Detect markdown line width, resolve relative URLs (#332)

~~this is semi-blocked by https://github.com/charmbracelet/glamour/pull/96, but behaviour isn't really worse than the previous behaviour (most links work, some are still broken)~~

#### testcase for link resolver
```
tea pr 332
tea checkout 332 && make install && tea pr 332
```

- [rel](./332)
- [abs](/gitea/tea/pulls/332)
- [full](https://gitea.com/gitea/tea/pulls/332)

Co-authored-by: Norwin Roosen <git@nroo.de>
Co-authored-by: 6543 <6543@obermui.de>
Reviewed-on: https://gitea.com/gitea/tea/pulls/332
Reviewed-by: 6543 <6543@obermui.de>
Reviewed-by: Andrew Thornton <art27@cantab.net>
Co-authored-by: Norwin <noerw@noreply.gitea.io>
Co-committed-by: Norwin <noerw@noreply.gitea.io>

vendor/github.com/yuin/goldmark/parser/link.go

@@ -2,7 +2,6 @@ package parser
 
 import (
 	"fmt"
-	"regexp"
 	"strings"
 
 	"github.com/yuin/goldmark/ast"
@@ -113,8 +112,6 @@ func (s *linkParser) Trigger() []byte {
 	return []byte{'!', '[', ']'}
 }
 
-var linkDestinationRegexp = regexp.MustCompile(`\s*([^\s].+)`)
-var linkTitleRegexp = regexp.MustCompile(`\s+(\)|["'\(].+)`)
 var linkBottom = NewContextKey()
 
 func (s *linkParser) Parse(parent ast.Node, block text.Reader, pc Context) ast.Node {
@@ -293,20 +290,17 @@ func (s *linkParser) parseLink(parent ast.Node, last *linkLabelState, block text
 func parseLinkDestination(block text.Reader) ([]byte, bool) {
 	block.SkipSpaces()
 	line, _ := block.PeekLine()
-	buf := []byte{}
 	if block.Peek() == '<' {
 		i := 1
 		for i < len(line) {
 			c := line[i]
 			if c == '\\' && i < len(line)-1 && util.IsPunct(line[i+1]) {
-				buf = append(buf, '\\', line[i+1])
 				i += 2
 				continue
 			} else if c == '>' {
 				block.Advance(i + 1)
 				return line[1:i], true
 			}
-			buf = append(buf, c)
 			i++
 		}
 		return nil, false
@@ -316,7 +310,6 @@ func parseLinkDestination(block text.Reader) ([]byte, bool) {
 	for i < len(line) {
 		c := line[i]
 		if c == '\\' && i < len(line)-1 && util.IsPunct(line[i+1]) {
-			buf = append(buf, '\\', line[i+1])
 			i += 2
 			continue
 		} else if c == '(' {
@@ -329,7 +322,6 @@ func parseLinkDestination(block text.Reader) ([]byte, bool) {
 		} else if util.IsSpace(c) {
 			break
 		}
-		buf = append(buf, c)
 		i++
 	}
 	block.Advance(i)

vendor/github.com/yuin/goldmark/parser/parser.go

@@ -138,6 +138,9 @@ type Context interface {
 	// Get returns a value associated with the given key.
 	Get(ContextKey) interface{}
 
+	// ComputeIfAbsent computes a value if a value associated with the given key is absent and returns the value.
+	ComputeIfAbsent(ContextKey, func() interface{}) interface{}
+
 	// Set sets the given value to the context.
 	Set(ContextKey, interface{})
 
@@ -252,6 +255,15 @@ func (p *parseContext) Get(key ContextKey) interface{} {
 	return p.store[key]
 }
 
+func (p *parseContext) ComputeIfAbsent(key ContextKey, f func() interface{}) interface{} {
+	v := p.store[key]
+	if v == nil {
+		v = f()
+		p.store[key] = v
+	}
+	return v
+}
+
 func (p *parseContext) Set(key ContextKey, value interface{}) {
 	p.store[key] = value
 }

vendor/github.com/yuin/goldmark/parser/raw_html.go

@@ -2,10 +2,11 @@ package parser
 
 import (
 	"bytes"
+	"regexp"
+
 	"github.com/yuin/goldmark/ast"
 	"github.com/yuin/goldmark/text"
 	"github.com/yuin/goldmark/util"
-	"regexp"
 )
 
 type rawHTMLParser struct {
@@ -67,8 +68,6 @@ func (s *rawHTMLParser) parseSingleLineRegexp(reg *regexp.Regexp, block text.Rea
 	return node
 }
 
-var dummyMatch = [][]byte{}
-
 func (s *rawHTMLParser) parseMultiLineRegexp(reg *regexp.Regexp, block text.Reader, pc Context) ast.Node {
 	sline, ssegment := block.Position()
 	if block.Match(reg) {
@@ -102,7 +101,3 @@ func (s *rawHTMLParser) parseMultiLineRegexp(reg *regexp.Regexp, block text.Read
 	}
 	return nil
 }
-
-func (s *rawHTMLParser) CloseBlock(parent ast.Node, pc Context) {
-	// nothing to do
-}