Add script to generate license summary

feeds/de-berlin.json

@@ -9,7 +9,11 @@
         {
             "name": "vbb",
             "type": "http",
-            "url": "https://www.vbb.de/fileadmin/user_upload/VBB/Dokumente/API-Datensaetze/gtfs-mastscharf/GTFS.zip"
+            "url": "https://www.vbb.de/fileadmin/user_upload/VBB/Dokumente/API-Datensaetze/gtfs-mastscharf/GTFS.zip",
+            "license": {
+                "spdx-identifier": "CC-BY-4.0",
+                "url": "https://www.vbb.de/vbb-services/api-open-data/datensaetze/"
+            }
         }
     ]
 }

feeds/lv.json

@@ -14,18 +14,27 @@
         {
             "name": "rigas-satiksme",
             "type": "http",
-            "url": "https://gtfs.pro/files/uran/improved-gtfs-satiksme.zip"
+            "url": "https://gtfs.pro/files/uran/improved-gtfs-satiksme.zip",
+            "license": {
+                "spdx-identifier": "CC0-1.0"
+            }
         },
         {
             "name": "rigas-saraksti",
             "type": "http",
-            "url": "https://gtfs.pro/files/uran/improved-gtfs-saraksti.zip"
+            "url": "https://gtfs.pro/files/uran/improved-gtfs-saraksti.zip",
+            "license": {
+                "spdx-identifier": "CC0-1.0"
+            }
         },
         {
             "name": "valsts-sia-autotransporta",
             "type": "http",
             "url": "https://www.atd.lv/sites/default/files/GTFS/gtfs-latvia-lv.zip",
-            "fix": true
+            "fix": true,
+            "license": {
+                "spdx-identifier": "CC0-1.0"
+            }
         }
     ]
 }

src/generate-attribution.py

@@ -0,0 +1,67 @@
+#!/usr/bin/env python3
+# SPDX-FileCopyrightText: 2024 Jonah Brüchert <[email protected]>
+#
+# SPDX-License-Identifier: AGPL-3.0-or-later
+
+import json
+import csv
+import io
+import transitland
+
+from pathlib import Path
+from metadata import *
+from zipfile import ZipFile
+
+
+if __name__ == "__main__":
+    feed_dir = Path("feeds/")
+
+    transitland_atlas = transitland.Atlas.load(Path("transitland-atlas/"))
+
+    attributions = []
+
+    for feed in feed_dir.glob("*.json"):
+        parsed = {}
+        with open(feed, "r") as f:
+            parsed = json.load(f)
+
+        region = Region(parsed)
+        for source in region.sources:
+            if (type(source) == TransitlandSource):
+                source = transitland_atlas.source_by_id(source)
+
+            attribution = {}
+
+            if source.license:
+                if source.license.spdx_identifier:
+                    attribution["spdx_identifier"] \
+                        = source.license.spdx_identifier
+                if source.license.url:
+                    attribution["url"] = source.license.url
+
+            attribution["copyright_holders"] = []
+
+            metadata_filename = feed.name
+            region_name = metadata_filename[:metadata_filename.rfind('.')]
+
+            feed_path = Path(f"out/{region_name}_{source.name}.gtfs.zip")
+
+            attribution["filename"] = feed_path.name
+
+            if not feed_path.exists():
+                print(f"Info: {feed_path} does not exist, skipping…")
+                continue
+
+            with ZipFile(feed_path) as z:
+                with z.open("agency.txt", "r") as a:
+                    with io.TextIOWrapper(a) as at:
+                        agencyreader = \
+                            csv.DictReader(at, delimiter=',', quotechar='"')
+                        for row in agencyreader:
+                            attribution["copyright_holders"] \
+                                .append(row["agency_name"])
+
+            attributions.append(attribution)
+
+    with open("out/license.json", "w") as outfile:
+        json.dump(attributions, outfile, indent=4, ensure_ascii=False)

src/metadata.py

@@ -2,7 +2,7 @@
 #
 # SPDX-License-Identifier: AGPL-3.0-or-later
 
-from typing import List
+from typing import List, Optional
 
 
 class Maintainer:
@@ -14,12 +14,25 @@ def __init__(self, parsed: dict):
         self.email = parsed["email"]
 
 
+class License:
+    spdx_identifier: Optional[str] = None
+    url: Optional[str] = None
+
+
 class Source:
     name: str
     fix: bool = False
+    license: License
 
     def __init__(self, parsed: dict = None):
+        self.license = License()
         if parsed:
+            if "license" in parsed:
+                if "spdx-identifier" in parsed["license"]:
+                    self.license.spdx_identifier = parsed["license"]["spdx-identifier"]
+                if "url" in parsed["license"]:
+                    self.license.url = parsed["license"]["url"]
+
             self.name = parsed["name"]
             if "fix" in parsed:
                 self.fix = bool(parsed["fix"])

src/transitland.py

@@ -26,5 +26,14 @@ def load(path: Path):
     def source_by_id(self, source: TransitlandSource) -> HttpSource:
         feed = self.by_id[source.transitland_atlas_id]
         http_source = HttpSource()
+        http_source.name = source.name
         http_source.url = feed["urls"]["static_current"]
+
+        if "license" in feed:
+            http_source.license = License()
+            if "spdx_identifier" in feed["license"]:
+                http_source.license.spdx_identifier = feed["license"]["spdx_identifier"]
+            if "url" in feed["license"]:
+                http_source.license.url = feed["license"]["url"]
+
         return http_source