114 lines (67 loc) · 6.14 KB

Jinjava Releases

Version 2.0.9 (Maven Central)

update truncate_html filter to support preserving words by default, with an additional parameter to chop words at length
added unique filter to remove duplicate objects from a sequence
add support for global trim_blocks, lstrip_blocks config settings

Version 2.0.8 (Maven Central)

Updated escape filter to apply to string representation of all objects, not just string instances
Reworked variable resolution to use juel rather than custom approach
Adding ability to track dependencies used in rendering templates
fix issue with handling escape sequences in quoted strings within block tags

Version 2.0.7 (Maven Central)

Propagate interpreter errors from child interpreters
Changed error level to WARN for invalid date format strings, invalid locale strings

Version 2.0.6 (Maven Central)

Fix ifchanged end tag name
Fix implementation of autoescape tag
Fix propagation of InterpretExceptions from tag render
doc updates

Version 2.0.5 (Maven Central)

Changed behavior of date filter to be same as datetimeformat filter
Fixed date format conversion for single-digit values, for %e, %k, %l
Updated Jinjavadoc to support isDeprecated flag, as well as arbitrary key-value metadata

Version 2.0.4 (Maven Central)

Fix issue with PyishDate.isoformat()

Version 2.0.3 (Maven Central)

Upgrading javassist, commons-lang3 library dependencies
Adding new factory method to Jinjava for creating a new JinjavaInterpreter instance

Version 2.0.2 (Maven Central)

Adding code snippets to jinjavadoc
Adding shuffle filter for collections

Version 2.0.1 (Maven Central)

Adding self-documenting feature to jinjava core
Updating addition expression operator ('+') to work with lists and dicts

Version 2.0.0 (Maven Central)

2.0.x requires JDK 8, as it contains some critical fixes to date formatting for certain languages (i.e. Finnish months)
The 2.0.x release has some significant refactorings in the parsing code: ** nests the .parse package under the existing .tree package ** consolidating the token scanner logic, updating the node tree parser
future updates will be able to detect more specific template syntax errors than was previously possible.
NodeList has been removed in favor of the native JDK LinkedList implementation

Version 1.0.9 (Maven Central)

changed how validation of object properties happens; previously, jinjava would add TemplateErrors for unknown properties at the context root (i.e. 'foo') and when the property existed but the value was null (i.e. 'foo.bar == null'). the new logic will only add an error when the property isn't found on the given base object.
creating richer exception hierarchy for more detailed and specific error messaging
the simple Jinjava.render(String, Map) method will now return a container exception containing all errors encountered during rendering, instead of simply the first one

Version 1.0.8 (Maven Central)

fixed issue with multiple includes of the same file in a template

Version 1.0.7 (Maven Central)

added logic to avoid include/import cycles in include and import tags

Version 1.0.6 (Maven Central)

fixed issue in strftimeformatter with format strings ending in literals
updated expression resolver to return blank strings rather than false for exception cases

Version 1.0.5 (Maven Central)

fixed issue in sort filter when using nested attribute

Version 1.0.4 (Maven Central)

Fix issue where HelperStringTokenizer could return a null last value
Added unless tag, for inverse-if functionality
Implemented from tag for importing specific named variables from another context

Version 1.0.3 (Maven Central)

Adding python methods dict.update() and date.isoformat() to analogous types

Version 1.0.2 (Maven Central)

Fix issue with closure scope lookups within macro functions

Version 1.0.1 (Maven Central)

Properly implement call tag syntax
Fix issue with declaring a list as a default method parameter value
Adding string split built-in filter

Version 1.0.0 (Maven Central)

Initial Public Release