ref: 016b839ff4ca335b4e13c52a0a264cf8da429150
parent: 44f7d8fd2573d566042c863c45aed426433cc6b4
author: Jeff Bezanson <jeff.bezanson@gmail.com>
date: Sat Aug 19 08:27:27 EDT 2017
some printing improvements - don't print shared references to cprims, and estimate their size better. previously you could get output like this: ``` (#0=#\a #0#) ``` Now it looks like: ``` (#\a #\a) ``` - print `#\ ` and `#\newline` instead of `#\space` and `#\linefeed`. these seem to be more standard.
--- a/print.c
+++ b/print.c
@@ -89,9 +89,6 @@
for(i=0; i < vector_size(v); i++)
print_traverse(vector_elt(v,i));
}
- else if (iscprim(v)) {
- mark_cons(v);
- }
else if (isclosure(v)) {
mark_cons(v);
function_t *f = (function_t*)ptr(v);
@@ -171,7 +168,7 @@
if (fl_isstring(v))
return (cv_len((cvalue_t*)ptr(v)) < SMALL_STR_LEN);
return (isfixnum(v) || isbuiltin(v) || v==FL_F || v==FL_T || v==FL_NIL ||
- v == FL_EOF);
+ v == FL_EOF || iscprim(v));
}
static int smallp(value_t v)
@@ -208,6 +205,8 @@
// get the width of an expression if we can do so cheaply
if (issymbol(v))
return u8_strwidth(symbol_name(v));
+ if (iscprim(v) && cp_class((cprim_t*)ptr(v)) == wchartype)
+ return 4;
return -1;
}
@@ -441,9 +440,13 @@
}
}
break;
- case TAG_CVALUE:
case TAG_CPRIM:
- if (v == UNBOUND) { outs("#<undefined>", f); break; }
+ if (v == UNBOUND)
+ outs("#<undefined>", f);
+ else
+ cvalue_print(f, v);
+ break;
+ case TAG_CVALUE:
case TAG_VECTOR:
case TAG_CONS:
if (print_circle_prefix(f, v)) break;
@@ -477,7 +480,7 @@
outc(']', f);
break;
}
- if (iscvalue(v) || iscprim(v))
+ if (iscvalue(v))
cvalue_print(f, v);
else
print_pair(f, v);
@@ -640,13 +643,13 @@
else if (wc == 0x07) outsn("alarm", f, 5);
else if (wc == 0x08) outsn("backspace", f, 9);
else if (wc == 0x09) outsn("tab", f, 3);
- else if (wc == 0x0A) outsn("linefeed", f, 8);
- //else if (wc == 0x0A) outsn("newline", f, 7);
+ //else if (wc == 0x0A) outsn("linefeed", f, 8);
+ else if (wc == 0x0A) outsn("newline", f, 7);
else if (wc == 0x0B) outsn("vtab", f, 4);
else if (wc == 0x0C) outsn("page", f, 4);
else if (wc == 0x0D) outsn("return", f, 6);
else if (wc == 0x1B) outsn("esc", f, 3);
- else if (wc == 0x20) outsn("space", f, 5);
+ //else if (wc == 0x20) outsn("space", f, 5);
else if (wc == 0x7F) outsn("delete", f, 6);
else if (iswprint(wc)) outs(seq, f);
else HPOS+=ios_printf(f, "x%04x", (int)wc);
--- a/string.c
+++ b/string.c
@@ -230,7 +230,7 @@
argcount("char.upcase", nargs, 1);
cprim_t *cp = (cprim_t*)ptr(args[0]);
if (!iscprim(args[0]) || cp_class(cp) != wchartype)
- type_error("char.upcase", "wchar", args[0]);
+ type_error("char.upcase", "wchar", args[0]);
return mk_wchar(towupper(*(int32_t*)cp_data(cp)));
}
value_t fl_char_downcase(value_t *args, u_int32_t nargs)
@@ -238,7 +238,7 @@
argcount("char.downcase", nargs, 1);
cprim_t *cp = (cprim_t*)ptr(args[0]);
if (!iscprim(args[0]) || cp_class(cp) != wchartype)
- type_error("char.downcase", "wchar", args[0]);
+ type_error("char.downcase", "wchar", args[0]);
return mk_wchar(towlower(*(int32_t*)cp_data(cp)));
}