{

// Could also use spi_flash_read() here, but the return values differ between ESP (SPI_FLASH_RESULT_OK) and ESP32 (ESP_OK)

if (ESP.flashRead(_flashBase + _flashOffset, (uint32_t *)_buffer, sizeof(_buffer)))

{

_bufferPos = 0;

}

else

{

// _bufferPos > sizeof() indicates error

_bufferPos = sizeof(_buffer) + 1;

}

{

_flashBase = base;

_flashOffset = 0 - sizeof(_buffer); // underflow intentionally so offset+pos = 0

_bufferPos = sizeof(_buffer);

{

// align(4), rounding down

_flashOffset = (offset >> 2) << 2;

// capture the new bufferPos before fill advances it

uint8_t newBufferPos = offset - _flashOffset;

fillBuffer();

_bufferPos = newBufferPos;

{

int retVal = peek();

++_bufferPos;

return retVal;

{

if (_bufferPos == sizeof(_buffer))

{

_flashOffset += sizeof(_buffer);

fillBuffer();

}

return _buffer[_bufferPos];