Bash expr index command - bash

Question

I am trying to get the index position using Bash 'expr index".
e.g.
$ echo `expr index "Info.out.2014-02-08:INFO|SID:sXfzRjbmKbwX7jyaW1sog7n|Browser[Mozilla/5.0 (Windows NT 6.1; WOW64; rv:26.0) Gecko/20100101 Firefox/26.0]" Mozilla`
I am trying to get the index position of the word "Mozilla", and then get the substring using index value.
The result I got back is 4. Is it the period after Info caus the issue? How do I fix this issue?
I followed the Advanced Bash scripting guide www.tldp.org/LDP/abs/html/‎. See section Table B-5. String Operations
expr index "$string" $substring Numerical position in $string of first character in $substring* that matches [0 if no match, first character counts as position 1]
I tried with something simple, and it works.
I am running bash in cygwin.
$ ./bash --version
GNU bash, version 4.1.10(4)-release (i686-pc-cygwin)
Copyright (C) 2009 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software; you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.
Thanks.

Answer 1

In general, you shouldn't be using expr index unless you have a very good reason to.
For instance, let's say you want to get the browser name.
s="Info.out.2014-02-08:INFO|SID:sXfzRjbmKbwX7jyaW1sog7n|Browser[Mozilla/5.0 (Windows NT 6.1; WOW64; rv:26.0) Gecko/20100101 Firefox/26.0]"
# strip everything up to and including the first instance of 'Browser['
browser="${s#*Browser[}"
# strip everything after the first ']', again, inclusive
browser="${browser%%]*}"
# ...and show the result...
echo "$browser"
This would return:
Mozilla/5.0 (Windows NT 6.1; WOW64; rv:26.0) Gecko/20100101 Firefox/26.0
If you really do want to know how many characters precede Mozilla, well, you can do that too:
s="Info.out.2014-02-08:INFO|SID:sXfzRjbmKbwX7jyaW1sog7n|Browser[Mozilla/5.0 (Windows NT 6.1; WOW64; rv:26.0) Gecko/20100101 Firefox/26.0]"
# strip everything after the first instance of 'Mozilla'
prefix=${s%%Mozilla*}
# count number of characters in the string
index=${#prefix}
# ...and show the result...
echo "$index"
This should return 61.
For the "why" and "how" of the above examples, see BashFAQ #73.
To split by | separators, by contrast, I'd personally choose to use read, as documented in BashFAQ #1:
s="Info.out.2014-02-08:INFO|SID:sXfzRjbmKbwX7jyaW1sog7n|Browser[Mozilla/5.0 (Windows NT 6.1; WOW64; rv:26.0) Gecko/20100101 Firefox/26.0]"
IFS='|' read -r _ _ browser _
echo "$browser"
...which would emit...
Browser[Mozilla/5.0 (Windows NT 6.1; WOW64; rv:26.0) Gecko/20100101 Firefox/26.0]

Answer 2

The expr index command searches through your first string looking the the first occurrence of any character from your second string. In this case, it is recognizing that the 'o' in the characters 'Mozilla' matches the 4th character in "Info.out..."
This using this as a test to see what happens. It will return 4 as the first match for 'd':
echo `expr index "abcdefghijklmnopqrstuvwxyz" xyzd`
This one should do what you want:
echo "Info.out.2014-02-08:INFO|SID:sXfzRjbmKbwX7jyaW1sog7n|Browser[Mozilla/5.0 (Windows NT 6.1; WOW64; rv:26.0) Gecko/20100101 Firefox/26.0]" | grep -o -b Mozilla
The echo puts your string into stdout, so it can be piped into grep.
The -b prints the byte offset of the string shown.
The -o ensures that only the matching portion gets printed.

Answer 3

GNU expr does not match a substring using index; rather, it looks for the first occurrence of any character from the second string in the first. Your example returns 4 because the 4th character of the string is "o", the first character in "Mozilla" that is found in "Info.out...".
There is no built-in function of this kind in either bash or expr, but you can indirectly get the index of a given substring by first removing the substring and everything after it from the original string, then computing the remaining length.
string="Info.out..."
substring=Mozilla
tmp=${string%%$substring*}
index=${#tmp}