Tag patterns

Many parts of libhxl (both the Python library and the command-line tools) use tag patterns to select specific columns in a HXL dataset. Tag patterns are similar to tags, but they allow excluding attributes (with -) as well as including attributes (with +). Here are some examples:

• #adm1+name — match any column tagged with #adm1 that has the attribute +name (possibly among others).
• #adm1-name — match any column tagged with #adm1 that does not have the attribute +name (though it may have others).
• #affected+f-children — match any column tagged with #affected that has the attribute +f but does not have the attribute +children (regardless of what other attributes may appear).

Examples

Tag pattern	Matches	Does not match
#org+code	#org+code #org+code+fts	#org #org+name
#org-code	#org #org+name+fr	#org+code #org+fts+code
#sector+cluster-code	#sector+cluster #sector+un+cluster	#sector+code #sector+un