potygen VS prql

OMG this is one of my favorite tools paid for it all the way back when it went out. Have used it so many times just to write documentation for things like:

https://github.com/ivank/potygen/blob/main/packages/potygen/...

ASCII is just so versatile and allows you to put nice graphics in places where one does not expect, making things more easily understandable.

I also wrote a parser (in typescript) for postgres (https://github.com/ivank/potygen), and it turned out quite the educational experience - Learned _a lot_ about the intricacies of SQL, and how to build parsers in general.

Turned out in webdev there are a lot of instances where you actually want a parser - legacy places where they used to save things in plane text for example, and I started seeing the pattern everywhere.

Where I would have reached for some monstrosity of a regex to solve this, now I just whip out a recursive decent parser and call it a day, takes surprisingly small amount of code! (https://github.com/dmaevsky/rd-parse)

I used to agree 100% with this sentiment, as dissatisfaction with available ORMs at the time (early days of doctrine in PHP) drove me to actually write my own. Turned out an amazing exercise in why orms are hard.

Anyway a few years later I was in a position to start things fresh with a new project so thought to myself, great lets try to do things right this time - so went all the way in the other direction - raw sql everywhere, with some great sql analyzer lib (https://github.com/ivank/potygen) that would strictly type and format with prettier all the queries - kinda plugged all the possible disadvantages of raw query usage and was a breeze to work with … for me.

What I learned was that ORMs have other purposes - they kinda force you to think about the data model (even if giving you fewer tools to do so) With the amount of docs and tutorials out there it allows even junior members of the team to feel confident about building the system. I’m pretty used to sql, and thinking in it and its abstractions is easy for me, but its a skill a lot of modern devs have not acquired with all of our document dbs and orms so it was really hard on them to switch from thinking in objects and the few ways orms allows you to link them, to thinking in tables and the vast amounts of operations and dependencies you can build with them. Indexable json fields, views, CTEs, window functions all that on top of the usual relation theory … it was quite a lot to learn.

And the thing is while you can solve a lot of problems with raw sql, orms usually have plugins and extensions that solve common problems, things like soft delete, i18n, logs and audit, etc. Its easy even if its far from simple. With raw sql you have to deal with all that yourself, and while it can be done and done cleanly, still require intuition about performance characteristics that a lot of new devs just don’t possess yet. You need to be an sql expert to solve those in a reasonable manner m, just an average dev could easily string along a few plugins and call it a day. Would it have great performance? Probably not. Would it hold some future pitfalls because they did not understand the underlying sql? Absolutely! But hay it will work, at least for a while. And to be fair they would easily do those mistakes with raw sql as well, but with far few resources to understand why it would fail, because orms fail in predictable ways and there is usually tons of relevant blog posts and such about how to fix it.

It just allows for an better learning curve - learn a bit, build, fail, learn more, fix, repeat. Whereas raw sql requires a big upfront “learn” cost, while still going through the “fail” step more often than not.

Now I’m trying out a fp query builder / ORM - elixir’s ecto with the hopes that it gives me the best of both worlds … time will tell.

> I completely attribute this to SQL being difficult or "backwards" to parse. I mean backwards in the way that in SQL you start with what you want first (the SELECT) rather than what you have and widdling it down.

> The turning point for me was to just accept SQL for what it is.

Or just write PRQL and compile it to SQL

https://github.com/PRQL/prql

Hey HN! We’ve built Pretzel, an open-source data exploration and visualization tool that runs fully in the browser and can handle large files (200 MB CSV on my 8gb MacBook air is snappy). It’s also reactive - so if, for example, you change a filter, all the data transform blocks after it re-evaluate automatically. You can try it here: https://pretzelai.github.io/ (static hosted webpage) or see a demo video here: https://www.youtube.com/watch?v=73wNEun_L7w

You can play with the demo CSV that’s pre-loaded (GitHub data of text-editor adjacent projects) or upload your own CSV/XLSX file. The tool runs fully in-browser—you can disconnect from the internet once the website loads—so feel free to use sensitive data if you like.

Here’s how it works: You upload a CSV file and then, explore your data as a series of successive data transforms and plots. For example, you might: (1) Remove some columns; (2) Apply some filters (remove nulls, remove outliers, restrict time range etc); (3) Do a pivot (i.e, a group-by but fancier); (4) Plot a chart; (5) Download the chart and the the transformed data. See screenshot: https://imgur.com/a/qO4yURI

In the UI, each transform step appears as a “Block”. You can always see the result of the full transform in a table on the right. The transform blocks are editable - for instance in the example above, you can go to step 2, change some filters and the reactivity will take care of re-computing all the cells that follow, including the charts.

We wanted Pretzel to run locally in the browser and be extremely performant on large files. So, we parse CSVs with the fastest CSV parser (uDSV: https://github.com/leeoniya/uDSV) and use DuckDB-Wasm (https://github.com/duckdb/duckdb-wasm) to do all the heavy lifting of processing the data. We also wanted to allow for chained data transformations where each new block operates on the result of the previous block. For this, we’re using PRQL (https://prql-lang.org/) since it maps 1-1 with chained data transform blocks - each block maps to a chunk of PRQL which when combined, describes the full data transform chain. (PRQL doesn’t support DuckDB’s Pivot statement though so we had to make some CTE based hacks).

There’s also an AI block: This is the only (optional) feature that requires an internet connection but we’re working on adding local model support via Ollama. For now, you can use your own OpenAI API key or use an AI server we provide (GPT4 proxy; it’s loaded with a few credits), specify a transform in plain english and get back the SQL for the transform which you can edit.

Our roadmap includes allowing API calls to create new columns; support for an SQL block with nice autocomplete features, and a Python block (using Pyodide to run Python in the browser) on the results of the data transforms, much like a jupyter notebook.

There’s two of us and we’ve only spent about a week coding this and fixing major bugs so there are still some bugs to iron out. We’d love for you to try this and to get your feedback!

> Looks like PRQL doesn't have a Go library so I guess they just really wanted something in Go?

There's some C bindings and the example in the README shows integration with Go:

https://github.com/PRQL/prql/tree/main/prqlc/bindings/prqlc-...

Can someone tell me why PRQL is better? I went here: https://github.com/PRQL/prql

It looks nice, but what's the strengths compared to SQL?

PRQL [1] is a compile-to-SQL relational querying language that puts FROM first.

[1] https://prql-lang.org

https://prql-lang.org/ might be an answer for this. As a cross-database pipelined language, it would allow RAG to be intermixed with the query, and the syntax may(?) be more reliable to generate