Hello, We’ve noticed that Tika is incorrectly detecting the file .cr3 as video/quicktime, other raw files are detected as image/tiff (including the .cr3’s predecessor the .cr2). I’ve uploaded a sample file here https://dropfiles.org/j8CS4Snr (that was taken from this review https://www.photographyblog.com/reviews/canon_eos_r10_review#google_vignette)
When we add a custom-mimetypes.xml file with a mime-type entry like this: <mime-type type="image/x-raw-canon"> <_comment>Canon raw image</_comment> <sub-class-of type="image/tiff"/> <glob pattern="*.crw"/> <glob pattern="*.cr2"/> <glob pattern="*.cr3"/> </mime-type> The .cr3 file is still identified as video/quicktime but when we add the below configuration Tika matches it to something close to what we want: <mime-type type="image/x-raw-canon3"> <_comment>Canon raw image</_comment> <sub-class-of type="video/quicktime"/> <glob pattern="*.cr3"/> </mime-type> But this won’t give us our desired output as we’re hoping to group all Canon raw images under the same mime-type. Do you have any ideas how to get this working? We’re using tika-core 2.7.0 in a Java 8 project. Thank you, Richard
