I also implemented a spreadsheet last year [0] in pure TypeScript, with the fun twist that formulas also update backwards. While the backwards root finding algorithm was challenging, I also found it incredibly humbling to discover how much complexity there is in the UX of the simple spreadsheet interface. Handling selection states, reactive updates, detecting cycles of dependency and gracefully recovering from them is a massive state machine programming challenge! Very fun project with a lot of depth!
I myself didn't hand roll my own parser but used Ohm-js [1] which I highly recommend if you want to parse a custom language in Javascript or TypeScript.
> One way of doing this is to keep track of all dependencies between the cells and trigger updates when necessary. Maintaining a dependency graph would give us the most efficient updates, but it’s often an overkill for a spreadsheet.
On that subject, figuring out the efficient way to do it is also a large engineering challenge, and is definitely not overkill but absolutely required for a modern spreadsheet implementation. There is a good description of how Excel does it in this famous paper "Build systems a la carte" paper, which interestingly takes on a spreadsheet as a build system [2].
#define MAXIN 128 // max cell input length
enum { EMPTY, NUM, LABEL, FORMULA }; // cell types
struct cell {
int type;
float val;
char text[MAXIN]; // raw user input
};
#define NCOL 26 // max number of columns (A..Z)
#define NROW 50 // max number of rows
struct grid {
struct cell cells[NCOL][NROW];
};
I doubt that 171 KB of static allocation would fly on an Apple II! I do wonder how they did memory allocation, it must have been tricky with all the fragmentation.
Cool article but I think the write-up no longer matches the actual code. Snippets in the article use *p->p a lot. The *p is a parser struct defined above as
So there's the missing p, even though it's no longer an int. So I presume the member variable was once known as pos but got renamed at some point. Some of the snippets did not get updated to match.
Are there good command-line interfaces for spreadsheets? I don't do anything super financially-important and I'd prefer to stay in the terminal for quick editing of things, especially if I can have Vi keybindings.
This was one of the projects students did when I helped teach APCS to high schoolers as a TEALS volunteer (FracCalc).
Some of the implementations went way overboard and it was so much fun to watch and to play a part.
Even as a “seasoned” developer I learned some tidbits talking through the ways to do (and not do) certain parts. When to store input raw vs processed, etc.
I’m genuinely worried that we’re the last generation who will study and appreciate this craft. Because now a kid learning to program will just say “Write me a terminal spreadsheet app in plain C.”
Kinda cool to see... TBH, I'd be more inclined to reach for Rust and Ratatui myslf over C + ncurses. I know this would likely be a much larger executable though.
With MS Edit resurrected similarly, I wonder how hard it would be to get a flushed out text based spreadsheet closer in function to MS Excel or Lotus 123 versions for DOS, but cross platform. Maybe even able to load/save a few different formats from CSV/TSV to XLSX (without OLE/COM embeds).
80 comments
I also implemented a spreadsheet last year [0] in pure TypeScript, with the fun twist that formulas also update backwards. While the backwards root finding algorithm was challenging, I also found it incredibly humbling to discover how much complexity there is in the UX of the simple spreadsheet interface. Handling selection states, reactive updates, detecting cycles of dependency and gracefully recovering from them is a massive state machine programming challenge! Very fun project with a lot of depth!
I myself didn't hand roll my own parser but used Ohm-js [1] which I highly recommend if you want to parse a custom language in Javascript or TypeScript.
> One way of doing this is to keep track of all dependencies between the cells and trigger updates when necessary. Maintaining a dependency graph would give us the most efficient updates, but it’s often an overkill for a spreadsheet.
On that subject, figuring out the efficient way to do it is also a large engineering challenge, and is definitely not overkill but absolutely required for a modern spreadsheet implementation. There is a good description of how Excel does it in this famous paper "Build systems a la carte" paper, which interestingly takes on a spreadsheet as a build system [2].
[0] https://victorpoughon.github.io/bidicalc/
[1] https://ohmjs.org/
[2] https://www.microsoft.com/en-us/research/wp-content/uploads/...
It's not overkill at all. In fact, it's absolutely necessary for all but the simplest toy examples.
*p->pa lot. The *p is a parser struct defined above as Notice there is nopmember within. Assume the author meant*p->pos? And indeed if you look at the code in github the parser struct is defined as So there's the missingp, even though it's no longer an int. So I presume the member variable was once known asposbut got renamed at some point. Some of the snippets did not get updated to match.Some of the implementations went way overboard and it was so much fun to watch and to play a part.
Even as a “seasoned” developer I learned some tidbits talking through the ways to do (and not do) certain parts. When to store input raw vs processed, etc.
Though I think the definition of the parser struct should be
based on the rest of the code.With MS Edit resurrected similarly, I wonder how hard it would be to get a flushed out text based spreadsheet closer in function to MS Excel or Lotus 123 versions for DOS, but cross platform. Maybe even able to load/save a few different formats from CSV/TSV to XLSX (without OLE/COM embeds).