fileconv 0.0.3

This CLI converts text an MS Word and MS Excel files to PDF files

fileconv

Convert your files to PDF's

:computer: How does it work?

fileconv gives you the possibility to conveniently convert your text, MS Word or MS Excel files to PDF-A2 standard, by using a simple and intuitive command line interface.

This program was made to convert digital born documents to PDF files, in order to have a long term archiving solution for all office documents. It is a free and open source alternative to proprietary converters and it is taking advantage of the extremely powerful Pywin32 and the pyfpdf packages.

This Project has been brought to life with the help of the AfZ (Archive of Contemporary History) at ETH Zürich.

Overview

The fileconv tool is mainly used as a fast solution for file conversion via the command line. The Conversion package of this library is also available for in depth use in your python code. In order to use its methods, simply import the fileconv package.

At the moment, the package supports files that are

• WORD
  • .doc
  • .docx
• EXCEL
  • .xls
  • .xlsx
• TEXT
  • .txt

Support for more file types will be addes in the future.

The package is now also able to convert PDF files back to .txt files.

Guide

The following shows how to get and use fileconv.

Installation

$ pip install fileconv

Consider that you might have to add the installation folder to your PATH.

If you would rather like to customize the code to your needs, grab a stable version under "Releases". All the files are extensively commented as well, in order to make the files more user friendly.

Usage

When in a terminal, specify:

$ fileconv --pdf path/to/dir --output path/to/output/dir

The program takes a directory, which is populated with at least one (!) file or subdirectory and takes an output location as a positional argument. It doesn't matter if your files are in a nested structure as shown in the example below:

input_dir
│
├── xy.xlsx
├── my_dir
│   ├── xy.doc
│   ├── ab.txt 
│   └── sub_dir
│        └── cd.docx
└── ...

The program 'walks' through any directory structure and places all the files in a single folder that was specified in the --output flag.

Depending on your directory structure, the amount of files and your CPU, the process can take some time. With 1000 directories, 15'000 files to process, from which 800 are MS Office conversions, the program needs about 2 minutes and 30 seconds to finish. To not let you worry, we have implemented a progress bar that looks like this while running:

5%|██████▌                                                             | 55/1003 [00:08<01:23, 11.38it/s]

This tool was based on the excellent tqdm library.

Convert PDF files back to text

In order to convert PDF's back to text, use the --txt flag as shown bellow:

$ fileconv --pdf path/to/dir --output path/to/output/dir

Note, that the larger the files are, the longer the conversion takes.

:bug: Common errors

1. interrupting the programm

The program has one weakness regarding its running time. The pywin32 library has to 'open' the according MS programs in the background in order to properly convert the files. This process is prone to an error, if the program is interrupted while running, i.e. by pressing CTRL + C. In this case, fileconv will prompt an error the next time it converts a MS file type, because the background process has not been closed. To prevent this, one can open a terminal and type the following commands:

$ python

or (depending on your OS and python versions)

$ python3

this will open the python interpreter

>>> import win32com.client
>>> win32com.client.Dispatch('Word.Application').Close()
>>> win32com.client.Dispatch('Excel.Application').Close()

This should prevent the error from occurring again.

2. cross platform compatibility

Even though this tool is intended and written to work on all operating systems, the pywin32 library seems only to be reliable on Mircrosoft Windows. UNIX based file systems, like MacOS and Linux tend to prompt errors during the library's installation. As referred in this issue thread the library is mostly developed on Windows and therefore causing instability when used on other systems. The library is probably working on fixes and is potentially fixing this error in future versions, which will be included in this package as well.

We are also looking for stable and reliable alternatives, but right now pywin32 seems to be the best option.