Fix LT-21839: Avoid crash and allow paste success with no ws

[jasonleenaylor]

Analyzing the data that caused the crash showed that in some edge
cases no writing system can be identified in MakeSubString.
This could cause an empty builder to have no writing system.
I couldn't identify any actual side effects in this edge case so
I left the assert, but allowed the paste to succeed

---

This change is 

[JakeOliver28]

Reviewable status: 0 of 1 files reviewed, 1 unresolved discussion (waiting on @jasonleenaylor)

---

Src/views/VwSelection.cpp line 5358 at r1 (raw file):

			}

			if(wsNew <= 0)

Could we just put the Assert and return inside of the first "if (wsNew <=0)" instead of having a duplicate if statement?

[jasonleenaylor]

Src/views/VwSelection.cpp line 5358 at r1 (raw file):

Previously, JakeOliver28 (Jake Oliver) wrote…

Could we just put the Assert and return inside of the first "if (wsNes <=0)" instead of having a duplicate if statement?

Good question. The value could have been changed inside of the first if statement, this covers the case where there is no m_qtsbProp and also if trying to get the writing system from those properties fails.

[jasonleenaylor]

Src/views/VwSelection.cpp line 5358 at r1 (raw file):

Previously, jasonleenaylor (Jason Naylor) wrote…

Good question. The value could have been changed inside of the first if statement, this covers the case where there is no m_qtsbProp and also if trying to get the writing system from those properties fails.

Hmm, I retract, we can still put this inside the first if.

[jasonleenaylor]

Src/views/VwSelection.cpp line 5358 at r1 (raw file):

Previously, jasonleenaylor (Jason Naylor) wrote…

Hmm, I retract, we can still put this inside the first if.

After drinking more coffee I retract my retraction, if we want it to 'return' under the correct conditions a second check is necessary even if we put it inside the first if. So this way reduces the nesting.

[JakeOliver28]

Reviewable status: 0 of 1 files reviewed, 1 unresolved discussion