- Fix
min_by
/max_by
with empty lists (`issue 151) - Fix reverse type for
null
type (issue 145)
- Fix issue where long types in py2 and
Decimal
types were not being evaluated as numbers (issue 125) - Handle numbers in scientific notation in
to_number()
function (issue 120) - Fix issue where custom functions would override the function table of the builtin function class (issue 133)
- Fix regression when using ordering comparators on strings (issue 124)
- Raise LexerError on invalid numbers (issue 98)
- Add support for custom functions (#100) (issue 100)
- Fix ZeroDivisionError for built-in function avg() on empty lists (#115) (issue 115)
- Properly handle non numerical ordering operators (#117) (issue 117)
Add support for new lines with tokens in an expression
Add support for JEP 9, which introduces "and" expressions, "unary" expressions, "not" expressions, and "paren" expressions
Fix issue with hardcoded path in
jp.py
executable (issue 90,
- Improve lexing performance (issue 84)
- Fix parsing error for multiselect lists (issue 86)
- Fix issue with escaping single quotes in literal strings (issue 85)
- Add support for providing your own dict cls to support ordered dictionaries (issue 94)
- Add map() function (issue 95)
- Rename
bin/jp
tobin/jp.py
- Fix issue with precedence when parsing wildcard projections
- Remove ordereddict and simplejson as py2.6 dependencies. These were never actually used in the jmespath code base, only in the unit tests. Unittests requirements are handled via requirements26.txt.
- Add support for JEP-12, raw string literals
- Support .whl files
- Implement JEP-10, slice projections
- Fix bug with filter projection parsing
- Add
to_array
function - Add
merge
function - Fix error messages for function argument type errors