M7: preserve ideographic spaces in edited paragraphs (#309)

Co-authored-by: Codex <noreply@openai.com>

crates/edit/src/tests.rs

@@ -316,6 +316,26 @@ fn japanese_paragraph_uses_unicode_type3_fallback() {
     assert_eq!(text::text_blocks(&reopened, 0).unwrap()[0].text, replacement);
 }
 
+#[test]
+fn japanese_paragraph_keeps_ideographic_spaces() {
+    if without_craft_fonts("japanese_paragraph_keeps_ideographic_spaces") {
+        return;
+    }
+    let mut doc = text_page("BT /F1 12 Tf 72 700 Td (Original paragraph) Tj ET");
+    let replacement = "　見本商会　　御中　";
+    let style = text::BlockStyle { width: Some(400.0), ..Default::default() };
+    text::rewrite_block(&mut doc, 0, 0, Some(replacement), &style).unwrap();
+    let reopened = reopen(&doc);
+    let lines = text::text_lines(&reopened, 0).unwrap();
+    assert_eq!(lines.len(), 1);
+    assert_eq!(lines[0].text, replacement);
+    assert_eq!(text::text_blocks(&reopened, 0).unwrap()[0].text, replacement);
+    // Restyling the saved paragraph without new text must also preserve its spacing.
+    let mut doc = reopened;
+    text::rewrite_block(&mut doc, 0, 0, None, &text::BlockStyle { size: Some(14.0), ..style }).unwrap();
+    assert_eq!(text::text_lines(&reopen(&doc), 0).unwrap()[0].text, replacement);
+}
+
 /// Without craft-fonts there is no Japanese face: editing in Japanese is a clear error that
 /// leaves the page untouched, never a panic; Latin edits work as before.
 #[test]

crates/edit/src/text.rs

@@ -800,13 +800,13 @@ fn group_blocks(lines: &[TextLine]) -> Vec<TextBlock> {
             } else {
                 b.text.push(' ');
             }
-            b.text.push_str(l.text.trim());
+            b.text.push_str(l.text.trim_matches(paragraph_separator));
             b.rect = [b.rect[0].min(l.rect[0]), b.rect[1].min(l.rect[1]), b.rect[2].max(l.rect[2]), b.rect[3].max(l.rect[3])];
             b.lines.push(i);
         } else {
             gap = None;
             blocks.push(TextBlock {
-                text: l.text.trim().to_string(),
+                text: l.text.trim_matches(paragraph_separator).to_string(),
                 rect: l.rect,
                 base_font: l.base_font.clone(),
                 size: l.size,
@@ -820,11 +820,17 @@ fn group_blocks(lines: &[TextLine]) -> Vec<TextBlock> {
     blocks
 }
 
+// An ideographic space carries intentional Japanese spacing, including paragraph indentation.
+// Keep other whitespace normalization unchanged.
+fn paragraph_separator(c: char) -> bool {
+    c.is_whitespace() && c != '\u{3000}'
+}
+
 /// Greedy word wrapping to `width` with `advance` giving a string's width.
 fn wrap(text: &str, width: f64, advance: impl Fn(&str) -> f64) -> Vec<String> {
     let mut out = Vec::new();
     let mut line = String::new();
-    for word in text.split_whitespace() {
+    for word in text.split(paragraph_separator).filter(|word| !word.is_empty()) {
         let candidate = if line.is_empty() { word.to_string() } else { format!("{line} {word}") };
         if !line.is_empty() && advance(&candidate) > width {
             out.push(std::mem::take(&mut line));
@@ -882,7 +888,7 @@ pub fn rewrite_block(doc: &mut Document, page: usize, block: usize, text: Option
     let streams = content_streams(doc, &p.dict);
     let (stream_obj, data) = streams.get(first.stream).cloned().ok_or_else(|| EditError::Invalid("the page's content changed".into()))?;
     let ops = parse(&data).ops;
-    let text = text.unwrap_or(&b.text).split_whitespace().collect::<Vec<_>>().join(" ");
+    let text = text.unwrap_or(&b.text).split(paragraph_separator).filter(|word| !word.is_empty()).collect::<Vec<_>>().join(" ");
     let o = &first.origin;
     let (font_name, old_size) = o.state.font.clone().ok_or_else(|| EditError::Invalid("the paragraph has no font".into()))?;
     let k = o.k.max(1e-6);
@@ -1105,3 +1111,17 @@ pub fn rewrite_block(doc: &mut Document, page: usize, block: usize, text: Option
     })?;
     Ok(LineEdit { substituted })
 }
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    #[test]
+    fn paragraph_wrapping_keeps_ideographic_spaces() {
+        let width = |s: &str| s.chars().count() as f64;
+        assert_eq!(wrap("A　　B C", 4.0, width), ["A　　B", "C"]);
+        assert_eq!(wrap("A\u{a0}B C", 3.0, width), ["A B", "C"]);
+        assert_eq!(wrap("A\u{2028}B C", 3.0, width), ["A B", "C"]);
+        assert_eq!(wrap(" \tA  B\r\nC ", 3.0, width), ["A B", "C"]);
+    }
+}

parity/acrobat-features.toml

@@ -1806,8 +1806,8 @@ milestone = "M7"
 status = "shipped"
 commands = ["edit.edit_text"]
 tools = ["text_edit"]
-evidence = ["crates/edit/src/tests.rs::paragraphs_are_found_and_rewrapped", "crates/edit/src/tests.rs::text_lines_are_found_and_replaced_in_place", "crates/ui-egui/tests/editing.rs::editing_existing_text_in_place", "crates/automation/tests/automation.rs::editing_existing_text_through_tools"]
-notes = "edited paragraphs are rewrapped to the box width with the font's own advance widths and the original line spacing, starting at the original text matrix; text state and fill colour restored after; a single line grows to the right margin"
+evidence = ["crates/edit/src/tests.rs::japanese_paragraph_keeps_ideographic_spaces", "crates/edit/src/tests.rs::paragraphs_are_found_and_rewrapped", "crates/edit/src/tests.rs::text_lines_are_found_and_replaced_in_place", "crates/ui-egui/tests/editing.rs::editing_existing_text_in_place", "crates/automation/tests/automation.rs::editing_existing_text_through_tools"]
+notes = "edited paragraphs are rewrapped to the box width with the font's own advance widths and the original line spacing, starting at the original text matrix; text state and fill colour restored after; a single line grows to the right margin; U+3000 ideographic spaces are preserved (including paragraph edges and repeated spaces); Japanese font-style fallback remains limited (#184)"
 
 [[feature]]
 id = "edit.text-move-resize"