using not case-sensitive symbols?

[ThomasWaldmann]

Usually I am all for case-sensitivity, because it avoids a lot of issues.

But considering the symbol names from espressif have a lot of SYMBOLS_LIKE_THAT, I find that rather ugly, so I am considering just ignoring case, so typing the same stuff in lowercase does also match.

Opinions?

[igrr]

Do you mean symbols in the "symbols in object file" sense or as in identifiers? If we have upper case symbols in object files somewhere, that's likely a bug which we should fix. If this is about assembly tokens, then not making them case sensitive makes sense, as this is how binutils works as well. FWIW, ULP examples use lower case, except for preprocessor macros.

[ThomasWaldmann]

I meant stuff in assembly source. Instructions and also long-named constants for specific SoC register addresses / memory areas. The latter is somehow convention to be uppercase, but still, typing / reading uppercase is not nice.

I had mostly looked at ulp.h at the time I wrote that. I am using the tech ref and other non-deprecated docs now.

Yes, that example already looks much better.

[ThomasWaldmann]

ok, so does that look ok?:

• mnemonics: ulp instructions: not case-sensitive
• mnemonics: assembler directives: not case-sensitive
• operands: registers: not case-sensitive
• operands: conditions (eq, ov, ge, ...): not case-sensitive
• labels / symbols: case-sensitive

[igrr]

Yes, this looks reasonable!

[ThomasWaldmann]

guess this is how it is currently implemented, closing.