geometry-io-ewkt

Part of the boost_geometry workspace — a Rust port of Boost.Geometry. Most users should depend on the facade crate, which re-exports this one; depend on this crate directly only for a slimmer build.

PostGIS Extended Well-Known Text (EWKT) reader and writer.

Not part of Boost.Geometry; follows the PostGIS manual, sections "4.1.3. WKT and WKB" and "4.2.1. PostGIS EWKB and EWKT", and the PostGIS reader in liblwgeom/lwin_wkt_lex.l. EWKT is the dialect ST_AsEWKT emits and ST_GeomFromEWKT accepts: OGC WKT plus an optional SRID=<digits>; prefix and the glued dimension-suffix spellings. ST_AsEWKT glues only M (POINTM(1 2 3)); PostGIS reads the glued Z, M, and ZM forms and the OGC-spaced POINT Z / POINT M / POINT ZM forms alike, and this crate reads all of them.

The geometry body is delegated to [geometry_io_wkt]. This crate scans the prefix, overwrites a glued suffix with spaces in place — a same-length rewrite, so every byte offset it reports indexes the string the caller passed — and hands the body over. [from_ewkt] therefore emits a [geometry_model::DynGeometry], and the six parse_* conveniences return concrete model types; each is paired with the spatial-reference id in an [Ewkt].

PostGIS treats SRID 0 as unknown ([Srid::UNKNOWN]) and omits the prefix for it. This crate reads SRID=0; as Some(Srid::UNKNOWN) and writes it back as SRID=0;; callers pass None to omit the prefix.

Every ordinate past the second is discarded, as in [geometry_io_wkt], and nothing above 2D is ever written — so a dimension suffix carries no information this model can hold.

Read and write a prefixed geometry

use geometry_io_ewkt::{Srid, from_ewkt, to_ewkt};

let e = from_ewkt("SRID=4326;POINTM(1 2 3)").unwrap();
assert_eq!(e.srid, Some(Srid::new(4326)));
assert_eq!(to_ewkt(&e.geometry, e.srid), "SRID=4326;POINT(1 2)");

License

BSL-1.0 — see LICENSE.