feed_processor_utils 0.0.0 → 0.0.1
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.
@@ -87,7 +87,7 @@ module FeedProcessorUtils
|
|
87
87
|
end.last
|
88
88
|
if largest_img
|
89
89
|
if is_news
|
90
|
-
image_urls << largest_img[:url]
|
90
|
+
image_urls << largest_img[:url] if largest_img[:dim] && have_minimum_size?(largest_img[:dim])
|
91
91
|
else
|
92
92
|
image_urls.unshift(largest_img[:url]) if largest_img[:dim] && dimensions_ok?(largest_img[:dim])
|
93
93
|
end
|
@@ -13,11 +13,13 @@ module FeedProcessorUtils
|
|
13
13
|
|
14
14
|
def parse_data(input)
|
15
15
|
input_doc = Nokogiri::HTML(input)
|
16
|
-
Hash[
|
16
|
+
parsed = Hash[
|
17
17
|
fields.map do |field_name, parsing_data|
|
18
18
|
[field_name, extract_field(input_doc, parsing_data)]
|
19
19
|
end
|
20
20
|
]
|
21
|
+
parse_lazy_images!(parsed[:lazy_image_tags]) if parsed[:lazy_image_tags]
|
22
|
+
parsed
|
21
23
|
end
|
22
24
|
|
23
25
|
def parse_url(url)
|
@@ -57,5 +59,15 @@ module FeedProcessorUtils
|
|
57
59
|
@config
|
58
60
|
end
|
59
61
|
|
62
|
+
def parse_lazy_images!(lazy_images)
|
63
|
+
# this gets rid of #{whatever} in sky sports articles
|
64
|
+
regex = /#\{(.+)\}/
|
65
|
+
lazy_images.map! do |lazy_image|
|
66
|
+
lazy_image.sub! regex do |full_match|
|
67
|
+
$1.to_s # this is the 'whatever' inside #{whatever}
|
68
|
+
end
|
69
|
+
end.compact!
|
70
|
+
end
|
71
|
+
|
60
72
|
end
|
61
73
|
end
|