Keeping a 1 x 1 table for every further `<Table>` was unbounded in the one
direction the budget exists to close. `<Table></Table>` is fifteen bytes, so
the 512 MiB a part may decompress to holds some 36 million of them, and a
`Cell` is 1,360 bytes: tens of gigabytes, with every table inside both caps.
The test I wrote said as much without my noticing -- its invariant was
`reserved <= MAX + tables`, and `tables` had no bound.
`budgeted_size` now returns `None` once the budget is spent and the table is
not read at all, so the cells one import reserves are at most the budget,
exactly. The invariant is now `reserved <= MAX_IMPORT_TABLE_CELLS`, with no
term that grows with the file.
The warnings needed the same treatment, for the same reason: a file of 400,000
tables each claiming one row and two columns got a line about every one of
them. At most 20 are named, then one line saying the rest were the same, and
the file is told once that its tables ran out of cells between them.
Review pointed out that capping each table leaves a file free to hold
many capped tables: three at the per-table limit claim 750,000 cells,
about 1 GB of Cell, with every table inside its own limit. The package
reader's limits count input bytes, which say nothing about how many
cells a row and column count claim.
budgeted_size clamps each table against what the import has left, and
the cells are reserved before Table::new so a cell's own story cannot
spend the budget the table around it is counting on. The budget is
400,000 cells: a Cell carries a whole story and is 1,360 bytes, so that
is about the 512 MiB one part of the file may already decompress to,
and the same count the spreadsheet importer collects at most.
A table is never read as nothing. One row and one column are kept, so
the table stays in the document and still says where its text went, and
a file pays one cell for a table it really does contain -- which its own
size already bounds. The warning says which limit cut the table down.
The arithmetic is tested rather than the allocation: 400,000 cells is
544 MB and three such tables are not a thing to ask CI to allocate. The
sequence test runs 32 claimed sizes, including usize::MAX, and holds the
invariant across all of them.
`Table::new` allocates `rows * cols` cells up front, and the counts reach it
from IDML and DOCX, where they are whatever the file claims. Measured here:
`size_of::<Cell>()` is 1,360 bytes, so the reported IDML case -- one `<Table>`
with 5,000 `<Row/>` and 5,000 `<Column/>`, about 1 KB zipped -- asks for 25M
cells, some 34 GB. That allocation does not fail politely: it aborts the
process through `handle_alloc_error`, which no `catch_unwind` can catch, so
every open document's unsaved work goes with it.
A table is clamped to InDesign's own limits, 10,000 rows and 1,000 columns,
and to 250,000 cells whatever its shape -- the row and column limits alone
would still allow 13.6 GB. Columns are clamped first and rows take what is
left, because a table is far more often long than wide.
The IDML importer walks the clamped slices and reports the difference.
Indexing its rows and columns by the file's own counts would be out of
bounds, which is how a 34 GB abort would otherwise have become a panic.
Rebased onto the post-split main: the constants and `clamped_size` live in
`doc`'s `table` module root, `Table::new` in `table/grid.rs`, and the IDML
table reader in `import/table.rs`.
The policy check now reads the compiled AppSearch and RegLocator rows for the
remembered-folder properties (registry root, key, value name and type for both
scopes) and runs the DIR, FOLDER, PATH and legacy exe actions in sequence order
against seeded values, so their precedence and the preset-folder guard cannot
regress. It also replays the wizard folder actions in sequence order (a
remembered folder beats the new-install default, a typed folder beats both) and
requires the machine restore action between the scope page and costing.
The pre-signing check matched any folder named Program Files, so it rejected
a valid MSI when Windows resolves Program Files elsewhere, accepted the 64-bit
root for an x86 package and accepted a different app name. Build the expected
folder from the package architecture's resolved root and the policy name.
The PDF exporter called Cache::get twice for every text frame it drew.
Each call rebuilds the cache signature of the whole story, which looks up
every frame of the story with Document::find (a linear walk over all
spreads), and the anchored-frame placements walk them again. For a story
threaded through many frames this is O(frames^2 x items) per export.
The document doesn't change during an export, so the exporter now keeps
the composition it got for each (story, page) and reuses it. The cache
still decides what to compose; the output is byte-identical.
Add examples/pdf_export.rs to designcraft-cli: File > Export PDF through
a headless session on the `designcraft-cli perf` stress document (200
pages, 1M characters in 40 threaded stories, 500 placed images). The
stress document generator moves unchanged from src/perf.rs to a small
library module (src/stress.rs) that the binary and the example share.
Ryzen 16 threads, Linux, 12 runs each, interleaved:
main median 3485 ms, min 3427 ms
patched median 2016 ms, min 1656 ms (-42 % median, -52 % min)
The exported PDF is byte-identical (12,075,744 bytes).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Real msiexec runs (per-user and all-users, silent and wizard) found two defects
that the metadata fixtures cannot see:
- "Only for me" published ALLUSERS=2, so FindRelatedProducts treated the run as
per-machine and skipped the existing per-user copy. An interactive upgrade
left two registered products and uninstall left files behind. The page now
leaves ALLUSERS unset for a per-user choice.
- Windows Installer resolves Program Files once, for the context it started in,
so choosing "Everyone" in the wizard installed into the user profile with a
machine-wide registration. The wizard now sets the new-install folder itself
after the choice (Program Files, Program Files (x86) for 32-bit, or the
per-user Programs folder). Silent installs are unchanged.
check-install-policy.ps1 now fails if either regresses.
* fix(ui): font menus grow back after a search with no matches
Every combo box is craft_ui::ComboBox (craft-libs pinned at 62bb0e1d), whose
menu follows its contents' height while open (egui#5225); clippy.toml bans
egui::ComboBox.
The Font menu splits into a header (search field, filters) and a list
(font_menu_header, font_menu_list). Font fields show it in a
craft_ui::combo::Menu and the style dialogs' Font Family combo boxes in
ComboBox::show_ui_with_header, so only the rows scroll and the search field
stays in place. The blank and Parent's Font rows sit in the header, above
the search.
* chore(deps): craft-libs 60d60d48 (dropdown fixes)
* chore(deps): craft-libs 4d7b82f7 (dropdown fixes)
* chore(deps): craft-libs 92d392cf (#387 merged)
* chore(xtask): pdf-deps expects craft-libs 92d392cf
Auto-Collapse Iconic Panels is off by default, so a flyout closes only from its icon or
another icon. The setting is saved under a new key, so preferences saved with the old
default start with it off, and Preferences > Interface > Panels shows it.
Split files share test helpers through pub(super)/pub(crate) test modules. The guard read only `mod` and `pub mod` as a test module, so it scanned their fixtures as production code. `attr.contains("test")` also matched `#[cfg(not(test))]`, so such a module was skipped; it is scanned now.
Hamza