1

Adding support for yelp

This commit is contained in:
Alex Yatskov 2015-09-23 12:03:33 +09:00
parent 5a7b773c84
commit 78b593ea13
4 changed files with 72 additions and 17 deletions

View File

@ -10655,6 +10655,14 @@
"Latitude": 35.299925,
"Longitude": 139.480876
},
"〒220-0005 神奈川県横浜市西区南幸2丁目12-6": {
"Latitude": 35.463411,
"Longitude": 139.6160962
},
"〒220-0005 神奈川県横浜市西区南幸2丁目15-1": {
"Latitude": 35.4647525,
"Longitude": 139.6187736
},
"〒220-0005西区南幸1-3-1 横浜モアーズ8F": {
"Latitude": 35.4672263,
"Longitude": 139.6221788
@ -10687,6 +10695,10 @@
"Latitude": 35.4492682,
"Longitude": 139.6353589
},
"〒231-0012 神奈川県横浜市中区相生町1丁目7": {
"Latitude": 35.4461686,
"Longitude": 139.6393198
},
"〒231-0014中区常盤町5-58-2": {
"Latitude": 35.4472157,
"Longitude": 139.6349617
@ -10695,6 +10707,30 @@
"Latitude": 35.4462405,
"Longitude": 139.6428978
},
"〒231-0023 神奈川県横浜市中区山下町 279": {
"Latitude": 35.4462742,
"Longitude": 139.6534105
},
"〒231-0023 神奈川県横浜市中区山下町114": {
"Latitude": 35.4424218,
"Longitude": 139.646289
},
"〒231-0023 神奈川県横浜市中区山下町130-11": {
"Latitude": 35.4420302,
"Longitude": 139.6456029
},
"〒231-0023 神奈川県横浜市中区山下町140": {
"Latitude": 35.4424663,
"Longitude": 139.6451224
},
"〒231-0023 神奈川県横浜市中区山下町189-9": {
"Latitude": 35.442644,
"Longitude": 139.6468862
},
"〒231-0023 神奈川県横浜市中区海岸通 1-1-4": {
"Latitude": 35.4484322,
"Longitude": 139.6447693
},
"〒231-0023中区山下町10 ホテルニューグランド本館5F": {
"Latitude": 35.444773,
"Longitude": 139.649593
@ -10715,6 +10751,10 @@
"Latitude": 35.4441466,
"Longitude": 139.6468555
},
"〒231-0041 神奈川県横浜市中区吉田町5番地4": {
"Latitude": 35.4459992,
"Longitude": 139.6322037
},
"〒231-0861中区元町1-31ラ・スピーガ元町001号室": {
"Latitude": 35.4410507,
"Longitude": 139.6503305

View File

@ -103,7 +103,6 @@ func (l *selector) locateStrings(doc *goquery.Document) ([]string, error) {
})
return strs, err
}
func (l *selector) locateString(doc *goquery.Document) (string, error) {

View File

@ -0,0 +1,31 @@
name = "yelp"
domains = ["www.yelp.co.jp"]
[index.items]
path = "a.biz-name"
attr = "href"
[index.next]
path = "a.next"
attr = "href"
[item.name]
path = "h1.biz-page-title"
[item.address]
path = "div.media-story address"
[item.count]
path = "span.review-count > span"
[item.props]
[item.props.overall]
accomodating = 1.0
affordable = 1.0
atmospheric = 1.0
delicious = 1.0
scale = 5.0
path = "#wrap > div.biz-country-jp > div > div.top-shelf > div > div.biz-page-header.clearfix > div.biz-page-header-left > div.biz-main-info.embossed-text-white > div.rating-info.clearfix > div.biz-rating.biz-rating-very-large.clearfix > div > i"
regEx = "^([0-9]*\\.?[0-9]+)"
attr = "title"

View File

@ -1,16 +1 @@
http://tabelog.com/en/kanagawa/rstLst/
http://www.tripadvisor.com/Restaurants-g298173-Yokohama_Kanagawa_Prefecture_Kanto.html
http://www.tripadvisor.com/Restaurants-g298172-Kawasaki_Kanagawa_Prefecture_Kanto.html
http://www.tripadvisor.com/Restaurants-g1021282-Sagamihara_Kanagawa_Prefecture_Kanto.html
http://www.tripadvisor.com/Restaurants-g1021277-Fujisawa_Kanagawa_Prefecture_Kanto.html
http://www.tripadvisor.com/Restaurants-g303156-Kamakura_Kanagawa_Prefecture_Kanto.html
http://www.tripadvisor.com/Restaurants-g298174-Yokosuka_Kanagawa_Prefecture_Kanto.html
http://www.tripadvisor.com/Restaurants-g1021278-Odawara_Kanagawa_Prefecture_Kanto.html
http://www.tripadvisor.com/Restaurants-g681222-Hiratsuka_Kanagawa_Prefecture_Kanto.html
http://www.tripadvisor.com/Restaurants-g298169-Atsugi_Kanagawa_Prefecture_Kanto.html
http://www.tripadvisor.com/Restaurants-g1021286-Yamato_Kanagawa_Prefecture_Kanto.html
http://www.tripadvisor.com/Restaurants-g1021279-Chigasaki_Kanagawa_Prefecture_Kanto.html
http://www.tripadvisor.com/Restaurants-g1021285-Hadano_Kanagawa_Prefecture_Kanto.html
http://www.yelp.co.jp/search?find_desc=&find_loc=%E6%A8%AA%E6%B5%9C