feat: cache IANA TLDs for faster lookups

src/validators/domain.py

@@ -1,20 +1,37 @@
 """Domain."""
 
 # standard
+import os
 from pathlib import Path
 import re
+from typing import Generator, Optional, Set, Union
 
 # local
 from .utils import validator
 
 
-def _iana_tld():
-    """Load IANA TLDs as a Generator."""
-    # source: https://data.iana.org/TLD/tlds-alpha-by-domain.txt
-    with Path(__file__).parent.joinpath("_tld.txt").open() as tld_f:
-        _ = next(tld_f)  # ignore the first line
-        for line in tld_f:
-            yield line.strip()
+class _TLDList:
+    """Read IANA TLDs, and optionally cache them."""
+
+    cache: Optional[Set[str]] = None
+
+    @classmethod
+    def read_tlds_from_file(cls) -> Generator[str, None, None]:
+        # Try the most common TLDs before opening the file
+        yield from ("COM", "ORG", "RU", "DE", "NET", "BR", "UK", "JP", "FR", "IT")
+        with Path(__file__).parent.joinpath("_tld.txt").open() as tld_f:
+            _ = next(tld_f)  # ignore the first line
+            for line in tld_f:
+                yield line.strip()
+
+    @classmethod
+    def tlds(cls) -> Union[Set[str], Generator[str, None, None]]:
+        if not cls.cache and os.environ.get("PYVLD_LOAD_TLD_TO_MEMORY") == "True":
+            cls.cache = set(_TLDList.read_tlds_from_file())
+        if cls.cache:
+            return cls.cache
+
+        return cls.read_tlds_from_file()
 
 
 @validator
@@ -56,7 +73,7 @@ def domain(
     if not value:
         return False
 
-    if consider_tld and value.rstrip(".").rsplit(".", 1)[-1].upper() not in _iana_tld():
+    if consider_tld and value.rstrip(".").rsplit(".", 1)[-1].upper() not in _TLDList.tlds():
         return False
 
     try: