From e54bd2ccb7d9b7c679701cf1c2eb8bdc9e9f95e5 Mon Sep 17 00:00:00 2001
From: NetterSilver <133535268+Silvertree2010@users.noreply.github.com>
Date: Thu, 15 Oct 2026 15:35:04 +0200
Subject: [PATCH] Read a newline in pasted HTML text as a space

Found while testing crates/engine/src/cmd/richclip.rs:257-283 (decode_text)

Co-Authored-By: Claude Haiku 5.5 <noreply@anthropic.com>
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
---
 crates/engine/src/cmd/richclip.rs | 53 ++++++++++++++++++++++++++++---
 crates/engine/src/tests.rs        | 19 +++++++++++
 2 files changed, 68 insertions(+), 4 deletions(-)

diff --git a/crates/engine/src/cmd/richclip.rs b/crates/engine/src/cmd/richclip.rs
index fc63314b..dc17474f 100644
--- a/crates/engine/src/cmd/richclip.rs
+++ b/crates/engine/src/cmd/richclip.rs
@@ -145,8 +145,13 @@ pub fn html_to_body(html: &str) -> Option<TextBody> {
             continue;
         }
         let end = bytes[i..].iter().position(|c| *c == b'<').map(|n| i + n).unwrap_or(bytes.len());
-        let text = decode_text(&html[i..end]);
+        let mut text = decode_text(&html[i..end]);
         i = end;
+        // A space at the start of a paragraph, or after a space, does not render in HTML.
+        let after_space = paragraphs.last().and_then(|p| p.runs.last()).is_none_or(|r| r.text.ends_with(' '));
+        if after_space {
+            text = text.trim_start_matches(' ').to_string();
+        }
         if text.is_empty() {
             continue;
         }
@@ -155,6 +160,7 @@ pub fn html_to_body(html: &str) -> Option<TextBody> {
     while paragraphs.last().is_some_and(Paragraph::is_empty) && paragraphs.len() > 1 {
         paragraphs.pop();
     }
+    trim_end_space(&mut paragraphs);
     if paragraphs.iter().all(Paragraph::is_empty) {
         return None;
     }
@@ -240,10 +246,19 @@ fn new_para(paragraphs: &mut Vec<Paragraph>) {
         return;
     }
     if paragraphs.last().is_some_and(|p| !p.is_empty()) {
+        trim_end_space(paragraphs);
         paragraphs.push(Paragraph::default());
     }
 }
 
+/// Whitespace at the end of a paragraph does not render in HTML.
+fn trim_end_space(paragraphs: &mut [Paragraph]) {
+    if let Some(run) = paragraphs.last_mut().and_then(|p| p.runs.last_mut()) {
+        let len = run.text.trim_end_matches(' ').len();
+        run.text.truncate(len);
+    }
+}
+
 fn push_run(paragraphs: &mut [Paragraph], runs: &mut usize, text: String, bold: bool, italic: bool, underline: bool) {
     if *runs >= MAX_RUNS {
         return;
@@ -263,14 +278,32 @@ fn push_run(paragraphs: &mut [Paragraph], runs: &mut usize, text: String, bold:
     *runs += 1;
 }
 
+/// Text with each run of HTML whitespace (spaces, tabs, line breaks) collapsed to one space.
+fn collapse_space(s: &str) -> String {
+    let mut out = String::with_capacity(s.len());
+    let mut space = false;
+    for c in s.chars() {
+        if matches!(c, ' ' | '\t' | '\n' | '\r') {
+            if !space {
+                out.push(' ');
+            }
+            space = true;
+        } else {
+            out.push(c);
+            space = false;
+        }
+    }
+    out
+}
+
 fn decode_text(raw: &str) -> String {
     let mut out = String::new();
     let mut rest = raw;
     while let Some(at) = rest.find('&') {
-        out.push_str(&rest[..at].replace(['\n', '\r'], ""));
+        out.push_str(&collapse_space(&rest[..at]));
         rest = &rest[at..];
         let Some(end) = rest.find(';') else {
-            out.push_str(&rest.replace(['\n', '\r'], ""));
+            out.push_str(&collapse_space(rest));
             break;
         };
         let entity = &rest[1..end];
@@ -293,7 +326,7 @@ fn decode_text(raw: &str) -> String {
         }
     }
     if !rest.is_empty() && !rest.starts_with('&') {
-        out.push_str(&rest.replace(['\n', '\r'], ""));
+        out.push_str(&collapse_space(rest));
     }
     out
 }
@@ -371,6 +404,18 @@ mod tests {
         assert!(!same_words(&body.text(), "other"));
     }
 
+    #[test]
+    fn a_newline_in_html_text_is_a_space() {
+        assert_eq!(decode_text("Hello\nworld"), "Hello world");
+        let body = html_to_body("<p>Hello\r\n  world</p>").unwrap();
+        assert_eq!(body.text(), "Hello world");
+        let body = html_to_body("<p>\n<b>Hello</b>\n<i>world</i>\n</p>\n<p>next</p>").unwrap();
+        assert_eq!(body.text(), "Hello world\nnext");
+        assert!(same_words(&body.text(), "Hello world\nnext"));
+        assert_eq!(body.paragraphs[0].runs[0].props.bold, Some(true));
+        assert_eq!(body.paragraphs[0].runs.last().unwrap().props.italic, Some(true));
+    }
+
     #[test]
     fn a_huge_fragment_is_refused() {
         let html = "x".repeat(MAX_HTML + 1);
diff --git a/crates/engine/src/tests.rs b/crates/engine/src/tests.rs
index dd64b08c..e1d0f179 100644
--- a/crates/engine/src/tests.rs
+++ b/crates/engine/src/tests.rs
@@ -249,6 +249,25 @@ fn bold_copy_is_html_and_pages_html_pastes_bold() {
     assert_ne!(plain.paragraphs[0].runs[0].props.bold, Some(true));
 }
 
+#[test]
+fn html_paste_with_a_source_newline_keeps_its_bold() {
+    let mut s = session();
+    s.execute("edit.paste", &json!({"text": "Hello world", "html": "<p><b>Hello</b>\nworld</p>"})).unwrap();
+    let body = s
+        .doc()
+        .unwrap()
+        .current_slide()
+        .unwrap()
+        .shapes
+        .iter()
+        .find(|sh| sh.text.as_ref().is_some_and(|t| t.text() == "Hello world"))
+        .unwrap()
+        .text
+        .clone()
+        .unwrap();
+    assert_eq!(body.paragraphs[0].runs[0].props.bold, Some(true), "{body:?}");
+}
+
 #[test]
 fn group_ungroup_roundtrip_positions() {
     let mut s = session();