| pkhuong | do you think slime could use the current readtable to know whether to treat, e.g., [] and {} as list delimiters? |
| Riastradh | Treat those as delimiters in what context, precisely? |
| pkhuong | Riastradh: indentation, mostly. |
| nyef | pkhuong: How would it know, other than by trying to READ them? |
| rudi | pkhuong: slime doesn't do indentation; you'd have to hack cl-indent.el |
| tritchey_ | is there anything that might be left around in an sbcl build that sh clean.sh wouldn't, err _clean_ that might cause problems if I moved the tree to a different machine? |
| rr-- | "{ up | down | lower | higher } in the call stack" |
| Zhivago | what's the difference between up and higher? |
| rr-- | this terminology assumes the convention that the call stack grow ... up? down? no convention? grows everyone has their own convention? |
| pkhuong | nyef: get-macro-character != nil looks like a decent heuristic. |
| nyef | minion: Advice on heuristic? |
| minion | #11953: Of course, this is a heuristic, which is a fancy way of saying that it doesn't work. |
| nyef | rr--: Interestingly, it assumes that activation records are provided as a -stack-. (Why yes, some people do heap-allocate function activation records.) pkhuong: Would you expect #\< to be a macro character in any custom readtable that can still parse most lisp code? What about funky syntaxen which can mix [ with } as a delimiter? There are interesting possibilities in terms of parsing lisp code via the reader with lispm-style rubout handling style hacks in the underlying streams... |
| pkhuong | nyef: xmlisp? (: |
| nyef | Was more thinking xml-mixed-mode, tbh. |
| karstensrage | lol I joined on another network and was the only one in there and I thought "did everyone just decide to give up lisp for good?" |
| rahul | you didn't get the memo? |
| nyef | Heck, even I've been doing some work in that-other-python. |
| rr-- | is there a special/specific term for the process/step whereby the reader converts symbol names to upper case |
| rahul | gasp |
| nyef | rr--: Maybe? It doesn't always happen, and the readers behavior is fairly well specified, so the spec might have a name for it. |
| rr-- | maybe something along the lines of 'canonicalization' (i am just making that up) |
| rahul | for presentational simplicity, the examples assume that the readtable case of the current readtable is :upcase. thta's all I can find about case |
| nyef | Well, the spec doesn't seem to have a name for the process, preferring to describe it longhand each time. (sections 2.2 and 23.1.2.) |
| rahul | yeah it's merely an 'effect' I'd call it case conversion althought the algo described in 2.2 for this is not really the right one, considering :invert |
| nyef | How so? |
| rahul | you need to preserve case until you get to the end of the token because you don't know if _all_ the chars are of the same case till you're done |
| nyef | Ahh. |