A zero value was used before to indicate invalid
data, and thus 0° latitude/longitude or 0 m altitude were
not displayed while being valid data. Use NaN to indicate
invalid data. Increase altitude precision by avoiding
intermediate float type
Details
Details
- Reviewers
bruns - Commits
- R286:c3fbdea4f7f3: Fix failing test of exiv gps data
Diff Detail
Diff Detail
- Repository
- R286 KFileMetaData
- Branch
- fix_zero_gps
- Lint
No Linters Available - Unit
No Unit Test Coverage - Build Status
Buildable 7001 Build 7019: arc lint + arc unit
Comment Actions
Can you create two test files, one with e.g. altitude == 0, and one without altitude?
| src/extractors/exiv2extractor.cpp | ||
|---|---|---|
| 298 | when you rewrite this as if (it != data.end()) {
auto ratio = it->value().toRational();
it = data.findKey(Exiv2::ExifKey("Exif.GPSInfo.GPSAltitudeRef"));
if ((ratio.second != 0) && (it != data.end())) {
auto altRef = it->value().toLong();
if (altRef) {
alt = -1.0 * ratio.first / ratio.second;
} else {
alt = 1.0 * ratio.first / ratio.second;
}
}
}does the unit test pass without special handling of the altitude fuzz? | |
| src/extractors/exiv2extractor.cpp | ||
|---|---|---|
| 298 | Yes, it does | |