본문 바로가기
R-크롤링

이미지 크롤링

by 미스터탁 2019. 12. 12.


keyword<-"박보영"
keyowrd<-iconv(keyword,from="CP949",to="UTF-8")
keyword2<-URLencode(keyword)

url<-paste0("https://search.daum.net/search?w=img&nil_search=btn&DA=NTB&enc=utf8&q=",keyword2)


b<-readLines(url,encoding='UTF-8')
library(stringr)
library(RJSONIO)

b2<-b[str_detect(b,"collection.meta = ")]
b3<-str_split(b2," collection.items = ")
b4<-str_split(b3[[1]][2],"; ")
b5<-b4[[1]][1]
b6<-str_sub(b5,3)


b7<-fromJSON(b6)
class(b7[[1]]$nf)


serach_url<-function(x){
  data<-x$nf
  data[names(data) %in% "imgur"]
  
}
umg_url<-sapply(b7,serach_url)




dir.create(paste0("D:\\인프런\\crawling\\",keyword))
setwd(paste0("D:\\인프런\\crawling\\",keyword))

for(i in 1:length(umg_url)){
  
 download.file(umg_url[i],paste0(keyword,i,'.jpg'), mode = 'wb')

}

반응형

댓글