Changes the container backend strategy to use SD catalog by default for all EPUBs. On first open or stale index, builds the catalog once by borrowing the framebuffer (which can be released since e-ink holds the loading popup image). Subsequent opens use the fast path with compact resident tables. Falls back to legacy in-RAM open if catalog build fails due to IoError or OOM.
Implement cover extraction for books stored in catalog.fibc omnibus format to avoid loading large (~400KB) book arenas into RAM. Uses catalog's pre-resolved cover entries with minimal memory footprint (~42KB resident). Includes arena management for catalog opening and ZIP entry extraction with configurable scratch buffers.
Implements a two-tier system for handling EPUB books: small books use in-RAM metadata (probe ladder with 48KB/72KB/96KB rungs), while large containers (e.g. webnovel omnibuses with 1500+ spine items) automatically fall back to SD-backed BookCatalog with persistent index. The catalog fast path reuses existing catalog.fibc files to skip probing on subsequent opens. Adds framebuffer borrowing for larger probe rungs and incremental chapter building for giant chapters.
Introduce methods to temporarily release the 48KB framebuffer during memory-intensive chapter builds, allowing the memory to be reused. Add font metadata query methods (hasGlyphMeta, ensureAdvance) that check glyph existence and fetch advance metrics without loading full bitmap data, avoiding expensive SD reads during layout measurement.
Replace fixed-size arena buffers with two-pass allocation strategy. First pass parses book into temporary full-size arena to determine actual memory footprint (5-40 KB vs 48 KB cap), then reallocates exact-sized buffers. This reduces peak memory usage by avoiding simultaneous allocation of all arenas with chapter-build scratch buffer (~100 KB) on constrained heap (~140 KB free). Stylesheet is also compacted to actual size, and all open-time transients are freed before per-chapter arenas are allocated.
Refactors book cover and metadata handling out of Epub class into a dedicated BookCoverUtils utility. Moves FootnoteEntry struct into reader activity directory. Simplifies cache clearing to use direct storage operations instead of Epub class.
Implements bidirectional conversion between KOReader XPath positions and FreeInkBook character offsets. Uses streaming SAX parsing to match chapter text accounting with ChapterLayout, enabling position sync without requiring DOM or pagination. Handles element ancestry tracking, text offset calculation, and spine index extraction.
Implements layout-independent bookmarks using character offsets within chapters instead of percentage-based positions. This allows bookmarks to restore to the exact position regardless of font size or orientation changes. Adds TOC navigation helpers, book progress calculation based on spine item sizes, and a NONE BidiBaseDir option for pre-reordered text. Legacy bookmarks without character offsets fall back to percentage-based positioning.
## Summary
* **What is the goal of this PR?** Fixes#2402. On the X3, the "Time to
Sleep"
picker's 5-minute side buttons were inverted (left increased, right
decreased)
and the on-screen legend didn't match the physical buttons.
* **What changes are included?**
* Flip the large-step (±5 min) direction on X3 so the left side button
decreases and the right increases, matching the layout. X4 is unchanged.
* Add a device-specific step-hint string
(`STR_SLEEP_TIMER_STEP_HINT_X3`)
selected via `gpio.deviceIsX3()`. X3 shows `+/-: 1 min Side buttons: 5
min`;
X4 keeps the original `Left/Right: 1 min Up/Down: 5 min`.
* Same fix applied to both slider pickers that read the raw up/down side
buttons: the Time-to-Sleep picker and the "Go to %" picker
(`EpubReaderPercentSelectionActivity`), each with its own X3 hint string
(`STR_SLEEP_TIMER_STEP_HINT_X3`, `STR_PERCENT_STEP_HINT_X3`).
## Additional Context
* **Root cause:** the X3's side buttons sit one on each edge of the
screen
(power on top), whereas the X4 has a vertical up/down rocker on the
right
edge. So `BTN_UP` is physically the *left* button on X3 but the *top*
button
on X4. The picker mapped up→+5 / down→−5 unconditionally, which reads
naturally on the X4 rocker but inverted on the X3's left/right buttons.
The
static legend ("Up/Down: 5 min") had the same X3-only mismatch.
* The codebase already detects the device at runtime and handles this
layout
difference elsewhere (e.g. `LyraTheme::drawSideButtonHints` draws "Up on
left,
Down on right" for X3), so this reuses the same `gpio.deviceIsX3()`
signal.
* Swept the codebase for other side-button sliders:
`ClockOffsetActivity` already
uses the `Next/Previous` (`ButtonNavigator`) abstraction, which resolves
to
left=decrement / right=increment on X3, so it needs no change. List/page
navigation (Wi-Fi, KOReader sync, BMP viewer) and the keyboard cursor
are also
unaffected. The two slider pickers above were the only ones using raw
up/down.
* **X4 is untouched** — same direction, same legend wording as before.
* **Tested on X3 hardware:** left side button now decreases, right
increases, and
the legend matches. X4 not yet tested on device (no unit on hand); its
behaviour and strings are unchanged from master.
* **Translations:** the new `STR_SLEEP_TIMER_STEP_HINT_X3` and
`STR_PERCENT_STEP_HINT_X3` were added to all 26 languages, but the
non-English
ones are AI-generated and would benefit from a native-speaker pass —
particularly Hebrew (RTL ordering with a leading `+/-`), Kazakh, and
Vietnamese.
---
### AI Usage
Did you use AI tools to help write this code? _**YES**_
---------
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
This PR moves us from the xteink openx4 SDK to the freeink sdk from
https://freeink.org. Out of the box there are NO changes needed in the
firmware to support this swap, it all magically works as is. However as
we support more than just the x3/x4 devices, this sdk allows us to pass
env vars into the build commands to include support for other devices.
As support for new hardware such as touch screens and bluetooth are
added the xteink builds decide at compile time if the libraries are used
or not. For example right now the freeinkui and icons libraries are in
the platform.io file but as they are not used anywhere, they won't be
included in the final build. Once the touch branch and sd themes branch
are merged in this sdk is required for them to function correctly. All
the docs for freeink are available at freeink.org/docs. x4/x3 is a
single binary build unlike other devices that will build unique binaries
for each device. Eventually we will want to remove a lot of the manual
isx3 type stuff from our firmware and go through the boardsupport api
the sdk provides as it will generalize everything into one common system
that any device can support. The upcoming touch branch does a lot of
this for us but this initial PR is JUST to get the sdk swapped over
without any code changes to show seamless integration without any
regressions.
## Improvements
* Moved the File Manager breadcrumb into the Contents card header for a
cleaner, more consistent interface.
* Updated the Wireless Transfer section of the User Guide with clearer
instructions.
* Battery status bar indicator no longer changes position when adding or
removing bookmarks.
## Performance
* Optimized path normalization for faster file handling.
* Significantly improved bookmark rendering by removing unnecessary
XPath lookups.
* Optimized dithered rectangle drawing (fillRectDither) using a
byte-aligned rendering implementation, improving display performance on
supported devices.
## Bug Fixes
* Fixed an issue where the Inverted Orientation label was incorrectly
combined with the Color Filter label for Geman localization.
* Fixed excessive ghosting on the X3 cover screen during sleep
## Summary
* **What is the goal of this PR?**
Replace the pixel-by-pixel `fillRectDither` implementation with a new
byte-aligned `fillRectImpl` that eliminates per-pixel
`rotateCoordinates` calls and Read-Modify-Write bitwise loops, yielding
a significant rendering speedup on ESP32 E-ink framebuffers.
Ported from @rhythmerc's crosspoint-reader fork commit 27ea625.
* **What changes are included?**
- **`GfxRenderer.cpp` — `fillRectDither` refactor:** The existing
`if/else if` chain is replaced with a `switch` statement that delegates
each `Color` case to the new `fillRectImpl<Color>()` template,
eliminating runtime branching.
- **`GfxRenderer.cpp` — new `fillRectImpl<C>()` template:** Core of the
optimization. Key behaviors:
- Clips the rectangle in logical space upfront.
- Rotates only **2 opposing corner points** (top-left and bottom-right)
into physical framebuffer space instead of rotating every pixel
individually.
- Derives physical-space `byteStart`/`byteEnd` and precomputes
`headMask` / `tailMask` for MSB-first partial-byte boundaries,
performing RMW only on the edge bytes.
- **Solid fills (`Black` / `White`):** Uses `memset` for all interior
full-byte runs per row — no per-pixel writes.
- **Dithered fills (`LightGray` / `DarkGray`):** Precomputes both parity
variants of `blackMask` (even/odd `py`) **outside** the row loop,
eliminating the previously re-evaluated 8-bit construction loop on every
physical row. Interior full bytes are then written with a single
`memset(whiteMask)`.
- Uses `if constexpr` throughout to dispatch on `Color` at compile time,
generating zero runtime branches per template instantiation.
- **`GfxRenderer.h`:** Declares the new private `fillRectImpl<Color>()`
template method with an explanatory doc-comment.
- **Explicit template instantiations** added for all four active `Color`
variants (`Black`, `White`, `LightGray`, `DarkGray`).
## Additional Context
* **Performance:** The primary motivation is ESP32 E-ink framebuffer
performance. The old path called `rotateCoordinates` and did a full RMW
for every single pixel in the rectangle. The new path calls
`rotateCoordinates` exactly **twice** per fill regardless of rectangle
size, then operates at byte granularity — a complexity reduction from
O(W×H) coordinate transforms to O(1).
* **Dither correctness:** The `blackMask` precomputation relies on the
dither pattern having period 2 in both logical X and Y, which makes the
per-row byte pattern repeat with period 2 in `py`. Reviewers should
verify the `lxBase`/`lyBase` derivations for all four orientations
(`Portrait`, `PortraitInverted`, `LandscapeClockwise`,
`LandscapeCounterClockwise`) match the inverse of `rotateCoordinates`.
* **Edge case — single-byte rows:** When `byteStart == byteEnd`, the
head and tail masks are ANDed together into a single `rectMask` to avoid
double-masking the same byte. This path should be tested with narrow
rectangles (width < 8px).
* **No behavioral change for `Color::Clear`:** The `Clear` case exits
early via `if constexpr` and is a no-op, matching the original behavior.
---
### AI Usage
While CrossPoint doesn't have restrictions on AI tools in contributing,
please be transparent about their usage as it
helps set the right context for reviewers.
Did you use AI tools to help write this code? _**< YES >**_
---------
Co-authored-by: Ryan Mercado <rmercado@firstdollar.com>
Force display resync on X3 when HALF refresh is requested to clear prior content before rendering. Add grayscale preconditioning for X3's UC81xx controller to even out single-pixel dithering artifacts that appear as speckle with its turbo BW waveform.
## Summary
fix#2414
### Root Cause
`updateBookmarkFlag()` was introduced in commit 1db1442 and is executed
on every render (every page turn). The function calls:
`ProgressMapper::toSavedProgress()`
→ `ChapterXPathResolver::findXPathForProgress()`
This path decompresses the current EPUB section content twice:
1. To count visible characters.
2. To resolve the corresponding XPath.
For larger sections (e.g. ~133 KB decompressed content), this adds
approximately **1 second of I/O overhead per page turn**, with the cost
increasing as chapter size grows.
### Fix
`updateBookmarkFlag()` only needs to determine whether a bookmark falls
within the currently displayed page range.
The required information is already available during rendering:
* `currentPage`
* `section->pageCount`
* `currentSpineIndex`
Instead of converting the current location to a saved progress object
(and resolving an XPath), the implementation now computes the current
page's progress range directly and compares bookmark percentages against
that range.
This is effectively the same percentage-based matching logic already
used as a fallback in `bookmarkMatchesProgress()` when XPath matching is
unavailable.
---
### AI Usage
While CrossPoint doesn't have restrictions on AI tools in contributing,
please be transparent about their usage as it
helps set the right context for reviewers.
Did you use AI tools to help write this code? _**< PARTIALLY >**_
Justin Mitchell