fix(parser,vectorstore): stop the first-request downloads in a warmed install

Three things still reached the network from a container whose models were
baked in:

- tiktoken fetched cl100k_base from openaipublic.blob.core.windows.net on
  every fresh container (its cache defaulted to /tmp), and token accounting
  calls it on every chat. prefetch_models now warms it too; the image sets
  TIKTOKEN_CACHE_DIR.
- The chunker loaded its tokenizer with Tokenizer.from_pretrained, which
  revalidates the revision with a HEAD request per process start and stalls
  for the etag timeout (10 s) when huggingface.co is unreachable. It now reads
  tokenizer.json from the hub cache first and only downloads on a miss; the
  repo-metadata read for models outside the registry does the same.
- tldextract fetched the public suffix list on the first web crawl; the
  bundled snapshot is used instead.

application/scripts/verify_offline.py exercises these paths (and docling's
conversion when the extra is installed) so an image can be checked with
docker run --network none.

tests/parser/test_tokenization.py

@@ -1,6 +1,9 @@
 """Chunk sizes must be counted in the embedding model's units, and splitting
 must never rewrite the text it splits."""
 
+import sys
+import types
+
 import pytest
 
 from application.parser import tokenization
@@ -298,3 +301,35 @@ class TestUnknownTokenCollapse:
         assert "".join(pieces) == text, "split must not lose or alter text"
         assert len(pieces) > 1, "a collapsed run must still be cut into pieces"
         assert all(counter.count(p) <= 20 for p in pieces)
+
+
+class TestTokenizerFile:
+    """The chunker's tokenizer must come from the hub cache without a network round trip."""
+
+    def test_cache_hit_makes_no_online_call(self, monkeypatch):
+        calls = []
+
+        def fake_download(repo, filename, local_files_only=False):
+            calls.append(local_files_only)
+            return "/cache/tokenizer.json"
+
+        fake_hub = types.ModuleType("huggingface_hub")
+        fake_hub.hf_hub_download = fake_download
+        monkeypatch.setitem(sys.modules, "huggingface_hub", fake_hub)
+        assert tokenization._tokenizer_file("org/model") == "/cache/tokenizer.json"
+        assert calls == [True]
+
+    def test_cache_miss_falls_back_to_online(self, monkeypatch):
+        calls = []
+
+        def fake_download(repo, filename, local_files_only=False):
+            calls.append(local_files_only)
+            if local_files_only:
+                raise FileNotFoundError("not cached")
+            return "/downloaded/tokenizer.json"
+
+        fake_hub = types.ModuleType("huggingface_hub")
+        fake_hub.hf_hub_download = fake_download
+        monkeypatch.setitem(sys.modules, "huggingface_hub", fake_hub)
+        assert tokenization._tokenizer_file("org/model") == "/downloaded/tokenizer.json"
+        assert calls == [True, False]