Move B105/B107 from global skips to location-specific nosec comments

Author: Copilot

pyproject.toml

@@ -500,11 +500,9 @@ ignore_missing_imports = true
 
 [tool.bandit]
 # Skip tests that produce known false positives or are accepted risks.
-# B105, B107: String/default value detected as hardcoded password.
-#             False positives: tokenizer special tokens and numeric config values.
 # B110: Try/Except/Pass. Accepted pattern in optional-dependency loading.
-# B202, B310, B311, B614: suppressed at each call site with # nosec comments.
-skips = ["B105", "B107", "B110"]
+# B105, B107, B202, B310, B311, B614: suppressed at each call site with # nosec comments.
+skips = ["B110"]
 
 [tool.pylint.main]
 disable = [

pythainlp/generate/thai2fit.py

@@ -97,7 +97,7 @@
     "emb_sz": 400,
     "n_hid": 1550,
     "n_layers": 4,
-    "pad_token": 1,
+    "pad_token": 1,  # nosec B105
     "qrnn": False,
     "tie_weights": True,
     "out_bias": True,

pythainlp/phayathaibert/core.py

@@ -120,7 +120,7 @@ def rm_useless_spaces(self, text: str) -> str:
         """
         return re.sub(" {2,}", " ", text)
 
-    def replace_spaces(self, text: str, space_token: str = "<_>") -> str:  # noqa: S107
+    def replace_spaces(self, text: str, space_token: str = "<_>") -> str:  # noqa: S107  # nosec B107
         """Replace spaces with _
         :param str text: text to replace spaces
         :return: text where all spaces replaced with _

pythainlp/tokenize/_utils.py

@@ -56,7 +56,7 @@ def rejoin_formatted_num(segments: list[str]) -> list[str]:
         is_span_beginning = pos >= match.start()
         token = segments[segment_idx]
         if is_span_beginning:
-            connected_token = ""
+            connected_token = ""  # nosec B105
             while pos < match.end() and segment_idx < len(segments):
                 connected_token += segments[segment_idx]
                 pos += len(segments[segment_idx])

pythainlp/translate/tokenization_small100.py

@@ -149,11 +149,11 @@ def __init__(
         vocab_file: str,
         spm_file: str,
         tgt_lang: Optional[str] = None,
-        bos_token: str = "<s>",  # noqa: S107
-        eos_token: str = "</s>",  # noqa: S107
-        sep_token: str = "</s>",  # noqa: S107
-        pad_token: str = "<pad>",  # noqa: S107
-        unk_token: str = "<unk>",  # noqa: S107
+        bos_token: str = "<s>",  # noqa: S107  # nosec B107
+        eos_token: str = "</s>",  # noqa: S107  # nosec B107
+        sep_token: str = "</s>",  # noqa: S107  # nosec B107
+        pad_token: str = "<pad>",  # noqa: S107  # nosec B107
+        unk_token: str = "<unk>",  # noqa: S107  # nosec B107
         language_codes: str = "m2m100",
         sp_model_kwargs: Optional[dict[str, str]] = None,
         num_madeup_words: int = 8,