使用 strsplit 模式匹配将字符串扩展到多列

Question

这是我第一次从 PDF 文档中抓取文本。我正在展示我认为对我正在做的数据最有用的格式，但我可能是错的。清理 PDF 文本后，我将其格式化为 tibble（如下）。

我尝试使用strsplit(dmt, \\s+) 将字符串拆分为三个单独的列，但这只是将所有内容完全分开。我曾使用str_squish() 来消除字符串中间文本部分的空格，但这对模式匹配没有帮助。

字符串的第一个数字部分有时以) 或number 结尾。这是我正在使用的：

dmt
# A tibble: 612 x 1
   datamatrixtest[,1]                                             
   <chr>                                                          
 1 110.05          Human Service Vehicle Inspection Reqd         6
 2 23.33(12)(b)    ATV-Fail/Stop for Law Enforce. Official       1
 3 23.33(6)(a)     ATV-Fail/Display Lighted Headlamp             1
 4 341.03          Oper Veh After Sus/Rev or Can of Reg      8,862
 5 341.04(1)       Non-Registration of Vehicle              10,125
 6 341.04(2)       Improper Registration of Vehicle              4
 7 341.15(1)       Fail/Display Vehicle License Plates       2,010
 8 341.15(1m)(a)   Fail/Attach Rear Regis. Decal/Tag             3
 9 341.15(1m)(b)   Fail/Attach Front Regis. Decal/Tag            2
10 341.15(2)       Improperly Attached License Plates            7
# ... with 602 more rows

理想情况下，我可以利用 strsplit 和精确的模式匹配将数据放入三个单独的列中。

dmt
# A tibble: 612 x 3
   statute         offense                                    cases
   <chr>           <chr>                                       <num>        
 1 110.05          Human Service Vehicle Inspection Reqd         6
 2 23.33(12)(b)    ATV-Fail/Stop for Law Enforce. Official       1
 3 23.33(6)(a)     ATV-Fail/Display Lighted Headlamp             1
 4 341.03          Oper Veh After Sus/Rev or Can of Reg      8,862
 5 341.04(1)       Non-Registration of Vehicle              10,125
 6 341.04(2)       Improper Registration of Vehicle              4
 7 341.15(1)       Fail/Display Vehicle License Plates       2,010
 8 341.15(1m)(a)   Fail/Attach Rear Regis. Decal/Tag             3
 9 341.15(1m)(b)   Fail/Attach Front Regis. Decal/Tag            2
10 341.15(2)       Improperly Attached License Plates            7

Answer 1

我假设您的数据基本上与所显示的一样，列之间有多个空格。换句话说，检查您的dmt 是否与我在下面创建的类似。在这种情况下，我们可以像这样使用\\s{2,} 将多于一个空格的任何部分上的每一行分开。如果您的数据不是这样，或者任何单个字段恰好包含多个空格，请使用dput 和head 提供示例，以便我们找到更精确的模式。

library(tidyverse)
dmt <- read_lines(
"110.05          Human Service Vehicle Inspection Reqd         6
23.33(12)(b)    ATV-Fail/Stop for Law Enforce. Official       1
23.33(6)(a)     ATV-Fail/Display Lighted Headlamp             1
341.03          Oper Veh After Sus/Rev or Can of Reg      8,862
341.04(1)       Non-Registration of Vehicle              10,125
341.04(2)       Improper Registration of Vehicle              4
341.15(1)       Fail/Display Vehicle License Plates       2,010
341.15(1m)(a)   Fail/Attach Rear Regis. Decal/Tag             3
341.15(1m)(b)   Fail/Attach Front Regis. Decal/Tag            2
1341.15(2)       Improperly Attached License Plates            7"
) %>%
  enframe(name = NULL, value = "line")

dmt %>%
  separate(line, c("statute", "offense", "cases"), sep = "\\s{2,}") %>%
  mutate(cases = cases %>% str_remove_all(",") %>% as.integer)
#> # A tibble: 10 x 3
#>    statute       offense                                 cases
#>    <chr>         <chr>                                   <int>
#>  1 110.05        Human Service Vehicle Inspection Reqd       6
#>  2 23.33(12)(b)  ATV-Fail/Stop for Law Enforce. Official     1
#>  3 23.33(6)(a)   ATV-Fail/Display Lighted Headlamp           1
#>  4 341.03        Oper Veh After Sus/Rev or Can of Reg     8862
#>  5 341.04(1)     Non-Registration of Vehicle             10125
#>  6 341.04(2)     Improper Registration of Vehicle            4
#>  7 341.15(1)     Fail/Display Vehicle License Plates      2010
#>  8 341.15(1m)(a) Fail/Attach Rear Regis. Decal/Tag           3
#>  9 341.15(1m)(b) Fail/Attach Front Regis. Decal/Tag          2
#> 10 1341.15(2)    Improperly Attached License Plates          7

^{由reprex package (v0.3.0) 于 2019 年 9 月 23 日创建}

Answer 2

根据您的数据，我们还可以通过定义不同的捕获组来使用tidyr::extract。

library(dplyr)
library(tidyr)

df %>%
   extract(datamatrixtest, into = c("statute", "offense", "cases"),
           regex = "(.*?)\\s(.*?)(\\d.*)") %>%
    mutate_all(trimws)


#   statute       offense                                 cases 
#   <chr>         <chr>                                   <chr> 
# 1 110.05        Human Service Vehicle Inspection Reqd   6     
# 2 23.33(12)(b)  ATV-Fail/Stop for Law Enforce. Official 1     
# 3 23.33(6)(a)   ATV-Fail/Display Lighted Headlamp       1     
# 4 341.03        Oper Veh After Sus/Rev or Can of Reg    8,862 
# 5 341.04(1)     Non-Registration of Vehicle             10,125
# 6 341.04(2)     Improper Registration of Vehicle        4     
# 7 341.15(1)     Fail/Display Vehicle License Plates     2,010 
# 8 341.15(1m)(a) Fail/Attach Rear Regis. Decal/Tag       3     
# 9 341.15(1m)(b) Fail/Attach Front Regis. Decal/Tag      2     
#10 1341.15(2)    Improperly Attached License Plates      7  

在这里，我们定义了三个组，第一个从文本的开头开始直到遇到第一个空格，第二个从第一个结束的地方开始直到遇到一个数字，第三个从数字到结尾的所有内容的句子。

数据

df <- structure(list(datamatrixtest = c("110.05          Human Service 
Vehicle Inspection Reqd         6", 
"23.33(12)(b)    ATV-Fail/Stop for Law Enforce. Official       1", 
"23.33(6)(a)     ATV-Fail/Display Lighted Headlamp             1", 
"341.03          Oper Veh After Sus/Rev or Can of Reg      8,862", 
"341.04(1)       Non-Registration of Vehicle              10,125", 
"341.04(2)       Improper Registration of Vehicle              4", 
"341.15(1)       Fail/Display Vehicle License Plates       2,010", 
"341.15(1m)(a)   Fail/Attach Rear Regis. Decal/Tag             3", 
"341.15(1m)(b)   Fail/Attach Front Regis. Decal/Tag            2", 
"1341.15(2)       Improperly Attached License Plates            7"
)), row.names = c(NA, -10L), class = c("tbl_df", "tbl", "data.frame"))