jpegxl: Handle various bits per sample values in JPEG XL files (#4738)

Handle bps other than 8, 16, 32, with appropriate conversion.

---------

Signed-off-by: Peter Kovář <peter.kovar@reflexion.tv>

Author: 1div0

src/jpegxl.imageio/jxlinput.cpp

@@ -221,6 +221,7 @@ JxlInput::open(const std::string& name, ImageSpec& newspec)
     JxlPixelFormat format;
     JxlDataType jxl_data_type;
     TypeDesc m_data_type;
+    uint32_t bits = 0;
 
     for (;;) {
         JxlDecoderStatus status = JxlDecoderProcessInput(m_decoder.get());
@@ -249,20 +250,21 @@ JxlInput::open(const std::string& name, ImageSpec& newspec)
             // Need to check how we can support bfloat16 if jpegxl supports it
             bool is_float = info.exponent_bits_per_sample > 0;
 
-            switch (info.bits_per_sample) {
-            case 8:
+            if (info.bits_per_sample <= 8) {
                 jxl_data_type = JXL_TYPE_UINT8;
                 m_data_type   = TypeDesc::UINT8;
-                break;
-            case 16:
+                bits          = 8;
+            } else if (info.bits_per_sample <= 16) {
                 jxl_data_type = is_float ? JXL_TYPE_FLOAT16 : JXL_TYPE_UINT16;
                 m_data_type   = is_float ? TypeDesc::HALF : TypeDesc::UINT16;
-                break;
-            case 32:
+                bits          = 16;
+            } else if (info.bits_per_sample <= 32) {
                 jxl_data_type = JXL_TYPE_FLOAT;
                 m_data_type   = TypeDesc::FLOAT;
-                break;
-            default: errorfmt("Unsupported bits per sample\n"); return false;
+                bits          = 32;
+            } else {
+                errorfmt("Unsupported bits per sample\n");
+                return false;
             }
 
             format = { m_channels, jxl_data_type, JXL_NATIVE_ENDIAN, 0 };
@@ -307,11 +309,9 @@ JxlInput::open(const std::string& name, ImageSpec& newspec)
                 return false;
             }
             if (buffer_size
-                != info.xsize * info.ysize * m_channels * info.bits_per_sample
-                       / 8) {
+                != info.xsize * info.ysize * m_channels * bits / 8) {
                 errorfmt("Invalid out buffer size {} {}\n", buffer_size,
-                         info.xsize * info.ysize * m_channels
-                             * info.bits_per_sample / 8);
+                         info.xsize * info.ysize * m_channels * bits / 8);
                 return false;
             }