| Age | Commit message (Collapse) | Author |
|---|
|
|
|
IAU_2015:xxxx transparently
|
|
|
|
|
|
Update database layout version number to 1.2 consequently
This new capability will be used by IAU planetary CRS (refs #2601)
|
|
None of our tables are indexed by a INTEGER PRIMARY KEY, but most of
them are by a (auth_name, code) primary key. Consequently they can
benefit from being created as WITHOUT ROWID tables
(https://sqlite.org/withoutrowid.html), which avoids an index to be
created on the rowid we don't use. WITHOUT ROWID is a feature added in
SQLite 3.8.2, so as our baseline is 3.11, we can use it.
This decreases the DB size from 7,749,632 to 7,229,440 bytes, without
any measurable consequence on performance.
|
|
Non-trivial updates:
- some vertical CRS are now encoded as DerivedVerticalCRS. e.g EPSG:8228
"NAVD88 height (ft)", with base EPSG:5703 "NAVD88 height". As we don't
have support in our PROJ db model for DerivedVerticalCRS, modify the
import script to 'resolve' the derivation up to the original datum.
- Method EPSG:1069 'Change of Vertical Unit' is no longer used. It is
replaced by a generic-purpose EPSG:1104 method that doesn't take any
conversion factor. And generic conversions EPSG:7812 and EPSG:7813 are
now used in concatenated operations, which require code changes as
well.
|
|
We never select by those columns, so don't set them. Reduce from 8.4 to
7.9 MB.
Upgrade the minor version of the database layout. (that database can
still be read by PROJ 8.0)
|
|
Most codes (especially *all* in EPSG) are integers. As integers are
stored as variable-length in the db, it is preferable to store them as
such when possible.
So use a special INTEGER_OR_TEXT data type. This is a
"non-standard" type declaration, but this is perfectly legal as SQLite is
loosely typed.
As this declaration contains the string INT, it is assigned INTEGER affinity.
Which means that values provided either as text (that contains integer value)
or integer will be stored as integers, whereas text values will be stored as
text. See paragraph 3 and 3.1 of https://www.sqlite.org/datatype3.html.
The "INTEGER_OR_TEXT" name is a hint for the user, and software like
GDAL (>= 3.3) to expose the column as string...
The effect of using this rather than TEXT is making the DB size go from
9 MB to 8.4.
There is no need to change the DATABASE.LAYOUT version numbering as this
is completely forward and backward compatible.
|
|
grid_transformation and other_transformation. No practical impact
|
|
- change foreign key check, so that it is enabled outside of the
transaction where we insert things, and can make the sqlite3
process fail in case of violations, without the postcheck done
in the autoconf build
- autoconf and cmake builds: simplification related to the above
(which also means that cmake builds now have the fkey check, which
was omitted until now)
|
|
conversion method in the PROJ namespace
Helps for example to use CRS definitions using the non-EPSG codified
Sinusoidal method.
|
|
* Correction grid NKG:ETRF14 to EPSG:7922
* Added NKG:ITRF_TO_NO GIE test
* Correction grid no_kv_NKGETRF14_EPSG7922_2000 added to grid_alternatives.sql
* proj_method 'velocity_grid' added in check_grid_alternatives_proj_method. NKG velocity grid added to grid_alternatives.sql
Co-authored-by: Even Rouault <even.rouault@spatialys.com>
|
|
That is the option to make SQLite reject misuses of double quotes instead
of single quotes. Most SQLite builds are forgiving, but some forks proposing
a CMake build system default to SQLITE_DQS=0
|
|
to other_transformation table
|
|
ellipsoidal height
|
|
handle dynamic vertical datums, and instanciate them properly from database
|
|
|
|
dynamic datums, but not yet used
|
|
Content mostly unchanged since v9.9
This update is "minimal" in that it mostly reflects the removal of the 'area'
table, replaced now by 'extent', 'scope' and 'usage'
Other new aspects of EPSG v10 are left aside.
|
|
This is in preparation for EPSG 9.8.11 import that supersedes a number of
grid transformation for US transformations, but the superseded and replacement
transformations don't operate on the same (source_crs, target_crs), which is
a bit weird. So in that situation, ignores the supersession.
|
|
|
|
|
|
|
|
Fixes #1913
AuthorityFactory::createBetweenGeodeticCRSWithDatumBasedIntermediates() issued
a complex SQL query that pushes the SQLite3 query plan optimizer to its limits.
Was working reasonably with sqlite 3.11, but not with later versions.
So put less constraints in the main query and do post-processing checks and
auxiliary requests to avoid such issues.
For some unknown reason, this slightly slows down a bit execution time of the
whole test_cpp_api binary (~ 10%), but couldn't come with something better,
despite trying many variations of the main SQL query. It seems that in the
general case the non-filter LEFT JOIN on the supersession table helped,
except on this EPSG:7842 case.
|
|
(gr3df97a.txt) and map them to new +proj=xyzgridshift
|
|
in the EPSG/grid_transformation table
|
|
|
|
grid_alternatives, fix related entries and simplify/robustify logic to deal with EPSG 'Geographic3D to GravityRelatedHeight' methods
|
|
Fixes #1867
|
|
CONSTRAINT CHECK
|
|
Populated from realization_epoch column from EPSG
The 'publication_date' naming is from OGC Topic 2, and hasn't been yet adopted
by the EPSG dataset.
See http://docs.opengeospatial.org/as/18-005r4/18-005r4.html , Annex G, clause 11
and https://32zn56499nov99m251h4e9t8-wpengine.netdna-ssl.com/wp-content/uploads/2019/09/EPSG-relational-data-model-changes_2019-09-18.pdf
|
|
This was introduced in 63857c92b271bbcd10df0a032304982011acb2a9. Due to
the fix done in the previous commit, we can mostly revert the above commit.
We just keep the added tests and the custom WGS 84<-->WGS 84 (Gxxxx) null
transformations.
|
|
a proj_create_vertical_crs_ex()
|
|
arbitrary number of steps (fixes #1632)
EPSG:9103 (NAD27 to ITRF2014 (1)) is now handled.
Note:EPSG:9104 (NAD27 to ITRF2014 (2)) is not currently, since it uses
for step EPSG:8861 (NAD83(HARN) to NAD83(FBN) (1))
an unsupported transformation method (NADCON5 (3D), EPSG:1075).
|
|
Currently very few transformations from/to WGS84 (Gxxxx) are registered
in the EPSG database, and there isn't even transformations between WGS84 EPSG:4326
and those ones. Consequently transformations to those realizations often
ended up as no-operation, whereas going through WGS84 EPSG:4326 will bring
more meaningful results. So register those EPSG:4326<-->WGS 84 (Gxxx)
null transformations, and when having WGS 84 (Gxxx) as source/target,
consider EPSG:4326 as an intermediate.
This change has no effect on the existing direct transformations
from/to WGS 84 (Gxxx).
|
|
|
|
- Transform conversion as a view, and when inserting into it, actually
insert into 3 tables: conversion_table, conversion_method and conversion_param,
so that method and parameter names are not repeated each time.
- Similarly for helmert_tranformation, insert into helmert_transformation_tabl
and coordinate_operation_method.
This reduces the db size from 6 344 704 bytes to 5 853 184 bytes, without
significant slowdown for queries.
|
|
|
|
|
|
to restrict and prioritize searches
|
|
transformations
|
|
This work mostly consists of:
- a C++ implementation of the ISO-19111:2018 / OGC Topic 2
"Referencing by coordinates" classes to represent Datums,
Coordinate systems, CRSs (Coordinate Reference Systems) and
Coordinate Operations.
- methods to convert between this C++ modeling and WKT1, WKT2
and PROJ string representations of those objects
- management and query of a SQLite3 database of CRS and Coordinate Operation definition
- a C API binding part of those capabilities
This is all-in-one squashed commit of the work of
https://github.com/OSGeo/proj.4/pull/1040
|
