1

I have a command "awk -F'[~|^]' 'BEGIN{RS=FS}/*searching string*/{print NR-1}' <<< *user defined string*", which will provide me the delimiters count based on the given search string. However the user defined string is not having any pre-defined delimiters. I got this command from this link of stackoverflow "Using multiple delimiters in awk"

But the above code is working only if the delimiters are in single character. When I am using "^|" as a delimiter it is not considering 2 characters as a single delimiter.

Please find a sample as per my above statement,


Input : a=name^|phone^|contact^|title^|address^|mobilenumber (delimiter may vary)
Searching String : title
Command : awk -F'[~|^]' 'BEGIN{RS=FS}/title/{print NR-1}' <<< $a
Output: 3

So could anyone please help me how i will fit all the delimiters (like - "~", "|", "^", "^|") in the above mentioned command.

user3040157
  • 85
  • 9

1 Answers1

1

You need to define each possible field separator individually. The '|' means OR, so it will not combine them. Instead you can escape the '|' and '^' characters with '\\' (double backslash), then separate characters (fields) with '|'

a='f1|f2^|f3~f4^f5'
awk -F'\\^\\||~|\\^|\\|' '/f1/{print NF-1}' <<< "${a}"
# outputs 4

So this will use '^|', '~', '^', and '|' as field separators, match the string 'f1', and print the number of fields-1.

644
  • 629
  • 4
  • 14
  • Btw if it's possible you should be using only one delimiter (usually a ':' or nullbyte for shell scripting). Data structures are far more important than the code. Really any other structure like CSV/XML/JSON/SQL would be a better option. – 644 May 14 '20 at 23:47