From 1748b27fcdba28b7a8b791e5e08220045d27fc38 Mon Sep 17 00:00:00 2001
From: Stephen Young <me@hownowstephen.com>
Date: Wed, 29 Mar 2023 16:46:55 -0400
Subject: [PATCH] Linting and fix paragraph prefixing (#11)

Fixes an issue with spans eating their siblings, and leading lines
before paragraphs
---
 regexp.go         |  6 +++---
 textplain_test.go | 18 ++++++++++++++++++
 tree.go           |  8 +++++++-
 3 files changed, 28 insertions(+), 4 deletions(-)

diff --git a/regexp.go b/regexp.go
index 9dc4703..bb0da3d 100644
--- a/regexp.go
+++ b/regexp.go
@@ -69,7 +69,7 @@ func NewRegexpConverter() Converter {
 			handler: func(t string, submatch []int) string {
 				href, value := strings.TrimSpace(t[submatch[4]:submatch[5]]), strings.TrimSpace(t[submatch[6]:submatch[7]])
 				var replace string
-				if strings.ToLower(href) == strings.ToLower(value) {
+				if strings.EqualFold(href, value) {
 					replace = value
 				} else if value != "" {
 					replace = fmt.Sprintf("%s ( %s )", value, href)
@@ -184,7 +184,7 @@ func (s *submatchReplacer) Replace(text string) string {
 		finalText += text[start:submatch[0]] + s.handler(text, submatch)
 		start = submatch[1]
 	}
-	return finalText + text[start:len(text)]
+	return finalText + text[start:]
 }
 
 // Convert returns a text-only version of supplied document in UTF-8 format with all HTML tags removed
@@ -251,7 +251,7 @@ func (t *RegexpConverter) Convert(document string, lineLength int) (string, erro
 	//  strip text ignored html. Useful for removing
 	//  headers and footers that aren't needed in the
 	//  text version
-	txt := t.ignoredHTML.ReplaceAllString(string(clean.Bytes()), "")
+	txt := t.ignoredHTML.ReplaceAllString(clean.String(), "")
 
 	//  strip out html comments
 	txt = t.comments.ReplaceAllString(txt, "")
diff --git a/textplain_test.go b/textplain_test.go
index 863f70a..618ac19 100644
--- a/textplain_test.go
+++ b/textplain_test.go
@@ -102,6 +102,24 @@ func TestWrappingSpans(t *testing.T) {
 			</p>`,
 			expect: "Test spans\n\ninbetween\n\nline 2\nagain",
 		},
+		{
+			name: "tables and spans",
+			body: `<table>
+						<tbody>
+							<tr>
+								<td>
+									<span>ID</span>
+									<p>ABC-1234</p>
+								</td>
+								<td>
+									<span>Date</span>
+									<p>Mar 29, 2023</p>
+								</td>
+							</tr>
+						</tbody>
+					</table>`,
+			expect: "ID\nABC-1234\n\nDate\nMar 29, 2023",
+		},
 	})
 }
 
diff --git a/tree.go b/tree.go
index 146f9b2..9c18372 100644
--- a/tree.go
+++ b/tree.go
@@ -81,6 +81,12 @@ func (t *TreeConverter) doConvert(n *html.Node) ([]string, error) {
 				if err != nil {
 					return nil, err
 				}
+				if len(parts) > 0 {
+					if p := strings.Trim(parts[len(parts)-1], " \t"); len(p) == 0 || p[len(p)-1] != '\n' {
+						parts = append(parts, "\n")
+					}
+				}
+
 				parts = append(parts, more...)
 				parts = append(parts, "\n\n")
 				continue
@@ -258,7 +264,7 @@ func (t *TreeConverter) wrapSpans(n *html.Node) (*html.Node, []string, error) {
 	for c = n; c != nil; c = c.NextSibling {
 
 		if c.Type == html.ElementNode && c.DataAtom != atom.Span {
-			return c, parts, nil
+			return c.PrevSibling, parts, nil
 		}
 
 		var span string