Update keywords for all archives

catalyst-cooperative · Sep 11, 2023 · 513369f · 513369f
1 parent 86ece0c
commit 513369f
Show file tree

Hide file tree

Showing 2 changed files with 45 additions and 2 deletions.
diff --git a/src/pudl/metadata/constants.py b/src/pudl/metadata/constants.py
@@ -259,6 +259,34 @@
         "government",
         "federal",
     ],
+    "msha": [
+        "msha",
+        "mshamines",
+        "mine safety and health administration",
+        "mines",
+        "mining",
+        "coal",
+        "metal",
+    ],
+    "phmsa": [
+        "phmsa",
+        "phmsagas",
+        "pipelines and hazardous materials safety administration",
+        "pipelines",
+        "natural gas",
+        "transmission",
+        "distribution",
+        "gathering",
+        "liquified natural gas",
+        "lng",
+        "underground natural gas storage",
+    ],
+    "eia_water": [
+        "eia thermoelectric cooling water",
+        "eia waterthermoelectric",
+        "cooling water",
+        "water usage",
+    ],
 }
 
 XBRL_TABLES = [

diff --git a/src/pudl/metadata/sources.py b/src/pudl/metadata/sources.py
@@ -50,11 +50,10 @@
                 [
                     "eia176",
                     "form 176",
+                    "natural gas",
                 ]
                 + KEYWORDS["eia"]
                 + KEYWORDS["us_govt"]
-                + KEYWORDS["fuels"]
-                + KEYWORDS["environment"]
             )
         ),
         "license_raw": LICENSES["us-govt"],
@@ -262,6 +261,9 @@
             "Monthly cooling water usage by generator and boiler. Data "
             "collected in conjunction with the EIA-860 and EIA-923."
         ),
+        "keywords": sorted(
+            set(KEYWORDS["eia"] + KEYWORDS["us_govt"] + KEYWORDS["eia_water"])
+        ),
         "license_raw": LICENSES["us-govt"],
         "license_pudl": LICENSES["cc-by-4.0"],
     },
@@ -523,6 +525,17 @@
             "as well as transaction information for short-term and long-term "
             "market-based power sales and cost-based power sales."
         ),
+        "keywords": sorted(
+            set(
+                [
+                    "ferceqr",
+                    "electric quarterly report",
+                ]
+                + KEYWORDS["ferc"]
+                + KEYWORDS["us_govt"]
+                + KEYWORDS["electricity"]
+            )
+        ),
         "license_raw": LICENSES["us-govt"],
         "license_pudl": LICENSES["cc-by-4.0"],
     },
@@ -535,6 +548,7 @@
             "mine (Active, Abandoned, NonProducing, etc.), the current owner and "
             "operating company, commodity codes and physical attributes of the mine."
         ),
+        "keywords": sorted(set(KEYWORDS["msha"] + KEYWORDS["us_govt"])),
         "license_raw": LICENSES["us-govt"],
         "license_pudl": LICENSES["cc-by-4.0"],
     },
@@ -548,6 +562,7 @@
             "pipeline mileage, facilities, commodities transported, miles by material, "
             "and installation dates."
         ),
+        "keywords": sorted(set(KEYWORDS["phmsa"] + KEYWORDS["us_govt"])),
         "license_raw": LICENSES["us-govt"],
         "license_pudl": LICENSES["cc-by-4.0"],
     },