How do I find this string?

Question

I'm trying to match these file names with the find command:

4702011-10-21CR719557-R85262-ALPHA.jar
4702011-10-21CR719557-R85262-BETA.jar

I'm reasonable familiar with Java-Perl regex syntax. Unfortunately, find only supports emacs and posix syntax.

Even after looking at the emacs syntax I cannot get a basic match to work.

find . -regex "^[[:digit:]]+" -- matches nothing
find . -regex "^[[:digit:]]+.*(ALPHA\|BETA)\.jar" -- matches nothing

This is pretty straightforward in Java and Perl, for example:

ls | grep -P "^\d+.*(ALPHA|BETA)\.jar"

works.

Answer 1

a regex for matching those files with either posix-awk or posix-extended regxtype would be:

find . -regextype posix-awk -regex ".*/[[:digit:]]+-[[:digit:]]+-[A-Z0-9]+-[A-Z0-9]+-(ALPHA|BETA)\.jar"

because find -regex does matches, not searches

you can specify regex type with -regextype. see the manpage of your find implementation to check what regex engines are supported.

sample (get the files no matter the dir depth):

root@smgw:/tmp# ls -1
4702011-10-21CR719557-R85262-ALPHA.jar
4702011-10-21CR719557-R85262-BETA.jar
one
root@smgw:/tmp# find . -regextype posix-awk -regex ".*/[[:digit:]]+-[[:digit:]]+-[A-Z0-9]+-[A-Z0-9]+-(ALPHA|BETA)\.jar"
./4702011-10-21CR719557-R85262-ALPHA.jar
./4702011-10-21CR719557-R85262-BETA.jar

move the jars around, make sure they're still returned

search them with an abs path

root@smgw:/tmp# find /tmp  -regextype posix-awk -regex ".*/[[:digit:]]+-[[:digit:]]+-[A-Z0-9]+-[A-Z0-9]+-(ALPHA|BETA)\.jar"
/tmp/4702011-10-21CR719557-R85262-ALPHA.jar
/tmp/4702011-10-21CR719557-R85262-BETA.jar

modified depth:

root@smgw:/tmp# mkdir -p x/y/z
root@smgw:/tmp# cp *^C
root@smgw:/tmp# mv *.jar x/y/z/
root@smgw:/tmp# find /tmp  -regextype posix-awk -regex ".*/[[:digit:]]+-[[:digit:]]+-[A-Z0-9]+-[A-Z0-9]+-(ALPHA|BETA)\.jar"
/tmp/x/y/z/4702011-10-21CR719557-R85262-ALPHA.jar
/tmp/x/y/z/4702011-10-21CR719557-R85262-BETA.jar

Answer 2

A regex using emacs style to match your pattern could be:

".*/[0-9]+-[0-9]+-[0-9A-Z]+-[0-9A-Z]+-\(ALPHA\|BETA\)\.jar$"

Below is a simple test I've made:

abarbosa@SPF-157:~/Temp/test$ ls -1
4702011-10-21CR719557-R85262-ALPHA.jar
4702011-10-21CR719557-R85262-BETA.jar
whatever.jar
abarbosa@SPF-157:~/Temp/test$ find . -regex ".*/[0-9]+-[0-9]+-[0-9A-Z]+-[0-9A-Z]+-\(ALPHA\|BETA\)\.jar$"
./4702011-10-21CR719557-R85262-BETA.jar
./4702011-10-21CR719557-R85262-ALPHA.jar

Answer 3

According to the man find, this is a match on the whole path, not a search. So, you must use something like this:

$ find . -regex .*/[0-9]+.*\.jar

You can also change the regular expression type to posix-awk, posix-basic, posix-extended, ... with the -regextype option.