Show HN: AlexCalc, a scientific calculator with LaTeX equation display
alexbarry.github.io
alexbarry.github.io
I wish I had something like this back when I actually wrote reports and stuff in LaTeX. Now I barely have any use case for my own calculator, except occasional unit conversions.
The use case that I had in mind for mine was electrical engineering students, where I had to often use complex numbers in polar form, in degrees, for circuit questions. My calculator at the time had no obvious way of entering polar complex numbers in degrees, I ended up typing things like `e^(i*angle/1r)`, where the r was "radian", and when the calculator was in degrees mode, "1 r" would evaluate to "180/pi".
It was also cumbersome to write long expressions for. It showed something like 24 characters per line, enough that I'd often get multi line expressions. These expressions would be full of brackets, and sometimes I'd hit enter and not be sure if I missed a bracket. (Hence my goal of having a LaTeX display to show you what you're entering).
Nitpick: in dark theme, the fraction line didn't invert to white in 1/sqrt(2)
Or is it Android? On my phone 1/sqrt(2) looks fine in "dark" and "very dark". If it doesn't for you, could you tell me what type of phone you have? I put some contact info in the "about" section of the app if you'd rather respond privately.
For dark mode on Android I just set the color and background-color attributes of the HTML display elements. (I didn't actually fully implement dark mode in the HTML version, though there are some commented out CSS files). But it's possible that something else is needed to work with MathJax in some cases.
I don't know how I'll fix this (I've never really debugged mobile web specific stuff before), but I'll look into it this weekend.
There must be a way to fix it, since I don't see the issue when going to https://www.mathjax.org/#demo and entering something like this: $$\frac{1}{\sqrt{2}}$$
EDIT: I actually have no idea why this would be happening, unless there is some sort of metadata that I should be providing on the page to tell chrome how to enable dark mode on my content. I don't think I'm messing with the styling on the web version at all.
This is chrome 97.0.4692.98
I'd also think we move this from HN to some issue tracker?
I thought that it might be useful to use this application to generate TeX, so one feature I'm working on is to fine tune the output format (e.g. choosing the multiplication symbol, instead of \cdot, maybe \times or \,).
Thanks for trying it!
Available on:
* web (mobile or desktop): https://alexbarry.github.io/AlexCalc/
* Android: https://play.google.com/store/apps/details?id=net.alexbarry....
Some neat features:
* LaTeX equation display (I hated counting brackets on my old graphing calculators back in college, now I wouldn't have to if I wasn't ~10 years late in making this)
* variables (you can store variables with `1 + 2 -> x`, then access them `x^2 - 3`)
* unit operations, and unit conversions, e.g. `10 km / 2 minutes to mph`, or more fancily: `sqrt((3 kOhm + j*100 mH * 1000 kHz)/(1 S + j * 1 nF * 1000 kHz))`
* complex numbers (specifically focused on making it easy to enter complex numbers in polar form, in degrees, e.g. `5 angle 90`, but press the "degree" button)
* can enter input either via button presses or typing, so copying and pasting inputs works well
The core calculator parsing, evaluating, and LaTeX generation is in C++, with CMake build files. Using emscripten[0], this compiles to webassembly for the web version, and I can also compile this for Android and use JNI.
The library converting my generated LaTeX code to graphics is "MathJax"[1], a really nice browser library that I loved working with both in the web version and in the Android WebView.
I'm happy to receive any feedback at all. One big weakness currently are the syntax error messages, I haven't yet implemented a way to point to a particular position where an error occurs. If there is some nice parsing library that I could use instead, I might consider switching to that if it isn't too difficult. Currently the parsing uses a bunch of (fairly simple) regular expressions (e.g. check for number, check for optional unit, check for binary operator, add to stack, collapse stack. This ends up with a tree of nodes, and then this tree can either be evaluated or converted to LaTeX.)
I'm also interested in hearing any strategies for a decent cross platform UI. I didn't want to just include the HTML UI in a WebView on Android, since I figured it wouldn't be a great user experience. Originally I thought it would be small enough that I could just write two separate copies... but it turns out that there is a lot more UI code than I expected (automatically insert multiplication symbols on button presses, input history, storing "recently used units" when the button is pressed, etc...)
Would it be feasible to support unicode variables? Having to spell the greek letters seems a bit clunky. If I type them directly it gives a parsing error. Notice that in modern LaTeX (say, xetex or luatex) you can do that. No need to type \'e to get é like we did in the nineties.
Assuming C++ standard regex supports unicode, then this should be easy enough. I never thought to try this. I'll see if I can implement it later today or this weekend. Do you think the spelled out "alpha" should map to the same thing as "<unicode alpha character>"?
Also, do you mind if I ask how you are entering Greek letters directly? (e.g. are you a Greek speaker, using a Greek keyboard?) I'd like to better understand how people would like to use this. I'm happy to accommodate whatever I can, and I'm relying on people asking for it directly since I don't have any sort of tracking implemented. (The android version doesn't even request any INTERNET permission (or any permission).)
I've setup my keyboard (on linux) so that the CAPS-LOCK key acts as a "dead greek" composer. If I press "CAPSLOCK+a" it types "α", and so on.
> Do you think the spelled out "alpha" should map to the same thing as "<unicode alpha character>"?
My unpopular opinion is that typing "alpha" should compute the product of the five values, equivalent to "a^2lph".
Cool! Is this because you type a lot of Greek (or math)?
> My unpopular opinion is that typing "alpha" should compute the product of the five values, equivalent to "a^2lph".
I wouldn't call it an unpopular opinion, that's how my physical graphing calculator worked, though it only had single letter variable names. The programmer in me likes my current way, though the (nearly long gone) engineering student cringes when I have to write things like "i*2*pi*f".
If people were mostly universally happy with single letter variable names then I could probably switch to this. If they were space separated then it would be a fairly simple change (except for the way units works, is "1 s" 1 times variable s, or 1 second? This is why I don't support 1 m/s to mean metre per second, it means "1 metre divided by variable s". 1 metre per second is "1 m s^-1").
I tried to compromise by making the UI automatically enter a multiplication symbol when entering tokens that aren't valid.
Perhaps if I added an additional input layer when typing then this would be feasible. Maybe something like pressing a special character, then typing "alpha" will convert it to the alpha character, and non greek letter variables could be wrapped in quotes or square brackets or something.
The one thing that would be lost is the ability to name a variable "x2". Maybe that's fine since "x_2" would still work.
As an update on unicode parsing, I still think it should be easy, but I'm running into issues. I quickly tried https://stackoverflow.com/a/11255698/9596600 but I'm seeing this:
terminate called after throwing an instance of 'std::regex_error'
what(): Unexpected escape character.
Just regular math, where greek letters abound. A colleague mocked me because I was "still" writing \varphi instead of φ, so I learned to type the greek letters directly.
> (...) cringes when I have to write things like "i2pi*f".
In LaTeX you would write 2πif, in a "LaTeX calculator" as yours I'd expect to be able to do so!
Right now 0.1+0.2-0.3 results = 2.775557562 \cdot 10^{-17}, which depending on your viewpoint is either expected or unexpected. Also (0.1+0.2)-0.3 is simplified to 0.1+0.2-0.3 for display, but results a different value = 5.551115123 \cdot 10^{-17}
I haven't had a chance yet to read the full paper, but it talks about fixing trig issues too. Would this fix my (unfortunate) result for `e^(ipi) + 1` to be exactly 0, too (instead of something ugly like (-210^-16 + j1.22*10^-16)).
And is there an implementation I could use?
At some point I plan on open sourcing this code, so if I never manage to fix this then maybe someone else might someday do it. Originally I figured I could finish this whole project in a few months, especially after figuring out webassembly and discovering MathJax. But it ended up taking way longer than that, with little things constantly popping up that I wanted to fix before I could even think about sharing it.
If you want to go this route, there are standardized decimal floating points, e.g. https://en.wikipedia.org/wiki/Decimal64_floating-point_forma... (or the 128 bit version).
> And is there an implementation I could use?
Afaik Boehms implementation lived only in the Android source tree, i.e. https://android.googlesource.com/platform/packages/apps/Exac... . So to use that you'd need to extract and port them, which might be some effort.
There are all sorts of different approaches for doing calculations on real numbers. I think Johanssons work is particularly impressive in this domain, but I'm not sure if that is much use for a calculator like this: https://fredrikj.net/
For choosing the desired output unit, press "alt" and the now shown "to units" button. Then enter the desired unit.