Right, the first link I posted above has info on how to use app passwords, etc. Although, that seems to only be necessary if two factor is enabled.

Is there way to car a string or pattern match a string’s head and rest? e.g. (car "string")

You can use string-ref.

Consider using string->list then work on the list instead.

@samth it can get head of string, but how to get rest of string?

I found there is function called string->immutable-string , but it still a string. (no car/head cdr/tail function provided)

@soegaard2 This works, thanks.

@chansey97 (define (car-str str) (substring str 0 1)) (define (cdr-str str) (substring str 1))

@badkins no i dont have two factor enabled

@zenspider how did you create an app specific password?

@badkins yes, but it can not be pattern matched.

Why not?

I would also suggest that usually processing strings character-by-character is not the right approach — string-ref (and similarly substring and string->list) doesn’t handle more complicated cases like combining characters.

sorry, I am not familiar with pattern matching. But you provide two functions, so we must use if else to check the return value.

Are you wanting something like Haskell’s pattern match on x:xs for strings?

@brentgordon146 https://support.google.com/accounts/answer/185833?hl=en

@badkins yes It would be nice, if we can do (match some-string [() ......] [(x xs) ......])

I wouldn’t be surprised if you’re able to create a custom matcher to handle what you want, but I’d probably just use: (cond [(empty-string? str) ... ] [else ... ])

You can do: (match s [(app list->string (list)) ...] [(app list->string (cons x xs)) ...])

they can only be used with two factor authentication accounts

eeewww :slightly_smiling_face:

You can also do: (define-match-expander str (syntax-rules () [(_ p) (app list->string p)])) (match s [(str (list)) ...] [(str (cons x xs)) ...])

@samth In your first example, the x and xs are bounded? it is surprising! Could you give me some referece of app ? I’m sorry, the keyword app is not friendly to search engine.

See the grammar here: https://docs.racket-lang.org/reference/match.html#(form._((lib._racket%2Fmatch..rkt)._match))

But again, I want to encourage you to think about whether character-by-character processing is really the right thing. For example, what is the first character in ":jp:"?

Because the result from string-ref is #\🇯

@samth Yes, finally will preprocess the string by using list->string, and use list instead.

I’m guessing @samth meant string->list above

yes

@badkins yes.

just typo..

@samth I am just writing a toy parser by functional way in racket, I don’t want to use cursor for the position of the string.

@samth What do you mean by For example, what is the first character in ":jp:"?  Does string->list support unicode?

What do you mean by “support unicode”? string-length of that string produces 2.

Flags (and many other emoji) are made up of multiple combining characters.

@samth OK, that’s not a problem.

@samth but wait, Does Emoji and Unicode overlap? For example:

Is this unicode?

I see this at https://home.unicode.org/

yes, all the characters you see anywhere are unicode

:grinning: <— this is emoji from Slack

yes, and it’s a unicode character

OMG! emojis are exactly unicode characters? I did not know before…

well, “character” is a complicated concept, and some emoji (like that flag) are made up of multiple characters (in some senses)

But I copy this face in slack, it give me ": grinning :"

that’s how you can type it in slack

@samth I used to think emoji were just ASCII characters, and the text renderer parse these characters between ::. Thank you for telling me this knowledge!

I copied a japanese flag from https://emojipedia.org/flag-japan/ to Emacs.

It return 2.

It’s amazing.

In slack, we just type :jp: