PROJ/test, branch 6.0

bonne: avoid division by zero

2019-04-01T20:18:18+00:00

Fixes https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=14010
Credit to OSS Fuzz

pj_gauss_ini(): fix division by zero

2019-04-01T20:06:03+00:00

Fixes https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=14003
Credit to OSS Fuzz

Merge pull request #1400 from OSGeo/backport-1396-to-6.0

2019-04-01T17:38:45+00:00

[Backport 6.0] Ossfuzz 13947 13948

tpeqd: avoid division by zero

2019-03-29T22:31:57+00:00

Fixes https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=13948
Credit to OSS Fuzz

unitconvert: prevent division by zero

2019-03-29T22:13:55+00:00

Fixes https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=13947
Credit to OSS Fuzz

createOperations(): improve behaviour with input CRS from WKT that lacks intermediate IDs (fixes #1343)

2019-03-29T12:30:57+00:00

createOperations(): improve BoundCRS<-->non-bound-CRS case

2019-03-28T16:33:26+00:00

Fixes #1388
Typically helps for
projinfo -s "+proj=longlat +ellps=GRS80 +towgs84=1,2,3 +type=crs" -t EPSG:4258
by researching operations from the pivot WGS84 implied by the towgs84 clause
to EPSG:4258.

ocea: fix behaviour when +alpha is specified

2019-03-28T10:58:59+00:00

The one-point case was completely broken with lat_0 being ignored.
I've fixed that, and also modified the alpha orientation, so that the angle
corresponds to the one of the 2 point case when going from point 1 to point 2,
similarly to what omerc does.
This was found rather experimentaly with the added test cases that try to
find equivalence between 1-point and 2-point cases.

Fixes #1379
and adresses https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=13930

omerc inverse: fix division

2019-03-26T14:49:42+00:00

Test case https://oss-fuzz.com/testcase-detail/5739351578771456 of
https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=13894
Credit to OSS Fuzz

pj_strerrno: enable system error messages

2019-03-25T18:12:03+00:00

HAVE_STRERROR is defined in proj_config.h.