Searching for a text in scanned files using OCR.
If the document has a table, then the program determines the structure of the table and recognizes the contents.
The original document (image/PDF) can be rotated to any angle.
The search is also performed in files of other formats, if these files are present in the target directory.
Supported formats: pdf/png/jpeg/jpg/tif/doc/docx/odt/xls/xlsx/ods/txt
Works on Windows, Linux

The program starts without any parameters. Uses settings from config file.
You need to set at least one parameter in search.conf: the 'search' parameter specifies what to search for.
If you do not specify a search directory, the current directory will be used.

1. python 3.9 (or later)
2. tesseract-ocr-w64-setup-v5.2.0.20220712.exe
   2.1. PATH environment variable: "D:\Tesseract" (specify the directory selected during installation)
   2.2. Add file with prefered language (ukr.traineddata) to "D:\Tesseract\tessdata"
   (for Linux: apt-get install tesseract-ocr)
3. pip install pytesseract
4. poppler and pdf2image
   4.1. install Poppler (release 22.12.0-0 or later) for Windows: https://github.com/oschwartz10612/poppler-windows/releases
   Extract Poppler to "D:\Poppler\poppler-22.12.0" (or any other folder) and add PATH:
   "D:\Poppler\poppler-22.12.0\Library\bin"
   4.2. pip install pdf2image
5. pip install filetype
6. for 'XLS/XLSX' file format:
   pip install pandas
   pip install xlrd
7. for 'DOCX' file format:
   pip install docx2python
8. pip install opencv
9. for 'DOC' file format:
   Download and unpack antiword to c:\antiword.
   Set the PATH and add Environment Variable:
   ANTIWORDHOME = c:\antiword
   PATH = c:\antiword (for Linux: apt-get install antiword)
10. pip install progress
11. for 'ods/odt' file format:
    pip install odfpy

For Windows: to generate exe file:

pyinstaller.exe --onefile search.py
search.exe

or run:

search.py

v3.3

• Svitlana Viblaia

This project is licensed under the MIT License.