Revision history for Perl module P



Bugs Fixed since 1.0.20
(for more info, see: 
https://rt.cpan.org/Dist/Display.html?Status=Resolved;Queue=P
)
rt#89046 rt#89047 rt#85050 rt#89051 rt#89052 rt#89054 rt#89055
rt#89056 rt#89057 rt#89058 rt#89059 rt#89060 rt#89063 rt#89064
rt#89065 rt#89066 rt#89067 rt#89069 rt#89071
----

Feature changes / additions:

*	- P now prints {…} instead of HASH(0x12356892) and […] for arrays
	  (brackets for type, and ellipses between them).
* - Hash prints are sorted on output to have non-refs sorted before
		refs
*	- switch parsing in object-mode calls expanded to allow specifying a
		max string length to print
* - type 'REF' is detected and printed
* - Detect recursive data structure references when expanding a single
		a single call and use '🔁 ' (U+1F500) (reverse rotating loop) to
		indicate such and resort to old style memory address listing.
* - instead of taking up 7 spaces with "(undef)", use the Unicode symbol
		for "there does not exist"  (∄) (U+2204), which looks like a backwards
		capital 'E' with a diagonal slash drawn through it.
* - Cleaned up prototypes for Pe and P -- both try to take an array argument.
		This should prevent any incoming args being turned into integers.
* - Pod cleanups
* - Fix for using 0x83 as "line feed" supression char.  Check if it is
		already encoded in unicode OR not.
* - Handle case of non-arabic numerals that match "\d", giving a warning
		message when trying to format them with sprintf's "%d".  
		
		This caused problems when trying to print shortened floats (only 2 
		digits to right of decimal point): "%.2f" complained about UTF8
		numeric chars not being numbers.  P detects such and prints the 
		result as a string instead of using "%f" or "%d".
* - Beef up tests in self-test to test instances of some of the bugs
		that also work as examples.


Dependency Change:
*  - execution depends on 'mem' due to having included unpublished
		 code from a Types module.