: {}
|
42
|
+
post_install_message:
|
43
|
+
rdoc_options: []
|
44
|
+
require_paths:
|
45
|
+
- lib
|
46
|
+
required_ruby_version: !ruby/object:Gem::Requirement
|
47
|
+
requirements:
|
48
|
+
- - ">="
|
49
|
+
- !ruby/object:Gem::Version
|
50
|
+
version: '0'
|
51
|
+
required_rubygems_version: !ruby/object:Gem::Requirement
|
52
|
+
requirements:
|
53
|
+
- - ">="
|
54
|
+
- !ruby/object:Gem::Version
|
55
|
+
version: '0'
|
56
|
+
requirements: []
|
57
|
+
rubyforge_project:
|
58
|
+
rubygems_version: 2.5.2
|
59
|
+
signing_key:
|
60
|
+
specification_version: 4
|
61
|
+
summary: Pustaka ruby untuk mengekstrak data acara kajian Islam dari berbagai situs
|
62
|
+
di Indonesia.
|
63
|
+
test_files: []
|