refactor: use Python's native string.Template formatting

Require successfully reading all the QubesDB keys found by string.Template.get_identifiers(), rather than requiring an explicit list of keys via "--key".

files/template-from-qubesdb.py

@@ -1,7 +1,6 @@
 #!/usr/bin/python3
 import argparse
-import subprocess
-from jinja2 import Template
+import string
 
 from qubesdb import QubesDB
 
@@ -14,6 +13,7 @@ def get_env(args):
             value = db.read(f"/vm-config/{key}")
             if not value or len(value) == 0:
                 raise KeyError(f"Could not read from QubesDB: {key}")
+            env[key] = (value or "").decode()
 
     finally:
         db.close()
@@ -23,17 +23,20 @@ def get_env(args):
 
 if __name__ == "__main__":
     parser = argparse.ArgumentParser(
-        description="Use -k/--key one or more times to read the specified key from QubesDB.  The input template will be written to the output file with those variables substituted."
+        description="The input template will be written to the output file, with each variable "
+        "replaced by its value from QubesDB.  If any variable cannot be read from QubesDB, "
+        "templating will fail.  See the documentation on Python's string.Template for details "
+        "(https://docs.python.org/3/library/string.html#template-strings)",
     )
-    parser.add_argument("-k", "--key", action="append")
     parser.add_argument("input")
     parser.add_argument("output")
     args = parser.parse_args()
-    env = get_env(args.key)
 
     with open(args.input) as input_f:
-        template = Template(input_f.read())
+        template = string.Template(input_f.read())
+
+    env = get_env(template.get_identifiers())  # type: ignore[attr-defined]
 
     with open(args.output, "w") as output_f:
-        rendered = template.render(**env)
+        rendered = template.substitute(env)
         output_f.write(rendered)