Row queries

libhxl current uses row queries for the hxlselect command-line script, and the with_rows() and without_rows() filters, but will likely extend them to other uses in the future. The query selects a row by combining a tag pattern with an operator and a value, like this:

org+name=unicef

This query selects every row where the value "unicef" (case-insensitive) appears in a column that matches the tag pattern "org+name" (note that the tag pattern omits the initial '#' — it is optional in queries and most other places that tag patterns appear.)

Operators

The following operators are available:

Operator	Description	Example
=	The value appears in a column matching the tag pattern.	org+name=unicef
!=	The value does not appear in a column matching the tag pattern.	status+code!=completed
<	The content of the column is numerically or lexically less than the value provided.	targeted<300
<=	The content of the column is numerically or lexically less than or equal to the value provided.	population+f<=20000
>	The content of the column is numerically or lexically great than the value provided.	population+f>20000
>=	The content of the column is numerically or lexically great than or equal to the value provided.	contact+name>=m
~	The content of the column matches the [regular expression](https://docs.python.org/2/howto/regex.html) provided.	activity.name~vaccination
!~	The content of the column does not match the regular expression provided.	status!~^(planned|completed)$

Notes

• Matching is always case-insensitive: org=UNICEF and org=unicef will match the same rows.
• The leading # is optional in the tag pattern: org=red cross and #org=red cross will match the same rows.
• Strings are always complete matches; regular expressions can match just part of the content (use ^ and $ to constrain to a full match).
• Leading and trailing whitespace is always removed, and remaining whitespace, normalised, before matching.