[TIKA-4781] Map EXIF GPS altitude to the existing geo:alt property (#2941)

The GeotagHandler mapped latitude, longitude and the GPS timestamp from
the EXIF GpsDirectory, but not the altitude, although tika-core defines
Geographic.ALTITUDE and metadata-extractor exposes the value in the same
directory. Consumers only got the raw description string
("GPS:GPS Altitude" = "227.4 metres").

Set geo:alt as a signed value honoring GPSAltitudeRef (1 = below sea
level = negative), formatted with the same DecimalFormat as latitude
and longitude. Analogous to TIKA-4425, which added the normalized GPS
timestamp from the same directory; see TIKA-2861 for the corresponding
ISO 6709 altitude handling in MP4/QuickTime.

The existing IMG_1034.heic fixture already carries an altitude, so
HeifParserTest now asserts it end to end; handler-level tests cover the
above and below sea level cases.

Author: dschmidt

tika-parsers/tika-parsers-standard/tika-parsers-standard-modules/tika-parser-image-module/src/main/java/org/apache/tika/parser/image/ImageMetadataExtractor.java

@@ -684,10 +684,10 @@ public boolean supports(Class<? extends Directory> directoryType) {
         }
 
         public void handle(Directory directory, Metadata metadata) throws MetadataException {
+            DecimalFormat geoDecimalFormat = new DecimalFormat(GEO_DECIMAL_FORMAT_STRING,
+                    new DecimalFormatSymbols(Locale.ENGLISH));
             GeoLocation geoLocation = ((GpsDirectory) directory).getGeoLocation();
             if (geoLocation != null) {
-                DecimalFormat geoDecimalFormat = new DecimalFormat(GEO_DECIMAL_FORMAT_STRING,
-                        new DecimalFormatSymbols(Locale.ENGLISH));
                 metadata.set(TikaCoreProperties.LATITUDE,
                         geoDecimalFormat.format(geoLocation.getLatitude()));
                 metadata.set(TikaCoreProperties.LONGITUDE,
@@ -697,6 +697,17 @@ public void handle(Directory directory, Metadata metadata) throws MetadataExcept
             if (gpsDate != null) {
                 metadata.set(Geographic.TIMESTAMP, gpsDate);
             }
+            Rational altitude = ((GpsDirectory) directory).getRational(GpsDirectory.TAG_ALTITUDE);
+            if (altitude != null) {
+                double metres = altitude.doubleValue();
+                //GPSAltitudeRef 1 means below sea level
+                Integer altitudeRef =
+                        ((GpsDirectory) directory).getInteger(GpsDirectory.TAG_ALTITUDE_REF);
+                if (altitudeRef != null && altitudeRef == 1) {
+                    metres = -metres;
+                }
+                metadata.set(TikaCoreProperties.ALTITUDE, geoDecimalFormat.format(metres));
+            }
         }
     }
 

tika-parsers/tika-parsers-standard/tika-parsers-standard-modules/tika-parser-image-module/src/test/java/org/apache/tika/parser/image/HeifParserTest.java

@@ -51,6 +51,7 @@ public void testSimple() throws Exception {
             assertEquals("image/heic", metadata.get(Metadata.CONTENT_TYPE));
             assertEquals("23.177917", metadata.get(Metadata.LATITUDE));
             assertEquals("113.394317", metadata.get(Metadata.LONGITUDE));
+            assertEquals("42.810337", metadata.get(Geographic.ALTITUDE));
 
             assertEquals("2018-02-05T07:11:43Z", metadata.get(Geographic.TIMESTAMP));
         }

tika-parsers/tika-parsers-standard/tika-parsers-standard-modules/tika-parser-image-module/src/test/java/org/apache/tika/parser/image/ImageMetadataExtractorTest.java

@@ -28,11 +28,13 @@
 import java.util.Locale;
 import java.util.TimeZone;
 
+import com.drew.lang.Rational;
 import com.drew.metadata.Directory;
 import com.drew.metadata.MetadataException;
 import com.drew.metadata.Tag;
 import com.drew.metadata.exif.ExifIFD0Directory;
 import com.drew.metadata.exif.ExifSubIFDDirectory;
+import com.drew.metadata.exif.GpsDirectory;
 import com.drew.metadata.jpeg.JpegCommentDirectory;
 import org.junit.jupiter.api.Test;
 import org.mockito.Mockito;
@@ -85,6 +87,28 @@ public void testExifHandlerParseDate() throws MetadataException {
                 "Should be ISO date without time zone");
     }
 
+    @Test
+    public void testGeotagHandlerAltitude() throws MetadataException {
+        GpsDirectory gps = Mockito.mock(GpsDirectory.class);
+        Mockito.when(gps.getRational(GpsDirectory.TAG_ALTITUDE)).thenReturn(new Rational(2274, 10));
+        Mockito.when(gps.getInteger(GpsDirectory.TAG_ALTITUDE_REF)).thenReturn(0);
+        Metadata metadata = new Metadata();
+
+        new ImageMetadataExtractor.GeotagHandler().handle(gps, metadata);
+        assertEquals("227.4", metadata.get(TikaCoreProperties.ALTITUDE));
+    }
+
+    @Test
+    public void testGeotagHandlerAltitudeBelowSeaLevel() throws MetadataException {
+        GpsDirectory gps = Mockito.mock(GpsDirectory.class);
+        Mockito.when(gps.getRational(GpsDirectory.TAG_ALTITUDE)).thenReturn(new Rational(2274, 10));
+        Mockito.when(gps.getInteger(GpsDirectory.TAG_ALTITUDE_REF)).thenReturn(1);
+        Metadata metadata = new Metadata();
+
+        new ImageMetadataExtractor.GeotagHandler().handle(gps, metadata);
+        assertEquals("-227.4", metadata.get(TikaCoreProperties.ALTITUDE));
+    }
+
     @Test
     public void testExifHandlerParseDateFallback() throws MetadataException {
         ExifIFD0Directory exif = Mockito.mock(ExifIFD0Directory.class);