marketplace_opportunity_scraper 0.0.2 → 0.0.3
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 +4 -4
- data/.rubocop.yml +1 -1
- data/.rubocop_todo.yml +1 -1
- data/Gemfile.lock +1 -1
- data/fixtures/cassettes/MarketplaceOpportunityScraper_Opportunity/_all/gets_data_that_is_not_on_the_homepage.yml +2698 -0
- data/fixtures/cassettes/MarketplaceOpportunityScraper_Opportunity/_all/gets_the_correct_opportunity_data.yml +2773 -0
- data/fixtures/cassettes/MarketplaceOpportunityScraper_Opportunity/_all/returns_all_open_opportunities.yml +2773 -0
- data/fixtures/cassettes/MarketplaceOpportunityScraper_Opportunity/_all/when_type_is_specified/returns_the_correct_opportunities.yml +1498 -0
- data/lib/marketplace_opportunity_scraper/opportunity.rb +3 -19
- data/lib/marketplace_opportunity_scraper/version.rb +1 -1
- metadata +1 -1
@@ -13,6 +13,7 @@ module MarketplaceOpportunityScraper
|
|
13
13
|
question_deadline
|
14
14
|
closing
|
15
15
|
description
|
16
|
+
expected_start_date
|
16
17
|
].freeze
|
17
18
|
|
18
19
|
attr_reader *ATTRIBUTES
|
@@ -62,10 +63,6 @@ module MarketplaceOpportunityScraper
|
|
62
63
|
['digital-outcomes', 'digital-specialists', 'user-research-participants']
|
63
64
|
end
|
64
65
|
|
65
|
-
def self.get_date(date)
|
66
|
-
Date.parse date.text.split(':').last
|
67
|
-
end
|
68
|
-
|
69
66
|
def self.opportunity_from_id(id)
|
70
67
|
url = BASE_URL + '/digital-outcomes-and-specialists/opportunities/' + id.to_s
|
71
68
|
page = mechanize.get(url)
|
@@ -82,6 +79,7 @@ module MarketplaceOpportunityScraper
|
|
82
79
|
published: Date.parse(text_from_label(page, 'Published')),
|
83
80
|
question_deadline: Date.parse(text_from_label(page, 'Deadline for asking questions')),
|
84
81
|
closing: Date.parse(text_from_label(page, 'Closing date for applications')),
|
82
|
+
expected_start_date: Date.parse(text_from_label(page, 'Latest start date')),
|
85
83
|
description: text_from_label(page, 'Summary of the work')
|
86
84
|
}
|
87
85
|
|
@@ -90,23 +88,9 @@ module MarketplaceOpportunityScraper
|
|
90
88
|
|
91
89
|
def self.opportunity_from_search_result(element)
|
92
90
|
title = element.at('.search-result-title')
|
93
|
-
important_metadata = element.search('ul.search-result-important-metadata li')
|
94
|
-
dates = element.search('ul.search-result-metadata')[1].search('li')
|
95
91
|
url = BASE_URL + title.at('a').attributes['href'].value
|
96
92
|
|
97
|
-
|
98
|
-
id: url.split('/').last.to_i,
|
99
|
-
title: title.text.strip,
|
100
|
-
url: url,
|
101
|
-
buyer: important_metadata[0].text.strip,
|
102
|
-
location: important_metadata[1].text.strip,
|
103
|
-
published: get_date(dates[0]),
|
104
|
-
question_deadline: get_date(dates[1]),
|
105
|
-
closing: get_date(dates[2]),
|
106
|
-
description: element.at('.search-result-excerpt').text.strip
|
107
|
-
}
|
108
|
-
|
109
|
-
new(attrs)
|
93
|
+
opportunity_from_id(url.split('/').last.to_i)
|
110
94
|
end
|
111
95
|
|
112
96
|
def self.text_from_label(page, label)
|