Markdown: textual numbers are mistaken for <pre> characters at margins

[rdipardo]

At the start of every line, LexMarkdown::Lex defaults to SCE_MARKDOWN_PRECHAR in the absence of any reserved markup token:

lexilla/lexers/LexMarkdown.cxx

Lines 341 to 344 in 31a5815

	else {
	precharCount = 0;
	sc.SetState(SCE_MARKDOWN_PRECHAR);
	}

This is supposed to set off indented code blocks by filling the leading white space. But a textual number with no leading white space is also filled as if the digits were <pre> characters:

Text:

The number:

338269006135764734700913562171

is prime

Styles:

{0}The number:{1}

{12}338269006135764734700913562171{1}

{0}is prime{1}

Every digit is captured because a subsequent code path increments the style region as if the number were an ordered list item:

lexilla/lexers/LexMarkdown.cxx

Lines 380 to 384 in 31a5815

	// Ordered list
	else if (IsADigit(sc.ch)) {
	int digitCount = 0;
	while (IsADigit(sc.GetRelative(++digitCount)))
	;

This patch adds a corrective fallback to SCE_MARKDOWN_DEFAULT as soon as a string of digits fails to match a list item:

0001-LexMarkdown-Fix-numbers-as-pre-char.diff.txt

[rdipardo]

These are the tests: Issue117-Tests.diff.txt

[nyamatongwe]

Included in 5.2.1 release.