I use xmllint in my command-line scripts. The default mode of xmllint does not support namespaces, so when I must deal with XML files with namespaces I use the --shell mode of xmllint. I noticed that the xpath command in the shell behaves differently than the *xmllint --xpath* on the command line. For example if I want to extract only the string portion of an attribute I use: xmllint --xpath "string(//node)" When I want to do the same in the shell, I must use the *xpath* command, because the *cat* command only supports node addresses. xpath "string(//node)" Unfortunately the xpath command gives an output like: Object is a string : http://purl.obolibrary.org/obo/DOID_8545... Is it possible to get a clean output with xmllint without using other tools? Gökçe
Attachment:
smime.p7s
Description: S/MIME Cryptographic Signature