HowdenParser 5.1.0__tar.gz → 5.2.0__tar.gz

{howdenparser-5.1.0 → howdenparser-5.2.0}/HowdenParser/parser.py

@@ -3,6 +3,9 @@ import logging
 import dotenv
 from typing import overload, Literal
 from inspect import signature, Signature
+from typing import Any
+import json
+import hashlib
 
 dotenv.load_dotenv()
 logger = logging.getLogger(__name__)
@@ -26,6 +29,30 @@ class BaseParser(ABC):
     def parse(self, text: str, include_pagenumbers: bool = False):
         pass
 
+    def make_serializable(self, obj: Any) -> Any:
+        if obj is None or isinstance(obj, (str, int, float, bool)):
+            return obj
+        if isinstance(obj, (list, tuple, set)):
+            return [self.make_serializable(v) for v in obj]
+        if isinstance(obj, dict):
+            return {k: self.make_serializable(v) for k, v in sorted(obj.items())}
+        if hasattr(obj, "__dict__"):
+            return self.make_serializable(vars(obj))
+        return f"{type(obj).__name__}:{repr(obj)}"
+
+    def compute_hash(self, parameter:dict) -> str:
+
+        attrs = {
+            k: v
+            for k, v in parameter.items()
+            if k != "hash"
+               and not k.startswith("_")
+               and k not in ("client", "provider")
+        }
+        serializable = self.make_serializable(attrs)
+        attrs_str = json.dumps(serializable, sort_keys=True, ensure_ascii=True)
+        hashed = hashlib.sha256(attrs_str.encode()).hexdigest()[:8]
+        return hashed
 
 class Parser(BaseParser):
     """Factory + registry interface for all parsers."""

{howdenparser-5.1.0 → howdenparser-5.2.0}/HowdenParser/parsers/llama_parser.py

@@ -9,6 +9,13 @@ class LlamaParser(BaseParser, name="llamaparser"):
                  hide_footers: bool, hide_headers: bool) -> None:
 
         super().__init__()
+        # Capture only input parameters
+        self._input_params = {
+            k: v
+            for k, v in locals().items()
+            if k not in ("self", "__class__", "__len__")
+        }
+
         logging.info("Configuring LlamaParser (lazy init enabled)...")
 
         # Store configuration only
@@ -20,7 +27,7 @@ class LlamaParser(BaseParser, name="llamaparser"):
         self.preserve_layout_alignment_across_pages = preserve_layout_alignment_across_pages
         self.hide_footers = hide_footers
         self.hide_headers = hide_headers
-
+        self.hashed = self.compute_hash(self._input_params)
         # DO NOT CREATE LlamaParse HERE
         # It creates asyncio objects bound to the wrong loop
         self._parser = None

{howdenparser-5.1.0 → howdenparser-5.2.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: HowdenParser
-Version: 5.1.0
+Version: 5.2.0
 Summary: A simple configuration manager with Pydantic and JSON export.
 License: MIT
 Keywords: config,configuration,pydantic,json

{howdenparser-5.1.0 → howdenparser-5.2.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "HowdenParser"
-version = "5.1.0"
+version = "5.2.0"
 description = "A simple configuration manager with Pydantic and JSON export."
 readme = "README.md"
 requires-python = ">=3.12,<3.14"