From 8facc4d3ef6ffed1e4006a7549a3c03a1a5b7d83 Mon Sep 17 00:00:00 2001
From: Lars Viklund <zao@zao.se>
Date: Fri, 19 Apr 2024 03:58:25 +0200
Subject: [PATCH] fix: enable Unicode separators and caret motions

As the runtime is going to support Unicode installation locations and
build directories, some UTF-8 text is going to reach the Lua side of
the project. This includes the script path, the user path, any paths
yielded from file searches and also imported character names from
accounts.

Care needs to be taken in many places where string operations are
performed as no longer does a byte necessarily correspond to a single
character and anything that truncates, reverses or otherwise slices
strings could need an audit.

This change fixes cursor movement in `EditControl`s with the arrow keys
as those historically used string matching and byte offsets. It also
ensures that the use of arbitrary Unicode codepoints as decimal and
thousands separators works correctly as the previous code used unaware
reversing and slicing.
---
 src/Classes/EditControl.lua | 31 ++++++++-----------------------
 src/Modules/Common.lua      | 12 +++++++-----
 2 files changed, 15 insertions(+), 28 deletions(-)

diff --git a/src/Classes/EditControl.lua b/src/Classes/EditControl.lua
index 63eedfa1e..2964057ad 100644
--- a/src/Classes/EditControl.lua
+++ b/src/Classes/EditControl.lua
@@ -7,6 +7,7 @@ local m_max = math.max
 local m_min = math.min
 local m_floor = math.floor
 local protected_replace = "*"
+local utf8 = require('lua-utf8')
 
 local function lastLine(str)
 	local lastLineIndex = 1
@@ -541,18 +542,10 @@ function EditClass:OnKeyDown(key, doubleClick)
 		if self.caret > 1 then
 			if ctrl then
 			-- Skip leading space, then jump word
-				while self.buf:sub(self.caret-1, self.caret-1):match("[%s%p]") do
-					if self.caret > 1 then
-						self.caret = self.caret - 1
-					end
-				end
-				while self.buf:sub(self.caret-1, self.caret-1):match("%w") do
-					if self.caret > 1 then
-						self.caret = self.caret - 1
-					end
-				end
+				self.caret = self.caret - #utf8.match(self.buf:sub(1, self.caret-1), "[%s%p]*$")
+				self.caret = self.caret - #utf8.match(self.buf:sub(1, self.caret-1), "%w*$")
 			else
-				self.caret = self.caret - 1
+				self.caret = utf8.next(self.buf, self.caret, -1) or 0
 			end
 			self.lastUndoState.caret = self.caret
 			self:ScrollCaretIntoView()
@@ -562,19 +555,11 @@ function EditClass:OnKeyDown(key, doubleClick)
 		self.sel = shift and (self.sel or self.caret) or nil
 		if self.caret <= #self.buf then
 			if ctrl then
-			-- Jump word, then skip trailing space, 
-				while self.buf:sub(self.caret, self.caret):match("%w") do
-					if self.caret <= #self.buf then
-						self.caret = self.caret + 1
-					end
-				end
-				while self.buf:sub(self.caret, self.caret):match("[%s%p]") do
-					if self.caret <= #self.buf then
-						self.caret = self.caret + 1
-					end
-				end
+			-- Jump word, then skip trailing space,
+				self.caret = self.caret + #utf8.match(self.buf:sub(self.caret), "^%w*")
+				self.caret = self.caret + #utf8.match(self.buf:sub(self.caret), "^[%s%p]*")
 			else
-				self.caret = self.caret + 1
+				self.caret = utf8.next(self.buf, self.caret, 1) or #self.buf + 1
 			end
 			self.lastUndoState.caret = self.caret
 			self:ScrollCaretIntoView()
diff --git a/src/Modules/Common.lua b/src/Modules/Common.lua
index 3c99c20d2..91877ebc8 100644
--- a/src/Modules/Common.lua
+++ b/src/Modules/Common.lua
@@ -26,6 +26,7 @@ common.curl = require("lcurl.safe")
 common.xml = require("xml")
 common.base64 = require("base64")
 common.sha1 = require("sha1")
+local utf8 = require('lua-utf8')
 
 -- Try to load a library return nil if failed. https://stackoverflow.com/questions/34965863/lua-require-fallback-error-handling
 function prerequire(...)
@@ -723,20 +724,21 @@ function formatNumSep(str)
 		end
 		local x, y, minus, integer, fraction = str:find("(-?)(%d+)(%.?%d*)")
 		if main.showThousandsSeparators then
-			integer = integer:reverse():gsub("(%d%d%d)", "%1"..main.thousandsSeparator):reverse()
+			rev1kSep = utf8.reverse(main.thousandsSeparator)
+			integer = utf8.reverse(utf8.gsub(utf8.reverse(integer), "(%d%d%d)", "%1"..rev1kSep))
 			-- There will be leading separators if the number of digits are divisible by 3
 			-- This checks for their presence and removes them
 			-- Don't use patterns here because thousandsSeparator can be a pattern control character, and will crash if used
 			if main.thousandsSeparator ~= "" then
-				local thousandsSeparator = string.find(integer, main.thousandsSeparator, 1, 2)
+				local thousandsSeparator = utf8.find(integer, rev1kSep, 1, 2)
 				if thousandsSeparator and thousandsSeparator == 1 then
-					integer = integer:sub(2)
+					integer = utf8.sub(integer, 2)
 				end
 			end
 		else
-			integer = integer:reverse():gsub("(%d%d%d)", "%1"):reverse()
+			integer = utf8.reverse(utf8.gsub(utf8.reverse(integer), "(%d%d%d)", "%1"))
 		end
-		return colour..minus..integer..fraction:gsub("%.", main.decimalSeparator)
+		return colour..minus..integer..utf8.gsub(fraction, "%.", main.decimalSeparator)
 	end)
 end