cottongin/crosspoint-reader
Archived
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 4 Wiki Activity

granular position tracking

Browse Source
This commit is contained in:
cottongin
2026-01-25 00:24:54 -05:00
parent fedc14bcb4
commit 91c8cc67ce
9 changed files with 345 additions and 61 deletions
Download Patch File Download Diff File Expand all files Collapse all files

79
lib/Epub/Epub/parsers/ChapterHtmlSlimParser.cpp
View File

@@ -332,6 +332,11 @@ void XMLCALL ChapterHtmlSlimParser::characterData(void* userData, const XML_Char
if (self->skipUntilDepth < self->depth) {
return;
}
// Capture byte offset of this character data for page position tracking
if (self->xmlParser) {
self->lastCharDataOffset = XML_GetCurrentByteIndex(self->xmlParser);
}
// Determine font style from depth-based tracking and CSS effective style
const bool isBold = self->boldUntilDepth < self->depth || self->effectiveBold;
@@ -477,17 +482,18 @@ void XMLCALL ChapterHtmlSlimParser::endElement(void* userData, const XML_Char* n
bool ChapterHtmlSlimParser::parseAndBuildPages() {
startNewTextBlock((TextBlock::Style)this->paragraphAlignment);
const XML_Parser parser = XML_ParserCreate(nullptr);
xmlParser = XML_ParserCreate(nullptr);
int done;
if (!parser) {
if (!xmlParser) {
Serial.printf("[%lu] [EHP] Couldn't allocate memory for parser\n", millis());
return false;
}
FsFile file;
if (!SdMan.openFileForRead("EHP", filepath, file)) {
XML_ParserFree(parser);
XML_ParserFree(xmlParser);
xmlParser = nullptr;
return false;
}
@@ -495,19 +501,24 @@ bool ChapterHtmlSlimParser::parseAndBuildPages() {
const size_t totalSize = file.size();
size_t bytesRead = 0;
int lastProgress = -1;
// Initialize offset tracking - first page starts at offset 0
currentPageStartOffset = 0;
lastCharDataOffset = 0;
XML_SetUserData(parser, this);
XML_SetElementHandler(parser, startElement, endElement);
XML_SetCharacterDataHandler(parser, characterData);
XML_SetUserData(xmlParser, this);
XML_SetElementHandler(xmlParser, startElement, endElement);
XML_SetCharacterDataHandler(xmlParser, characterData);
do {
void* const buf = XML_GetBuffer(parser, 1024);
void* const buf = XML_GetBuffer(xmlParser, 1024);
if (!buf) {
Serial.printf("[%lu] [EHP] Couldn't allocate memory for buffer\n", millis());
XML_StopParser(parser, XML_FALSE); // Stop any pending processing
XML_SetElementHandler(parser, nullptr, nullptr); // Clear callbacks
XML_SetCharacterDataHandler(parser, nullptr);
XML_ParserFree(parser);
XML_StopParser(xmlParser, XML_FALSE); // Stop any pending processing
XML_SetElementHandler(xmlParser, nullptr, nullptr); // Clear callbacks
XML_SetCharacterDataHandler(xmlParser, nullptr);
XML_ParserFree(xmlParser);
xmlParser = nullptr;
file.close();
return false;
}
@@ -516,10 +527,11 @@ bool ChapterHtmlSlimParser::parseAndBuildPages() {
if (len == 0 && file.available() > 0) {
Serial.printf("[%lu] [EHP] File read error\n", millis());
XML_StopParser(parser, XML_FALSE); // Stop any pending processing
XML_SetElementHandler(parser, nullptr, nullptr); // Clear callbacks
XML_SetCharacterDataHandler(parser, nullptr);
XML_ParserFree(parser);
XML_StopParser(xmlParser, XML_FALSE); // Stop any pending processing
XML_SetElementHandler(xmlParser, nullptr, nullptr); // Clear callbacks
XML_SetCharacterDataHandler(xmlParser, nullptr);
XML_ParserFree(xmlParser);
xmlParser = nullptr;
file.close();
return false;
}
@@ -537,27 +549,33 @@ bool ChapterHtmlSlimParser::parseAndBuildPages() {
done = file.available() == 0;
if (XML_ParseBuffer(parser, static_cast<int>(len), done) == XML_STATUS_ERROR) {
Serial.printf("[%lu] [EHP] Parse error at line %lu:\n%s\n", millis(), XML_GetCurrentLineNumber(parser),
XML_ErrorString(XML_GetErrorCode(parser)));
XML_StopParser(parser, XML_FALSE); // Stop any pending processing
XML_SetElementHandler(parser, nullptr, nullptr); // Clear callbacks
XML_SetCharacterDataHandler(parser, nullptr);
XML_ParserFree(parser);
if (XML_ParseBuffer(xmlParser, static_cast<int>(len), done) == XML_STATUS_ERROR) {
Serial.printf("[%lu] [EHP] Parse error at line %lu:\n%s\n", millis(), XML_GetCurrentLineNumber(xmlParser),
XML_ErrorString(XML_GetErrorCode(xmlParser)));
XML_StopParser(xmlParser, XML_FALSE); // Stop any pending processing
XML_SetElementHandler(xmlParser, nullptr, nullptr); // Clear callbacks
XML_SetCharacterDataHandler(xmlParser, nullptr);
XML_ParserFree(xmlParser);
xmlParser = nullptr;
file.close();
return false;
}
} while (!done);
XML_StopParser(parser, XML_FALSE); // Stop any pending processing
XML_SetElementHandler(parser, nullptr, nullptr); // Clear callbacks
XML_SetCharacterDataHandler(parser, nullptr);
XML_ParserFree(parser);
XML_StopParser(xmlParser, XML_FALSE); // Stop any pending processing
XML_SetElementHandler(xmlParser, nullptr, nullptr); // Clear callbacks
XML_SetCharacterDataHandler(xmlParser, nullptr);
XML_ParserFree(xmlParser);
xmlParser = nullptr;
file.close();
// Process last page if there is still text
if (currentTextBlock) {
makePages();
// Set the content offset for the final page
if (currentPage) {
currentPage->firstContentOffset = static_cast<uint32_t>(currentPageStartOffset);
}
completePageFn(std::move(currentPage));
currentPage.reset();
currentTextBlock.reset();
@@ -570,8 +588,15 @@ void ChapterHtmlSlimParser::addLineToPage(std::shared_ptr<TextBlock> line) {
const int lineHeight = renderer.getLineHeight(fontId) * lineCompression;
if (currentPageNextY + lineHeight > viewportHeight) {
// Set the content offset for the page being completed
if (currentPage) {
currentPage->firstContentOffset = static_cast<uint32_t>(currentPageStartOffset);
}
completePageFn(std::move(currentPage));
// Start new page - offset will be set when first content is added
currentPage.reset(new Page());
currentPageStartOffset = lastCharDataOffset; // Use offset from when content was parsed
currentPageNextY = 0;
}
@@ -587,6 +612,8 @@ void ChapterHtmlSlimParser::makePages() {
if (!currentPage) {
currentPage.reset(new Page());
// Use offset captured during character data parsing
currentPageStartOffset = lastCharDataOffset;
currentPageNextY = 0;
}

5
lib/Epub/Epub/parsers/ChapterHtmlSlimParser.h
View File

@@ -54,6 +54,11 @@ class ChapterHtmlSlimParser {
bool effectiveBold = false;
bool effectiveItalic = false;
bool effectiveUnderline = false;
// Byte offset tracking for position restoration after re-indexing
XML_Parser xmlParser = nullptr; // Store parser for getting current byte index
size_t currentPageStartOffset = 0; // Byte offset when current page was started
size_t lastCharDataOffset = 0; // Byte offset of last character data (captured during parsing)
void updateEffectiveInlineStyle();
void startNewTextBlock(TextBlock::Style style);