XLSX: printOptions gridLines="1" is ignored, so printed cell gridlines are missing entirely

[developer0hye]

Summary

<printOptions gridLines="1"/> tells Excel to print the worksheet gridlines across the used
range, independently of any <border> styling. office2pdf does not read the attribute, so a
sheet that relies on printed gridlines for its structure comes out with no rules at all.

Measured

NumberFormatTests sets <printOptions gridLines="1" .../> on both worksheets.

	GT (Excel)	office2pdf
cell gridlines, all four sides, every cell in the used range	present, ~0.5pt solid hairline	absent
pages carrying them	2-7 (337 rows × 4 columns)	none
explicit <border> rules on the same pages	none declared	none drawn

Nothing else on those pages differs in stroke: the fixture declares no cell borders, so the
entire ruling of the sheet comes from this one attribute.

Reproduction

cargo run -p office2pdf-cli -- tests/fixtures/xlsx/poi/NumberFormatTests.xlsx -o out.pdf
mutool draw -F trace -o - out.pdf 2 | grep -ci 'stroke_path\|fill_path'   # no gridline rules

Expected

With printOptions@gridLines set, every cell in the printed range is ruled on all four
sides with Excel's gridline hairline, under any explicit border styling.

Actual

The attribute is ignored and no gridlines are drawn, leaving the column structure of a
formatting-test grid ambiguous.

Found in the 2026-07-29 full-corpus visual audit at c1adfbd (native Office GT, 150 DPI, mutool-trace geometry).

Visual comparison

Compare

GT left, office2pdf right, 150 DPI, same page.

[developer0hye]

Implementation note for whoever picks this up.

The attribute cannot be read through umya. PrintOptions in the pinned
fork exposes only get_horizontal_centered and get_vertical_centered:

$ grep -n "pub fn get_" .../umya-spreadsheet/src/structs/print_options.rs
17:    pub fn get_horizontal_centered(&self) -> &bool {
28:    pub fn get_vertical_centered(&self) -> &bool {

gridLines, gridLinesSet and headings are parsed off the element by
set_attributes but never surfaced, so sheet.get_print_options() cannot
answer the question this issue asks. That is why nothing in
crates/office2pdf/src/parser/ references the attribute today — a grep for
gridLines across the parser returns nothing.

Two ways forward, and the repo already has a precedent for the second:

1. Patch the fork to expose the getters (the workspace already carries
   [patch.crates-io] umya-spreadsheet = { git = ... }).
2. Extract it from the raw package XML, which is what
   xlsx_fit_to_page.rs::sheets_fitting_to_page(data) and
   xlsx_cond_fmt_raw.rs::extract_cond_fmt_hints(data) already do for
   attributes umya does not expose. Both take the original &[u8] and return a
   per-sheet map, and XlsxParser::parse already threads that shape through.

Option 2 matches the existing pattern and keeps the change inside this repo.

Ground truth is reproducible locally on Windows. Native Excel 16.0 exports
NumberFormatTests.xlsx to 15 pages at A4; page 2's printed grid shows up in
mutool draw -F trace as paired vertical rules at x = 54.0, 76.1, 197.0,
307.9, … , i.e. one per column boundary across the used range, which is the
ruling this issue says is missing.

Note #623 (printOptions@headings) is the same attribute element and the same
umya gap, but a separate root cause — it needs the row/column gutters drawn and
the grid inset to make room, which this issue does not.

[developer0hye]

Fixed by #721. printOptions gridLines (with the ECMA-376 gridLinesSet conjunction, read by a scanner that ignores customSheetViews' nested printOptions) now prints Excel's gridlines as 1pt pure-black boundary-anchored [B, B+1] bands through the #619 border machinery — measured off this fixture's own native GT, which shows black fills (not gray strokes), explicit borders owning their boundaries, and cell fills suppressing all four adjacent segments.

Fresh GT comparison on the Tests sheet: every rule lands on GT's exact bands with fill-edge suppression reproduced (before: zero rules). Remaining deviations on the audited page each have their own open issue: #658 (row track), #717 (Letter default), #623 (heading gutters), #718 (formatted empty column), #719 (preserved spaces), #657 (left inset).