ne1uno · June 7, 2025 13:20
diff --git a/parse-??.red b/parse-??.red
 get red-lang from red-lang.org

 takeoff on a red-lang gitter post from 2018 about parse-?? custom tracing
 un-comment at do-parse	to see the demo parse-trace and parse/trace :callback
 gui is from a simplified parse tab in my unpublished ne1 clean-script tool

 cGPT suggested some utf color markers for succeed/fail that improve readability
 difficult to imagine how to make trace debug parse rules without showing everything
 
 	;credits,
 	; 2018 post in gitter/red/parse @endo64, @gltewalt, @gltewalt & others for parse ideas
 	; splitters I think from live code enhanced on gist or @toomasv dx2 demo?

 the examples are really to demo the custom trace and maybe not great parse examples
 the sliders and area size calculations are a bit sloppy, please don't blame the sources
 they need a limit so they can't be dropped in the wrong place 
 resizing doesn't account for scroll bars propeely so it's a bit of an adventure

 Red [
 	Title: 		"parse tracing"
 	File:		%parse-??.red
 	Version:	0.1
 	Needs:  'View
 ]


 	; system/view/debug?: on  ; does step by step print of layout
 	
 line-end:	"^/" ; crlf in area in windows seems to work better? otherwise ^M or ^/
 random/seed now/time/precise  


 			;parse gui is setup to run a function live to demonstrate the parse trace
 			; a top row of buttons to clear areas, parse and quit
 			; p1..p4 area/text with 3 horizontal splitters between them
 			; there is no error reporting in the virtual function used in the demo
 			; they may just silently fail with typo or syntax errors, be warned.

 P-def: random 4
 		; select 1st 2nd 3rd nth example, drop-list to choose in ui however many
 		; each block of P-drop is 3 strings.
 		; [[{title}{function}{data}]...] , output of function will be to p3. trace to p4
 		; <1> message for p3 <2> function spec & body for p1 <3> data to parse
 		; form gets rendered and errors on some char literal is why charset reduce works
 		; there few no error checks on the function syntax in this demo
 		; if you see nothing happening when you click parse that may be one reason why
 		; p1 area is not a full editor so has the normal area editing, it's a demo...
 		; p2 area data needs to be loadable as string or block
 		; the examples don't have to use parse, it is normal red language code
 		; it might've been simpler to read/write example scripts to/from disk

 P-drop: [ 
 	[{parse-block-demo}  ;<P-d1> simpler test,   
 	{["function spec & body body to parse data from P2 area"
 	data [series! block!] "from P2 area"
 	/local out spc rule et ;tag val not sure how parse words propagate
 	][ ;could be dangerous, don't do anything stupid
 	out: clear []
 	print ['data mold data]

 	rule: [
 		some [
 			copy tag set-word!
 			copy val [integer! | string! | set-word! | word!]
 			(append/only out reduce [tag val])
 		]
 	]

 	; et: parse-trace data rule	;default-trace prints to console
 	; et: parse data rule		 ;no-trace
 	et: parse-?? data rule 45  ;n=depth custom trace redirect enabled

 	reduce [et out]
 	]}
 	{[name: "Alice" age: 30 logged-in: true]}
 	]

 	; converts string like "2 +3 -4 'cmt" to [2 + 3 - 4] ignoring 'cmt not much error checking
 	[{parse-mlstring-code}  ;<P-d2>
 	{["function spec & body body to parse data from P2 area"
 	data [series! block!] "from P2 area"
 	/local out rule spcs et lf digit sign ncmt v vars i x 
 	][ ;could be dangerous, don't do anything stupid
 	out: clear []
 	;data: "1 +3 'cmt"
 	print ['data mold data]
 	
 	;crlf? area renders char literal
 	;note: ^^+-	 as tab, ^^+/ will embed a newline string or comment
 	
 	spcs: charset reduce [space tab newline]
 	lf: charset reduce [newline]
 	digit: charset "0123456789."
 	sign: charset "+-"
 	ncmt: complement charset reduce ["'" newline]

 	rule: [
 			some [
 				some [spcs | end]
 				| [copy v [some [sign]] (append out v)]
 				| [copy v [some [digit]] (append out v)]
 				| ["'" any ncmt any [lf | end]]  ; skip comment to end of line
 			]
 	]

 	; et: parse-trace data rule	;default-trace prints to console
 	; et: parse data rule		 ;no-trace
 	et: parse-?? data rule 45  ;n=depth custom trace redirect enabled

 	;rework the strings
 	print ["math ans=" attempt [math probe vars: collect [
 		foreach [i x] probe out [
 				keep reduce [to-float i]
 				keep either x [first to-block x][continue]
 			  ]
 		]]]
 	print reduce [";ans=" do vars]
 	append p3/text " p3/text output functest " ;don't try to embed ^-^/
 	reduce [et out]
 	]}
 	 {{5 +8 'cmt}} ;not a math replacement, only sums signed numbers as a demo
 	]
 							
 	[{parse-string-char-counter}  ;<P-d4> ; show skip if not a letter
 	{["function spec & body body to parse data from P2 area"
 	data [series!] "from P2 area"
 	/local out spc rule et letters counts ;ch cur
 	][ ;could be dangerous, don't do anything stupid
 	out: clear []
 	print ['data mold trim data]
 	
 	letters: charset "abcdefghijklmnopqrstuvwxyz"
 	counts: make map! []

 	rule: [
 		some [
 			copy ch [letters | skip]
 			(
 				either all [ch curr: counts/:ch] [counts/:ch: curr + 1][counts/:ch: 1]
 			)
 			;end
 		]
 	]
 	et: parse-?? data rule 45  ;n=depth custom trace redirect enabled
 	probe append out counts
 	reduce [et out]
 	]}
 	{"all 5"}
 	]

 	[{parse-string-block-code}  ;<P-d4> empty template
 	{["function spec & body body to parse data from P2 area"
 	data [string! block!] "from P2 area"
 	/local out spc rule et ;var
 	][ ;could be dangerous, don't do anything stupid
 	out: clear []
 	print ['data mold trim data]
 	
 	spc: charset reduce [space tab]
 	rule: [
 		some [copy var [word! | string! | end] (append out var)]
 	] ;
 	et: parse-?? data rule 45  ;n=depth custom trace redirect enabled

 	reduce [et out]
 	]}
 	{[data block "data string"]
 	}
 	]
 ]

 ;for cli use bypass gui area/text
 p4a: none
 to-area: no

 init-areas: func [
 	P-def "integer block of P-drop to use"
 ][
 	p1/text: P-drop/:P-def/2
 	p2/text: P-drop/:P-def/3
 	p3/text: rejoin ["parsed results ^M^/^-" 
 							P-drop/:P-def/1
 						]
 	p4/text: {parse trace ^M^/}
 ]

 p-indent:  make string! 30
 parse-skips: 0 ;0: all 1: fails 2: etc  TBD

 					; can override while testing print also to console
 					;line endings tricky in windows area sometimes
 					
 print-et: function ["mimic print echo to etrace area, caller should reduce blk?"
 	blk [string! block!] ;
 	/local x
 ][
 	; if block? blk [
 		; repeat x length? blk [
 			; blk/x: mold blk/x
 		; ]
 		; blk: join/with blk "^/"
 	; ]
 		print blk   ;option to echo to console or not (only if don't redirect print)
 		
 		;etrace/text: rejoin [etrace/text mold blk line-end]  ;included brackets & quotes
 		
 		if to-area [ ;area on windows needs crnl?
 			;p4a/text: rejoin [p4a/text replace/all blk "^/" "^M^/"]
 			;p4a/text: rejoin [p4a/text mold blk]
 			p4a/text: rejoin [p4a/text to-string blk newline]
 			
 		]	;newline, trace adds newline between match & input
 ]
 	
 on-parse-compact: function [
 	"Compact and clear parse/trace event printer"
 	event [word!] 
 	match? [logic!] 
 	rule [block!] 
 	input [series!] 
 	stack [block!]
 ][
 	switch event [
 		push  [
 			; print-et [p-indent] ;"-->"  ;need no nl option print-et
 			append p-indent "  "
 		]
 		pop	  [
 			clear back back tail p-indent
 			; print-et [p-indent "<--"]  ;is on line by itself, just adds noise really
 		] 
 		fetch [
 			print-et rejoin [
 			   p-indent "R: " mold/flat/part rule 40 line-end 	; newline added by caller
 			   p-indent "I: " either parse-skips = 0  ["  At:    " mold/flat/part input 30][""]
 			]
 		]
 		match [
 			print-et rejoin [
 				p-indent either match? ["✅"] ["❌"] ;green & red on console, newline added
 				;not sure why need to repeat rule, is 2 lines above
 			   ; either parse-skips = 0 [mold/flat/part rule 60][""] ; newline caller added, skip detail
 			]
 		]
 		end [
 			print-et reduce ["Done: " mold match?]
 		]
 	]
 	true
 ]

 						;\gist\repos\repo-red\Oldes\Red-master\environment\functions.red
 						;the default callback in parse-trace I think
 on-parse-event: func [
 	event	[word!]   "Trace events: push, pop, fetch, match, iterate, paren, end"
 	match?	[logic!]  "Result of last matching operation"
 	rule	[block!]  "Current rule at current position"
 	input	[series!] "Input series at next position to match"
 	stack	[block!]  "Internal parse rules stack"
 	return: [logic!]  "TRUE: continue parsing, FALSE: stop and exit parsing"
 ][
 	switch event [
 		push  [
 			print [p-indent "-->"]
 			append p-indent "  "
 		]
 		pop	  [
 			clear back back tail p-indent
 			print [p-indent "<--"]
 		]
 		fetch [
 			print [
 				p-indent "match:" mold/flat/part rule 50 line-end
 				p-indent "input:" mold/flat/part input 50 p-indent
 			]
 		]
 		match [print [p-indent "==>" pick ["matched" "not matched"]  match?]]
 		end   [print ["return:" match?]]
 	]
 	true
 ]

 parse-??: function [{Wrapper for parse/trace using ?? hook w/o the ?? hook} 
 	input [series!] 
 	rules [block!] 
 	/case 
 	/part 
 	/skips sps
 	limit [integer!] 
 	;return: [logic! block!]
 ][
 	parse-skips: any [sps 1] ;0 all, 1 less detail, RSN
 	
 	clear p-indent ;no longer works to indent red 22B
 	parse-event: :on-parse-compact  ; :on-parse-event-mod	;on-parse-?? option TBD

 	either case [
 		print ["parse/case/trace input rules :on-parse-??"]
 		parse/case/trace input rules :parse-event
 	][
 		either part [
 			print ["parse/part/trace input rules :on-parse-??"]
 			parse/part/trace input rules limit :parse-event
 		][
 			print-et  reduce ["parse/trace input rules :on-parse-??"]
 			parse/trace input rules :parse-event
 		]
 	]
 ]


 ;; June, 2018 https://rebol.tech/gitter.im/red/parse/2018/#msg5b164af7edd06f7d150e7faa
 ;; endo64> I saw "??" usage in Respectech's http://video.respectech.com:8080/tutorial/r3/index.r3 tutorial
 ;;    We don't have ?? in Red/Parse, is it handy? 
 ;; Should we consider to add it, or parse-trace is already do what ?? does?
 ;; gltewalt> Looks like [pindent] was a todo, to put in non-global?
 ;; greggirwin> Thanks @gltewalt 
 ;; parse-?? "dog" [ ?? "d" ?? [ "i" | "o" ] ?? "g" ?? ]



 ;?? FIXME, I think the idea was ?? would be replaced in the rules or trigger more info?
 ; @gltewalt went on to work on parse caddy but it didn't show traces or have ??
 

 do-parse: does [
 		rule: [["d" | "c"] [ "o" | "a" ]  [ "g" | "t" ]]

 		parse-trace "dog" rule 

 		parse-?? "cat" rule
 	]

 ; do-parse	; un-comment to show compact trace to console
 ; halt


 						; cfg stuff lifted from larger multi panel layout
 						; not all of it was needed for this parse trace demo
 			
 ; not sure of the utility of this allocation, lifted from a red editor script
 ; I've also heard there's a 32k area/text limit, they seem to work ok for small demo's
 p1script: make string! 30000  
 p2script: make string! 60000  
 p3script: make string! 40000  
 p4script: make string! 90000  


 out: copy ""	;replaced later but here incase of compiler confusion
 f: func [][]	;replaced later
 fs: func [][]	;replaced later


 	init-pos: 	 50x5		;offset upper left of screen
 	window-size: 760x560	;x650 was too close for some laptop x700 & taskbar

 	T-size:	window-size - 5x30	  ;room for status with n rows of buttons

 	H-size:	as-pair (T-size/x - 12) (7)	;hoiz splitter init

 	;parse area inits. not sure how <>100% works
 	P1-size: as-pair (T-size/x - 12)  to-integer (T-size/y * 30%)
 	P2-size: as-pair (T-size/x - 12)  to-integer (T-size/y * 10%)
 	P3-size: as-pair (T-size/x - 12)  to-integer (T-size/y * 25%)
 	P4-size: as-pair (T-size/x - 12)  to-integer (T-size/y * 25%)


 a_color: 240.240.240
 hvs_coloron: gray
 hvs_coloroff: blue


 afont: make font! [ ;ui area
 	   name: "Consolas"    ;
 		size: 11
 		color: coal
 	]  
 ifont: make font! [ ;ui buttons
 	   name: "Consolas"    ;
 		size: 9
 		color: coal
 	]  


 gview: layout [
 	title "parse Text quad Pane"
 	size window-size
 	backdrop ivory

 	style btn: button 92x17  center font ifont pewter	;~9 char
 	style text: text  font afont  snow pewter
 	style drop-list: drop-list font afont snow pewter	;~9 char
 	style area: area a_color font afont
 	style splitter: base 5x5 hvs_coloron loose  ;make I-beam cursor


 		across
 		pad -2x-4
 		btn "cls P2-data" [
 					 clear p1/text
 				   ]
 		btn "cls P3-results" [
 					 clear p3/text
 				   ]
 		btn "cls P4-trace" [
 					 clear p4/text
 				   ]
 		pad 20x0 
 		text 90x16 "examples:"
 		drop-list 30x16 data [] [
 				;print [face/selected]
 			]
 		on-created [ ;wonder can this fail if areas not created yet?
 			repeat ix length? P-drop [append face/data form ix]
 			face/selected: P-def
 			init-areas face/selected
 		]
 		on-change [
 			init-areas face/selected
 		]
    
    
    
 		btn "Parse"  [	; p1 func rule p2 data p3 results p4 trace
 						;build dynamic function and run it
 						;need a way to trap error, 
 			; just as likely to get syntax error that won't point to users function or data
 			; could check structure & open/close balance but way beyond trace demo already
 			
 			either 
 				;all [attempt [load to-string first blk] attempt [load to-string second blk]]
 				all [attempt [blk:  to-block p1/text]  attempt [first blk] attempt [second blk]]
 			[
 				f: function first blk second blk
 			][
 				
 				print "p1 function has malformed blocks?" ;FIXME
 				blk:  to-block p1/text
 				print [mold first blk newline mold second blk]
 			] 
 			; probe spec-of :f
 			; probe reflect :f 'body
 			; ? f
 			; ?? f
 			; probe find spec-of :f 'local  ; check all setwords if use function


 			;check debug cb or don't bother echo
 			if no [
 				append p3/text rejoin [ newline "******" newline
 					 'f-s mold spec-of :f newline 'f-b	mold body-of :f  newline
 					 'data mold p2/text  
 					 newline newline
 			]]

 			px: copy ";in case attempt fails" 
 			append p3/text rejoin [
 				newline "------- mold" newline
 				 ;ok for string or block litteral in data
 				 mold attempt [px: f attempt [load p2/text]]  
 				 newline "-------form " newline
 				 form px
 				 newline newline	;last line in area sometimes truncated
 			]

 			
 		] 
 		;add cd no-trace or spinbox verbose
 		;add  cb string/block

 		pad 24x0 
 		btn  "QuitAll"  [quit] ;same compiled?

 		btn 38x15 "Exit"  [print [";do system/options/script"] unview] ;prints to history


 	
 		return
 		pad -2x-4

 		below
 		p1: area  P1-size p1script  react [
 				face/size/x: face/parent/size/x - 10
 			]
 		
 		splitter-hp1: splitter H-size react [
 				face/size/x: face/parent/size/x - 10
 			]
 			on-down [ofsx1: face/offset/x system/view/auto-sync?: off] 
 			on-up [system/view/auto-sync?: on] 
 			on-drag [
 				face/offset/x: ofsx1 
 				p1/size/y: face/offset/y - 10 
 				p2/offset/y: face/offset/y + 10 
 				p2/size/y: face/parent/size/y - face/offset/y - 2 ;20 
 				show [p1 p2 face]
 			]  on-over [face/color: either event/away? [hvs_coloron][hvs_coloroff]] 

 		p2: area P2-size p2script react [
 				face/size/x: face/parent/size/x
 			]

 		splitter-hp2: splitter H-size  react [
 				face/size/x: face/parent/size/x - 10
 			]
 			on-down [ofsx2: face/offset/x system/view/auto-sync?: off] 
 			on-up [system/view/auto-sync?: on] 
 			on-drag [
 				face/offset/x: ofsx2 
 				p2/size/y: face/offset/y - 10 
 				p3/offset/y: face/offset/y + 10 
 				p3/size/y: face/parent/size/y - face/offset/y - 2 ;20 
 				show [p2 p3 face]
 			]  on-over [face/color: either event/away? [hvs_coloron][hvs_coloroff]] 

 		p3: area P3-size p3script  react [
 				face/size/x: face/parent/size/x - 10
 			]

 		splitter-hp3: splitter H-size react [
 				face/size/x: face/parent/size/x
 			]
 			on-down [ofsx3: face/offset/x system/view/auto-sync?: off] 
 			on-up [system/view/auto-sync?: on] 
 			on-drag [
 				face/offset/x: ofsx3 
 				p3/size/y: face/offset/y - 10 
 				p4/offset/y: face/offset/y + 10 
 				p4/size/y: face/parent/size/y - face/offset/y - 2 ;20 
 				show [p3 p4 face]
 			]  on-over [face/color: either event/away? [hvs_coloron][hvs_coloroff]] 

 		p4: area P4-size p4script  react [
 				face/size/x: face/parent/size/x - 10 
 			]
 	]

 		; copy to above print-et if compiling
 		p4a: p4			;trace area object
 		to-area: on		; set toarea to do print-et redirect

 	;don't use /no-wait compiled or w/o opening console first
 	probe init-pos  ;open console interpreted
 	
 	view/no-wait/flags/options gview ['resize][offset: init-pos]



 comment { don't debug the comments
 }
	get red-lang from red-lang.org

	takeoff on a red-lang gitter post from 2018 about parse-?? custom tracing
	un-comment at do-parse to see the demo parse-trace and parse/trace :callback
	gui is from a simplified parse tab in my unpublished ne1 clean-script tool

	cGPT suggested some utf color markers for succeed/fail that improve readability
	difficult to imagine how to make trace debug parse rules without showing everything

	;credits,
	; 2018 post in gitter/red/parse @endo64, @gltewalt, @gltewalt & others for parse ideas
	; splitters I think from live code enhanced on gist or @toomasv dx2 demo?

	the examples are really to demo the custom trace and maybe not great parse examples
	the sliders and area size calculations are a bit sloppy, please don't blame the sources
	they need a limit so they can't be dropped in the wrong place
	resizing doesn't account for scroll bars propeely so it's a bit of an adventure

	Red [
	Title: "parse tracing"
	File: %parse-??.red
	Version: 0.1
	Needs: 'View
	]


	; system/view/debug?: on ; does step by step print of layout

	line-end: "^/" ; crlf in area in windows seems to work better? otherwise ^M or ^/
	random/seed now/time/precise


	;parse gui is setup to run a function live to demonstrate the parse trace
	; a top row of buttons to clear areas, parse and quit
	; p1..p4 area/text with 3 horizontal splitters between them
	; there is no error reporting in the virtual function used in the demo
	; they may just silently fail with typo or syntax errors, be warned.

	P-def: random 4
	; select 1st 2nd 3rd nth example, drop-list to choose in ui however many
	; each block of P-drop is 3 strings.
	; [[{title}{function}{data}]...] , output of function will be to p3. trace to p4
	; <1> message for p3 <2> function spec & body for p1 <3> data to parse
	; form gets rendered and errors on some char literal is why charset reduce works
	; there few no error checks on the function syntax in this demo
	; if you see nothing happening when you click parse that may be one reason why
	; p1 area is not a full editor so has the normal area editing, it's a demo...
	; p2 area data needs to be loadable as string or block
	; the examples don't have to use parse, it is normal red language code
	; it might've been simpler to read/write example scripts to/from disk

	P-drop: [
	[{parse-block-demo} ;<P-d1> simpler test,
	{["function spec & body body to parse data from P2 area"
	data [series! block!] "from P2 area"
	/local out spc rule et ;tag val not sure how parse words propagate
	][ ;could be dangerous, don't do anything stupid
	out: clear []
	print ['data mold data]

	rule: [
	some [
	copy tag set-word!
	copy val [integer! \| string! \| set-word! \| word!]
	(append/only out reduce [tag val])
	]
	]

	; et: parse-trace data rule ;default-trace prints to console
	; et: parse data rule ;no-trace
	et: parse-?? data rule 45 ;n=depth custom trace redirect enabled

	reduce [et out]
	]}
	{[name: "Alice" age: 30 logged-in: true]}
	]

	; converts string like "2 +3 -4 'cmt" to [2 + 3 - 4] ignoring 'cmt not much error checking
	[{parse-mlstring-code} ;<P-d2>
	{["function spec & body body to parse data from P2 area"
	data [series! block!] "from P2 area"
	/local out rule spcs et lf digit sign ncmt v vars i x
	][ ;could be dangerous, don't do anything stupid
	out: clear []
	;data: "1 +3 'cmt"
	print ['data mold data]

	;crlf? area renders char literal
	;note: ^^+- as tab, ^^+/ will embed a newline string or comment

	spcs: charset reduce [space tab newline]
	lf: charset reduce [newline]
	digit: charset "0123456789."
	sign: charset "+-"
	ncmt: complement charset reduce ["'" newline]

	rule: [
	some [
	some [spcs \| end]
	\| [copy v [some [sign]] (append out v)]
	\| [copy v [some [digit]] (append out v)]
	\| ["'" any ncmt any [lf \| end]] ; skip comment to end of line
	]
	]

	; et: parse-trace data rule ;default-trace prints to console
	; et: parse data rule ;no-trace
	et: parse-?? data rule 45 ;n=depth custom trace redirect enabled

	;rework the strings
	print ["math ans=" attempt [math probe vars: collect [
	foreach [i x] probe out [
	keep reduce [to-float i]
	keep either x [first to-block x][continue]
	]
	]]]
	print reduce [";ans=" do vars]
	append p3/text " p3/text output functest " ;don't try to embed ^-^/
	reduce [et out]
	]}
	{{5 +8 'cmt}} ;not a math replacement, only sums signed numbers as a demo
	]

	[{parse-string-char-counter} ;<P-d4> ; show skip if not a letter
	{["function spec & body body to parse data from P2 area"
	data [series!] "from P2 area"
	/local out spc rule et letters counts ;ch cur
	][ ;could be dangerous, don't do anything stupid
	out: clear []
	print ['data mold trim data]

	letters: charset "abcdefghijklmnopqrstuvwxyz"
	counts: make map! []

	rule: [
	some [
	copy ch [letters \| skip]
	(
	either all [ch curr: counts/:ch] [counts/:ch: curr + 1][counts/:ch: 1]
	)
	;end
	]
	]
	et: parse-?? data rule 45 ;n=depth custom trace redirect enabled
	probe append out counts
	reduce [et out]
	]}
	{"all 5"}
	]

	[{parse-string-block-code} ;<P-d4> empty template
	{["function spec & body body to parse data from P2 area"
	data [string! block!] "from P2 area"
	/local out spc rule et ;var
	][ ;could be dangerous, don't do anything stupid
	out: clear []
	print ['data mold trim data]

	spc: charset reduce [space tab]
	rule: [
	some [copy var [word! \| string! \| end] (append out var)]
	] ;
	et: parse-?? data rule 45 ;n=depth custom trace redirect enabled

	reduce [et out]
	]}
	{[data block "data string"]
	}
	]
	]

	;for cli use bypass gui area/text
	p4a: none
	to-area: no

	init-areas: func [
	P-def "integer block of P-drop to use"
	][
	p1/text: P-drop/:P-def/2
	p2/text: P-drop/:P-def/3
	p3/text: rejoin ["parsed results ^M^/^-"
	P-drop/:P-def/1
	]
	p4/text: {parse trace ^M^/}
	]

	p-indent: make string! 30
	parse-skips: 0 ;0: all 1: fails 2: etc TBD

	; can override while testing print also to console
	;line endings tricky in windows area sometimes

	print-et: function ["mimic print echo to etrace area, caller should reduce blk?"
	blk [string! block!] ;
	/local x
	][
	; if block? blk [
	; repeat x length? blk [
	; blk/x: mold blk/x
	; ]
	; blk: join/with blk "^/"
	; ]
	print blk ;option to echo to console or not (only if don't redirect print)

	;etrace/text: rejoin [etrace/text mold blk line-end] ;included brackets & quotes

	if to-area [ ;area on windows needs crnl?
	;p4a/text: rejoin [p4a/text replace/all blk "^/" "^M^/"]
	;p4a/text: rejoin [p4a/text mold blk]
	p4a/text: rejoin [p4a/text to-string blk newline]

	] ;newline, trace adds newline between match & input
	]

	on-parse-compact: function [
	"Compact and clear parse/trace event printer"
	event [word!]
	match? [logic!]
	rule [block!]
	input [series!]
	stack [block!]
	][
	switch event [
	push [
	; print-et [p-indent] ;"-->" ;need no nl option print-et
	append p-indent " "
	]
	pop [
	clear back back tail p-indent
	; print-et [p-indent "<--"] ;is on line by itself, just adds noise really
	]
	fetch [
	print-et rejoin [
	p-indent "R: " mold/flat/part rule 40 line-end ; newline added by caller
	p-indent "I: " either parse-skips = 0 [" At: " mold/flat/part input 30][""]
	]
	]
	match [
	print-et rejoin [
	p-indent either match? ["✅"] ["❌"] ;green & red on console, newline added
	;not sure why need to repeat rule, is 2 lines above
	; either parse-skips = 0 [mold/flat/part rule 60][""] ; newline caller added, skip detail
	]
	]
	end [
	print-et reduce ["Done: " mold match?]
	]
	]
	true
	]

	;\gist\repos\repo-red\Oldes\Red-master\environment\functions.red
	;the default callback in parse-trace I think
	on-parse-event: func [
	event [word!] "Trace events: push, pop, fetch, match, iterate, paren, end"
	match? [logic!] "Result of last matching operation"
	rule [block!] "Current rule at current position"
	input [series!] "Input series at next position to match"
	stack [block!] "Internal parse rules stack"
	return: [logic!] "TRUE: continue parsing, FALSE: stop and exit parsing"
	][
	switch event [
	push [
	print [p-indent "-->"]
	append p-indent " "
	]
	pop [
	clear back back tail p-indent
	print [p-indent "<--"]
	]
	fetch [
	print [
	p-indent "match:" mold/flat/part rule 50 line-end
	p-indent "input:" mold/flat/part input 50 p-indent
	]
	]
	match [print [p-indent "==>" pick ["matched" "not matched"] match?]]
	end [print ["return:" match?]]
	]
	true
	]

	parse-??: function [{Wrapper for parse/trace using ?? hook w/o the ?? hook}
	input [series!]
	rules [block!]
	/case
	/part
	/skips sps
	limit [integer!]
	;return: [logic! block!]
	][
	parse-skips: any [sps 1] ;0 all, 1 less detail, RSN

	clear p-indent ;no longer works to indent red 22B
	parse-event: :on-parse-compact ; :on-parse-event-mod ;on-parse-?? option TBD

	either case [
	print ["parse/case/trace input rules :on-parse-??"]
	parse/case/trace input rules :parse-event
	][
	either part [
	print ["parse/part/trace input rules :on-parse-??"]
	parse/part/trace input rules limit :parse-event
	][
	print-et reduce ["parse/trace input rules :on-parse-??"]
	parse/trace input rules :parse-event
	]
	]
	]


	;; June, 2018 https://rebol.tech/gitter.im/red/parse/2018/#msg5b164af7edd06f7d150e7faa
	;; endo64> I saw "??" usage in Respectech's http://video.respectech.com:8080/tutorial/r3/index.r3 tutorial
	;; We don't have ?? in Red/Parse, is it handy?
	;; Should we consider to add it, or parse-trace is already do what ?? does?
	;; gltewalt> Looks like [pindent] was a todo, to put in non-global?
	;; greggirwin> Thanks @gltewalt
	;; parse-?? "dog" [ ?? "d" ?? [ "i" \| "o" ] ?? "g" ?? ]



	;?? FIXME, I think the idea was ?? would be replaced in the rules or trigger more info?
	; @gltewalt went on to work on parse caddy but it didn't show traces or have ??


	do-parse: does [
	rule: [["d" \| "c"] [ "o" \| "a" ] [ "g" \| "t" ]]

	parse-trace "dog" rule

	parse-?? "cat" rule
	]

	; do-parse ; un-comment to show compact trace to console
	; halt


	; cfg stuff lifted from larger multi panel layout
	; not all of it was needed for this parse trace demo

	; not sure of the utility of this allocation, lifted from a red editor script
	; I've also heard there's a 32k area/text limit, they seem to work ok for small demo's
	p1script: make string! 30000
	p2script: make string! 60000
	p3script: make string! 40000
	p4script: make string! 90000


	out: copy "" ;replaced later but here incase of compiler confusion
	f: func [][] ;replaced later
	fs: func [][] ;replaced later


	init-pos: 50x5 ;offset upper left of screen
	window-size: 760x560 ;x650 was too close for some laptop x700 & taskbar

	T-size: window-size - 5x30 ;room for status with n rows of buttons

	H-size: as-pair (T-size/x - 12) (7) ;hoiz splitter init

	;parse area inits. not sure how <>100% works
	P1-size: as-pair (T-size/x - 12) to-integer (T-size/y * 30%)
	P2-size: as-pair (T-size/x - 12) to-integer (T-size/y * 10%)
	P3-size: as-pair (T-size/x - 12) to-integer (T-size/y * 25%)
	P4-size: as-pair (T-size/x - 12) to-integer (T-size/y * 25%)


	a_color: 240.240.240
	hvs_coloron: gray
	hvs_coloroff: blue


	afont: make font! [ ;ui area
	name: "Consolas" ;
	size: 11
	color: coal
	]
	ifont: make font! [ ;ui buttons
	name: "Consolas" ;
	size: 9
	color: coal
	]


	gview: layout [
	title "parse Text quad Pane"
	size window-size
	backdrop ivory

	style btn: button 92x17 center font ifont pewter ;~9 char
	style text: text font afont snow pewter
	style drop-list: drop-list font afont snow pewter ;~9 char
	style area: area a_color font afont
	style splitter: base 5x5 hvs_coloron loose ;make I-beam cursor


	across
	pad -2x-4
	btn "cls P2-data" [
	clear p1/text
	]
	btn "cls P3-results" [
	clear p3/text
	]
	btn "cls P4-trace" [
	clear p4/text
	]
	pad 20x0
	text 90x16 "examples:"
	drop-list 30x16 data [] [
	;print [face/selected]
	]
	on-created [ ;wonder can this fail if areas not created yet?
	repeat ix length? P-drop [append face/data form ix]
	face/selected: P-def
	init-areas face/selected
	]
	on-change [
	init-areas face/selected
	]



	btn "Parse" [ ; p1 func rule p2 data p3 results p4 trace
	;build dynamic function and run it
	;need a way to trap error,
	; just as likely to get syntax error that won't point to users function or data
	; could check structure & open/close balance but way beyond trace demo already

	either
	;all [attempt [load to-string first blk] attempt [load to-string second blk]]
	all [attempt [blk: to-block p1/text] attempt [first blk] attempt [second blk]]
	[
	f: function first blk second blk
	][

	print "p1 function has malformed blocks?" ;FIXME
	blk: to-block p1/text
	print [mold first blk newline mold second blk]
	]
	; probe spec-of :f
	; probe reflect :f 'body
	; ? f
	; ?? f
	; probe find spec-of :f 'local ; check all setwords if use function


	;check debug cb or don't bother echo
	if no [
	append p3/text rejoin [ newline "******" newline
	'f-s mold spec-of :f newline 'f-b mold body-of :f newline
	'data mold p2/text
	newline newline
	]]

	px: copy ";in case attempt fails"
	append p3/text rejoin [
	newline "------- mold" newline
	;ok for string or block litteral in data
	mold attempt [px: f attempt [load p2/text]]
	newline "-------form " newline
	form px
	newline newline ;last line in area sometimes truncated
	]


	]
	;add cd no-trace or spinbox verbose
	;add cb string/block

	pad 24x0
	btn "QuitAll" [quit] ;same compiled?

	btn 38x15 "Exit" [print [";do system/options/script"] unview] ;prints to history



	return
	pad -2x-4

	below
	p1: area P1-size p1script react [
	face/size/x: face/parent/size/x - 10
	]

	splitter-hp1: splitter H-size react [
	face/size/x: face/parent/size/x - 10
	]
	on-down [ofsx1: face/offset/x system/view/auto-sync?: off]
	on-up [system/view/auto-sync?: on]
	on-drag [
	face/offset/x: ofsx1
	p1/size/y: face/offset/y - 10
	p2/offset/y: face/offset/y + 10
	p2/size/y: face/parent/size/y - face/offset/y - 2 ;20
	show [p1 p2 face]
	] on-over [face/color: either event/away? [hvs_coloron][hvs_coloroff]]

	p2: area P2-size p2script react [
	face/size/x: face/parent/size/x
	]

	splitter-hp2: splitter H-size react [
	face/size/x: face/parent/size/x - 10
	]
	on-down [ofsx2: face/offset/x system/view/auto-sync?: off]
	on-up [system/view/auto-sync?: on]
	on-drag [
	face/offset/x: ofsx2
	p2/size/y: face/offset/y - 10
	p3/offset/y: face/offset/y + 10
	p3/size/y: face/parent/size/y - face/offset/y - 2 ;20
	show [p2 p3 face]
	] on-over [face/color: either event/away? [hvs_coloron][hvs_coloroff]]

	p3: area P3-size p3script react [
	face/size/x: face/parent/size/x - 10
	]

	splitter-hp3: splitter H-size react [
	face/size/x: face/parent/size/x
	]
	on-down [ofsx3: face/offset/x system/view/auto-sync?: off]
	on-up [system/view/auto-sync?: on]
	on-drag [
	face/offset/x: ofsx3
	p3/size/y: face/offset/y - 10
	p4/offset/y: face/offset/y + 10
	p4/size/y: face/parent/size/y - face/offset/y - 2 ;20
	show [p3 p4 face]
	] on-over [face/color: either event/away? [hvs_coloron][hvs_coloroff]]

	p4: area P4-size p4script react [
	face/size/x: face/parent/size/x - 10
	]
	]

	; copy to above print-et if compiling
	p4a: p4 ;trace area object
	to-area: on ; set toarea to do print-et redirect

	;don't use /no-wait compiled or w/o opening console first
	probe init-pos ;open console interpreted

	view/no-wait/flags/options gview ['resize][offset: init-pos]



	comment { don't debug the comments
	}