WIP: dhatu_metadata

vidyut-data/scripts/.python-version

@@ -0,0 +1 @@
+3.13

vidyut-data/scripts/fetch_dhatu_metadata.py

@@ -0,0 +1,41 @@
+"""Creates dhatu metadata based on data from ashtadhyayi.com.
+
+Usage:
+
+    uv run fetch_dhatu_metadata.py
+"""
+
+
+import json
+import pprint
+import urllib.request
+from vidyut.lipi import transliterate, Scheme
+
+
+def load_metadata() -> dict:
+    url = "https://github.com/ashtadhyayi-com/data/raw/refs/heads/master/dhatu/data.txt"
+    f = urllib.request.urlopen(url)
+    return json.load(f)
+
+
+data = load_metadata()
+dhatus = data["data"]
+
+for dhatu in dhatus:
+    artha_en = dhatu["artha_english"]
+    artha_hi = dhatu["artha_hindi"]
+    code = dhatu["baseindex"]
+    karma = dhatu["karma"]
+    pada = dhatu["pada"]
+    settva = dhatu["settva"]
+
+    assert karma in {"S", "A", "D", '-'}, karma
+    assert pada in {"P", "A", "U", '-'}, pada
+    assert settva in {"S", "A", "V", '-'}, settva
+
+    if karma == '-':
+        assert karma == pada == settva == '-'
+        pprint.pprint(dhatu)
+        continue
+
+

vidyut-data/scripts/pyproject.toml

@@ -0,0 +1,9 @@
+[project]
+name = "scripts"
+version = "0.1.0"
+description = "Add your description here"
+readme = "README.md"
+requires-python = ">=3.13"
+dependencies = [
+    "vidyut>=0.3.1",
+]