fix(Textbox): implemente a fix for the style shifting issues on new lines

[asturur]

Motivation

not sure this is 100% safe and needs to be tested rigorously.
When using get2dCursorLocation ignoring wrapping, the missingnewlineoffset is not influent, because is a wrapping only issue. get2dCursorLocation is used in insertNewStyleBlock to determine the style to copy, and in that case we have to ignore the specific of splitByGrapheme because we are making counts on unwrapped text lines.

But what about all the other times we call missingnewlineoffset outside the get2dCursorLocation? that we need to find out.

It would be probably better to just have get2dCursorLocation call a specific function that is different, or have get2dCursorLocation different for textbox.

Note:

As pointed out by @ShaMan123 this is a bug from 5.x not a regression

Description

Changes

Gist

In Action

[github-actions]

Build Stats

file / KB (diff)	bundled	minified
fabric	915.655 (+0.139)	305.470 (+0.008)

[github-actions]

Coverage after merging fix-style-issue into master will be

82.97%

Coverage Report

File	Stmts	Branches	Funcs	Lines	Uncovered Lines
index.node.ts	7.69%	100%	0%	14.29%	17, 20, 23, 35, 38, 41
src
ClassRegistry.ts	100%	100%	100%	100%
Collection.ts	94.71%	94.64%	86.67%	97.09%	101, 104, 207–208, 233–234
CommonMethods.ts	96.55%	87.50%	100%	100%	9
Intersection.ts	100%	100%	100%	100%
Observable.ts	87.76%	85.29%	87.50%	89.58%	134–135, 160–161, 32–33, 41, 50, 78, 89
Point.ts	100%	100%	100%	100%
Shadow.ts	98.36%	95.65%	100%	100%	178
cache.ts	97.06%	90%	100%	100%	57
config.ts	75%	66.67%	66.67%	82.76%	130, 138, 138, 138, 138, 138–140, 151–153
constants.ts	100%	100%	100%	100%
src/Pattern
Pattern.ts	92.31%	91.89%	90%	93.55%	118, 129, 138, 31, 94
src/brushes
BaseBrush.ts	100%	100%	100%	100%
CircleBrush.ts	0%	0%	0%	0%	108, 108, 108, 110, 112, 114–116, 118–121, 128–129, 13, 136, 138, 23–24, 32–36, 40–44, 51–54, 62–66, 68, 76, 76, 76, 76, 76–77, 79, 79, 79–82, 84, 92–93, 95, 97–99
PatternBrush.ts	97.06%	87.50%	100%	100%	21
PencilBrush.ts	91.06%	82.35%	100%	93.81%	122–123, 152, 152–154, 176, 176, 276, 280, 285–286, 68–69, 84–85
SprayBrush.ts	0%	0%	0%	0%	107, 107, 107, 107, 107–108, 110–111, 118–119, 121, 123–127, 136, 140–141, 141, 149, 149, 149–152, 154–157, 161–162, 164, 166–169, 17, 172, 179, 18, 180, 182, 184–185, 187, 194–195, 197–198, 20, 201, 201, 208, 208, 21, 212, 22, 22, 22–24, 28, 32, 39, 46, 53, 60, 67, 84–86, 94–96, 98–99
src/canvas
Canvas.ts	79.12%	77.68%	83.05%	79.61%	1000, 1003–1004, 1004, 1004, 1006, 1014, 1014, 1014–1016, 1016, 1016, 1023–1024, 1032–1033, 1033, 1033–1034, 1040, 1042, 1073–1075, 1078–1079, 1083–1084, 1197–1199, 1202–1203, 1276, 1395, 1517, 157, 182, 292–293, 296–300, 305, 328–329, 334–339, 359, 359, 359–360, 360, 360–361, 369, 37, 374–375, 375, 375–376, 378, 387, 393–394, 394, 394, 41, 437, 445, 449, 449, 449–450, 452, 535–536, 536, 536–537, 543, 543, 543–545, 565, 567, 567, 567–568, 568, 568, 571, 571, 571–572, 575, 584–585, 587–588, 590, 590–591, 593–594, 606–607, 607, 607–608, 610–615, 621, 628, 665, 665, 665, 667, 669–674, 680, 686, 686, 686–687, 689, 692, 697, 710, 738, 738, 796–797, 797, 797–798, 800, 803–804, 804, 804–805, 807–808, 811, 811–813, 816–817, 887, 899, 906, 927, 959, 980–981, 997–998, 998, 998–999
CanvasOptions.ts	100%	100%	100%	100%
SelectableCanvas.ts	94.44%	92.28%	94.23%	96.18%	1099, 1101, 1103–1104, 301, 478–479, 481–482, 482, 482, 531–532, 593–594, 647–649, 681, 686–687, 714–715, 899, 899–900, 903, 923, 923, 972, 980
StaticCanvas.ts	96.78%	93.09%	100%	98.53%	1030, 1040, 1092–1093, 1096, 1131–1132, 1208, 1217, 1217, 1221, 1221, 1268–1269, 185–186, 202, 570, 582–583, 913–914, 914, 914–915
StaticCanvasOptions.ts	100%	100%	100%	100%
TextEditingManager.ts	84.31%	71.43%	91.67%	88%	17–18, 18, 18–19, 19, 19
src/canvas/DOMManagers
CanvasDOMManager.ts	95.52%	70%	100%	100%	21–22, 29
StaticCanvasDOMManager.ts	100%	100%	100%	100%
util.ts	86.67%	80.56%	83.33%	93.94%	14, 26, 63–64, 67, 67, 74, 93–94
src/color
Color.ts	94.96%	91.67%	96.30%	96.05%	233, 258–259, 267–268, 48
color_map.ts	100%	100%	100%	100%
constants.ts	100%	100%	100%	100%
util.ts	85.71%	76.92%	100%	89.74%	55–56, 56, 58, 58, 58–59, 61–62, 89
src/controls
Control.ts	93.33%	87.88%	91.67%	97.78%	175, 240, 327, 327, 362
changeWidth.ts	100%	100%	100%	100%
commonControls.ts	100%	100%	100%	100%
controlRendering.ts	81.63%	78%	100%	84.78%	106, 111, 121, 121, 45, 50, 61, 61, 65–72, 81–82
drag.ts	100%	100%	100%	100%
fireEvent.ts	88.89%	75%	100%	100%	13
polyControl.ts	6.06%	0%	0%	11.43%	103, 117, 117, 117, 117, 117, 119–122, 122, 125, 132, 17, 25–29, 29, 29, 29, 29, 29, 29, 29, 50–56, 56, 56, 56, 56, 58, 63–64, 66, 76, 82–84, 84, 86, 89–90, 90, 90, 90, 90, 92, 97
rotate.ts	19.57%	12.50%	50%	21.43%	41, 45, 51, 51, 51–52, 55–57, 59, 59, 59, 59, 59–61, 61, 61–63, 65, 65, 65–67, 67, 67–68, 73, 73, 73–74, 76, 78, 80–81
scale.ts	93.57%	92.94%	100%	93.67%	129–130, 132–134, 148–149, 181–183, 42
scaleSkew.ts	78.79%	64.29%	100%	85.71%	27, 29, 29, 29, 31, 33, 35
skew.ts	91.03%	79.31%	100%	97.67%	131–132, 163–164, 171, 177, 179
util.ts	100%	100%	100%	100%
wrapWithFireEvent.ts	100%	100%	100%	100%
wrapWithFixedAnchor.ts	100%	100%	100%	100%
src/env
browser.ts

[asturur]

I'll wait resolution of #9192 in order to write a test for this

[asturur]

needs a different test to show what is fixing.
Is probably better to first commit the test and make snapshot with current master and the have this PR fix the snapshots

[ShaMan123]

hard to review this thing.
there is still confusion around usage of wrapped/unwrapped as we discussed.
good luck

[ShaMan123]

what is this, looks cool

[asturur]

this is to avoid that we run 2 fabricJS tests in the same browser window and we mix up the copy paste status that is stored globally.
Or at least it solved that for me since code that was behaving normally when using my hands wasn't behaving the same when using the tests in parallel.
Should affect ONLY this file here.

[asturur]

hard to review this thing. there is still confusion around usage of wrapped/unwrapped as we discussed. good luck

I can add more explanations.

Every new line added with \n adds a char, but doesn't take a space in the style.
This is represented with the function: missingNewlineOffset

If the text is wrapping inside the textbox, the \n is not in the line and so we do not have this newLineOffset to subtract, the function returns 0 while all the other time returns 1.

Now get2dCursorLocation is used both on wrapped and unwrapped lines and that is done with the skipWrapping boolean.
If skipWrapping is ON, we shouldn't search for wrapped lines in that missingNewlineOffset and we short cut to return the standard value of 1.

Now yes there is still confusion about that and text could be way better, but is not going to be refactored live there is too much risks of bugs and breakage and text is one of the most important features for apps.

The text code is confusing, but this change is made in a way that is not confusing to review because is focused on the exact problem that is causing styling disruption.