st_intersection (with missing y) produces "degenerate" polygons

[jguelat]

Describe the bug I'm using the st_intersection() function on a single sf object with overlapping polygons to extract the geometries with and without overlap. The polygons of the overlapping parts are OK, but the non-overlapping polygons almost always have some zero-width "part".

To Reproduce Using the attached data, polygons 1 and 3 have a zero-width part while polygon 2 is OK. testinter.zip

library(sf)
pols <- st_read("testinter.gpkg", "pols")
inter <- st_intersection(pols)
par(mfrow = c(1, 3))
plot(st_geometry(inter[1,]))
plot(st_geometry(inter[2,]))
plot(st_geometry(inter[3,]))

Additional context Thanks a lot for all your work on the sf package!

sessionInfo()

R version 4.4.2 (2024-10-31 ucrt) Platform: x86_64-w64-mingw32/x64 Running under: Windows 10 x64 (build 19045)

Matrix products: default

locale: [1] LC_COLLATE=German_Switzerland.utf8 LC_CTYPE=German_Switzerland.utf8 LC_MONETARY=German_Switzerland.utf8 [4] LC_NUMERIC=C LC_TIME=German_Switzerland.utf8

time zone: Europe/Zurich tz

attached base packages: [1] stats graphics utils datasets grDevices methods base

other attached packages: [1] sf_1.0-19

loaded via a namespace (and not attached): [1] compiler_4.4.2 magrittr_2.0.3 class_7.3-23 tools_4.4.2 DBI_1.2.3 units_0.8-5
[7] proxy_0.4-27 Rcpp_1.0.14 KernSmooth_2.23-26 grid_4.4.2 e1071_1.7-16 classInt_0.4-11

sf::sf_extSoftVersion()

      GEOS           GDAL         proj.4 GDAL_with_GEOS     USE_PROJ_H           PROJ 
  "3.12.2"        "3.9.3"        "9.4.1"         "true"         "true"        "9.4.1"