Webserver:
- Remove MD5 hash computation from file listings (caused EAGAIN errors)
- Implement JSON batching (2KB) with pacing for file listings
- Simplify sendContentSafe() flow control
Memory:
- Cache QR codes on server start instead of regenerating per render
- Optimize WiFi scan: vector deduplication, 20 network limit, early scanDelete()
- Fix 48KB cover buffer leak when navigating Home -> File Transfer
EPUB Reader:
- Fix errant underlining by flushing partWordBuffer before style changes
- Text before styled inline elements (e.g., <a> with CSS underline) no longer
incorrectly receives the element's styling
Flash Screen:
- Fix version string buffer overflow (30 -> 50 char limit)
- Use half refresh for cleaner display
- Adjust pre_flash.py timing for half refresh completion
EPUB rendering improvements:
- Add margin-left/padding-left CSS parsing for block indentation
- Add vertical bar and italic styling for blockquotes
- Add left margin indentation for list items (ol/ul)
- Fix ordered lists showing bullets instead of numbers
- Fix nested <p> inside <li> causing marker on separate line
End-of-book improvements:
- Add "Start Over" option to wrap to first page when pressing next
- Show "Start Over" button hint on finished book prompt
This commit completes a series of fixes addressing dictionary crashes,
memory issues, and UI/UX improvements.
Memory & Stability (from previous checkpoints):
- Add uncompressed dictionary (.dict) support to avoid decompression
memory issues with large dictzip chunks (58KB -> direct read)
- Implement chunked on-demand HTML parsing for large definitions,
parsing pages as user navigates rather than all at once
- Refactor TextBlock/ParsedText from std::list to std::vector,
reducing heap allocations by ~12x per TextBlock and eliminating
crashes from repeated page navigation due to heap fragmentation
- Limit cached pages to MAX_CACHED_PAGES (4) with re-parse capability
for backward navigation beyond the cache window
UI/Layout Fixes (this commit):
- Restore DictionaryMargins.h for proper orientation-aware button
hint space (front buttons: 45px, side buttons: 50px)
- Add side button hints to definition screen with proper "<" / ">"
labels for page navigation
- Add side button hints to word selection screen ("UP"/"DOWN" labels,
borderless, small font, 2px edge margin)
- Add side button hints to dictionary menu ("< Prev", "Next >")
- Fix double-button press bug when loading new chunks by checking
forward navigation availability after parsing instead of page count
- Add drawSideButtonHints() drawBorder parameter for minimal hints
- Add drawTextRotated90CCW() for LandscapeCCW text orientation
- Move page indicator up to avoid bezel cutoff
Reduces heap fragmentation by ~12x fewer allocations per TextBlock.
This fixes crashes when repeatedly navigating dictionary pages.
- Replace std::list with std::vector in TextBlock members
- Replace splice() with move+erase in ParsedText::extractLine()
- Use index-based access in hyphenateWordAtIndex()
On ESP32-C3 with USB CDC, Serial.printf() blocks indefinitely when USB
is not connected. This caused device freezes when booted without USB.
Solution: Call Serial.setTxTimeoutMs(0) after Serial.begin() to make
all Serial output non-blocking.
Also added if (Serial) guards to high-traffic logging paths in
EpubReaderActivity as belt-and-suspenders protection.
Includes documentation of the debugging process and Serial call inventory.
Also applies clang-format to fix pre-existing formatting issues.
- Add global high contrast mode flag in BitmapHelpers
- Modify quantization thresholds for high contrast rendering
- Update ditherer classes (Atkinson, Floyd-Steinberg) for contrast mode
- Add displayContrast setting with persistence
- Add "High Contrast" toggle in display settings
- Apply high contrast mode on startup from saved settings
Manual implementation of upstream PR #525
Added flushPartWordBuffer() function and call it before <br/> handling
to fix issue where preceding word was incorrectly wrapped to new line
Add ZipFile::fillUncompressedSizes() for single-pass ZIP central directory
scan with hash-based target matching.
Also apply clang-format fixes for CI.
Shadow Slave results:
- buildBookBin: 506s → 35s
- Total indexing: 8.7min → 50s
Three optimizations for EPUBs with many chapters (e.g. 2768 chapters):
1. OPF idref→href lookup: Build sorted hash index during manifest parsing,
use binary search during spine resolution. Reduces ~4min to ~30-60s.
2. TOC href→spineIndex lookup: Build sorted hash index in beginTocPass(),
use binary search in createTocEntry(). Reduces ~4min to ~30-60s.
3. ZIP central-dir cursor: Resume scanning from last position instead of
restarting from beginning. Reduces ~8min to ~1-3min.
All optimizations only activate for large EPUBs (≥400 spine items).
Small books use unchanged code paths.
Memory impact: ~33KB + ~39KB temporary during indexing, freed after.
Expected total: ~17min → ~3-5min for Shadow Slave (2768 chapters).
Also adds phase timing logs for performance measurement.
The unordered_map with 2768 string keys (~100KB+) was causing OOM crashes
at beginTocPass() on ESP32-C3's limited ~380KB RAM.
Reverted createTocEntry() to use original O(n) spine file scan instead.
Kept the safe spineToTocIndex vector in buildBookBin() (only ~5.5KB).
Replace O(n²) lookups with O(n) preprocessing:
1. createTocEntry(): Build href->spineIndex map once in beginTocPass()
instead of scanning spine file for every TOC entry
2. buildBookBin(): Build spineIndex->tocIndex vector in single pass
instead of scanning TOC file for every spine entry
For 2768-chapter EPUBs, this reduces:
- TOC pass: from ~7.6M file reads to ~5.5K reads
- buildBookBin: from ~7.6M file reads to ~5.5K reads
Memory impact: ~80KB for href map (acceptable trade-off for 10x+ speedup)
Remove the call to loadAllFileStatSlims() which pre-loads all ZIP central
directory entries into memory. For EPUBs with 2000+ chapters (like webnovels),
this exhausts the ESP32-C3's ~380KB RAM and causes abort().
The existing loadFileStatSlim() function already handles individual lookups
by scanning the central directory per-file when the cache is empty. This is
O(n*m) instead of O(n), but prevents memory exhaustion.
Fixes#134
## Summary
- Rewrite OpdsParser to stream parsing instead of full content
- Fix OOM due to big http xml response
Closes#385
---
### 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
* Include superscripts and subscripts in fonts
## Additional Context
* Original change came from
https://github.com/crosspoint-reader/crosspoint-reader/pull/248
---
### 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: cor <cor@pruijs.dev>
* origin:
fix: truncate chapter names that are too long (#422)
feat: dict based Hyphenation (#305)
fix: render U+FFFD replacement character instead of ? (#366)
fix: Invert colors on home screen cover overlay when recent book is selected (#390)
Adds KOReader Sync support (#232)
feat: Change keyboard "caps" to "shift" & Wrap Keyboard (#377)
fix: XTC 1-bit thumb BMP polarity inversion (#373)
## Summary
* Adds (optional) Hyphenation for English, French, German, Russian
languages
## Additional Context
* Included hyphenation dictionaries add approximately 280kb to the flash
usage (German alone takes 200kb)
* Trie encoded dictionaries are adopted from hypher project
(https://github.com/typst/hypher)
* Soft hyphens (and other explicit hyphens) take precedence over
dict-based hyphenation. Overall, the hyphenation rules are quite
aggressive, as I believe it makes more sense on our smaller screen.
---------
Co-authored-by: Dave Allie <dave@daveallie.com>
The current behavior of rendering `?` for an unknown Unicode character
can be hard to distinguish from a typo. Use the standard Unicode
"replacement character" instead, that's what it's designed for:
https://en.wikipedia.org/wiki/Specials_(Unicode_block)
I'm making this PR as a draft because I'm not sure I did everything that
was needed to change the character set covered by the fonts. Running
that script is in its own commit. If this is proper, I'll rebase/squash
into one commit and un-draft.
Co-authored-by: Maeve Andrews <maeve@git.mail.maeveandrews.com>
## Summary
- Adds KOReader progress sync integration, allowing CrossPoint to sync
reading positions with other
KOReader-compatible devices
- Stores credentials securely with XOR obfuscation
- Uses KOReader's partial MD5 document hashing for cross-device book
matching
- Syncs position via percentage with estimated XPath for compatibility
# Features
- Settings: KOReader Username, Password, and Authenticate options
- Sync from chapters menu: "Sync Progress" option appears when
credentials are configured
- Bidirectional sync: Can apply remote progress or upload local progress
---------
Co-authored-by: Dave Allie <dave@daveallie.com>
## Summary
* **What is the goal of this PR?** (e.g., Implements the new feature for
file uploading.)
* **What changes are included?**
- Fix inverted colors in Continue Reading cover image for 1-bit XTC
files
## Additional Context
* Add any other information that might be helpful for the reviewer
(e.g., performance implications, potential risks,
specific areas to focus on).
- Fix `grayValue = pixelBit ? 0 : 255` → `grayValue = pixelBit ? 255 :
0` in `lib/Xtc/Xtc.cpp`
- The thumb BMP generation had inverted polarity compared to cover BMP
generation
- bit=0 should be black, bit=1 should be white (matching the BMP palette
order)
- Update misleading comment about XTC polarity
---
### 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**_
