Significally increased re-parsing performance on code typing
It's not possible anymore to start identifier with digit, like $123abc
Hexadecimal and binary numbers with underscores now being parsed right.
Fixed parsing problem for concatenation of string and fractional number.
Fixed parsing for simple dereferences like $$array_ref[1].
Added requirement for 14 platform, no more buggy usages in outdated products.
Fixed and tested interactions with perl process.
Typeglobs for built-ins are not annotated as unused anymore.
Fixed package file creation in specialized IDEs.
Here-doc end marker, pod and formats are not being indented on block tabbing.
Here-doc end marker is not being indented on live-template insertion.
Tricky use vars declarations now being handled correctly. use vars @SOMELIST still not working. Also fixed declaration of hashes, code refs and typeglobs.
Fixed bug with tail LF in convert string to here-doc intention.
Fixed bug with lookup elements caching for subs.
Cleaned up extended regex parsing.
Replacement part of extended regex /x is now parsed properly.
Fixed bug with broken line splitting in other languages.
Fixed bug with print OF <<EOM; parsing.
Suppressed livetemplates inside a string
Suppressed string to namespace resolution for top-level namespaces.
Pod block now may starts with =cut.
Fixed an error creating new file outside of the module.
Implemented distinction of array index from characters group in regexes.
Improved parser recovery before compound statements.
Fixed scoping problem for if/elsif/else compound statement.
Fixed autocompletion behaviour problem with global variables auto-completion with short main:: form.
Fixed problems with packages refactoring.
Fixed recovery after annotations.
Fixed POSIX character group classes within regex with bracketed delimiters.