johannes-wolf / rpnspire Goto Github PK

Currently, there is a problem deleting multibyte characters, as they are removed byte per byte which is bogus.

The current @bind function is not very userfriendly.

See options menu/UIMenus col 3 items

I can't seem to do anything related to variables. It always either complains about it not being defined or, if I try to define it, shows "Internal Error" Error!!!

I do not have ndless installed and I am on version 5.3.2.129.

Thanks!

Refactor the group of dispatch* functions that handle RPN input.

Storing expressions as trees would make it easier analyzing and modifying expressions. Pattern matching an expression for a scheme would pre pretty easy, too.

Pressing the (-) key causes the entire expression to be negated. This is contrary to usual HP RPN-style exponentiation, where the (-) key should only negate the input after the EE. I suppose you can work around this by pushing your signifcand to the stack first and then multiplying by your value after using the 10^x key.

Using rpn.tns from build #306 on a CXII CAS (OS 6.0.3.374).

The stack of the RPNInput tests contains garbage (some self or the like). This is not the case on the device.

When filtering small top menus recursive, they can contain lots of items.
As they currently do not resize vertically, they are cumbersome to navigate.

Got no idea yet why this happens.

When attempting to open the document on my TI-Nspire CX CAS, I get a message saying "An error was found in the format of this document". Is this only compatible with the CX II? I noticed all of the screenshots are on that device

Hello! I would like the ability to force the recieved answer in scientific/enginering notation when the output is not a large number that needs to be approximated. I would also like to be able to change the number of significant digits in that output.

Does not close recursive, calls action twice.

Goal: Come up with a better syntax that is also easy to type using the handheld keyboard.
Idea: Support fx(,,,) in infix & rpn mode and use the stack to fill up unset arguments. fx(123,) would use 123 as first and stack-top as second argument. That would make simplify various calls too.

Entering (+) (or (+-) etc.) results in an emty stack entry.

Move the rest of the shortcut handling to the new keybind manager class.

See title.
Call invalidate on the focused view after calling arrow handlers.

The expression solve(1^(x/12)=y,x) is transformed to y=x which is wrong.

Left/Right movement is not possible after certain yet unknown actions.

Multiple bugs:

• Calling root2 does not use 2 as argument count.
• Pressing the root key should call root2.
• root should default to 2 arguments.