CODEWITHSUNDEEP
rliststrsplit
Liam Young
Liam Young

Reputation: 11

R split string and keep section

I have a string containing the starting lineup (extracted from the web) for a rugby game, it looks like this:

 "Crusaders: 15 David Havili, 14 Seta Tamanivalu, 13 Jack Goodhue, 12 Ryan Crotty, 11 George Bridge, 10 Richie Mo’unga, 9 Bryn Hall, 8 Kieran Read, 7 Matt Todd, 6 Heiden Bedwell-Curtis, 5 Sam Whitelock (c), 4 Scott Barrett, 3 Owen Franks, 2 Codie Taylor, 1 Joe MoodyReplacements: 16 Sam Anderson-Heather, 17 Tim Perry, 18 Michael Alaalatoa, 19 Luke Romano, 20 Pete Samu, 21 Mitchell Drummond, 22 Mitchell Hunt, 23 Braydon Ennor"

What I want is essentially a table with two columns, one being the player's number, and the other being the player's name. e.g.

position     name
1            Joe Moody
2            Codie Taylor
3            Owen Franks
4            Scott Barrett
...          ...

For all players.

I've tried using strsplit, splitting by the "," however the problem becomes the first player:

"Crusaders: 15 David Havili"

and the number 1 and 16 merge 

"1 Joe MoodyReplacements: 16 Sam Anderson-Heather".

Any ideas?

Upvotes: 0

Views: 90

Answers (2)

Elio Diaz
Elio Diaz

Reputation: 600

Using stringr::str_match_all() and some regex you can find and extract all matches, being careful to use non-greedy (?) operator and matching end of line where there is no comma:

library(dplyr)
library(stringr)
ea <- "Crusaders: 15 David Havili, 14 Seta Tamanivalu, 13 Jack Goodhue, 12 Ryan Crotty, 11 George Bridge, 10 Richie Mo’unga, 9 Bryn Hall, 8 Kieran Read, 7 Matt Todd, 6 Heiden Bedwell-Curtis, 5 Sam Whitelock (c), 4 Scott Barrett, 3 Owen Franks, 2 Codie Taylor, 1 Joe MoodyReplacements: 16 Sam Anderson-Heather, 17 Tim Perry, 18 Michael Alaalatoa, 19 Luke Romano, 20 Pete Samu, 21 Mitchell Drummond, 22 Mitchell Hunt, 23 Braydon Ennor"
ea <- unlist(strsplit(ea, "Replacements: "))

tibble(jersey = str_match_all(ea, "\\d+") %>% unlist(),
player = str_match_all(ea, "(?<=\\d\\s).*?(?=.$|,)") %>% unlist())

# A tibble: 23 x 2
   jersey player               
   <chr>  <chr>                
 1 15     David Havili         
 2 14     Seta Tamanivalu      
 3 13     Jack Goodhue         
 4 12     Ryan Crotty          
 5 11     George Bridge

Upvotes: 0

Maurits Evers
Maurits Evers

Reputation: 50678

I agree with @HongOoi's comment; it's best to take a step back and ensure that data is imported in a more sensible way. That said, here is a post-hoc hacky solution. Not sure how well this generalises, if at all.

ss <-  "Crusaders: 15 David Havili, 14 Seta Tamanivalu, 13 Jack Goodhue, 12 Ryan Crotty, 11 George Bridge, 10 Richie Mo’unga, 9 Bryn Hall, 8 Kieran Read, 7 Matt Todd, 6 Heiden Bedwell-Curtis, 5 Sam Whitelock (c), 4 Scott Barrett, 3 Owen Franks, 2 Codie Taylor, 1 Joe MoodyReplacements: 16 Sam Anderson-Heather, 17 Tim Perry, 18 Michael Alaalatoa, 19 Luke Romano, 20 Pete Samu, 21 Mitchell Drummond, 22 Mitchell Hunt, 23 Braydon Ennor"


library(tidyverse)
data.frame(ss = ss) %>%
    mutate(ss = str_replace(ss, "Replacements", "")) %>%   # Remove "Replacements"
    mutate(ss = str_split(ss, "(,|:) ")) %>%               # Split on "," or ":"
    unnest() %>%
    separate(ss, c("position", "name"), sep = "(?<=\\d)\\s", fill = "right") %>%
    filter(!is.na(name))                                   # Remove the first "Crusaders" line
#   position                  name
#1        15          David Havili
#2        14       Seta Tamanivalu
#3        13          Jack Goodhue
#4        12           Ryan Crotty
#5        11         George Bridge
#6        10        Richie Mo’unga
#7         9             Bryn Hall
#8         8           Kieran Read
#9         7             Matt Todd
#10        6 Heiden Bedwell-Curtis
#11        5     Sam Whitelock (c)
#12        4         Scott Barrett
#13        3           Owen Franks
#14        2          Codie Taylor
#15        1             Joe Moody
#16       16  Sam Anderson-Heather
#17       17             Tim Perry
#18       18     Michael Alaalatoa
#19       19           Luke Romano
#20       20             Pete Samu
#21       21     Mitchell Drummond
#22       22         Mitchell Hunt
#23       23         Braydon Ennor

Upvotes: 1

Related Questions