Skip to content

World's first open-source enterprise world model.

A non-breaking hyphen in a Word document is read as a hyphen, so the numbers on either side of it stay apart (#1031)

Maya Kid6b0a1d7e7f6a
diff --git a/crates/utopia-ingest/src/parsers.rs b/crates/utopia-ingest/src/parsers.rs
index d9bbad1a..7e017740 100644
--- a/crates/utopia-ingest/src/parsers.rs
+++ b/crates/utopia-ingest/src/parsers.rs
@@ -372,6 +372,13 @@ pub(crate) fn docx_xml_to_text(
Some(c) => c.0.push(' '),
None => out.push(' '),
},
+ // Word 的不断行连字符(Ctrl+Shift+-)不是 w:t 里的字,是一个元素:页面上照样画出
+ // 连字符,只是不在这里折行。丢了它,2024‑01‑15 读成 20240115,010‑62345678 读成
+ // 01062345678。写成普通连字符,日期和号码才认得出来
+ "w:noBreakHyphen" => match cell.as_mut() {
+ Some(c) => c.0.push('-'),
+ None => out.push('-'),
+ },
_ => {}
},
Ok(Event::End(e)) => match e.name().as_ref() {
diff --git a/crates/utopia-ingest/tests/docx_breaks.rs b/crates/utopia-ingest/tests/docx_breaks.rs
index ef6292e3..7250ba66 100644
--- a/crates/utopia-ingest/tests/docx_breaks.rs
+++ b/crates/utopia-ingest/tests/docx_breaks.rs
@@ -57,3 +57,16 @@ fn paragraph_breaks_remain_newlines_and_formatting_runs_join() {
let text = read("<w:p><w:r><w:t>first</w:t><w:tab/><w:t>line</w:t></w:r></w:p><w:p><w:r><w:t>second</w:t></w:r></w:p>");
assert_eq!(text.trim(), "first line\nsecond");
}
+
+#[test]
+fn a_non_breaking_hyphen_keeps_the_numbers_on_either_side_apart() {
+ let nbh = "<w:noBreakHyphen/>";
+ let text = read(&format!(
+ r#"<w:p><w:r><w:t xml:space="preserve">合同于 2024</w:t>{nbh}<w:t>01</w:t>{nbh}<w:t xml:space="preserve">15 签订,电话 010</w:t></w:r><w:r>{nbh}</w:r><w:r><w:t>62345678。</w:t></w:r></w:p>"#
+ ));
+ assert_eq!(text.trim(), "合同于 2024-01-15 签订,电话 010-62345678。");
+ let text = read(&table(&format!(
+ "<w:t>2024</w:t>{nbh}<w:t>01</w:t>{nbh}<w:t>15</w:t>"
+ )));
+ assert!(text.contains("| Sample | 2024-01-15 |"), "{text}");
+}