starred/markdig
1
0
Fork 0
mirror of https://github.com/xoofx/markdig.git synced 2026-02-04 05:44:50 +00:00
Code Issues 690 Packages Projects Releases 20 Wiki Activity

fix escaped characters in LinkInline

Browse Source
This commit is contained in:
Ruud Poutsma
2020-10-23 22:47:39 +02:00
parent 7b20299d2b
commit 6506e4594c
5 changed files with 268 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
src/Markdig.Tests/RoundtripSpecs/Roundtrip.md
View File

@@ -63,4 +63,4 @@ In order:
- Newline struct itself
- handling newlines
- should newlines be supported?
- Example 207, 209: Special-casing certain edgecases
- Example 207, 209, 291: Special-casing certain edgecases

259
src/Markdig/Helpers/LinkHelper.cs
View File

@@ -418,7 +418,9 @@ namespace Markdig.Helpers
public static bool TryParseInlineLinkWhitespace(
ref StringSlice text,
out string link,
out string unescapedLink,
out string title,
out string unescapedTitle,
out char titleEnclosingCharacter,
out SourceSpan linkSpan,
out SourceSpan titleSpan,
@@ -436,7 +438,9 @@ namespace Markdig.Helpers
bool isValid = false;
var c = text.CurrentChar;
link = null;
unescapedLink = null;
title = null;
unescapedTitle = null;
linkSpan = SourceSpan.Empty;
titleSpan = SourceSpan.Empty;
@@ -454,7 +458,7 @@ namespace Markdig.Helpers
text.TrimStart();
whitespaceBeforeLink = new SourceSpan(sourcePosition, text.Start - 1);
var pos = text.Start;
if (TryParseUrl(ref text, out link, out urlHasPointyBrackets))
if (TryParseUrlWhitespace(ref text, out link, out unescapedLink, out urlHasPointyBrackets))
{
linkSpan.Start = pos;
linkSpan.End = text.Start - 1;
@@ -482,7 +486,7 @@ namespace Markdig.Helpers
{
isValid = true;
}
else if (TryParseTitle(ref text, out title, out titleEnclosingCharacter))
else if (TryParseTitleWhitespace(ref text, out title, out unescapedTitle, out titleEnclosingCharacter))
{
titleSpan.Start = pos;
titleSpan.End = text.Start - 1;
@@ -612,6 +616,108 @@ namespace Markdig.Helpers
return isValid;
}
public static bool TryParseTitleWhitespace<T>(ref T text, out string title, out string unescapedTitle, out char enclosingCharacter) where T : ICharIterator
{
bool isValid = false;
var buffer = StringBuilderCache.Local();
var unescaped = new StringBuilder();
enclosingCharacter = '\0';
// a sequence of zero or more characters between straight double-quote characters ("), including a " character only if it is backslash-escaped, or
// a sequence of zero or more characters between straight single-quote characters ('), including a ' character only if it is backslash-escaped, or
var c = text.CurrentChar;
if (c == '\'' || c == '"' || c == '(')
{
enclosingCharacter = c;
var closingQuote = c == '(' ? ')' : c;
bool hasEscape = false;
// -1: undefined
// 0: has only spaces
// 1: has other characters
int hasOnlyWhiteSpacesSinceLastLine = -1;
while (true)
{
c = text.NextChar();
if (c == '\r' || c == '\n')
{
if (hasOnlyWhiteSpacesSinceLastLine >= 0)
{
if (hasOnlyWhiteSpacesSinceLastLine == 1)
{
break;
}
hasOnlyWhiteSpacesSinceLastLine = -1;
}
buffer.Append(c);
unescaped.Append(c);
if (c == '\r' && text.PeekChar() == '\n')
{
buffer.Append('\n');
unescaped.Append(c);
}
continue;
}
if (c == '\0')
{
break;
}
if (c == closingQuote)
{
if (hasEscape)
{
buffer.Append(closingQuote);
unescaped.Append(closingQuote);
hasEscape = false;
continue;
}
// Skip last quote
text.NextChar();
isValid = true;
break;
}
if (hasEscape && !c.IsAsciiPunctuation())
{
buffer.Append('\\');
unescaped.Append('\\');
}
if (c == '\\')
{
hasEscape = true;
unescaped.Append('\\');
continue;
}
hasEscape = false;
if (c.IsSpaceOrTab())
{
if (hasOnlyWhiteSpacesSinceLastLine < 0)
{
hasOnlyWhiteSpacesSinceLastLine = 1;
}
}
else if (c != '\n' && c != '\r' && (c != '\r' && text.PeekChar() != '\n'))
{
hasOnlyWhiteSpacesSinceLastLine = 0;
}
buffer.Append(c);
unescaped.Append(c);
}
}
title = isValid ? buffer.ToString() : null;
unescapedTitle = isValid ? unescaped.ToString() : null;
buffer.Length = 0;
return isValid;
}
public static bool TryParseUrl<T>(T text, out string link) where T : ICharIterator
{
return TryParseUrl(ref text, out link, out _);
@@ -757,6 +863,155 @@ namespace Markdig.Helpers
return isValid;
}
public static bool TryParseUrlWhitespace<T>(ref T text, out string link, out string unescapedLink, out bool hasPointyBrackets, bool isAutoLink = false) where T : ICharIterator
{
bool isValid = false;
hasPointyBrackets = false;
var buffer = StringBuilderCache.Local();
var unescaped = new StringBuilder();
unescapedLink = null;
var c = text.CurrentChar;
// a sequence of zero or more characters between an opening < and a closing >
// that contains no line breaks, or unescaped < or > characters, or
if (c == '<')
{
bool hasEscape = false;
do
{
c = text.NextChar();
if (!hasEscape && c == '>')
{
text.NextChar();
hasPointyBrackets = true;
isValid = true;
break;
}
if (!hasEscape && c == '<')
{
break;
}
if (hasEscape && !c.IsAsciiPunctuation())
{
buffer.Append('\\');
unescaped.Append('\\');
}
if (c == '\\')
{
hasEscape = true;
unescaped.Append('\\');
continue;
}
if (c.IsNewLine())
{
break;
}
hasEscape = false;
buffer.Append(c);
unescaped.Append(c);
} while (c != '\0');
}
else
{
// a nonempty sequence of characters that does not start with <, does not include ASCII space or control characters,
// and includes parentheses only if (a) they are backslash-escaped or (b) they are part of a
// balanced pair of unescaped parentheses that is not itself inside a balanced pair of unescaped
// parentheses.
bool hasEscape = false;
int openedParent = 0;
while (true)
{
// Match opening and closing parenthesis
if (c == '(')
{
if (!hasEscape)
{
openedParent++;
}
}
if (c == ')')
{
if (!hasEscape)
{
openedParent--;
if (openedParent < 0)
{
isValid = true;
break;
}
}
}
if (!isAutoLink)
{
if (hasEscape && !c.IsAsciiPunctuation())
{
buffer.Append('\\');
unescaped.Append('\\');
}
// If we have an escape
if (c == '\\')
{
hasEscape = true;
c = text.NextChar();
unescaped.Append('\\');
continue;
}
hasEscape = false;
}
if (IsEndOfUri(c, isAutoLink))
{
isValid = true;
break;
}
if (isAutoLink)
{
if (c == '&')
{
if (HtmlHelper.ScanEntity(text, out _, out _, out _) > 0)
{
isValid = true;
break;
}
}
if (IsTrailingUrlStopCharacter(c) && IsEndOfUri(text.PeekChar(), true))
{
isValid = true;
break;
}
}
buffer.Append(c);
unescaped.Append(c);
c = text.NextChar();
}
if (openedParent > 0)
{
isValid = false;
}
}
link = isValid ? buffer.ToString() : null;
unescapedLink = isValid ? unescaped.ToString() : null;
buffer.Length = 0;
return isValid;
}
[MethodImpl(MethodImplOptions.AggressiveInlining)]
private static bool IsTrailingUrlStopCharacter(char c)
{

4
src/Markdig/Parsers/Inlines/LinkInlineParser.cs
View File

@@ -223,7 +223,9 @@ namespace Markdig.Parsers.Inlines
if (LinkHelper.TryParseInlineLinkWhitespace(
ref text,
out string url,
out string unescapedUrl,
out string title,
out string unescapedTitle,
out char titleEnclosingCharacter,
out SourceSpan linkSpan,
out SourceSpan titleSpan,
@@ -240,9 +242,11 @@ namespace Markdig.Parsers.Inlines
{
WhitespaceBeforeUrl = wsBeforeLink,
Url = HtmlHelper.Unescape(url), // TODO: RTP: unescape
UnescapedUrl = unescapedUrl, // TODO: RTP: unescape
UrlHasPointyBrackets = urlHasPointyBrackets,
WhitespaceAfterUrl = wsAfterLink,
Title = HtmlHelper.Unescape(title), // TODO: RTP: unescape
UnescapedTitle = unescapedTitle,
TitleEnclosingCharacter = titleEnclosingCharacter,
WhitespaceAfterTitle = wsAfterTitle,
IsImage = openParent.IsImage,

7
src/Markdig/Renderers/Normalize/Inlines/LinkInlineRenderer.cs
View File

@@ -58,14 +58,14 @@ namespace Markdig.Renderers.Normalize.Inlines
{
renderer.Write('<');
}
renderer.Write(link.Url);
renderer.Write(link.UnescapedUrl);
if (link.UrlHasPointyBrackets)
{
renderer.Write('>');
}
renderer.Write(link.WhitespaceAfterUrl);
if (!string.IsNullOrEmpty(link.Title))
if (!string.IsNullOrEmpty(link.UnescapedTitle))
{
var open = link.TitleEnclosingCharacter;
var close = link.TitleEnclosingCharacter;
@@ -74,7 +74,8 @@ namespace Markdig.Renderers.Normalize.Inlines
close = ')';
}
renderer.Write(open);
renderer.Write(link.Title.Replace(@"""", @"\""")); // TODO: RTP: should this always be done?
//renderer.Write(link.Title.Replace(@"""", @"\""")); // TODO: RTP: should this always be done?
renderer.Write(link.UnescapedTitle);
renderer.Write(close);
renderer.Write(link.WhitespaceAfterTitle);
}

2
src/Markdig/Syntax/Inlines/LinkInline.cs
View File

@@ -45,6 +45,7 @@ namespace Markdig.Syntax.Inlines
/// Gets or sets the URL.
/// </summary>
public string Url { get; set; }
public string UnescapedUrl { get; internal set; }
public bool UrlHasPointyBrackets { get; set; }
@@ -60,6 +61,7 @@ namespace Markdig.Syntax.Inlines
/// Gets or sets the title.
/// </summary>
public string Title { get; set; }
public string UnescapedTitle { get; internal set; }
public char TitleEnclosingCharacter { get; set; }