Shell Programming and Scripting

Code:

[root@#### ~]# grep "[:alpha:]" regexp
asdh1987
4yuop8pou
sam99917c00l
Akoold0g8
data
sample data
NetWorkManager781C00l

It is missing out

Code:

981towm
1234oqn

Which also happen to be Alpha numeric. Any ideas.

Thanks
Syed

---------- Post updated at 11:50 AM ---------- Previous update was at 11:17 AM ----------

Code:

[root@#### ~]# grep -Ei "[a-z][0-9]|[0-9][a-z]" regexp
asdh1987
dog897you
981towm
1234oqn
4yuop8pou
sam99917c00l
Akoold0g8
NetWorkManager781C00l

and for some reason [:alpha:] and [:alnum:] are not yeilding the desired result.

Note: I have not used [:alpha:] and [:alnum:] in the past.

Hi there,
please make it a point to read all the replies to the thread carefully before asking further questions. Your question has already been answered by Don Cragun in post #2 of this thread.

Have you gone through all my posts before coming to a conclusion.Please understand the question and then try to answer. Anyways, I figured out a solution my self.

Dear maverick_here,
Maybe you should go back and reread your posts. The title of this thread is grep and alphanumerics, not grep alphabetics and numerics. I pointed out that the RE you were using was matching and printing any line that contained any one of the characters ':', 'a', 'l', 'n', 'u', or 'm' rather than any alphanumeric character. You responded that the RE I suggested [[:alnum:]] also printed the lines:

Code:

data
sample data

and said that they are only lower case (which is not true since' 'is not a lowercase letter) and shouldn't matter since you said you wanted to match alphanumeric characters (and lowercase letters are alphanumeric characters).

The RE that you say works: [a-z][0-9]|[0-9][a-z], is not looking for an alphanumeric character; it is an ERE looking for a lowercase letter immediately followed by a decimal digit or a decimal digit immediately followed by a lowercase letter assuming that you are using a locale that is using a codeset that is a superset of ASCII. (But, since you used grep's -i option, grep will ignore case when looking for alphabetic characters.) If you are using an EBCDIC codeset, [a-z] will also match several characters that are not lowercase alphabetics, and if you are in a locale in which alphabetic characters includes letters with accents, [a-z] will not find the accented letters. If you had said that you wanted a portable ERE that would match a lowercase letter adjacent to a digit in the current locale in either order or an ERE that would match any letter adjacent to a digit in the current locale, I would have suggested:

Code:

grep -E '[[:lower:]][[:digit:]]|[[:digit:]][[:lower:]]' regexp

or

Code:

grep -E '[[:alpha:]][[:digit:]]|[[:digit:]][[:alpha:]]' regexp

respectively, but you never requested that.