Include all TIFF string bytes in StringValue instances

[pdgmdm]

Hi, sometimes string meta data contains actually a byte array.
I tried to get the data as byte array using
public static byte[]? GetByteArray(this Directory directory, int tagType)
Now it appears the meta data reader creates a StringValue by terminating on first null character.
Is there a work-around to force the tag to be read as byte array ?

[drewnoakes]

When a value is believed to be a string, it's stored as a StringValue. Internally this is a byte[] and an optional Encoding. This allows the user to reinterpret the raw values in different ways if needed. If you want the raw bytes, just access the StringValue.Bytes property.

In your case, it sounds like the value you want contains a \0 which is causing the read of that StringValue to end earlier than you'd like. This might be due to use of the GetNullTerminatedStringValue method.

If the bytes returned by StringValue.Bytes aren't what you're expecting, then it really depends upon the specific tag. Can you provide a sample image that exhibits this behaviour, and explain which tag in particular you're looking at, and what you're like to see instead?

[pdgmdm]

Yes you are right: it is a tiff file and I see in the TifReader.cs that strings are always trimmed to the first null character:

        private static void ProcessTag(ITiffHandler handler, int tagId, in int valueOffset, int componentCount, TiffDataFormatCode formatCode, IndexedReader reader)
        {
            switch (formatCode)
            {
                case TiffDataFormatCode.Undefined:
                {
                    // this includes exif user comments
                    handler.SetByteArray(tagId, reader.GetBytes(valueOffset, componentCount));
                    break;
                }
                case TiffDataFormatCode.String:
                {
                    handler.SetString(tagId, reader.GetNullTerminatedStringValue(valueOffset, componentCount));
                    break;
                }

Would be nice if StringValue would contain the complete byte[] array, and the trimming to first null character would be done when converting the StringValue to a string with ToString(). This would allow access to the full byte[] for cases where binary data is stored in string tags

[drewnoakes]

Agreed. Did you want to make a pull request to update the TIFF reader? I'm working towards getting the next release out, and could include this in that release.