Compare commits

..
Author SHA1 Message Date
Dylan M. Taylor 623c30efa9 xlsx: use shared ST_Xstring encoding and decoding 2026-11-16 19:42:17 -05:00
8 changed files with 128 additions and 112 deletions
Generated
+24 -6
View File
@@ -932,6 +932,17 @@ dependencies = [
"thiserror 2.0.21",
]
[[package]]
name = "craft-drawingml"
version = "0.1.0"
source = "git+https://github.com/storytold/craft-libs?rev=bb57d422d0bc842cbde430955825553ab7b976f4#bb57d422d0bc842cbde430955825553ab7b976f4"
dependencies = [
"craft-xml",
"kurbo",
"log",
"serde",
]
[[package]]
name = "craft-fonts-build"
version = "0.1.0"
@@ -998,11 +1009,6 @@ dependencies = [
"thiserror 2.0.21",
]
[[package]]
name = "craft-pdf"
version = "0.1.0"
source = "git+https://github.com/storytold/craft-libs?rev=bc358686389892a3827e64fc26bedc278ef1e299#bc358686389892a3827e64fc26bedc278ef1e299"
[[package]]
name = "craft-print"
version = "0.1.0"
@@ -1132,6 +1138,18 @@ dependencies = [
"thiserror 2.0.21",
]
[[package]]
name = "craft-xlsx"
version = "0.1.0"
source = "git+https://github.com/storytold/craft-libs?rev=bb57d422d0bc842cbde430955825553ab7b976f4#bb57d422d0bc842cbde430955825553ab7b976f4"
dependencies = [
"craft-drawingml",
"craft-opc",
"craft-sheet",
"craft-xml",
"thiserror 2.0.21",
]
[[package]]
name = "craft-xml"
version = "0.1.0"
@@ -2311,7 +2329,6 @@ dependencies = [
name = "gridcraft-pdf"
version = "0.19.1"
dependencies = [
"craft-pdf",
"craft-stream-codecs",
"harfrust",
"sha2 0.11.1",
@@ -2385,6 +2402,7 @@ dependencies = [
"craft-offcrypto",
"craft-opc",
"craft-xls",
"craft-xlsx",
"craft-xml",
"gridcraft-core",
"gridcraft-formula",
+1 -1
View File
@@ -32,7 +32,6 @@ gridcraft-math = { path = "crates/math" }
# craft-update: app-specific pretend versions apply to startup and manual checks.
craft-update = { git = "https://github.com/storytold/craft-libs", rev = "bb57d422d0bc842cbde430955825553ab7b976f4", version = "=0.1.0", features = ["egui"] }
# PDF RunLength encoding; this revision also fixes the vendored JBIG2 dependency for Git consumers.
craft-pdf = { git = "https://github.com/storytold/craft-libs", rev = "bc358686389892a3827e64fc26bedc278ef1e299", version = "0.1.0", default-features = false }
craft-stream-codecs = { git = "https://github.com/storytold/craft-libs", rev = "f7e0d7fd5577e102b4634bcbe062e649eb68165f", version = "0.1.0" }
# Compound files ([MS-CFB]) and password-encrypted OOXML packages ([MS-OFFCRYPTO]).
craft-cfb = { git = "https://github.com/storytold/craft-libs", rev = "bb57d422d0bc842cbde430955825553ab7b976f4", version = "0.1.0" }
@@ -46,6 +45,7 @@ craft-csv = { git = "https://github.com/storytold/craft-libs", rev = "bb57d422d0
craft-gridcraft-format = { git = "https://github.com/storytold/craft-libs", rev = "9a6d894200221fcfedc274870c30394d404f859d", version = "0.1.0" }
# Excel 97-2003 (.xls, BIFF8) workbooks; `crates/xlsx/src/xls.rs` maps them onto the model.
craft-xml = { git = "https://github.com/storytold/craft-libs", rev = "bb57d422d0bc842cbde430955825553ab7b976f4", version = "0.1.0" }
craft-xlsx = { git = "https://github.com/storytold/craft-libs", rev = "bb57d422d0bc842cbde430955825553ab7b976f4", version = "0.1.0" }
craft-xls = { git = "https://github.com/storytold/craft-libs", rev = "bb57d422d0bc842cbde430955825553ab7b976f4", version = "0.1.0" }
# foundation
-1
View File
@@ -10,7 +10,6 @@ readme = "README.md"
workspace = true
[dependencies]
craft-pdf.workspace = true
craft-stream-codecs.workspace = true
harfrust = "=0.12.0"
skrifa = { workspace = true }
+10 -7
View File
@@ -10,7 +10,7 @@ use skrifa::{
};
use unicode_bidi::BidiInfo;
use crate::{CID_FONT, Error, Font, TextDirection, n, needs_cid, pdf_string, pdf_utf16_string};
use crate::{CID_FONT, Error, Font, TextDirection, dict_obj, n, needs_cid, pdf_string, pdf_utf16_string, stream_obj};
include!(concat!(env!("OUT_DIR"), "/arabic_fonts.rs"));
@@ -490,7 +490,7 @@ impl Subset {
}
/// Writes the subset fonts into `out` and returns their page resource entries.
pub(crate) fn write_objects(&self, next: &mut u32, out: &mut craft_pdf::classic::Writer) -> Result<String, Error> {
pub(crate) fn write_objects(&self, next: &mut u32, out: &mut Vec<u8>, offsets: &mut Vec<usize>) -> String {
let mut resources = String::new();
for (batch, glyphs) in self.glyphs.chunks(255).enumerate() {
let id = *next;
@@ -507,7 +507,7 @@ impl Subset {
let _ = write!(names, "/g{code} ");
let _ = write!(widths, "{} ", n(glyph.width));
let body = glyph_procedure(glyph);
out.stream(*next, "", body.as_bytes())?;
stream_obj(out, offsets, *next, "", body.as_bytes());
*next += 1;
bounds[0] = bounds[0].min(glyph.bounds[0]);
bounds[1] = bounds[1].min(glyph.bounds[1]);
@@ -526,8 +526,11 @@ impl Subset {
cmap.push_str("endbfchar\n");
}
cmap.push_str("endcmap\nCMapName currentdict /CMap defineresource pop\nend\nend\n");
out.stream(cmap_id, "", cmap.as_bytes())?;
out.object(id,
stream_obj(out, offsets, cmap_id, "", cmap.as_bytes());
dict_obj(
out,
offsets,
id,
&format!(
"<< /Type /Font /Subtype /Type3 /Name /GridCraftArabic /FontBBox [{} {} {} {}] /FontMatrix [0.001 0 0 0.001 0 0] /CharProcs << {procs}>> /Encoding << /Type /Encoding /Differences [1 {names}] >> /FirstChar 1 /LastChar {} /Widths [{widths}] /Resources << >> /ToUnicode {cmap_id} 0 R >>",
n(bounds[0]),
@@ -536,10 +539,10 @@ impl Subset {
n(bounds[3]),
glyphs.len()
),
)?;
);
let _ = write!(resources, "/A{} {id} 0 R ", batch + 1);
}
Ok(resources)
resources
}
}
+88 -29
View File
@@ -1,4 +1,4 @@
//! PDF 1.7 drawing for GridCraft print/export, with shared object and xref framing.
//! A tiny, self-contained PDF 1.7 writer for GridCraft's print and PDF export.
//!
//! - The 14 standard fonts (no embedding) with WinAnsiEncoding, plus approximate width tables
//! ([`text_width`]) so callers can measure, align and wrap text. Characters outside
@@ -52,7 +52,7 @@ impl TextDirection {
}
}
/// A PDF export could not preserve the requested text or serialize its object envelope.
/// A PDF export could not preserve the requested text.
#[derive(Clone, Debug)]
pub struct Error(String);
@@ -63,12 +63,6 @@ impl std::fmt::Display for Error {
}
impl std::error::Error for Error {}
impl From<craft_pdf::classic::Error> for Error {
fn from(error: craft_pdf::classic::Error) -> Self {
Self(error.to_string())
}
}
use std::collections::HashMap;
use std::collections::hash_map::DefaultHasher;
use std::fmt::Write as _;
@@ -1007,6 +1001,49 @@ fn pdf_text_string(text: &str) -> String {
s
}
/// Starts object `id` at the current end of `out` and records its offset for the xref table.
fn begin_obj(out: &mut Vec<u8>, offsets: &mut Vec<usize>, id: u32) {
let i = id as usize;
if offsets.len() <= i {
offsets.resize(i + 1, 0);
}
if let Some(slot) = offsets.get_mut(i) {
*slot = out.len();
}
let _ = writeln!(Utf8(out), "{id} 0 obj");
}
/// Writes `id 0 obj <body> endobj`.
fn dict_obj(out: &mut Vec<u8>, offsets: &mut Vec<usize>, id: u32, body: &str) {
begin_obj(out, offsets, id);
out.extend_from_slice(body.as_bytes());
out.extend_from_slice(b"\nendobj\n");
}
/// Writes a stream object: `dict` (the dictionary entries without `<<`, `>>` and `/Length`, may be
/// empty) and `data` as the stream body.
fn stream_obj(out: &mut Vec<u8>, offsets: &mut Vec<usize>, id: u32, dict: &str, data: &[u8]) {
begin_obj(out, offsets, id);
out.extend_from_slice(b"<<");
if !dict.is_empty() {
out.push(b' ');
out.extend_from_slice(dict.as_bytes());
}
let _ = write!(Utf8(out), " /Length {} >>\nstream\n", data.len());
out.extend_from_slice(data);
out.extend_from_slice(b"\nendstream\nendobj\n");
}
/// Lets `write!` append text to a byte buffer.
struct Utf8<'a>(&'a mut Vec<u8>);
impl std::fmt::Write for Utf8<'_> {
fn write_str(&mut self, s: &str) -> std::fmt::Result {
self.0.extend_from_slice(s.as_bytes());
Ok(())
}
}
impl PdfDoc {
pub fn new() -> PdfDoc {
PdfDoc::default()
@@ -1060,8 +1097,7 @@ impl PdfDoc {
/// numbered during writing (fonts, then per page its new images, content stream and page
/// dictionary), with the catalog, page tree and info dictionary last.
///
/// Returns an error if the Arabic font is unusable, text exceeds a size limit, or the
/// shared envelope writer refuses object counts, offsets or an allocation.
/// Returns an error if the Arabic font is unusable or Arabic text exceeds a size limit.
pub fn finish(mut self) -> Result<Vec<u8>, Error> {
if let Some(error) = self.pages.iter().find_map(|page| page.error.clone()) {
return Err(error);
@@ -1095,7 +1131,9 @@ impl PdfDoc {
let contents: Vec<String> = self.pages.iter_mut().map(|p| std::mem::take(&mut p.content)).collect();
let page_links: Vec<Vec<LinkAnnotation>> = self.pages.iter_mut().map(|p| std::mem::take(&mut p.links)).collect();
let reserve = contents.iter().fold(4096usize, |acc, c| acc.saturating_add(c.len()).saturating_add(1024));
let mut out = craft_pdf::classic::Writer::with_capacity(reserve)?;
let mut out: Vec<u8> = Vec::with_capacity(reserve);
out.extend_from_slice(b"%PDF-1.7\n%\xE2\xE3\xCF\xD3\n");
let mut offsets: Vec<usize> = vec![0; 4];
let mut kids = Vec::new();
let mut annotations = Vec::new();
let page_count = self.pages.len();
@@ -1105,34 +1143,42 @@ impl PdfDoc {
let _ = write!(font_dict, "/F{} {} 0 R ", i + 1, id);
if i == CID_FONT {
let (cid_id, desc_id) = (id + 1, id + 2);
out.object(
dict_obj(
&mut out,
&mut offsets,
*id,
&format!(
"<< /Type /Font /Subtype /Type0 /BaseFont /{CID_BASE_FONT} /Encoding /UniGB-UTF16-H /DescendantFonts [{cid_id} 0 R] >>"
),
)?;
out.object(cid_id,
);
dict_obj(
&mut out,
&mut offsets,
cid_id,
&format!(
"<< /Type /Font /Subtype /CIDFontType0 /BaseFont /{CID_BASE_FONT} /CIDSystemInfo << /Registry (Adobe) /Ordering (GB1) /Supplement 4 >> /FontDescriptor {desc_id} 0 R /DW {CID_WIDTH} >>"
),
)?;
out.object(desc_id,
);
dict_obj(
&mut out,
&mut offsets,
desc_id,
&format!(
"<< /Type /FontDescriptor /FontName /{CID_BASE_FONT} /Flags 6 /FontBBox [-25 -254 1000 880] /ItalicAngle 0 /Ascent 880 /Descent -120 /CapHeight 880 /StemV 93 >>"
),
)?;
);
continue;
}
let name = Font::ALL.get(i).copied().unwrap_or_default().base_name();
let enc = if name.starts_with("Symbol") { "" } else { " /Encoding /WinAnsiEncoding" };
out.object(*id, &format!("<< /Type /Font /Subtype /Type1 /BaseFont /{name}{enc} >>"))?;
dict_obj(&mut out, &mut offsets, *id, &format!("<< /Type /Font /Subtype /Type1 /BaseFont /{name}{enc} >>"));
}
}
// Images already written, by content hash: (page, image, object id). A hash match is
// confirmed by comparing the pixels before an object is shared.
let mut written: HashMap<u64, Vec<(usize, usize, u32)>> = HashMap::new();
for (pi, ((page, content), links)) in self.pages.iter().zip(contents).zip(page_links).enumerate() {
let arabic_fonts = page.arabic.write_objects(&mut next, &mut out)?;
let arabic_fonts = page.arabic.write_objects(&mut next, &mut out, &mut offsets);
let page_fonts = format!("{font_dict}{arabic_fonts}");
let mut xobjects = String::new();
for (k, img) in page.images.iter().enumerate() {
@@ -1158,13 +1204,13 @@ impl PdfDoc {
let _ = write!(dict, " /SMask {id} 0 R");
}
dict.push_str(" /Filter /RunLengthDecode");
out.stream(img_id, &dict, &run_length(&img.rgb))?;
stream_obj(&mut out, &mut offsets, img_id, &dict, &run_length(&img.rgb));
if let (Some(id), Some(a)) = (smask_id, &img.alpha) {
let dict = format!(
"/Type /XObject /Subtype /Image /Width {} /Height {} /ColorSpace /DeviceGray /BitsPerComponent 8 /Filter /RunLengthDecode",
img.width, img.height
);
out.stream(id, &dict, &run_length(a))?;
stream_obj(&mut out, &mut offsets, id, &dict, &run_length(a));
}
written.entry(img.hash).or_default().push((pi, k, img_id));
img_id
@@ -1175,7 +1221,7 @@ impl PdfDoc {
let content_id = next;
let page_id = next + 1;
next += 2;
out.stream(content_id, "", content.as_bytes())?;
stream_obj(&mut out, &mut offsets, content_id, "", content.as_bytes());
drop(content);
let xo = if xobjects.is_empty() { String::new() } else { format!(" /XObject << {xobjects}>>") };
let mut link_refs = String::new();
@@ -1189,7 +1235,9 @@ impl PdfDoc {
annotations.push((id, link));
}
let annots = if link_refs.is_empty() { String::new() } else { format!(" /Annots [{link_refs}]") };
out.object(
dict_obj(
&mut out,
&mut offsets,
page_id,
&format!(
"<< /Type /Page /Parent 2 0 R /MediaBox [0 0 {} {}] /Resources << /Font << {}>>{} >> /Contents {} 0 R{} >>",
@@ -1200,7 +1248,7 @@ impl PdfDoc {
content_id,
annots
),
)?;
);
kids.push(page_id);
}
for (id, link) in annotations {
@@ -1213,11 +1261,16 @@ impl PdfDoc {
format!("/Dest [{page_id} 0 R /XYZ {} {} null]", n(x), n(height - y))
}
};
out.object(id, &format!("<< /Type /Annot /Subtype /Link /Rect [{} {} {} {}] /Border [0 0 0] {action} >>", n(x0), n(y0), n(x1), n(y1)))?;
dict_obj(
&mut out,
&mut offsets,
id,
&format!("<< /Type /Annot /Subtype /Link /Rect [{} {} {} {}] /Border [0 0 0] {action} >>", n(x0), n(y0), n(x1), n(y1)),
);
}
out.object(1, "<< /Type /Catalog /Pages 2 0 R >>")?;
dict_obj(&mut out, &mut offsets, 1, "<< /Type /Catalog /Pages 2 0 R >>");
let kids_s: Vec<String> = kids.iter().map(|k| format!("{k} 0 R")).collect();
out.object(2, &format!("<< /Type /Pages /Kids [{}] /Count {} >>", kids_s.join(" "), kids.len()))?;
dict_obj(&mut out, &mut offsets, 2, &format!("<< /Type /Pages /Kids [{}] /Count {} >>", kids_s.join(" "), kids.len()));
let creator = if self.info.creator.is_empty() { "GridCraft" } else { &self.info.creator };
let mut info = format!("<< /Producer {} /Creator {}", pdf_text_string("GridCraft PDF"), pdf_text_string(creator));
if !self.info.title.is_empty() {
@@ -1227,9 +1280,15 @@ impl PdfDoc {
let _ = write!(info, " /Author {}", pdf_text_string(&self.info.author));
}
info.push_str(" >>");
out.object(3, &info)?;
dict_obj(&mut out, &mut offsets, 3, &info);
out.finish(1, Some(3)).map_err(Error::from)
let xref = out.len();
let _ = write!(Utf8(&mut out), "xref\n0 {}\n0000000000 65535 f \n", next);
for off in offsets.iter().skip(1) {
let _ = writeln!(Utf8(&mut out), "{off:010} 00000 n ");
}
let _ = write!(Utf8(&mut out), "trailer\n<< /Size {} /Root 1 0 R /Info 3 0 R >>\nstartxref\n{}\n%%EOF\n", next, xref);
Ok(out)
}
}
+1 -4
View File
@@ -1136,10 +1136,7 @@ fn explicit_rtl_standard_text_reorders_without_an_arabic_font() {
let page = doc.add_page(300.0, 100.0);
page.text_with_direction(Point::new(10.0, 30.0), 12.0, Font::Helvetica, [0, 0, 0], source, TextDirection::RightToLeft);
assert!(page.content.contains("(! ABC) Tj"), "{}", page.content);
let mut objects = craft_pdf::classic::Writer::with_capacity(0).unwrap();
let mut next = 4;
assert!(page.arabic.write_objects(&mut next, &mut objects).unwrap().is_empty());
assert_eq!(next, 4, "Latin-only RTL text must not allocate Arabic font objects");
assert!(page.arabic.write_objects(&mut 4, &mut Vec::new(), &mut Vec::new()).is_empty());
assert_eq!(text_width_with_direction(Font::Helvetica, 12.0, source, TextDirection::RightToLeft), text_width(Font::Helvetica, 12.0, source));
let pdf = doc.finish().unwrap();
assert_eq!(utf16_entry(&pdf, "/ActualText").unwrap(), source);
+1
View File
@@ -14,6 +14,7 @@ workspace = true
test-support = []
[dependencies]
craft-xlsx.workspace = true
gridcraft-core = { workspace = true }
gridcraft-formula = { workspace = true }
gridcraft-math = { workspace = true }
+3 -64
View File
@@ -360,6 +360,7 @@ pub fn parse_streaming(xml: &[u8], split: &[&str], cb: &mut dyn FnMut(El) -> Res
// ---------------------------------------------------------------- writing helpers
// XML 1.0 byte escaping is shared; spreadsheet string and phonetic policy stays here.
#[cfg(test)]
use craft_xml::is_xml_char;
#[cfg(test)]
use craft_xml::push_esc;
@@ -402,70 +403,8 @@ pub fn push_a1(o: &mut String, c: gridcraft_core::CellRef) {
push_u64(o, u64::from(c.row) + 1);
}
/// ST_Xstring encoding for cell text: characters XML can't carry and carriage returns become
/// `_xHHHH_`, and literal `_xHHHH_` sequences are protected with `_x005F_`. The result still needs
/// [`esc`].
pub fn encode_xstring(s: &str) -> String {
let mut o = String::with_capacity(s.len());
let chars: Vec<char> = s.chars().collect();
for (i, &c) in chars.iter().enumerate() {
if c == '_' && looks_like_escape(&chars, i) {
o.push_str("_x005F_");
} else if !is_xml_char(c) || (c.is_control() && !matches!(c, '\t' | '\n')) {
o.push_str(&format!("_x{:04X}_", c as u32 & 0xFFFF));
} else {
o.push(c);
}
}
o
}
fn looks_like_escape(chars: &[char], i: usize) -> bool {
chars.get(i + 1) == Some(&'x') && chars.get(i + 6) == Some(&'_') && (i + 2..i + 6).all(|k| chars.get(k).is_some_and(char::is_ascii_hexdigit))
}
fn escape_unit(chars: &[char], i: usize) -> Option<u32> {
let hex: String = chars.get(i + 2..i + 6)?.iter().collect();
u32::from_str_radix(&hex, 16).ok()
}
/// The character for a high surrogate escape at `i` (value `high`) followed directly by a low surrogate escape.
fn surrogate_pair(chars: &[char], i: usize, high: u32) -> Option<char> {
if !(0xD800..=0xDBFF).contains(&high) || chars.get(i + 7) != Some(&'_') || !looks_like_escape(chars, i + 7) {
return None;
}
let low = escape_unit(chars, i + 7).filter(|l| (0xDC00..=0xDFFF).contains(l))?;
char::from_u32(0x10000 + ((high - 0xD800) << 10) + (low - 0xDC00))
}
/// Inverse of [`encode_xstring`].
pub fn decode_xstring(s: &str) -> String {
if !s.contains("_x") {
return s.to_string();
}
let chars: Vec<char> = s.chars().collect();
let mut o = String::with_capacity(s.len());
let mut i = 0;
while i < chars.len() {
if chars[i] == '_' && looks_like_escape(&chars, i) {
let unit = escape_unit(&chars, i);
if let Some(ch) = unit.and_then(char::from_u32) {
o.push(ch);
i += 7;
continue;
}
// A high surrogate escape followed by a low surrogate escape is one UTF-16 pair.
if let Some(ch) = unit.and_then(|high| surrogate_pair(&chars, i, high)) {
o.push(ch);
i += 14;
continue;
}
}
o.push(chars[i]);
i += 1;
}
o
}
/// Shared ST_Xstring encoding/decoding; XML escaping remains a separate operation.
pub use craft_xlsx::{decode_xstring, encode_xstring};
/// `<t>` element with `xml:space="preserve"` when needed.
pub fn t_el(s: &str) -> String {