opendataloader-pdf 0.0.0__py3-none-any.whl → 0.0.5__py3-none-any.whl

opendataloader_pdf/wrapper.py

@@ -9,7 +9,7 @@ _JAR_NAME = "opendataloader-pdf-cli.jar"
 
 def run(
     input_path: str,
-    output_folder: str,
+    output_folder: str = None,
     password: str = None,
     to_markdown: bool = False,
     to_annotated_pdf: bool = False,

{opendataloader_pdf-0.0.0.dist-info → opendataloader_pdf-0.0.5.dist-info}/METADATA

@@ -1,91 +1,67 @@
-Metadata-Version: 2.4
-Name: opendataloader-pdf
-Version: 0.0.0
-Summary: A Python wrapper for the opendataloader-pdf Java CLI.
-Home-page: https://github.com/opendataloader-project/opendataloader-pdf
-Author: opendataloader-project
-Author-email: open.dataloader@hancom.com
-License: MPL-2.0
-Classifier: Programming Language :: Python :: 3
-Classifier: Operating System :: OS Independent
-Requires-Python: >=3.7
-Description-Content-Type: text/markdown
-Dynamic: author
-Dynamic: author-email
-Dynamic: classifier
-Dynamic: description
-Dynamic: description-content-type
-Dynamic: home-page
-Dynamic: license
-Dynamic: requires-python
-Dynamic: summary
-
-# Opendataloader PDF Python Wrapper
-
-This package is a Python wrapper for the `opendataloader-pdf` Java command-line tool.
-
-It allows you to process PDF files and convert them to JSON or Markdown format directly from Python.
-
-## Prerequisites
-
-- Java 11 or higher must be installed and available in your system's PATH.
-
-## Installation
-
-```bash
-pip install opendataloader-pdf
-```
-
-## Usage
-
-Here is a basic example of how to use the wrapper:
-
-```python
-import opendataloader_pdf
-import json
-
-try:
-    # Process a PDF and get the structured JSON output from stdout
-    json_output_str = opendataloader_pdf.run(input_path="path/to/your/document.pdf")
-
-    # The output is a JSON string, so you can parse it
-    if json_output_str:
-        data = json.loads(json_output_str)
-        print(f"Successfully processed {data['file name']}.")
-
-    # Example: Generate Markdown and annotated PDF in a specific folder
-    opendataloader_pdf.run(
-        input_path="path/to/your/document.pdf",
-        output_folder="path/to/output",
-        generate_markdown=True,
-        generate_pdf=True,
-        keep_line_breaks=True
-    )
-    print("Successfully generated Markdown and annotated PDF.")
-
-except FileNotFoundError:
-    print("Error: Input file not found or Java is not installed correctly.")
-except Exception as e:
-    print(f"An error occurred: {e}")
-```
-
-## Function: `run()`
-
-The main function to process PDFs.
-
-### Parameters
-
-- `input_path` (str): Path to the input PDF file or folder. **(Required)**
-- `output_folder` (str, optional): Path to the output folder. Defaults to the input folder.
-- `password` (str, optional): Password for the PDF file.
-- `generate_markdown` (bool, optional): If `True`, generates a Markdown output file. Defaults to `False`.
-- `generate_pdf` (bool, optional): If `True`, generates an annotated PDF output file. Defaults to `False`.
-- `keep_line_breaks` (bool, optional): If `True`, keeps line breaks in the output. Defaults to `False`.
-- `find_hidden_text` (bool, optional): If `True`, finds hidden text in the PDF. Defaults to `False`.
-- `html_in_markdown` (bool, optional): If `True`, uses HTML in the Markdown output. Defaults to `False`.
-- `add_image_to_markdown` (bool, optional): If `True`, adds images to the Markdown output. Defaults to `False`.
-- `debug` (bool, optional): If `True`, prints all messages from the CLI to the console during execution. Defaults to `False`.
-
-### Returns
-
-- `str`: The standard output from the CLI tool. When processing a single file without other output formats specified, this will be the JSON content. When `debug=True`, this will be the combined stdout and stderr from the CLI tool.
+Metadata-Version: 2.4
+Name: opendataloader-pdf
+Version: 0.0.5
+Summary: A Python wrapper for the opendataloader-pdf Java CLI.
+Home-page: https://github.com/opendataloader-project/opendataloader-pdf
+Author: opendataloader-project
+Author-email: open.dataloader@hancom.com
+License: MPL-2.0
+Classifier: Programming Language :: Python :: 3
+Classifier: Operating System :: OS Independent
+Requires-Python: >=3.7
+Description-Content-Type: text/markdown
+Dynamic: author
+Dynamic: author-email
+Dynamic: classifier
+Dynamic: description
+Dynamic: description-content-type
+Dynamic: home-page
+Dynamic: license
+Dynamic: requires-python
+Dynamic: summary
+
+# Opendataloader PDF Python Wrapper
+
+This package is a Python wrapper for the `opendataloader-pdf` Java command-line tool.
+
+It allows you to process PDF files and convert them to JSON or Markdown format directly from Python.
+
+## Prerequisites
+
+- Java 11 or higher must be installed and available in your system's PATH.
+
+## Installation
+
+```bash
+pip install opendataloader-pdf
+```
+
+## Usage
+
+Here is a basic example of how to use:
+
+```python
+import opendataloader_pdf
+
+opendataloader_pdf.run("path/to/document.pdf", to_markdown=True)
+
+# If you don’t specify an output_folder,
+# the output data will be saved in the same directory as the input document.
+```
+
+## Function: `run()`
+
+The main function to process PDFs.
+
+### Parameters
+
+- `input_path` (str): Path to the input PDF file or folder. **(Required)**
+- `output_folder` (str, optional): Path to the output folder. Defaults to the input folder.
+- `password` (str, optional): Password for the PDF file.
+- `to_markdown` (bool, optional): If `True`, generates a Markdown output file. Defaults to `False`.
+- `to_annotated_pdf` (bool, optional): If `True`, generates an annotated PDF output file. Defaults to `False`.
+- `keep_line_breaks` (bool, optional): If `True`, keeps line breaks in the output. Defaults to `False`.
+- `find_hidden_text` (bool, optional): If `True`, finds hidden text in the PDF. Defaults to `False`.
+- `html_in_markdown` (bool, optional): If `True`, uses HTML in the Markdown output. Defaults to `False`.
+- `add_image_to_markdown` (bool, optional): If `True`, adds images to the Markdown output. Defaults to `False`.
+- `debug` (bool, optional): If `True`, prints all messages from the CLI to the console during execution. Defaults to `False`.

{opendataloader_pdf-0.0.0.dist-info → opendataloader_pdf-0.0.5.dist-info}/RECORD

@@ -1,7 +1,7 @@
 opendataloader_pdf/LICENSE,sha256=rxdbnZbuk8IaA2FS4bkFsLlTBNSujCySHHYJEAuo334,15921
 opendataloader_pdf/NOTICE.md,sha256=Uxc6sEbVz2hfsDinzzSNMtmsjx9HsQUod0yy0cswUwg,562
 opendataloader_pdf/__init__.py,sha256=T5RV-dcgjNCm8klNy_EH-IgOeodcPg6Yc34HHXtuAmQ,44
-opendataloader_pdf/wrapper.py,sha256=ahbxo7YYunsNt3w66eXn_oa0XKy1LLfoA5GQcSuMvvM,4407
+opendataloader_pdf/wrapper.py,sha256=c9UnFHw0PaCd_FEwFasoxOZVuGM_owTMnrkNmPSVfH4,4414
 opendataloader_pdf/THIRD_PARTY/THIRD_PARTY_LICENSES.md,sha256=QRYYiXFS2zBDGdmWRo_SrRfGhrdRBwhiRo1SdUKfrQo,11235
 opendataloader_pdf/THIRD_PARTY/THIRD_PARTY_NOTICES.md,sha256=pB2ZitFM1u0x3rIDpMHsLxOe4OFNCZRqkzeR-bfpFzE,8911
 opendataloader_pdf/THIRD_PARTY/licenses/Apache-2.0.txt,sha256=z8d0m5b2O9McPEK1xHG_dWgUBT6EfBDz6wA0F7xSPTA,11358
@@ -13,8 +13,8 @@ opendataloader_pdf/THIRD_PARTY/licenses/LICENSE-JJ2000.txt,sha256=itSesIy3XiNWgJ
 opendataloader_pdf/THIRD_PARTY/licenses/MIT.txt,sha256=JPCdbR3BU0uO_KypOd3sGWnKwlVHGq4l0pmrjoGtop8,1078
 opendataloader_pdf/THIRD_PARTY/licenses/MPL-2.0.txt,sha256=CGF6Fx5WV7DJmRZJ8_6w6JEt2N9bu4p6zDo18fTHHRw,15818
 opendataloader_pdf/THIRD_PARTY/licenses/Plexus Classworlds License.txt,sha256=ZQuKXwVz4FeC34ApB20vYg8kPTwgIUKRzEk5ew74-hU,1937
-opendataloader_pdf/jar/opendataloader-pdf-cli.jar,sha256=jUVcuDXxNBFweOQbusLebCX1AZ3l3ssns4nJtSWsRTI,22116373
-opendataloader_pdf-0.0.0.dist-info/METADATA,sha256=y-T7UA9NodfcETSd2tX5FJdXMNArmpgXZG22wINIHRo,3382
-opendataloader_pdf-0.0.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-opendataloader_pdf-0.0.0.dist-info/top_level.txt,sha256=xee0qFQd6HPfS50E2NLICGuR6cq9C9At5SJ81yv5HkY,19
-opendataloader_pdf-0.0.0.dist-info/RECORD,,
+opendataloader_pdf/jar/opendataloader-pdf-cli.jar,sha256=m4pZYjMlwvnDEQ4-K_FYaIzSSYCTFuD68hc7hGY_nhk,22114429
+opendataloader_pdf-0.0.5.dist-info/METADATA,sha256=A3R7j4EZPgVFlfNvsFtVUjWqkir0z59H3i6L1TdI444,2352
+opendataloader_pdf-0.0.5.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+opendataloader_pdf-0.0.5.dist-info/top_level.txt,sha256=xee0qFQd6HPfS50E2NLICGuR6cq9C9At5SJ81yv5HkY,19
+opendataloader_pdf-0.0.5.dist-info/RECORD,,