网页链接在 r 网页抓取中不起作用 - 提出空白单元格

Question

我有一个可以抓取网页及其初始嵌套链接的工作代码，但我想单击第二级以抓取更多内容，因此我根据第一级的成功编写了我认为正确的方法。我没有得到任何值，只有标题存在。对此我将不胜感激。

library(rvest)
library(dplyr)
library(purrr)

link = "http://ufcstats.com/statistics/events/completed?page=all"
page = read_html(link)

name = page %>% html_nodes(".b-link_style_black") %>% html_text()
name_links <- page %>% html_nodes(".b-link_style_black") %>% html_attr("href")

fight_info = function(name_link) {
  fight_page = read_html(name_link)
  tibble(
    date = fight %>% html_nodes(".b-list__box-list-item:nth-child(1)") %>% html_text(),
      fight_event = fight %>% html_nodes(".b-list__box-list-item:nth-child(1)") %>% html_attr("href"),
        Fighter1 = fight_page %>% html_nodes(".b-fight-details__person-status_style_green+ .b-fight-details__person-text .b-fight-details__person-link") %>% html_text(),
        Fighter2 = fight_page %>% html_nodes(".b-fight-details__person-status_style_gray+ .b-fight-details__person-text .b-fight-details__person-link") %>% html_text(),
        Referee = fight_page %>% html_nodes(".b-fight-details__label+ span") %>% html_text(),
        F1Control = fight_page %>% html_nodes(".b-fight-details__table-col:nth-child(10) .b-fight-details__table-text:nth-child(1)") %>% html_text(),
        F2Control = fight_page %>% html_nodes(".b-fight-details__table-col:nth-child(10) .b-fight-details__table-text+ .b-fight-details__table-text") %>% html_text(),
        F1ControlRD = fight_page %>% html_nodes(".b-fight-details__table-col:nth-child(10) .b-fight-details__table-text:nth-child(1)") %>% html_text(),
        F2ControlRD = fight_page %>% html_nodes(".b-fight-details__table-col:nth-child(10) .b-fight-details__table-text+ .b-fight-details__table-text") %>% html_text(),
        Weightclass = fight_page %>% html_nodes(".b-fight-details__fight-title") %>% html_text()
  ) -> t
  return(t)
}

df <- map_dfr(name_links, fight_info)

Answer 1

除了我在 cmets 中已经提到的代码中的较小问题之外，使用稍微不同的方法实现所需结果的一个选项是通过 html_table 一次提取所有信息，然后进行一些数据清理：

棘手的部分是拆分像Fighter 这样的单元格，其中包含我使用string::str_split 的两行，并将这些行放入我使用tidyr::unnest_wider 的不同列中：

library(rvest)
library(dplyr)
library(purrr)
library(stringr)
library(tidyr)

link = "http://ufcstats.com/statistics/events/completed?page=all"
page = read_html(link)

name = page %>% html_nodes(".b-link_style_black") %>% html_text()
name_links <- page %>% html_nodes(".b-link_style_black") %>% html_attr("href")

get_fight_info = function(name_link) {
  fight_page <- read_html(name_link) 
  
  fight_info <- fight_page %>% html_table() %>% .[[1]]
  fight_info$date <- fight_page %>% html_node(".b-list__box-list-item:nth-child(1)") %>% html_text()
  fight_info$event <- fight_page %>% html_node(".b-list__box-list-item:nth-child(2)") %>% html_text()
  
  fight_info <- fight_info %>% 
    mutate(across(c("Fighter", "Str", "Method"), ~ str_split(.x, "\\n\\n\\s+\\n")),
           date = str_remove(date, "^\\s+Date:\\s+"),
           event = str_remove(event, "^\\s+Location:\\s+"))
  
  fight_info %>% 
    unnest_wider(Fighter, names_sep = "") %>% 
    unnest_wider(Str, names_sep = "") %>% 
    unnest_wider(Method, names_sep = "") %>% 
    mutate(across(everything(), ~ str_remove_all(.x, "\\n") %>% str_trim()))
}

fight_info <- map_dfr(name_links[1:3], get_fight_info)

fight_info
#> # A tibble: 35 x 15
#>    `W/L` Fighter1  Fighter2 Kd    Str1  Str2  Td    Sub   `Weight class` Method1
#>    <chr> <chr>     <chr>    <chr> <chr> <chr> <chr> <chr> <chr>          <chr>  
#>  1 win   Islam Ma… Thiago … 0   … 61    13    3   … 1   … Lightweight    SUB    
#>  2 win   Miesha T… Marion … 0   … 65    25    3   … 0   … Women's Banta… KO/TKO 
#>  3 win   Mateusz … Jeremy … 0   … 1     0     1   … 1   … Lightweight    SUB    
#>  4 win   Rodolfo … Dustin … 0   … 49    67    3   … 1   … Middleweight   SUB    
#>  5 win   Billy Qu… Gabriel… 1   … 100   57    3   … 1   … Featherweight  KO/TKO 
#>  6 win   Daniel R… Preston… 0   … 45    23    0   … 0   … Welterweight   KO/TKO 
#>  7 win   Amanda L… Montser… 1   … 7     3     0   … 0   … Women's Straw… KO/TKO 
#>  8 win   Sergey M… Khalid … 0   … 50    33    6   … 0   … Bantamweight   U-DEC  
#>  9 win   Malcolm … Francis… 0   … 46    28    2   … 0   … Flyweight      U-DEC  
#> 10 win   Rodrigo … Alan Ba… 0   … 49    60    0   … 0   … Heavyweight    KO/TKO 
#> # … with 25 more rows, and 5 more variables: Method2 <chr>, Round <chr>,
#> #   Time <chr>, date <chr>, event <chr>