World's first open-source enterprise world model.
Rust agent-memorybitemporalgraphragknowledge-baseknowledge-graphllmontologypgvectorpostgresqlragrustself-hostedsemantic-searchtemporal-knowledge-graphworld-model
A number in a spreadsheet is read with the 15 significant digits Excel keeps, not the binary tail of a computed value (#1032)
Maya Kid468ef0f94139
diff --git a/crates/utopia-ingest/src/parsers.rs b/crates/utopia-ingest/src/parsers.rsindex 7e017740..e3fbc828 100644--- a/crates/utopia-ingest/src/parsers.rs+++ b/crates/utopia-ingest/src/parsers.rs@@ -756,6 +756,7 @@ pub fn spreadsheet(bytes: &[u8]) -> anyhow::Result<String> {let text = match c {Data::Empty => String::new(),Data::DateTime(d) => excel_date(d),+ Data::Float(f) => excel_number(*f),other => other.to_string(),};(text, 1, 0)@@ -824,6 +825,22 @@ fn excel_date(d: &calamine::ExcelDateTime) -> String {}}+/// 数字格按 Excel 认的精度写:15 位有效数字。+///+/// Excel 比较、显示一个数都只看 15 位有效数字,公式的缓存值却按双精度的 17 位写进文件:+/// `=0.1+0.2` 存成 0.30000000000000004,`=1.1*1.1-1` 存成 0.21000000000000019,Excel 里+/// 看到的是 0.3 和 0.21。原样写出来,正文、引文和抽出来的值都带着这截二进制的尾巴。先舍到+/// 15 位有效数字,再写最短的十进制;本来就干净的数(19.9、1200、45306)一个字不变+fn excel_number(value: f64) -> String {+ if !value.is_finite() {+ return value.to_string();+ }+ format!("{value:.14e}")+ .parse::<f64>()+ .unwrap_or(value)+ .to_string()+}+/// Decode before conversion so legacy HTML encodings remain supported.pub fn html(bytes: &[u8]) -> anyhow::Result<String> {Ok(crate::html::page_to_markdown(&plain_text(bytes), None)?)diff --git a/crates/utopia-ingest/tests/spreadsheet_numbers.rs b/crates/utopia-ingest/tests/spreadsheet_numbers.rsnew file mode 100644index 00000000..8419beae--- /dev/null+++ b/crates/utopia-ingest/tests/spreadsheet_numbers.rs@@ -0,0 +1,66 @@+//! 电子表格里的数按 Excel 认的精度读:15 位有效数字。+//!+//! 公式的缓存值是按双精度的 17 位写进文件的:`=0.1+0.2` 存成 `<v>0.30000000000000004</v>`,+//! Excel 里看到的却是 0.3。这里照 Excel 写出来的样子拼一份 xlsx,值都是缓存值的写法,+//! 数字格式是常规(没有 `s`)。++use std::io::{Cursor, Write};++fn read(values: &[(&str, &str)]) -> String {+ let rows: String = values+ .iter()+ .enumerate()+ .map(|(i, (label, value))| {+ let r = i + 2;+ format!(+ r#"<row r="{r}"><c r="A{r}" t="inlineStr"><is><t>{label}</t></is></c><c r="B{r}"><v>{value}</v></c></row>"#+ )+ })+ .collect();+ let mut zip = zip::ZipWriter::new(Cursor::new(Vec::new()));+ for (part, xml) in [+ ("[Content_Types].xml", r#"<Types xmlns="http://schemas.openxmlformats.org/package/2006/content-types"><Default Extension="rels" ContentType="application/vnd.openxmlformats-package.relationships+xml"/><Default Extension="xml" ContentType="application/xml"/><Override PartName="/xl/workbook.xml" ContentType="application/vnd.openxmlformats-officedocument.spreadsheetml.sheet.main+xml"/><Override PartName="/xl/worksheets/sheet1.xml" ContentType="application/vnd.openxmlformats-officedocument.spreadsheetml.worksheet+xml"/></Types>"#.to_string()),+ ("_rels/.rels", r#"<Relationships xmlns="http://schemas.openxmlformats.org/package/2006/relationships"><Relationship Id="rId1" Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/officeDocument" Target="xl/workbook.xml"/></Relationships>"#.to_string()),+ ("xl/workbook.xml", r#"<workbook xmlns="http://schemas.openxmlformats.org/spreadsheetml/2006/main" xmlns:r="http://schemas.openxmlformats.org/officeDocument/2006/relationships"><sheets><sheet name="计算" sheetId="1" r:id="rId1"/></sheets></workbook>"#.to_string()),+ ("xl/_rels/workbook.xml.rels", r#"<Relationships xmlns="http://schemas.openxmlformats.org/package/2006/relationships"><Relationship Id="rId1" Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/worksheet" Target="worksheets/sheet1.xml"/></Relationships>"#.to_string()),+ ("xl/worksheets/sheet1.xml", format!(r#"<worksheet xmlns="http://schemas.openxmlformats.org/spreadsheetml/2006/main"><sheetData><row r="1"><c r="A1" t="inlineStr"><is><t>项目</t></is></c><c r="B1" t="inlineStr"><is><t>金额</t></is></c></row>{rows}</sheetData></worksheet>"#)),+ ] {+ zip.start_file(part, zip::write::SimpleFileOptions::default()).unwrap();+ zip.write_all(xml.as_bytes()).unwrap();+ }+ utopia_ingest::parse("计算.xlsx", &zip.finish().unwrap().into_inner())+ .unwrap()+ .text+}++#[test]+fn a_computed_number_reads_with_the_digits_excel_keeps() {+ let text = read(&[+ ("合计", "0.30000000000000004"),+ ("增长率", "0.21000000000000019"),+ ("差额", "-0.30000000000000004"),+ ]);+ for row in ["| 合计 | 0.3 |", "| 增长率 | 0.21 |", "| 差额 | -0.3 |"] {+ assert!(text.contains(row), "{row}: {text}");+ }+}++#[test]+fn a_number_that_was_already_clean_reads_as_before() {+ let text = read(&[+ ("含税金额", "1395.0527999999999"),+ ("单价", "19.899999999999999"),+ ("数量", "1200"),+ ("编号", "123456789012345000"),+ ("比例", "0.125"),+ ]);+ for row in [+ "| 含税金额 | 1395.0528 |",+ "| 单价 | 19.9 |",+ "| 数量 | 1200 |",+ "| 编号 | 123456789012345000 |",+ "| 比例 | 0.125 |",+ ] {+ assert!(text.contains(row), "{row}: {text}");+ }+}