Remove tinyxml2 dependency replace with expat parsers (#9)

2025-12-13 19:36:01 +11:00
parent d450f362d1
commit c7a32fe41f
17 changed files with 696 additions and 253 deletions
@@ -0,0 +1,161 @@
+#include "ContentOpfParser.h"
+
+#include <HardwareSerial.h>
+#include <ZipFile.h>
+
+bool ContentOpfParser::setup() {
+  parser = XML_ParserCreate(nullptr);
+  if (!parser) {
+    Serial.printf("[%lu] [COF] Couldn't allocate memory for parser\n", millis());
+    return false;
+  }
+
+  XML_SetUserData(parser, this);
+  XML_SetElementHandler(parser, startElement, endElement);
+  XML_SetCharacterDataHandler(parser, characterData);
+  return true;
+}
+
+bool ContentOpfParser::teardown() {
+  if (parser) {
+    XML_ParserFree(parser);
+    parser = nullptr;
+  }
+  return true;
+}
+
+size_t ContentOpfParser::write(const uint8_t data) { return write(&data, 1); }
+
+size_t ContentOpfParser::write(const uint8_t* buffer, const size_t size) {
+  if (!parser) return 0;
+
+  const uint8_t* currentBufferPos = buffer;
+  auto remainingInBuffer = size;
+
+  while (remainingInBuffer > 0) {
+    void* const buf = XML_GetBuffer(parser, 1024);
+
+    if (!buf) {
+      Serial.printf("[%lu] [COF] Couldn't allocate memory for buffer\n", millis());
+      XML_ParserFree(parser);
+      parser = nullptr;
+      return 0;
+    }
+
+    const auto toRead = remainingInBuffer < 1024 ? remainingInBuffer : 1024;
+    memcpy(buf, currentBufferPos, toRead);
+
+    if (XML_ParseBuffer(parser, static_cast<int>(toRead), remainingSize == toRead) == XML_STATUS_ERROR) {
+      Serial.printf("[%lu] [COF] Parse error at line %lu: %s\n", millis(), XML_GetCurrentLineNumber(parser),
+                    XML_ErrorString(XML_GetErrorCode(parser)));
+      XML_ParserFree(parser);
+      parser = nullptr;
+      return 0;
+    }
+
+    currentBufferPos += toRead;
+    remainingInBuffer -= toRead;
+    remainingSize -= toRead;
+  }
+
+  return size;
+}
+
+void XMLCALL ContentOpfParser::startElement(void* userData, const XML_Char* name, const XML_Char** atts) {
+  auto* self = static_cast<ContentOpfParser*>(userData);
+  (void)atts;
+
+  if (self->state == START && (strcmp(name, "package") == 0 || strcmp(name, "opf:package") == 0)) {
+    self->state = IN_PACKAGE;
+    return;
+  }
+
+  if (self->state == IN_PACKAGE && (strcmp(name, "metadata") == 0 || strcmp(name, "opf:metadata") == 0)) {
+    self->state = IN_METADATA;
+    return;
+  }
+
+  if (self->state == IN_METADATA && strcmp(name, "dc:title") == 0) {
+    self->state = IN_BOOK_TITLE;
+    return;
+  }
+
+  if (self->state == IN_PACKAGE && (strcmp(name, "manifest") == 0 || strcmp(name, "opf:manifest") == 0)) {
+    self->state = IN_MANIFEST;
+    return;
+  }
+
+  if (self->state == IN_PACKAGE && (strcmp(name, "spine") == 0 || strcmp(name, "opf:spine") == 0)) {
+    self->state = IN_SPINE;
+    return;
+  }
+
+  // TODO: Support book cover
+  // if (self->state == IN_METADATA && (strcmp(name, "meta") == 0 || strcmp(name, "opf:meta") == 0)) {
+  // }
+
+  if (self->state == IN_MANIFEST && (strcmp(name, "item") == 0 || strcmp(name, "opf:item") == 0)) {
+    std::string itemId;
+    std::string href;
+
+    for (int i = 0; atts[i]; i += 2) {
+      if (strcmp(atts[i], "id") == 0) {
+        itemId = atts[i + 1];
+      } else if (strcmp(atts[i], "href") == 0) {
+        href = self->baseContentPath + atts[i + 1];
+      }
+    }
+
+    self->items[itemId] = href;
+    return;
+  }
+
+  if (self->state == IN_SPINE && (strcmp(name, "itemref") == 0 || strcmp(name, "opf:itemref") == 0)) {
+    for (int i = 0; atts[i]; i += 2) {
+      if (strcmp(atts[i], "idref") == 0) {
+        self->spineRefs.emplace_back(atts[i + 1]);
+        break;
+      }
+    }
+    return;
+  }
+}
+
+void XMLCALL ContentOpfParser::characterData(void* userData, const XML_Char* s, const int len) {
+  auto* self = static_cast<ContentOpfParser*>(userData);
+
+  if (self->state == IN_BOOK_TITLE) {
+    self->title.append(s, len);
+    return;
+  }
+}
+
+void XMLCALL ContentOpfParser::endElement(void* userData, const XML_Char* name) {
+  auto* self = static_cast<ContentOpfParser*>(userData);
+  (void)name;
+
+  if (self->state == IN_SPINE && (strcmp(name, "spine") == 0 || strcmp(name, "opf:spine") == 0)) {
+    self->state = IN_PACKAGE;
+    return;
+  }
+
+  if (self->state == IN_MANIFEST && (strcmp(name, "manifest") == 0 || strcmp(name, "opf:manifest") == 0)) {
+    self->state = IN_PACKAGE;
+    return;
+  }
+
+  if (self->state == IN_BOOK_TITLE && strcmp(name, "dc:title") == 0) {
+    self->state = IN_METADATA;
+    return;
+  }
+
+  if (self->state == IN_METADATA && (strcmp(name, "metadata") == 0 || strcmp(name, "opf:metadata") == 0)) {
+    self->state = IN_PACKAGE;
+    return;
+  }
+
+  if (self->state == IN_PACKAGE && (strcmp(name, "package") == 0 || strcmp(name, "opf:package") == 0)) {
+    self->state = START;
+    return;
+  }
+}