Adds HAL-level support for the Xteink X3 (SSD1677 controller, 792x528
display). Includes device detection, display initialization, button
mapping, power management, and theme adjustments for the X3 form factor.
## Purpose
This PR includes some preparatory changes that are needed for an
upcoming performant CJK font feature. The changes have no impact on
render time and heap allocation for latin text. **Despite this, I think
these changes stand on their own as a better font
compression/decompression implementation.**
## Summary
- Font decompressor rewrite: Replaced the 4-slot LRU group cache with a
two-tier system — a page buffer (glyphs prewarmed before rendering
begins) and a hot-group fallback (last decompressed group retained for
non-prewarmed
glyphs).
- Byte-aligned compressed bitmap format: Glyph bitmaps within compressed
groups are now stored row-padded rather than tightly packed before
DEFLATE compression, improving compression ratios by making identical
pixel rows produce
identical byte patterns. Glyphs are compacted back to packed format on
demand at render time. Reduces flash size by 155 KB.
- Page prewarm system: Added `Page::collectText` and
`Page::getDominantStyle` to extract per-style glyph requirements before
rendering, and `GfxRenderer::prewarmFontCache` to pre-decompress only
the groups needed for the dominant style
— eliminating mid-render decompression for the common case.
- UTF-8 robustness fixes: `utf8NextCodepoint` now validates continuation
bytes and returns a replacement glyph on malformed input;
`ChapterHtmlSlimParser` correctly preserves incomplete multi-byte
sequences across word-buffer flush
boundaries rather than splitting them.
---
### 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**_ Architecture and
design was done by me, refined a bit by Claude. Code mostly by Claude,
but not entirely.
## Summary
**What is the goal of this PR?**
### Problem
Inter-word gap widths were computed as two separately-snapped integers:
```cpp
gap = getSpaceWidth(fontId, style); // fp4::toPixel(spaceAdvance)
gap += getSpaceKernAdjust(fontId, leftCp, rightCp, style); // fp4::toPixel(kern1 + kern2)
```
Because `fp4::toPixel(a) + fp4::toPixel(b)` can differ from
`fp4::toPixel(a + b)` by +/-1 pixel when the fractional parts straddle a
rounding boundary, each inter-word space could be one pixel wider or
narrower than the correct value. This affected line-break width
decisions and word-position accumulation across the whole paragraph
layout pipeline.
### Fix
Replaces `getSpaceKernAdjust()` with `getSpaceAdvance(fontId, leftCp,
rightCp, style)`, which combines the space glyph advance and both
flanking kern values (`kern(leftCp, ' ')` + `kern(' ', rightCp)`) into a
single fixed-point sum before the snap:
```cpp
return fp4::toPixel(spaceAdvanceFP + kern(leftCp, ' ') + kern(' ', rightCp));
```
This is the same single-snap pattern already used by `getTextAdvanceX`
for word widths.
### Changes
- **`GfxRenderer`**: Replaces `getSpaceKernAdjust()` with
`getSpaceAdvance()`. `getSpaceWidth()` is retained for the
single-space-word case in `measureWordWidth` where no adjacent-word kern
context is available.
- **`ParsedText`**: All four call sites (`computeLineBreaks`,
`computeHyphenatedLineBreaks`, and both loops in `extractLine`) updated
to use `getSpaceAdvance()`. The now-redundant `spaceWidth`
pre-computation and parameter are removed from all three internal layout
functions.
---
### 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 to analyze for
correctness**_
## Summary
**What is the goal of this PR?**
Hopefully fixes#1182.
_Note: I think letterforms got a "heavier" appearance after #1098, which
makes this more noticeable. The current version of this PR reverts the
change to add `--force-autohint` for Bookerly, which to me seems to
bring the font back to a more aesthetic and consistent weight._
#### Problem
Character spacing was uneven in certain words. The word "drew" in
Bookerly was the clearest example: a visible gap between `d` and `r`,
while `e` and `w` appeared tightly condensed. The root cause was
twofold:
1. **Integer-only glyph advances.** `advanceX` was stored as a `uint8_t`
of whole pixels, sourced from FreeType's hinted `advance.x` (which
grid-fits to integers). A glyph whose true advance is 15.56px was stored
as 16px -- an error of +0.44px per character that compounds across a
line.
2. **Floor-rounded kerning.** Kern adjustments were converted with
`math.floor()`, which systematically over-tightened negative kerns. A
kern of -0.3px became -1px -- a 0.7px over-correction that visibly
closed gaps.
Combined, these produced the classic symptom: some pairs too wide,
others too tight, with the imbalance varying per word.
#### Solution: fixed-point accumulation with 1/16-pixel resolution, for
sub-pixel precision during text layout
All font metrics now use a "fixed-point 4" format -- 4 fractional bits
giving 1/16-pixel (0.0625px) resolution. This is implemented with plain
integer arithmetic (shifts and adds), requiring no floating-point on the
ESP32.
**How it works:**
A value like 15.56px is stored as the integer `249`:
```
249 = 15 * 16 + 9 (where 9/16 = 0.5625, closest to 0.56)
```
Two storage widths share the same 4 fractional bits:
| Field | Type | Format | Range | Use |
|-------|------|--------|-------|-----|
| `advanceX` | `uint16_t` | 12.4 | 0 -- 4095.9375 px | Glyph advance
width |
| `kernMatrix` | `int8_t` | 4.4 | -8.0 -- +7.9375 px | Kerning
adjustment |
Because both have 4 fractional bits, they add directly into a single
`int32_t` accumulator during layout. The accumulator is only snapped to
the nearest whole pixel at the moment each glyph is rendered:
```cpp
int32_t xFP = fp4::fromPixel(startX); // pixel to 12.4: startX << 4
for each character:
xFP += kernFP; // add 4.4 kern (sign-extends into int32_t)
int xPx = fp4::toPixel(xFP); // snap to nearest pixel: (xFP + 8) >> 4
render glyph at xPx;
xFP += glyph->advanceX; // add 12.4 advance
```
Fractional remainders carry forward indefinitely. Rounding errors stay
below +/- 0.5px and never compound.
#### Concrete example: "drew" in Bookerly
**Before** (integer advances, floor-rounded kerning):
| Char | Advance | Kern | Cursor | Snap | Gap from prev |
|------|---------|------|--------|------|---------------|
| d | 16 px | -- | 33 | 33 | -- |
| r | 12 px | 0 | 49 | 49 | ~2px |
| e | 13 px | -1 | 60 | 60 | ~0px |
| w | 22 px | -1 | 72 | 72 | ~0px |
The d-to-r gap was visibly wider than the tightly packed `rew`.
**After** (12.4 advances, 4.4 kerning, fractional accumulation):
| Char | Advance (FP) | Kern (FP) | Accumulator | Snap | Ink start | Gap
from prev |
|------|-------------|-----------|-------------|------|-----------|---------------|
| d | 249 (15.56px) | -- | 528 | 33 | 34 | -- |
| r | 184 (11.50px) | 0 | 777 | 49 | 49 | 0px |
| e | 208 (13.00px) | -8 (-0.50px) | 953 | 60 | 61 | 1px |
| w | 356 (22.25px) | -4 (-0.25px) | 1157 | 72 | 72 | 0px |
Spacing is now `0, 1, 0` pixels -- nearly uniform. Verified on-device:
all 5 copies of "drew" in the test EPUB produce identical spacing,
confirming zero accumulator drift.
#### Changes
**Font conversion (`fontconvert.py`)**
- Use `linearHoriAdvance` (FreeType 16.16, unhinted) instead of
`advance.x` (26.6, grid-fitted to integers) for glyph advances
- Encode kern values as 4.4 fixed-point with `round()` instead of
`floor()`
- Add `fp4_from_ft16_16()` and `fp4_from_design_units()` helper
functions
- Add module-level documentation of fixed-point conventions
**Font data structures (`EpdFontData.h`)**
- `EpdGlyph::advanceX`: `uint8_t` to `uint16_t` (no memory cost due to
existing struct padding)
- Add `fp4` namespace with `constexpr` helpers: `fromPixel()`,
`toPixel()`, `toFloat()`
- Document fixed-point conventions
**Font API (`EpdFont.h/cpp`, `EpdFontFamily.h/cpp`)**
- `getKerning()` return type: `int8_t` to `int` (to avoid truncation of
the 4.4 value)
**Rendering (`GfxRenderer.cpp`)**
- `drawText()`: replace integer cursor with `int32_t` fixed-point
accumulator
- `drawTextRotated90CW()`: same accumulator treatment for vertical
layout
- `getTextAdvanceX()`, `getSpaceWidth()`, `getSpaceKernAdjust()`,
`getKerning()`: convert from fixed-point to pixel at API boundary
**Regenerated all built-in font headers** with new 12.4 advances and 4.4
kern values.
#### Memory impact
Zero additional RAM. The `advanceX` field grew from `uint8_t` to
`uint16_t`, but the `EpdGlyph` struct already had 1 byte of padding at
that position, so the struct size is unchanged. The fixed-point
accumulator is a single `int32_t` on the stack.
#### Test plan
- [ ] Verify "drew" spacing in Bookerly at small, medium, and large
sizes
- [ ] Verify uppercase kerning pairs: AVERY, WAVE, VALUE
- [ ] Verify ligature words: coffee, waffle, office
- [ ] Verify all built-in fonts render correctly at each size
- [ ] Verify rotated text (progress bar percentage) renders correctly
- [ ] Verify combining marks (accented characters) still position
correctly
- [ ] Spot-check a full-length book for any layout regressions
---
### 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, Claude Opus 4.6
helped figure out a non-floating point approach for sub-pixel error
accumulation**_
## Summary
* **What is the goal of this PR?** (e.g., Implements the new feature for
file uploading.)
* **What changes are included?**
Conrgegate the changes of #1074 , #1013 , and extended upon #911 by
@lkristensen
New function implemented in GfxRenderer.cpp
```C++
std::vector<std::string> GfxRenderer::wrappedText(const int fontId, const char* text, const int maxWidth,
const int maxLines, const EpdFontFamily::Style style) const
```
Applied logic to all uses in Lyra, Lyra Extended, and base theme
(continue reading card as pointed out by @znelson
## Additional Context
---
### 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: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com>
## Summary
**What is the goal of this PR?**
Improved typesetting, including
[kerning](https://en.wikipedia.org/wiki/Kerning) and
[ligatures](https://en.wikipedia.org/wiki/Ligature_(writing)#Latin_alphabet).
**What changes are included?**
- The script to convert built-in fonts now adds kerning and ligature
information to the generated font headers.
- Epub page layout calculates proper kerning spaces and makes ligature
substitutions according to the selected font.
## Additional Context
- I am not a typography expert.
- The implementation has been reworked from the earlier version, so it
is no longer necessary to omit Open Dyslexic, and kerning data now
covers all fonts, styles, and codepoints for which we include bitmap
data.
- Claude Opus 4.6 helped with a lot of this.
- There's an included test epub document with lots of kerning and
ligature examples, shown in the photos.
**_After some time to mature, I think this change is in decent shape to
merge and get people testing._**
After opening this PR I came across #660, which overlaps in adding
ligature support.
---
### 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, Claude Opus 4.6**_
---------
Co-authored-by: Cursor <cursoragent@cursor.com>
## Summary
**What is the goal of this PR?**
Consolidated repeated logic to fall back to REPLACEMENT_GLYPH.
---
### 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? _**NO**_
## Summary
* This PR fixes decomposed diacritic handling end-to-end:
- Hyphenation: normalize common Latin base+combining sequences to
precomposed codepoints before Liang pattern matching, so decomposed
words hyphenate correctly
- Rendering: correct combining-mark placement logic so non-spacing marks
are attached to the preceding base glyph in normal and rotated text
rendering paths, with corresponding text-bounds consistency updates.
- Hyphenation around non breaking space variants have been fixed (and
extended)
- Hyphenation of terms that already included of hyphens were fixed to
include Liang pattern application (eg "US-Satellitensystem" was
*exclusively* broken at the existing hyphen)
## Additional Context
* Before
<img width="800" height="480" alt="2"
src="https://github.com/user-attachments/assets/b9c515c4-ab75-45cc-8b52-f4d86bce519d"
/>
* After
<img width="480" height="800" alt="fix1"
src="https://github.com/user-attachments/assets/4999f6a8-f51c-4c0a-b144-f153f77ddb57"
/>
<img width="800" height="480" alt="fix2"
src="https://github.com/user-attachments/assets/7355126b-80c7-441f-b390-4e0897ee3fb6"
/>
* Note 1: the hyphenation fix is not a 100% bullet proof implementation.
It adds composition of *common* base+combining sequences (e.g. O +
U+0308 -> Ö) during codepoint collection. A complete solution would
require implementing proper Unicode normalization (at least NFC,
possibly NFKC in specific cases) before hyphenation and rendering,
instead of hand-mapping a few combining marks. That was beyond the scope
of this fix.
* Note 2: the render fix should be universal and not limited to the
constraints outlined above: it properly x-centers the compund glyph over
the previous one, and it uses at least 1pt of visual distance in y.
Before:
<img width="478" height="167" alt="Image"
src="https://github.com/user-attachments/assets/f8db60d5-35b1-4477-96d0-5003b4e4a2a1"
/>
After:
<img width="479" height="180" alt="Image"
src="https://github.com/user-attachments/assets/1b48ef97-3a77-475a-8522-23f4aca8e904"
/>
* This should resolve the issues described in #998
---
### 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**_
## Summary
* ``renderChar`` checked ``is2Bit`` on every pixel inside the inner
loop, even though the value is constant for the lifetime of a single
glyph
* Moved the branch above both loops so each path (2-bit antialiased /
1-bit monochrome) runs without a per-pixel conditional
* Eliminates redundant work in the two inner loops that render font
glyphs to the frame buffer, targeting ``renderChar`` and
``drawTextRotated90CW`` in ``GfxRenderer.cpp``
## Additional Context
* Measured on device using a dedicated framebuffer benchmark (no display
refresh). 100 repetitions of "The quick brown fox jumps".
| Test | Before | After | Change |
|-----------------|-----------------|-----------------|---------|
| drawText UI12 | 1,337 µs/call | 1,024 µs/call | −23%|
| drawText Bookerly14 | 2.174 µs / call | 1,847 µs/call | −15% |
---
### 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**_ Claude did
the analysis and wrote the benchmarks
## Summary
* **What is the goal of this PR?**
I flashed the last revision before commit 76f1c921, and chapter indexing
worked without any crashes.
After applying 76f1c921, the same chapter consistently triggered a
device reboot during indexing.
The affected chapter contains inline equation images surrounded by
styled (bold/italic) text that includes special math/symbol characters.
## Additional Context
Prior to 76f1c921, both `getTextAdvanceX()` and `getSpaceWidth()` always
measured text using `EpdFontFamily::REGULAR`, regardless of the actual
style.
Commit 76f1c921 improved correctness by passing the active style so
spacing is calculated using the actual bold/italic font variant.
However, bold and italic variants have narrower Unicode coverage than
the regular font. When a character exists in the regular font but not in
the selected styled variant, `pdFont::getGlyph()` returns `nullptr`.
The updated measurement functions did not check for this and immediately
dereferenced the pointer:
`width += font.getGlyph(cp, style)->advanceX; // nullptr->advanceX`
Because `advanceX` is located at byte offset 2 within `EpdGlyph`,
dereferencing a null pointer caused the CPU to attempt a load from
address `0x00000002`, resulting in a RISC-V:
Load access fault
MCAUSE = 5
MTVAL = 2
## Fix
Added null-safety checks to both `getTextAdvanceX()` and
`getSpaceWidth()`, following the same pattern used in the rendering
path:
If the glyph is missing in the selected style → fall back to the
replacement glyph.
If the replacement glyph is also unavailable → treat the character as
zero-width.
This preserves the improved style-correct spacing while preventing
crashes.
No behavioral changes occur for characters that are supported by the
selected font variant.
---
### AI Usage
Did you use AI tools to help write this code? _**< YES >**_
I encounter this bug while testing 1.1.0 RC.
I pasted the serial log to Claude, which identify the bug and fixed it.
I can confirm now the chapter in question is indexed and loaded
correctly.
## Summary
* What is the goal of this PR?
- Allow users to create custom sleep screen images with standard tools
(ImageMagick, GIMP, etc.) that render cleanly on the e-ink display
without dithering artifacts. Previously, avoiding dithering required
non-standard 2-bit BMPs that no standard image editor can produce. ( see
issue #931 )
* What changes are included?
- Add 4-bit BMP format support to Bitmap.cpp (standard format, widely
supported by image tools)
- Auto-detect "native palette" images: if a BMP has ≤4 palette entries
and all luminances map within ±21 of the display's native gray levels
(0, 85, 170, 255), skip dithering entirely and direct-map pixels
- Clarify pixel processing strategy with three distinct paths:
error-diffusion dithering, simple quantization, or direct mapping
- Add scripts/generate_test_bmps.py for generating test images across
all supported BMP formats
## Additional Context
* The e-ink display has 4 native gray levels. When a BMP already uses
exactly those levels, dithering adds noise to what should be clean
output. The native palette detection uses a ±21 tolerance (~10%) to
handle slight rounding from color space conversions in image tools.
Users can now create a 4-color grayscale BMP with (imagemagic example):
```
convert input.png -colorspace Gray -colors 4 -depth
```
---
### 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**_
## Summary
**What is the goal of this PR?**
`hasPrintableChars` does a pass over text before rendering. It looks up
glyphs in the font and measures dimensions, returning early if the text
results in zero size.
This additional pass doesn't offer any benefit over moving straight to
rendering the text, because the rendering loop already gracefully
handles missing glyphs. This change saves an extra pass over all
rendered text.
Note that both `hasPrintableChars` and `renderChar` replace missing
glyphs with `glyph = getGlyph(REPLACEMENT_GLYPH)`, so there's no
difference for characters which are not present in the font.
---
### 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? _**NO**_
## Summary
* GfxRender did handle horizontal and vertical lines but had a TODO for
arbitrary lines.
* Added integer based Bresenham line drawing
## Additional Context
---
### 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? _**NO**_
## Summary
**What is the goal of this PR?**
This change fixes an issue I noticed while reading where occasionally,
especially in italics, some words would have too much space between
them. The problem was that word width calculations were including any
negative X overhang, and combined with a space before the word, that can
lead to an inconsistently large space.
## Additional Context
Screenshots of some problematic text:
| In CrossPoint 1.0 | With this change |
| -- | -- |
| <img
src="https://github.com/user-attachments/assets/87bf0e4b-341f-4ba9-b3ea-38c13bd26363"
width="400" /> | <img
src="https://github.com/user-attachments/assets/bf11ba20-c297-4ce1-aa07-43477ef86fc2"
width="400" /> |
---
### 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? _**NO**_
## Summary
**What is the goal of this PR?**
Compress reader font bitmaps to reduce flash usage by 30.7%.
**What changes are included?**
- New `EpdFontGroup` struct and extended `EpdFontData` with
`groups`/`groupCount` fields
- `--compress` flag in `fontconvert.py`: groups glyphs (ASCII base group
+ groups of 8) and compresses each with raw DEFLATE
- `FontDecompressor` class with 4-slot LRU cache for on-demand
decompression during rendering
- `GfxRenderer` transparently routes bitmap access through
`getGlyphBitmap()` (compressed or direct flash)
- Uses `uzlib` for decompression with minimal heap overhead.
- 48 reader fonts (Bookerly, NotoSans 12-18pt, OpenDyslexic) regenerated
with compression; 5 UI fonts unchanged
- Round-trip verification script (`verify_compression.py`) runs as part
of font generation
## Additional Context
## Flash & RAM
| | baseline | font-compression | Difference |
|--|--------|-----------------|------------|
| Flash (ELF) | 6,302,476 B (96.2%) | 4,365,022 B (66.6%) | -1,937,454 B
(-30.7%) |
| firmware.bin | 6,468,192 B | 4,531,008 B | -1,937,184 B (-29.9%) |
| RAM | 101,700 B (31.0%) | 103,076 B (31.5%) | +1,376 B (+0.5%) |
## Script-Based Grouping (Cold Cache)
Comparison of uncompressed baseline vs script-based group compression
(4-slot LRU cache, cleared each page). Glyphs are grouped by Unicode
block (ASCII, Latin-1, Latin Extended-A, Combining Marks, Cyrillic,
General Punctuation, etc.) instead of sequential groups of 8.
### Render Time
| | Baseline | Compressed (cold cache) | Difference |
|---|---|---|---|
| **Median** | 414.9 ms | 431.6 ms | +16.7 ms (+4.0%) |
| **Pages** | 37 | 37 | |
### Memory Usage
| | Baseline | Compressed (cold cache) | Difference |
|---|---|---|---|
| **Heap free (median)** | 187.0 KB | 176.3 KB | -10.7 KB |
| **Heap free (min)** | 186.0 KB | 166.5 KB | -19.5 KB |
| **Largest block (median)** | 148.0 KB | 128.0 KB | -20.0 KB |
| **Largest block (min)** | 148.0 KB | 120.0 KB | -28.0 KB |
### Cache Effectiveness
| | Misses/page | Hit rate |
|---|---|---|
| **Compressed (cold cache)** | 2.1 | 99.85% |
------
### 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**_
Implementation was done by Claude Code (Opus 4.6) based on a plan
developed collaboratively. All generated font headers were verified with
an automated round-trip decompression test. The firmware was compiled
successfully but has not yet been tested on-device.
---------
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
## Summary
**What is the goal of this PR?**
Several methods in GfxRenderer were doing a `count()` followed by `at()`
on the fonts map, effectively doing the same map lookup unnecessarily.
This can be avoided by doing a single `find()` and reusing the iterator.
---
### 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? _**NO**_
## Summary
- Add embedded image support to EPUB rendering with JPEG and PNG
decoders
- Implement pixel caching system to cache decoded/dithered images to SD
card for faster re-rendering
- Add 4-level grayscale support for display
## Changes
### New Image Rendering System
- Add `ImageBlock` class to represent an image with its cached path and
display dimensions
- Add `PageImage` class as a new `PageElement` type for images on pages
- Add `ImageToFramebufferDecoder` interface for format-specific image
decoders
- Add `JpegToFramebufferConverter` - JPEG decoder with Bayer dithering
and scaling
- Add `PngToFramebufferConverter` - PNG decoder with Bayer dithering and
scaling
- Add `ImageDecoderFactory` to select appropriate decoder based on file
extension
- Add `getRenderMode()` to GfxRenderer for grayscale render mode queries
### Dithering and Grayscale
- Implement 4x4 Bayer ordered dithering for 4-level grayscale output
- Stateless algorithm works correctly with MCU block decoding
- Handles scaling without artifacts
- Add grayscale render mode support (BW, GRAYSCALE_LSB, GRAYSCALE_MSB)
- Image decoders and cache renderer respect current render mode
- Enables proper 4-level e-ink grayscale when anti-aliasing is enabled
### Pixel Caching
- Cache decoded/dithered images to `.pxc` files on SD card
- Cache format: 2-bit packed pixels (4 pixels per byte) with
width/height header
- On subsequent renders, load directly from cache instead of re-decoding
- Cache renderer supports grayscale render modes for multi-pass
rendering
- Significantly improves page navigation speed for image-heavy EPUBs
### HTML Parser Integration
- Update `ChapterHtmlSlimParser` to process `<img>` tags and extract
images from EPUB
- Resolve relative image paths within EPUB ZIP structure
- Extract images to cache directory before decoding
- Create `PageImage` elements with proper scaling to fit viewport
- Fall back to alt text display if image processing fails
### Build Configuration
- Add `PNG_MAX_BUFFERED_PIXELS=6402` to support up to 800px wide images
### Test Script
- Generate test EPUBs with annotated JPEG and PNG images
- Test cases cover: grayscale (4 levels), centering, scaling, cache
performance
## Test plan
- [x] Open EPUB with JPEG images - verify images display with proper
grayscale
- [x] Open EPUB with PNG images - verify images display correctly and no
crash
- [x] Navigate away from image page and back - verify faster load from
cache
- [x] Verify grayscale tones render correctly (not just black/white
dithering)
- [x] Verify large images are scaled down to fit screen
- [x] Verify images are centered horizontally
- [x] Verify page serialization/deserialization works with images
- [x] Verify images rendered in landscape mode
## Test Results
[png](https://photos.app.goo.gl/5zFUb8xA8db3dPd19)
[jpeg](https://photos.app.goo.gl/SwtwaL2DSQwKybhw7)
---
### AI Usage
Did you use AI tools to help write this code? _**< YES >**_
---------
Co-authored-by: Matthías Páll Gissurarson <mpg@mpg.is>
Co-authored-by: Dave Allie <dave@daveallie.com>
## Summary
* Definition and use of a central LOG function, that can later be
extended or completely be removed (for public use where debugging
information may not be required) to save flash by suppressing the
-DENABLE_SERIAL_LOG like in the slim branch
* **What changes are included?**
## Additional Context
* By using the central logger the usual:
```
#include <HardwareSerial.h>
...
Serial.printf("[%lu] [WCS] Obfuscating/deobfuscating %zu bytes\n", millis(), data.size());
```
would then become
```
#include <Logging.h>
...
LOG_DBG("WCS", "Obfuscating/deobfuscating %zu bytes", data.size());
```
You do have ``LOG_DBG`` for debug messages, ``LOG_ERR`` for error
messages and ``LOG_INF`` for informational messages. Depending on the
verbosity level defined (see below) soe of these message types will be
suppressed/not-compiled.
* The normal compilation (default) will create a firmware.elf file of
42.194.356 bytes, the same code via slim will create 42.024.048 bytes -
170.308 bytes less
* Firmware.bin : 6.469.984 bytes for default, 6.418.672 bytes for slim -
51.312 bytes less
### 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? _NO_
---------
Co-authored-by: Xuan Son Nguyen <son@huggingface.co>
## Summary
Continue my changes to introduce the HAL infrastructure from
https://github.com/crosspoint-reader/crosspoint-reader/pull/522
This PR touches quite a lot of files, but most of them are just name
changing. It should not have any impacts to the end behavior.
## Additional Context
My plan is to firstly add this small shim layer, which sounds useless at
first, but then I'll implement an emulated driver which can be helpful
for testing and for development.
Currently, on my fork, I'm using a FS driver that allow "mounting" a
local directory from my computer to the device, much like the `-v` mount
option on docker. This allows me to quickly reset `.crosspoint`
directory if anything goes wrong. I plan to upstream this feature when
this PR get merged.
---
### 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? NO
## Summary
Ref https://github.com/crosspoint-reader/crosspoint-reader/pull/737
This PR further reduce ~25ms from rendering time, testing inside the
Setting screen:
```
master:
[68440] [GFX] Time = 73 ms from clearScreen to displayBuffer
PR:
[97806] [GFX] Time = 47 ms from clearScreen to displayBuffer
```
And in extreme case (fill the entire screen with black or gray color):
```
master:
[1125] [ ] Test fillRectDither drawn in 327 ms
[1347] [ ] Test fillRect drawn in 222 ms
PR:
[1334] [ ] Test fillRectDither drawn in 225 ms
[1455] [ ] Test fillRect drawn in 121 ms
```
Note that
https://github.com/crosspoint-reader/crosspoint-reader/pull/737 is NOT
applied on top of this PR. But with 2 of them combined, it should reduce
from 47ms --> 42ms
## Details
This PR based on the fact that function calls are costly if the function
is small enough. For example, this simple call:
```
int rotatedX = 0;
int rotatedY = 0;
rotateCoordinates(x, y, &rotatedX, &rotatedY);
```
Generated assembly code:
<img width="771" height="215" alt="image"
src="https://github.com/user-attachments/assets/37991659-3304-41c3-a3b2-fb967da53f82"
/>
This adds ~10 instructions just to prepare the registers prior to the
function call, plus some more instructions for the function's
epilogue/prologue. Inlining it removing all of these:
<img width="1471" height="832" alt="image"
src="https://github.com/user-attachments/assets/b67a22ee-93ba-4017-88ed-c973e28ec914"
/>
Of course, this optimization is not magic. It's only beneficial under 3
conditions:
- The function is small, not in size, but in terms of effective
instructions. For example, the `rotateCoordinates` is simply a jump
table, where each branch is just 3-4 inst
- The function has multiple input arguments, which requires some move to
put it onto the correct place
- The function is called very frequently (i.e. critical path)
---
### 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? **NO**
## Summary
This PR optimizes the `fillRectDither` function, making it as fast as a
normal `fillRect`
Testing code:
```cpp
{
auto start_t = millis();
renderer.fillRectDither(0, 0, renderer.getScreenWidth(), renderer.getScreenHeight(), Color::LightGray);
auto elapsed = millis() - start_t;
Serial.printf("[%lu] [ ] Test fillRectDither drawn in %lu ms\n", millis(), elapsed);
}
{
auto start_t = millis();
renderer.fillRect(0, 0, renderer.getScreenWidth(), renderer.getScreenHeight(), true);
auto elapsed = millis() - start_t;
Serial.printf("[%lu] [ ] Test fillRect drawn in %lu ms\n", millis(), elapsed);
}
```
Before:
```
[1125] [ ] Test fillRectDither drawn in 327 ms
[1347] [ ] Test fillRect drawn in 222 ms
```
After:
```
[1065] [ ] Test fillRectDither drawn in 238 ms
[1287] [ ] Test fillRect drawn in 222 ms
```
## Visual validation
Before:
<img width="415" height="216" alt="Screenshot 2026-02-07 at 01 04 19"
src="https://github.com/user-attachments/assets/5802dbba-187b-4d2b-a359-1318d3932d38"
/>
After:
<img width="420" height="191" alt="Screenshot 2026-02-07 at 01 36 30"
src="https://github.com/user-attachments/assets/3c3c8e14-3f3a-4205-be78-6ed771dcddf4"
/>
## Details
The original version is quite slow because it does quite a lot of
computations. A single pixel needs around 20 instructions just to know
if it's black or white:
<img width="1170" height="693" alt="Screenshot 2026-02-07 at 00 15 54"
src="https://github.com/user-attachments/assets/7c5a55e7-0598-4340-8b7b-17307d7921cb"
/>
With the new, templated and more light-weight approach, each pixel takes
only 3-4 instructions, the modulo operator is translated into bitwise
ops:
<img width="1175" height="682" alt="Screenshot 2026-02-07 at 01 47 51"
src="https://github.com/user-attachments/assets/4ec2cf74-6cc0-4b5b-87d5-831563ef164f"
/>
---
### 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? **NO**
## Summary
* **What is the goal of this PR?**
The goal of this PR is to deliver a fix for or at least mitigate the
impact of the issue described in #561
* **What changes are included?**
This PR includes a new option "Sunlight Fading Fix" under "Settings ->
Display".
When set to ON, we will disable the displays analog supply voltage after
every update and turn it back on before the next update.
## Additional Context
* Until now, I was only able to do limited testing because of limited
sunlight at my location, but the fix seems to be working. I'll also
attach a pre-built binary based on 0.16.0 (current master) with the fix
applied to the linked ticket, as building this fix is a bit annoying
because the submodule open-x4-sdk also needs an update.
* [PR in
open-x4-sdk](https://github.com/open-x4-epaper/community-sdk/pull/15)
needs to be merged first, we also need to add another commit to this
here PR, updating this dependency.
* I decided to hide this behind a default-OFF option. While I'm not
really concerned that this fix might potentially damage the display,
someone more knowledgeable on E-Ink technology could maybe have a look
at this.
* There's a binary attached in the linked issue, if someone has the
required sunlight to test this in-depth.
---
### 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? _**NO**_
---------
Co-authored-by: Dave Allie <dave@daveallie.com>
## Summary
### What is the goal of this PR?
- Visual UI overhaul
- UI theme selection
### What changes are included?
- Added a setting "UI Theme": Classic, Lyra
- The classic theme is the current Crosspoint theme
- The Lyra theme implements these mockups:
https://www.figma.com/design/UhxoV4DgUnfrDQgMPPTXog/Lyra-Theme?node-id=2003-7596&t=4CSOZqf0n9uQMxDt-0
by Discord users yagofarias, ruby and gan_shu
- New functions in GFXRenderer to render rounded rectangles, greyscale
fills (using dithering) and thick lines
- Basic UI components are factored into BaseTheme methods which can be
overridden by each additional theme. Methods that are not overridden
will fallback to BaseTheme behavior. This means any new
features/components in CrossPoint only need to be developed for the
"Classic" BaseTheme.
- Additional themes can easily be developed by the community using this
foundation
## Additional Context
- Only the Home, Library and main Settings screens have been implemented
so far, this will be extended to the transfer screens and chapter
selection screen later on, but we need to get the ball rolling somehow
:)
- Loading extra covers on the home screen in the Lyra theme takes a
little more time (about 2 seconds), I added a loading bar popup (reusing
the Indexing progress bar from the reader view, factored into a neat UI
component) but the popup adds ~400ms to the loading time.
- ~~Home screen thumbnails will need to be generated separately for each
theme, because they are displayed in different sizes. Because we're
using dithering, displaying a thumb with the wrong size causes the
picture to look janky or dark as it does on the screenshots above. No
worries this will be fixed in a future PR.~~ Thumbs are now generated
with a size parameter
- UI Icons will need to be implemented in a future PR.
---
### 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**_
This is not a vibe coded PR. Copilot was used for autocompletion to save
time but I reviewed, understood and edited all generated code.
---------
Co-authored-by: Dave Allie <dave@daveallie.com>
## Summary
* **What is the goal of this PR?**
- Adds basic CSS parsing to EPUBs and determine the CSS rules when
rendering to the screen so that text is styled correctly. Currently
supports bold, underline, italics, margin, padding, and text alignment
## Additional Context
- My main reason for wanting this is that the book I'm currently
reading, Carl's Doomsday Scenario (2nd in the Dungeon Crawler Carl
series), relies _a lot_ on styled text for telling parts of the story.
When text is bolded, it's supposed to be a message that's rendered
"on-screen" in the story. When characters are "chatting" with each
other, the text is bolded and their names are underlined. Plus, normal
emphasis is provided with italicizing words here and there. So, this
greatly improves my experience reading this book on the Xteink, and I
figured it was useful enough for others too.
- For transparency: I'm a software engineer, but I'm mostly frontend and
TypeScript/JavaScript. It's been _years_ since I did any C/C++, so I
would not be surprised if I'm doing something dumb along the way in this
code. Please don't hesitate to ask for changes if something looks off. I
heavily relied on Claude Code for help, and I had a lot of inspiration
from how [microreader](https://github.com/CidVonHighwind/microreader)
achieves their CSS parsing and styling. I did give this as good of a
code review as I could and went through everything, and _it works on my
machine_ 😄
### Before
### After
---
### AI Usage
Did you use AI tools to help write this code? **YES**, Claude Code
jpirnay
