My work uses a custom database with addresses and other personal information. In my role I only have access to an html version.

I frequently need to make letters based on this information, but currently I'm expected to manually copy the information.

The information I need is spread through multiple pages, but fortunately the URLs are page specific and not record specific, meaning that I am dealing with the same five or so URLs for each record.

I'm sure I could use a scraping tool to achieve this task in a few clicks. What would people recommend that would not require installation onto my work machine (although browser extensions are acceptable)?

1 Answer 1


If you are able to run software from a USB stick then take a look at WinPython.

  • Free, Gratis & Open Source (MIT Licence)
  • 64 or 32 bit Windows
  • non-invasive: WinPython lives entirely in its own directory, without any OS installation
  • customizable: add your missing packages,
  • do your own version: a winpython-creator kit is made available for you
  • Runs out of the box(*) on any Windows 8+ with 2GB Ram (Jupyter Notebook will require a recent browser)
  • Requests is included in all but the Zero edition so your web fetching is taken care of
  • If you add scrapy you can write spiders to do web scraping.
  • You can start from the Zero edition (<28 MB download, 100 MB installed on USB key) and add requests or use one of the full featured editions (431-450 MB download, 2.7 GB installed on USB Key).

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.