HN Explorer

Show HN: A terminal spreadsheet editor with Vim keybindings (github.com)

126 points by garritfra ↗ HN
While speccing out this spreadsheet tool, I realized that I never had to think about the keybindings. It all just came naturally from Vim. Normal/insert/visual modes, hjkl navigation, dd/yy/p, :w, :q. The usual muscle memory works.

It supports CSV/TSV import and export, and a native .cell format that preserves formulas. The formula engine handles SUM, AVERAGE, COUNT, MIN, MAX, and IF with range references.

The codebase is a Cargo workspace: a pure cell-sheet-core library (no TUI dependency) and a cell-sheet-tui crate on top of ratatui. Early days, but it's usable.

To try it out: cargo install cell-sheet-tui

Feedback of any kind is greatly appreciated!

33 comments

[ 3.7 ms ] story [ 58.6 ms ] thread
[dead] SilentEditor
[flagged]
garritfra
Thanks!

Honestly, the current implementations are pretty naive — they pass the tests and feel snappy on the small sheets I work with, but they'd buckle pretty quickly under real load. Most of what you're asking about is already on the tracker; I opened a batch of issues citing your comment as the prompt.

Recalculation. Right now it's a full recalc on every edit: recalculate collects all formula cells, computes in-degrees across the whole formula set, topo-sorts, and evaluates top to bottom. The dirty flag gets propagated by mark_dirty but isn't actually used to prune work. It's also re-parsing every formula from its raw string on every pass. Two issues cover this: #8 introduces a batch boundary so paste/fill/CSV import trigger one recalc instead of N, and #7 adds criterion benches so we can actually tell whether the parser, the BFS, or the topo sort is the hotspot before optimizing. AST caching on Cell is the obvious next step once #7 confirms parsing dominates.

https://github.com/garritfra/cell/issues/8 https://github.com/garritfra/cell/issues/7

Dependency tracking. The bigger smell is in extract_deps — a range like SUM(A1:A1000) literally enumerates 1000 cell positions into the dep graph, with a HashSet per cell on each side. Fine at hundreds of cells, a disaster at hundreds of thousands. Range expansion is one of the bench cases in #7; the proper fix (interval-keyed deps so ranges stay first-class instead of fanning out) doesn't have its own issue yet — I should open one, since #7 only measures the problem.

Undo/redo. This is the worst offender right now. UndoEntry only had a single-cell variant until very recently; #12 added MultiCellEdit, but #13 tracks two destructive paths I missed — visual-mode d and p/P paste — that still don't push undo entries at all. #9 is the broader coalescing story (one dd = one undo, CSV import = one undo, etc.), tied to the batch mechanism from #8 so a single transaction produces a single undo entry. sort_by_column is also non-undoable today and belongs in that bucket.

https://github.com/garritfra/cell/issues/13 https://github.com/garritfra/cell/issues/9

Larger CSVs. Storage is HashMap<CellPos, Cell> — fine sparse but with overhead per cell; for very wide imports a column-oriented or arena layout would pay off. I haven't profiled it though, so this is speculative; the dependency-graph blowup will hurt before raw storage does. #7 includes a 100k-row CSV load case to put numbers on it.

And #10 is the meta-issue to lift all of this out of source comments and into actual architecture docs, which I probably should have done before posting.

https://github.com/garritfra/cell/issues/10

So: nothing here scales today, but the architecture splits cleanly enough that none of it needs a rewrite — AST caching, dirty-set recalc, range-aware deps, and grouped undo are the four threads, and most have issues attached.

WillAdams
Was initially hopeful that this would rather than have A..B..C... and 1..2..3... for columns and rows would instead have one creating categories à la Javelin/Lotus Improv/Quantrix/Flexisheet --- give me that, and have a dynamic system for displaying a pane of formulae and one would have a _very_ nice tool.
adilkhanovkz
Dimensions over A1/B2 is the right move. Rust is a good base for the engine — formula propagation in dimensional models needs to be fast. Curious whether the bigger challenge is the engine or the UI layer
marcyb5st
Pretty cool project! Congrats.

QQ: does it support programmatic cell access/modifications?

Eg `cell file.cell --write A2 "42"` or `cell file.cell --read "=SUM(A1:A10)"`? Couldn't surmise that from the glimpse I gave the README, but it would be pretty cool for scripting stuff.

bytejanitor
Looks very nice at first glance.

I have been looking for something like this for a long time. Maybe this has some potential to become somewhat popular.

Theaetetus
This is great. I often want a low-latency, minimal-feeling option for this sort of thing. (And I get to use my Vim muscle memory less and less often these days!)

Really, thanks for making and sharing this; so far, I feel calm and happy when I'm using it.

ximm
I was working on something similar a while back (https://github.com/xi/spreadsheet) but ended up not using it. I just didn't end up being the first tool that came to mind when I was reaching for a spreadsheet. Not sure why.

I will definitely try out your tool and check if it works better to me.

Two random thoughts:

- In excel I often use the dragging feature, i.e. use a formula like $A2 + B$1 and then drag it in both directions. Does your tool have something similar?

- Another nice feature are visualizations. In my tool I added a bar() function that renders a value between 0 and 1 as a bar in the cell. Not sure how flexible that is, but I did have some fun with it.

rpassos
I've been using https://github.com/medialab/xan for anything csv. It's the best tool there is for reading and analyzing csv files, but not for writing them. Something like cell-sheet-tui would fit my workflow very well, since I am using neovim to edit csv files manually (as well as some csv merging features offered by xan to simplify the process).

The reason why I find your software promising is because dealing with csv in plain text is very disorienting with the inconsistent horizontal cell space. A csv-aware text editor would help a lot there with two issues in particular:

- not having to set commas manually, especially not having to count them between empty cells

- being able to navigate using h/j/k/l between cells instead of having to get creative with vim-motions on plain text

OP, if you're reading this:

I would greatly appreciate if this tool could focus on the writing experience and less with anything formula engine related, since xan is already a very performant tool for that, even at large data sets. Replacing or even competing with xan seems very difficult in my eyes, considering both tools are targeting a similar audience (rust, minimalist, and performance), which makes me think cell could focus more on the writing experience, like I said.

There are some bugs I countered right away and I'll get to open issues in due time. Thanks for making this and I hope I can replace neovim with cell for anything csv in the future!

ranger_danger
> It's the best tool there is for reading and analyzing csv files

It does not appear to support the file types I regularly work with: https://www.fda.gov/medical-devices/medical-device-reporting...

These files have a field delimiter of '|' with a line delimiter of '\r\n', as individual fields also contain '\n' characters.

I prefer visidata as the program stays running while searching and navigating through large files, making subsequent searches and filtering much faster, plus editing/writing is also supported, as well as Excel and other formats I regularly use.

jzer0cool
Looks great something I want to check out. For comparison anyone share some built-in tools to pretty print-only the csv? Usually what I need to start but formatting is quite off.
unlikelyusernam
forgive me if I'm missing something obvious, but is there any plan to add jumping to a specific row? Usually when I scroll and find the row I'm looking for, I'll just nG there (where n is the row I found), which in vi-esque editors will jump to that specific line.
reagle
visidata is vim influenced.
Keyframe
Lotus 1-2-3 flashbacks kicking in. I'll try it out for sure.
freedomben
Very neat, a tool like this is badly needed. Also cool to see it's in rust (that means I can hack on it).

To be really useful to me though (and I would imagine others), it will need to be able to at least view (but probably edit) xlsx and/or open document format spreadsheets. I know that's a huge task, but it would add immense value to the project. I'm not asking for it and I don't feel "entitled" to it (lest the inevitable replies of "add it youself" come flying in), just trying to provide some helpful feedback.

Question: is this a project you see yourself staying committed to long term? (Obviously life events happen and things change, but at least for now, do you enjoy it? Do you use it as part of your workflow? Do you feel like it's something you'll want to own long term?)

tpoacher
i'll try this next time i'm on my laptop, but for now, how does it compare to visidata (which is a long standing peoject already?)
daft_pink
It’s really amazing how spreadsheet started in the terminal and have come full circle back to the terminal.

This is fantastic thanks for your work

pca2
Very cool, I was just wishing for a tool like this today. A small feature request I would love to see: Support for custom delimiters. I sometimes work with CSVs that actually use pipe (|) as the delimiter to avoid issues with cells that contain commas as content.
k2enemy
Looks terrific! I have a suggestion as you continue to add features and think about scripting. sc-im has lua external functions, but you cannot pass a range of cells, only the value of one cell. This severely limits the usefulness of external lua scripts. If/when you add scripting, hopefully you can overcome this limitation. If so, you'll have at least one more user!
immccc
The kind of stuff that once you see it, you think "how didn't come the idea up to me?" :)

Only thing is that it would be great to have it as a vim plugin, so that after opening a .csv file or similar, you can make use of the spreadsheet tooling.

nanis
This looks good, but, I noticed:

> Supported Functions (v1) > SUM, AVERAGE, COUNT, MIN, MAX, IF

I would recommend switching to numerically stable algorithms for adding numbers and taking their averages sooner rather than later.

Looking at the code[1], it looks like you are simply summing the numbers and dividing by count. If that is indeed the case, please switch to a stable average[2]. See also [3].

Additionally, I tried to test the executable[4] on Windows, but there is something wrong with input. Using Windows Terminal Version: 1.24.10921.0, characters are doubled on input. Regular Vim compiled using Visual C++ does not have this issue.

When I press `:`, I get `::` in the COMMAND line. Then, every key I press gets doubled. This means I cannot `:q!`. :-)

[1]: https://github.com/garritfra/cell/blob/99b0afc7a60a871b27452...

[2]: https://www.nu42.com/2015/03/how-you-average-numbers.html

[3]: https://en.wikipedia.org/wiki/Algorithms_for_calculating_var...

[4]: https://github.com/garritfra/cell/releases/tag/v0.2.0

sev_verso
Interesting. Last time I dealt with currency in SQL (which was well over a decade ago admittedly), we used a fixed point format for those fields.

What do you recon Excel does, floating point math for regular entries and lossless point math for currency? Or just does not rely on floats at all?

teo_zero
The equal operator should work on strings, too. It would allow formulas like IF(A1=C3,A2,0)
solarized
hi, did your claude.md files always works consistently?