R extracting part of a string

Question

I need extract a part of a string from a address variable. My data looks like this

"                                                                      
[45] "Matara Road, Habaraduwa | Talpe, Unawatuna, Galle GL 80630, Sri Lanka "                                   
[46] "Jungle Beach Road, Buonavista | Rumassala, Unawatuna, Galle 80600, Sri Lanka "                            
[47] "10 Church Street | inside the Fort, Galle, Sri Lanka "                                                    
[48] "78 Mile Post Matara Road Mihiripenna, Unawatuna, Galle 80615, Sri Lanka "                                 
[49] "No: 288 Galle Road | Dadella, Galle 80000, Sri Lanka "                                                    
[50] "Matara Road, Koggala, Galle, Sri Lanka "  

I want to extract the city from this string, which in this case should be "Galle". The only pattern I can think of is that it appears before "Sri lanka". Or the city is in between "," and ", Sri Lanka". Here is the code that I used

gsub("\\.s*|(, Sri Lanka).*", "", a)

However using this code I get the following results.

[45] "Matara Road, Habaraduwa | Talpe, Unawatuna, Galle GL 80630"                                   
[46] "Jungle Beach Road, Buonavista | Rumassala, Unawatuna, Galle 80600"                            
[47] "10 Church Street | inside the Fort, Galle"                                                    
[48] "78 Mile Post Matara Road Mihiripenna, Unawatuna, Galle 80615"                                 
[49] "No: 288 Galle Road | Dadella, Galle 80000"                                                    
[50] "Matara Road, Koggala, Galle" 

Is there anyway to keep only the city

Answer 1

n <- c(
     "Matara Road, Habaraduwa | Talpe, Unawatuna, Galle GL 80630, Sri Lanka "       ,
     "Jungle Beach Road, Buonavista | Rumassala, Unawatuna, Galle 80600, Sri Lanka ",
     "10 Church Street | inside the Fort, Galle, Sri Lanka "                        ,
     "78 Mile Post Matara Road Mihiripenna, Unawatuna, Galle 80615, Sri Lanka "     ,
     "No: 288 Galle Road | Dadella, Galle 80000, Sri Lanka "                        ,
     "Matara Road, Koggala, Galle, Sri Lanka " )

First, you want to extract the cityname with the possible statename and the possible zip code>

m <- sub('.*, (.*), Sri Lanka *$', '\\1', n)

m is now:

[1] "Galle GL 80630" "Galle 80600" "Galle" "Galle 80615" "Galle 80000" "Galle"

Extract the zip codes

l <- sub(' \\d{5} *$', '', m )

l is:

[1] "Galle GL" "Galle" "Galle" "Galle" "Galle" "Galle"

Finally, extract the state abbreviation

sub('( \\w{2})$', '', l)

[1] "Galle" "Galle" "Galle" "Galle" "Galle" "Galle"

Answer 2

I would use strsplit instead:

line  <- "Matara Road, Habaraduwa | Talpe, Unawatuna, Galle GL"
array <- strsplit(line,",")[[1]]
city  <- array[length(array)-1]

Try it!

to get rid of the numbers just take city and remove them with gsub. Hope it helps!

Answer 3

You can write a function to split the string at commas and take second last element which is usually the city name.

myfunction=function(x)
{
    x=strsplit(x,",")[[1]][length(unlist(strsplit(x,",")))-1]
    x=gsub("[[:digit:]]","",x )
}

This function does the job. Additionally,it then removes any number/digit.

Now use it in lapply function to get desired output

lapply(x,myfunction)