We witnessed the extreme tediousness of copying & pasting among different platforms when math researchers read and compose papers & computable documents. This issue comes from incompatibilities among math equations in image form and major math languages: LaTeX & Wolfram.
What it does
MathFlow automates the workflow from image equations directly to Wolfram & LaTeX codes. Simply capture a screenshot and codes will pop up. There is also an option to append them to a computable document along with a user comment.
How we built it
We use Swift to develop the user interface (macOS application). It will call an executable written in C, which combines MathPix (from image to LaTeX) & Wolfram API (from LaTeX to Wolfram) to properly generate the codes. As a key component of the C program, we use regular expression to handle incompatibilities b/w MathPix & Wolfram API.
Challenges we ran into
MathPix output and Wolfram API input have different LaTeX expression grammar. We have to use an interpreter (regular expressions) to bridge them together.
Accomplishments that we're proud of
With regular expression, we mainly address the compatibility issue between LaTeX & Wolfram which allows a "fuzzy mapping" from a typesetting language (displayable) to a semantic language (computable).
We lessen the burden of math researchers to a significant extent by considering the habits of most math researchers.
What we learned
- macOS application development using Swift.
- Extensive C network programming and string manipulation.
- Researchers-friendly UI design to simplify workflow.
What's next for Math Flow
- Support more math languages.
- Make LaTeX interpreter (regular expression) more accurate.
- Improve UI design based on user feedback.