Merge pull request #12 from KnowWhereGraph/develop

Support integrating against standalone geometries

README.md

@@ -11,32 +11,41 @@ Rather than relying on geosparql functions (which in turn rely on geosparql supp
 
 This breaks the reliance on the need for the graph database to support s2 indexing and instead make use of the predicate index from the pre-materialized spatial relations.
 
+## Cell Generation and Integration
+
+There are two tools:
+
+1. s2.py: This generates the S2 cell structure at a desired layer. For example, generating cells at level 3 and 4.
+2. integrate.py: This performs s2 integrations against existing geometries. These may be your own geometries, they may be the output of the s2 tool. 
+
 ## Running
 
 ### Docker
 
-Docker should be used to generate the s2 coverings, which can be done by running the following from the root folder.
+The project dependencies can be difficult to install; docker images are provided so that the code can be run in different environments without needing to install dependencies. Rather than offering a docker image for each cript, both scripts are included in the image and they can be called externally
+
+#### Generating S2 Cells
 
 ```bash
 git clone https://github.com/KnowWhereGraph/s2-coverings.git
 cd s2-coverings
 docker run -v ./:/s2 ghcr.io/knowwheregraph/s2-coverings:main python3 src/s2.py --level <level>
 ```
 
-Cell integration can be disabled by adding the `--ni` flag to the command,
+#### S2 Integration
+
 
 ```bash
-docker run -v ./:/s2 ghcr.io/knowwheregraph/s2-coverings:main python3 src/s2.py --level <level> --ni
+docker run -v ./:/s2 ghcr.io/knowwheregraph/s2-coverings:main python3 src/integrate.py --path <path to geometries>
 ```
 
-A complete list of options can be found by running
+A complete list of options can be found by running the help command on each tool. For example,
 ```bash
 python3 src/s2.py --help
 options:
   -h, --help            show this help message and exit
   --level LEVEL         Level at which the s2 cells are generated for
   --format [FORMAT]     The format to write the RDF in. Options are xml, n3, turtle, nt, pretty-xml, trix, trig, nquads, json-ld, hext
-  --ni [NI]             When used, s2 integration is disabled
   --compressed [COMPRESSED]
                         use the S2 hierarchy to write a compressed collection of relations at various levels
 ```

requirements.txt

@@ -1,8 +1,8 @@
-fiona==1.9.5
+fiona==1.10.1
 pytest==8.3.2
-rasterio==1.3.10
-rasterstats==0.19.0
-rdflib==7.0.0
+rasterio==1.4.3
+rasterstats==0.20.0
+rdflib==7.1.1
 s2sphere==0.2.5
-Shapely==2.0.5
+Shapely==2.0.6
 SPARQLWrapper==2.0.0

src/integrate.py

@@ -0,0 +1,24 @@
+from __future__ import annotations
+
+import argparse
+
+from lib.integrator import Integrator
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--path",
+        help="Path to the folder with triples being integrated",
+        type=str,
+        nargs="?",
+        default="./output",
+    )
+    parser.add_argument(
+        "--compressed",
+        help="use the S2 hierarchy to write a compressed collection of relations at various levels",
+        type=bool,
+        nargs="?",
+        default=True,
+    )
+    args = parser.parse_args()
+    Integrator(args.compressed, args.path)

src/lib/integrator.py

@@ -18,12 +18,18 @@ class Integrator:
     Abstraction over the process for integrating s2 cells together with spatial relations.
     """
 
-    def __init__(self, compressed: bool):
+    def __init__(self, compressed: bool, folder: str | Path):
+        """
+        Creates a new Integrator
+
+        :param compressed: Whether the triples are compressed or not
+        :param folder: Path to the folder where the triples are
+        """
         if compressed:
             print(
                 "Compression is on. Relations will be compressed using the S2 hierarchy..."
             )
-        data_path = Path("./output/")
+        data_path = Path(folder)
         output_folder = f"./output/{data_path.stem}"
 
         if compressed:

src/s2.py

@@ -144,20 +144,6 @@ def get_vertex_polygon(cell: S2Cell) -> Polygon:
         nargs="?",
         default="ttl",
     )
-    parser.add_argument(
-        "--ni",
-        help="When used, s2 integration is disabled",
-        nargs="?",
-        const=1,
-        type=int,
-    )
-    parser.add_argument(
-        "--compressed",
-        help="use the S2 hierarchy to write a compressed collection of relations at various levels",
-        type=bool,
-        nargs="?",
-        default=True,
-    )
     args = parser.parse_args()
 
     level = args.level
@@ -190,7 +176,3 @@ def get_vertex_polygon(cell: S2Cell) -> Polygon:
         )
         cell_id_integers = [parent.id() for parent in parents]
         level -= 1
-
-    # Handle integration
-    if not args.ni:
-        Integrator(args.compressed)

tests/test_kwg_ont.py

@@ -12,5 +12,6 @@ def test_generate_cell_iri():
     """
     cell_id_int = 288230376151711744
     cell_id = S2CellId(cell_id_int)
-    assert generate_cell_iri(cell_id) == URIRef("http://stko-kwg.geog.ucsb.edu/lod/resource/s2.level1"
-                                                ".288230376151711744")
+    assert generate_cell_iri(cell_id) == URIRef(
+        "http://stko-kwg.geog.ucsb.edu/lod/resource/s2.level1" ".288230376151711744"
+    )