From 3c548ab524e30267e03d5050b30f75d709027bdc Mon Sep 17 00:00:00 2001
From: Petter Rodhelind <petter.rodhelind@gmail.com>
Date: Sun, 21 Feb 2021 18:31:28 +0100
Subject: Fix newlines by parsing paragraphs better.

---
 parser.go | 29 +++++++++++++++++++++--------
 1 file changed, 21 insertions(+), 8 deletions(-)

diff --git a/parser.go b/parser.go
index 8bae5a9..58ce9bf 100644
--- a/parser.go
+++ b/parser.go
@@ -87,17 +87,18 @@ func parseNode(n *html.Node, p *post) {
 				}
 			}
 		}
+
+		// parse paragraphs
+		if n.Data == "p" {
+			parseParagraph(n, p)
+			p.Content += "\n\n"
+			return
+		}
 	}
 
-	// fetch all pure text elements
+	// fetch all pure text elements outside proper post paragraphs
 	if n.Type == html.TextNode {
-		nl := "\n" // newlines
-		if n.Parent.Type == html.ElementNode && n.Parent.Data == "p" {
-			nl = "\n\n"
-		}
-		if n.Data != "..." {
-			p.Content += strings.Trim(n.Data, " ") + nl
-		}
+		p.Content += n.Data + "\n"
 	}
 
 	// loop deeper
@@ -105,3 +106,15 @@ func parseNode(n *html.Node, p *post) {
 		parseNode(c, p)
 	}
 }
+
+func parseParagraph(n *html.Node, p *post) {
+	// fetch all pure text elements
+	if n.Type == html.TextNode {
+		p.Content += n.Data
+	}
+
+	// loop deeper
+	for c := n.FirstChild; c != nil; c = c.NextSibling {
+		parseParagraph(c, p)
+	}
+}
-- 
cgit v1.2.3