Class: Crawler
- Inherits:
-
Object
- Object
- Crawler
- Defined in:
- lib/gsv_downloader/crawler.rb
Instance Method Summary collapse
- #crawl(panoID) ⇒ Object
-
#initialize(area_validator, db) ⇒ Crawler
constructor
A new instance of Crawler.
- #start(pano_ids) ⇒ Object
Constructor Details
#initialize(area_validator, db) ⇒ Crawler
Returns a new instance of Crawler.
6 7 8 9 10 11 12 |
# File 'lib/gsv_downloader/crawler.rb', line 6 def initialize (area_validator, db) @metadata_downloader = MetaDataDownloader.new @stats = Statistics.new @db = db @area_validator = area_validator @max = 200 end |
Instance Method Details
#crawl(panoID) ⇒ Object
21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 |
# File 'lib/gsv_downloader/crawler.rb', line 21 def crawl(panoID) @db.mark_to_crawl(panoID) @metadata_downloader.download(panoID) do | response| json = extract_json(response) panoID = json["Location"]["panoId"] @stats.count unless @db.crawled?(panoID) @db.mark_as_crawled(panoID) # if inside the area if @area_validator.call(json) @db.add_pano(panoID, response) #TODO to save to save or whatever # for each valid and new link, crawl it extract_valid_links(json) do |link_id| crawl(link_id) end end end end end |
#start(pano_ids) ⇒ Object
14 15 16 17 18 19 |
# File 'lib/gsv_downloader/crawler.rb', line 14 def start(pano_ids) pano_ids.each do |pano_id| crawl(pano_id) end @metadata_downloader.start() end |