How can I build a beginner-friendly language checker?

Start much smaller than trying to analyze all of Rust. A typical compiler front end is built in stages: lexing turns characters into tokens, parsing turns tokens into a syntax tree, and semantic analysis checks meaning. For a portfolio project, you could report syntax errors, undeclared variables, or simple type mismatches. That would be a valuable compiler-front-end project, even if it does not generate executable code. Crafting Interpreters is an excellent free resource for learning these ideas.

You would mainly be building the front end of a compiler: tokenization, parsing, an AST, and semantic checks. It becomes a complete compiler when it also translates the program into another form, such as bytecode, an intermediate representation, or machine code. Pick a small language or subset rather than Rust, and use a standard introductory compiler text alongside an implementation project.

Learn about grammars, parsing, abstract syntax trees, and Backus–Naur Form. It is also worth understanding the difference between a compiler, an interpreter, and a linter. A manageable first project would be checking whether a small fragment of a language, such as a C loop, follows the grammar. Focus on recognizing valid structure before worrying about translating it into machine code.

“Errors” can mean several different things. Checking formatting and syntax is a reasonable starting point, while checking types, return values, and variable usage requires semantic analysis and is more advanced. Checking whether a program does what the author intended is a much harder problem. Writing a recursive-descent interpreter for a tiny language is one approachable way to learn the basics from scratch.