pembot 0.1.7__tar.gz → 0.1.9__tar.gz

{pembot-0.1.7 → pembot-0.1.9}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pembot
-Version: 0.1.7
+Version: 0.1.9
 Summary: A Python Package to convert PEM blog content to usseful information by leveraging LLMs
 Author-email: cyto <aryan_sidhwani@protonmail.com>
 License-Expression: MIT

pembot-0.1.9/pembot/.git/COMMIT_EDITMSG

@@ -0,0 +1 @@
+added DeepSeek-OCR as a model option to use spaces

{pembot-0.1.7 → pembot-0.1.9}/pembot/.git/logs/HEAD

@@ -19,3 +19,5 @@ c3cc0da3d955ecec0f865c46c030a0c073697495 e6adbc3c373070269f97ef82d4f63027d7878f6
 e6adbc3c373070269f97ef82d4f63027d7878f67 81d01e1c63d48b096c77aae83471d42272ca9fce cyto <silverstone965@gmail.com> 1758890262 +0530	commit: cyto/fixed a file already open error and wrote to file if myfile is excel type and output_dir is given
 81d01e1c63d48b096c77aae83471d42272ca9fce e89cb4f5af158d26dcff5eed03dba6671a818739 cyto <silverstone965@gmail.com> 1758891589 +0530	commit: cyto/made the excel convertor compatible with suffix input from both system file and with bytes + manual suffix input
 e89cb4f5af158d26dcff5eed03dba6671a818739 e3c62c141fc65ef2be0095c49b23e06263f0b734 cyto <silverstone965@gmail.com> 1758892053 +0530	commit: silly willy mistake
+e3c62c141fc65ef2be0095c49b23e06263f0b734 d440b20aae1265dabbd3ddaafb24c35e40e3ab3c cyto <silverstone965@gmail.com> 1759127664 +0530	commit: cyto/fixed the excel file input bug, in the conversion method; added exceptions where there is invalid input
+d440b20aae1265dabbd3ddaafb24c35e40e3ab3c 8d58ec13be39949ecfe7211b42c56acd2a83dc72 cyto <silverstone965@gmail.com> 1761392079 +0530	commit: added DeepSeek-OCR as a model option to use spaces

{pembot-0.1.7 → pembot-0.1.9}/pembot/.git/logs/refs/heads/main

@@ -19,3 +19,5 @@ c3cc0da3d955ecec0f865c46c030a0c073697495 e6adbc3c373070269f97ef82d4f63027d7878f6
 e6adbc3c373070269f97ef82d4f63027d7878f67 81d01e1c63d48b096c77aae83471d42272ca9fce cyto <silverstone965@gmail.com> 1758890262 +0530	commit: cyto/fixed a file already open error and wrote to file if myfile is excel type and output_dir is given
 81d01e1c63d48b096c77aae83471d42272ca9fce e89cb4f5af158d26dcff5eed03dba6671a818739 cyto <silverstone965@gmail.com> 1758891589 +0530	commit: cyto/made the excel convertor compatible with suffix input from both system file and with bytes + manual suffix input
 e89cb4f5af158d26dcff5eed03dba6671a818739 e3c62c141fc65ef2be0095c49b23e06263f0b734 cyto <silverstone965@gmail.com> 1758892053 +0530	commit: silly willy mistake
+e3c62c141fc65ef2be0095c49b23e06263f0b734 d440b20aae1265dabbd3ddaafb24c35e40e3ab3c cyto <silverstone965@gmail.com> 1759127664 +0530	commit: cyto/fixed the excel file input bug, in the conversion method; added exceptions where there is invalid input
+d440b20aae1265dabbd3ddaafb24c35e40e3ab3c 8d58ec13be39949ecfe7211b42c56acd2a83dc72 cyto <silverstone965@gmail.com> 1761392079 +0530	commit: added DeepSeek-OCR as a model option to use spaces

{pembot-0.1.7 → pembot-0.1.9}/pembot/.git/logs/refs/remotes/origin/main

@@ -18,3 +18,5 @@ c3cc0da3d955ecec0f865c46c030a0c073697495 e6adbc3c373070269f97ef82d4f63027d7878f6
 e6adbc3c373070269f97ef82d4f63027d7878f67 81d01e1c63d48b096c77aae83471d42272ca9fce cyto <silverstone965@gmail.com> 1758890281 +0530	update by push
 81d01e1c63d48b096c77aae83471d42272ca9fce e89cb4f5af158d26dcff5eed03dba6671a818739 cyto <silverstone965@gmail.com> 1758891600 +0530	update by push
 e89cb4f5af158d26dcff5eed03dba6671a818739 e3c62c141fc65ef2be0095c49b23e06263f0b734 cyto <silverstone965@gmail.com> 1758892065 +0530	update by push
+e3c62c141fc65ef2be0095c49b23e06263f0b734 d440b20aae1265dabbd3ddaafb24c35e40e3ab3c cyto <silverstone965@gmail.com> 1759127680 +0530	update by push
+d440b20aae1265dabbd3ddaafb24c35e40e3ab3c 8d58ec13be39949ecfe7211b42c56acd2a83dc72 cyto <silverstone965@gmail.com> 1761392127 +0530	update by push

pembot-0.1.9/pembot/.git/objects/8d/58ec13be39949ecfe7211b42c56acd2a83dc72

@@ -0,0 +1 @@
+x
��Mj�0@�u�ٗ�ѯc(����	F3��Բ��z���~��q-e���}�*�L�<%o��!���O�40KD<{�f�]�f�Dj]�B9�!���TOٳ�[����Wxm��{�u�1���B��µ������a�g�͡�`�RC"*�}����>.@

pembot-0.1.9/pembot/.git/refs/heads/main

@@ -0,0 +1 @@
+8d58ec13be39949ecfe7211b42c56acd2a83dc72

pembot-0.1.9/pembot/.git/refs/remotes/origin/main

@@ -0,0 +1 @@
+8d58ec13be39949ecfe7211b42c56acd2a83dc72

{pembot-0.1.7 → pembot-0.1.9}/pembot/AnyToText/convertor.py

@@ -5,7 +5,6 @@ from pembot.pdf2markdown.extract import MarkdownPDFExtractor
 import os
 import pandas as pd
 from typing import Literal, Union
-import tempfile
 from datetime import datetime, date
 from tabulate import tabulate
 
@@ -36,6 +35,10 @@ class Convertor():
         self.input_filepath= None
         self.output= ""
         self.suffix= suffix
+        self.file_bytes= file_bytes
+
+        if (file_bytes and not suffix) or (file_bytes and not file_type):
+            raise Exception("wrong use of convertor library, have to give suffix and file_type along with the file_bytes")
 
 
         if model_name is None:
@@ -44,7 +47,7 @@ class Convertor():
 
         # file_type can be pdf, excel, etc.
         if file_bytes and suffix:
-            with tempfile.TemporaryDirectory() as dp:
+            with TemporaryDirectory() as dp:
                 output_dir = Path(dp)
                 myfile = output_dir / f"input{suffix}"
                 myfile.write_bytes(file_bytes)
@@ -102,42 +105,51 @@ class Convertor():
         markdown_output = []
 
         file_suffix= ''
-        try:
-            if not input_filepath.exists():
-                file_suffix= self.suffix
-            else:
-                file_suffix = input_filepath.suffix.lower()
-
-            current_engine: PandasReadEngineType = excel_ods_engine
 
-            if file_suffix in ['.xls', '.xlsx', '.ods']:
-                if file_suffix == '.ods':
-                    if current_engine is None:
-                        current_engine = 'odf'
-                    elif current_engine != 'odf':
-                        print(f"Warning: Specified engine '{current_engine}' may not be optimal for ODS. Forcing 'odf'.")
-                        current_engine = 'odf'
-
-                excel_file = pd.ExcelFile(input_filepath, engine=current_engine)
-                if not excel_file.sheet_names:
-                    return f"Warning: File '{input_filepath.name}' contains no sheets."
-
-                for sheet_name in excel_file.sheet_names:
-                    df = excel_file.parse(sheet_name)
-                    markdown_output.append(f"## {sheet_name}\n")
+        try:
+            with TemporaryDirectory() as dp:
+                dp_path= Path(dp)
+
+                if input_filepath is None:
+                    file_suffix= self.suffix
+                    if file_suffix and self.file_bytes:
+                        input_filepath= dp_path / ("my_excel_file" + file_suffix)
+                        input_filepath.write_bytes(self.file_bytes)
+                    else:
+                        raise Exception("no input file name, and no file bytes either")
+                elif input_filepath.exists():
+                    file_suffix = input_filepath.suffix.lower()
+
+                current_engine: PandasReadEngineType = excel_ods_engine
+
+                if file_suffix in ['.xls', '.xlsx', '.ods']:
+                    if file_suffix == '.ods':
+                        if current_engine is None:
+                            current_engine = 'odf'
+                        elif current_engine != 'odf':
+                            print(f"Warning: Specified engine '{current_engine}' may not be optimal for ODS. Forcing 'odf'.")
+                            current_engine = 'odf'
+
+                    excel_file = pd.ExcelFile(input_filepath, engine=current_engine)
+                    if not excel_file.sheet_names:
+                        return f"Warning: File '{input_filepath.name}' contains no sheets."
+
+                    for sheet_name in excel_file.sheet_names:
+                        df = excel_file.parse(sheet_name)
+                        markdown_output.append(f"## {sheet_name}\n")
+                        markdown_table = tabulate(df, headers='keys', tablefmt='pipe')
+                        markdown_output.append(markdown_table)
+                        markdown_output.append("\n")
+
+                    return "\n".join(markdown_output)
+
+                elif file_suffix == '.csv':
+                    df = pd.read_csv(input_filepath)
                     markdown_table = tabulate(df, headers='keys', tablefmt='pipe')
-                    markdown_output.append(markdown_table)
-                    markdown_output.append("\n")
+                    return markdown_table
 
-                return "\n".join(markdown_output)
-
-            elif file_suffix == '.csv':
-                df = pd.read_csv(input_filepath)
-                markdown_table = tabulate(df, headers='keys', tablefmt='pipe')
-                return markdown_table
-
-            else:
-                return f"Error: Unsupported file type: '{file_suffix}'. Please provide a CSV, XLS, XLSX, or ODS file."
+                else:
+                    return f"Error: Unsupported file type: '{file_suffix}'. Please provide a CSV, XLS, XLSX, or ODS file."
 
         except ImportError as ie:
             if 'odfpy' in str(ie).lower() and file_suffix == '.ods':
@@ -183,10 +195,10 @@ if __name__ == '__main__':
         #     conv= Convertor(file_bytes= imgpdf.read(), suffix= ".pdf", file_type= "pdf")
         #     print(conv.output)
 
-        # print("Test 2: JD pdf, bytes")
-        # with open("/home/cyto/dev/pembotdir/jds/PM Trainee.pdf", "rb") as imgpdf:
-        #     conv= Convertor(file_bytes= imgpdf.read(), suffix= ".pdf", file_type= "pdf")
-        #     print(conv.output)
+        print("Test 2: balance sheet, bytes")
+        with open("/home/cyto/Downloads/balance_sheet_2023-24_final.xlsx", "rb") as xl:
+            conv= Convertor(file_bytes= xl.read(), suffix= ".xlsx", file_type= "excel")
+            print(conv.output)
 
         print("Test 3: excel schedule, bytes")
         with open("/home/cyto/Downloads/Assignment schedule.xlsx", "rb") as imgpdf:
@@ -194,13 +206,13 @@ if __name__ == '__main__':
             print(conv.output)
 
         # without bytes example:
-        print("Test 4: scanned pdf, path")
-        conv= Convertor(myfile= Path('/home/cyto/Documents/scanned.pdf'), output_dir= Path('/home/cyto/Documents'))
-        print(conv.output)
-
-        print("Test 5: schedule excel, path")
-        conv= Convertor(myfile= Path('/home/cyto/Downloads/Assignment schedule.xlsx'), output_dir= Path('/home/cyto/Downloads'))
-        print(conv.output)
+        # print("Test 4: scanned pdf, path")
+        # conv= Convertor(myfile= Path('/home/cyto/Documents/scanned.pdf'), output_dir= Path('/home/cyto/Documents'))
+        # print(conv.output)
+        #
+        # print("Test 5: schedule excel, path")
+        # conv= Convertor(myfile= Path('/home/cyto/Downloads/Assignment schedule.xlsx'), output_dir= Path('/home/cyto/Downloads'))
+        # print(conv.output)
     except FileNotFoundError as fe:
         print("file not found, modify the driver code to get sample files to test:\n\n", fe)
     except Exception as e:

{pembot-0.1.7 → pembot-0.1.9}/pembot/__init__.py

@@ -1,6 +1,6 @@
 """
 A Python Package to convert PEM blog content to usseful information by leveraging LLMs
 """
-__version__ = '0.1.7'
+__version__ = '0.1.9'
 from .main import save_to_json_file, make_query
 __all__ = ["save_to_json_file", "make_query"]

{pembot-0.1.7 → pembot-0.1.9}/pembot/config/config.yaml

@@ -2,4 +2,4 @@ OUTPUT_DIR: /home/cyto/dev/pembotdir
 PAGE_DELIMITER: ___________________________ NEXT PAGE ___________________________
 app:
   name: pembot
-  version: 0.1.7
+  version: 0.1.9

{pembot-0.1.7 → pembot-0.1.9}/pembot/pdf2markdown/extract.py

@@ -88,6 +88,9 @@ class MarkdownPDFExtractor(PDFExtractor):
 
             # zerogpu public
             self.nclient= Client("deepak-mehta/ocr-simplify", hf_token= os.getenv('HF_TOKEN', ''))
+        elif "DeepSeek-OCR" in self.MODEL_NAME:
+            # zerogpu private
+            self.dclient= Client("deepak-mehta/deepseek-ocr", hf_token= os.getenv('HF_TOKEN', ''))
 
 
         self.markdown_content= ""
@@ -150,7 +153,7 @@ class MarkdownPDFExtractor(PDFExtractor):
             )
             # print("response :", response)
             return response.text
-        elif 'nanonet' in model_name:
+        elif 'nanonet' in model_name or 'DeepSeek-OCR' in model_name:
 
             result= ""
             try:
@@ -159,24 +162,33 @@ class MarkdownPDFExtractor(PDFExtractor):
                     print("file name: ", temp_file.name)
                     gr_image= handle_file(temp_file.name)
                     print("gr image : ", gr_image)
-                    result = self.nclient.predict(
-                        # model_name="Nanonets-OCR-s",
-                        # text= prompt,
-                        gr_image,
-                        # max_new_tokens=max_new_tokens,
-                        # temperature=0.6,
-                        # top_p=0.9,
-                        # top_k=50,
-                        # repetition_penalty=1.2,
-
-                        # prithiv model
-                        # api_name="/generate_image"
-
-                        max_new_tokens,
-
-                        # spaces zerogpu
-                        api_name="/predict"
-                    )
+                    if 'nanonet' in model_name:
+                        result = self.nclient.predict(
+                            # model_name="Nanonets-OCR-s",
+                            # text= prompt,
+                            gr_image,
+                            # max_new_tokens=max_new_tokens,
+                            # temperature=0.6,
+                            # top_p=0.9,
+                            # top_k=50,
+                            # repetition_penalty=1.2,
+
+                            # prithiv model
+                            # api_name="/generate_image"
+
+                            max_new_tokens,
+
+                            # spaces zerogpu
+                            api_name="/predict"
+                        )
+                    else:
+                        result = self.dclient.predict(
+                            file_input=gr_image,
+                            prompt_type= 'markdown',
+
+                            # spaces zerogpu
+                            api_name="/predict"
+                        )
                     print("ocr'd: ", result[:100] + "...")
             except Exception as e:
                 print("Error during nanonet inference", e)

{pembot-0.1.7 → pembot-0.1.9}/pembot/requirements.txt

@@ -45,7 +45,7 @@ pandas==2.3.0
 pathlib==1.0.1
 pdfminer.six==20250506
 pdfplumber==0.11.7
-pembot==0.1.7
+pembot==0.1.9
 pillow==11.3.0
 primp==0.15.0
 pyasn1==0.6.1

pembot-0.1.7/pembot/.git/COMMIT_EDITMSG

@@ -1 +0,0 @@
-silly willy mistake

pembot-0.1.7/pembot/.git/refs/heads/main

@@ -1 +0,0 @@
-e3c62c141fc65ef2be0095c49b23e06263f0b734

pembot-0.1.7/pembot/.git/refs/remotes/origin/main

@@ -1 +0,0 @@
-e3c62c141fc65ef2be0095c49b23e06263f0b734