HowdenParser 3.1.0__tar.gz → 4.0.1__tar.gz

{howdenparser-3.1.0 → howdenparser-4.0.1}/HowdenParser/parser.py

@@ -62,7 +62,9 @@ class Parser(BaseParser):
             result_type: str,
             mode: bool,
             preserve_layout_alignment_across_pages: bool,
-            merge_tables_across_pages_in_markdown: bool
+            merge_tables_across_pages_in_markdown: bool,
+            hide_footers: bool,
+            hide_headers: bool,
     ) -> "LlamaParser": ...
 
     @overload
@@ -237,13 +239,18 @@ class LlamaParser(BaseParser, name="llamaparser"):
                  mode: bool,
                  provider_and_model: str,
                  merge_tables_across_pages_in_markdown: bool,
-                 preserve_layout_alignment_across_pages: bool) -> None:
+                 preserve_layout_alignment_across_pages: bool,
+                 hide_footers: bool,
+                 hide_headers: bool
+                 ) -> None:
         logging.info("Initializing LlamaParser...")
         self.result_type = result_type
         self.mode = mode
         self.provider_and_model = provider_and_model
         self.merge_tables_across_pages_in_markdown = merge_tables_across_pages_in_markdown
         self.preserve_layout_alignment_across_pages = preserve_layout_alignment_across_pages
+        self.hide_footers=hide_footers
+        self.hide_headers=hide_headers
 
         from llama_parse import LlamaParse, ResultType
         if result_type.lower() in ("md", "markdown"):
@@ -255,7 +262,9 @@ class LlamaParser(BaseParser, name="llamaparser"):
                                   result_type=result_type,
                                   premium_mode=self.mode,
                                   merge_tables_across_pages_in_markdown=self.merge_tables_across_pages_in_markdown,
-                                  preserve_layout_alignment_across_pages=self.preserve_layout_alignment_across_pages)
+                                  preserve_layout_alignment_across_pages=self.preserve_layout_alignment_across_pages,
+                                  hide_footers=self.hide_footers,
+                                  hide_headers=self.hide_headers)
 
     def parse(self, file_path: Path) -> str:
         documents = self._parser.load_data(str(file_path))

{howdenparser-3.1.0 → howdenparser-4.0.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: HowdenParser
-Version: 3.1.0
+Version: 4.0.1
 Summary: A simple configuration manager with Pydantic and JSON export.
 License: MIT
 Keywords: config,configuration,pydantic,json
@@ -11,6 +11,7 @@ Classifier: License :: OSI Approved :: MIT License
 Classifier: Programming Language :: Python :: 3
 Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3.13
+Provides-Extra: dev
 Requires-Dist: fitz (>=0.0.1.dev2,<0.0.2)
 Requires-Dist: hf-xet (>=1.1.7,<2.0.0)
 Requires-Dist: langchain (>=0.3.27,<0.4.0)
@@ -18,13 +19,11 @@ Requires-Dist: llama-parse (>=0.6.58,<0.7.0)
 Requires-Dist: mistralai (>=1.9.3,<2.0.0)
 Requires-Dist: pdf2image (>=1.17.0,<2.0.0)
 Requires-Dist: pypdf2 (>=3.0.1,<4.0.0)
+Requires-Dist: toml (>=0.10.2,<0.11.0) ; extra == "dev"
 Requires-Dist: torch (>=2.8.0,<3.0.0)
 Requires-Dist: torchaudio (>=2.8.0,<3.0.0)
 Requires-Dist: torchvision (>=0.23.0,<0.24.0)
 Requires-Dist: transformers (>=4.55.2,<5.0.0)
-Project-URL: Documentation, https://github.com/yourusername/config
-Project-URL: Homepage, https://github.com/yourusername/config
-Project-URL: Repository, https://github.com/yourusername/config
 Description-Content-Type: text/markdown
 
 # OCR & LLM Parser

howdenparser-4.0.1/pyproject.toml

@@ -0,0 +1,48 @@
+[project]
+name = "HowdenParser"
+version = "4.0.1"
+description = "A simple configuration manager with Pydantic and JSON export."
+readme = "README.md"
+requires-python = ">=3.12,<3.14"
+authors = [
+    { name = "JesperThoftIllemannJ", email = "jesper.jaeger@howdendanmark.dk" },
+]
+keywords = [
+    "config",
+    "configuration",
+    "pydantic",
+    "json",
+]
+homepage = "https://github.com/yourusername/config"
+repository = "https://github.com/yourusername/config"
+documentation = "https://github.com/yourusername/config"
+dependencies = [
+    "mistralai>=1.9.3,<2.0.0",
+    "llama-parse>=0.6.58,<0.7.0",
+    "langchain>=0.3.27,<0.4.0",
+    "transformers>=4.55.2,<5.0.0",
+    "fitz>=0.0.1.dev2,<0.0.2",
+    "pypdf2>=3.0.1,<4.0.0",
+    "pdf2image>=1.17.0,<2.0.0",
+    "torch>=2.8.0,<3.0.0",
+    "torchvision>=0.23.0,<0.24.0",
+    "torchaudio>=2.8.0,<3.0.0",
+    "hf-xet>=1.1.7,<2.0.0",
+]
+
+[project.license]
+text = "MIT"
+
+[project.optional-dependencies]
+dev = [
+    "toml>=0.10.2,<0.11.0",
+]
+
+[build-system]
+requires = [
+    "poetry-core>=2.0.0,<3.0.0",
+]
+build-backend = "poetry.core.masonry.api"
+
+[tool.poetry.group.dev.dependencies]
+tomli-w = "^1.2.0"

howdenparser-3.1.0/pyproject.toml

@@ -1,39 +0,0 @@
-[project]
-name = "HowdenParser"
-description = ""
-readme = "README.md"
-requires-python = ">=3.12,<3.14"
-dependencies = [ "mistralai (>=1.9.3,<2.0.0)", "llama-parse (>=0.6.58,<0.7.0)", "langchain (>=0.3.27,<0.4.0)", "transformers (>=4.55.2,<5.0.0)", "fitz (>=0.0.1.dev2,<0.0.2)", "pypdf2 (>=3.0.1,<4.0.0)", "pdf2image (>=1.17.0,<2.0.0)", "torch (>=2.8.0,<3.0.0)", "torchvision (>=0.23.0,<0.24.0)", "torchaudio (>=2.8.0,<3.0.0)", "hf-xet (>=1.1.7,<2.0.0)",]
-[[project.authors]]
-name = "JesperThoftIllemannJ"
-email = "jesper.jaeger@howdendanmark.dk"
-
-[build-system]
-requires = [ "poetry-core>=2.0.0,<3.0.0",]
-build-backend = "poetry.core.masonry.api"
-
-[tool.poetry]
-name = "HowdenParser"
-version = "3.1.0"
-description = "A simple configuration manager with Pydantic and JSON export."
-authors = [ "JesperThoftIllemannJ <jesper.jaeger@howdendanmark.dk>",]
-readme = "README.md"
-license = "MIT"
-keywords = [ "config", "configuration", "pydantic", "json",]
-homepage = "https://github.com/yourusername/config"
-repository = "https://github.com/yourusername/config"
-documentation = "https://github.com/yourusername/config"
-[[tool.poetry.packages]]
-include = "HowdenParser"
-
-[tool.poetry.dependencies.torch]
-source = "pypi"
-
-[tool.poetry.dependencies.torchvision]
-source = "pypi"
-
-[tool.poetry.dependencies.torchaudio]
-source = "pypi"
-
-[tool.poetry.group.dev.dependencies]
-toml = "^0.10.2"