Great work! As a side track, it led me to dive into the history of the manufacturing company of Breda trains. Originally founded in Milan late 1800s by Ernesto Breda for locomotives, expanded in the war products during the wars, and went through nationalization, fusion to become AnsaldoBreda and later bough by Japanese to become Hitachi Rail Italy.
Later to get banned from bidding on the contract for the replacement vehicles because the trains were so mediocre (although somehow still better than the earlier Boeings) and the company so brazenly corrupt.
Meanwhile SF runs 1800s–early 1900s era Milan trams on their heritage line. Not built by Breda, because of course.
Yet another appreciation story for Miguel’s mega tutorial. In 2017 I used it to create our wedding site and learn a bit of web dev (my background is in data science). To motivate me to actually do it I used the strategy the fund the then occurring refactoring of the tutorial. I am still very fond and proud of that first time I actually went and funded some open source effort, it gives you back more than you might expect
For the curious on how it works (not mentioned in the readme), it uses pymupdf and a precise mapping of all information in area coordinates, as such the document layout is hard coded.
When layout changes this breaks but layout changes on this sort of documents do not happen often (I think). Also code is very clean and it serms straightforward to fix.
This kind of code is maybe something that can be generated from an LLM/agent? (It would be easy to write checks)
Besides the practical value for those who might need it, I think it is possibly interesting for others to look at this approach.
I am a bit biased on the content ;), but note that this is a talk that is not focused on teaching the language but more on what does it mean to learn a second language, especially a niche one - like Nim - coming from Python. It does shows some code examples and highlights some features though. I like official documentation (tutorials) but Nim basics is another good resource.
When I first learned Nim I remember the thing I struggled the most with was the mutability of parameters when passing them. Another big thing I would say is letting go of python’s dict and embrace types. In place of dict I usually consider the following options: tuples (which are great and named), objects, tables (from stdlib, very similar to dicts but with homogenous types for keys and values) and JsonNode (as a final option if the others do not work).
Thomas Snyder, the creator of this blog (who gathered a number of contributors throughout the years; recently slowed down the content) is a puzzle grandmaster and recently mind behind the popular LinkedIn queens game. Lots of great content on this website.
https://en.wikipedia.org/wiki/Hitachi_Rail_Italy