Find All XML Tags from the CLI

These bash one liners served me well when exploring raw HTML and XML. Once building an ingestion / cleaning pipeline I would output more detailed statistics about tag usage.

HTML

For HTML, you may want to use a looser regex like this one. The regex is searching for anything starting with <. It allows the tag to not to be closed.

ag "</?([^> ]*)?>?" path/to/html/ -io \
    | sed -n -e 's/^.*:<?\/?//p' \
    | sort --unique

XML

For XML, finding the end tags (e.g. </EndTag>) may be sufficient. I didn’t consider start tags they can have attributes (e.g. <StartTag Attribute=X>). This assumes the XML is already decently formatted.

ag  "</.*?>" path/to/xml/ -oi \
    | sed -n -e 's/^.*://p' \
    | sort --unique

Notes:

• ag is the silver searcher

• sed is removing the line number from the result. (e.g. ag returns 123:<br />)

• sort --unique sorts the lines and with the --unique option discards the duplicates