关于s[s.length,1]为什么返回""

ROR群(15917997)里有朋友问到一个问题
s="hello"
s[s.length,1]为什么返回"",而不是返回nil或其他

群里展开了讨论,我翻了翻ruby源代码,试着做出解释:
首先查了API,
string[integer, integer]会执行

rb_str_substr(str, NUM2LONG(argv[0]), NUM2LONG(argv[1]));

rb_str_substr是这样的:

VALUE
rb_str_substr(str, beg, len)
    VALUE str;
    long beg, len;
{
    VALUE str2;

    if (len < 0) return Qnil;
    if (beg > RSTRING(str)->len) return Qnil;
    if (beg < 0) {
	beg += RSTRING(str)->len;
	if (beg < 0) return Qnil;
    }
    if (beg + len > RSTRING(str)->len) {
	len = RSTRING(str)->len - beg;
    }
    if (len < 0) {
	len = 0;
    }
    if (len == 0) {
	str2 = rb_str_new5(str,0,0);
    }
    else if (len > sizeof(struct RString)/2 &&
	beg + len == RSTRING(str)->len && !FL_TEST(str, STR_ASSOC)) {
	str2 = rb_str_new4(str);
	str2 = str_new3(rb_obj_class(str2), str2);
	RSTRING(str2)->ptr += RSTRING(str2)->len - len;
	RSTRING(str2)->len = len;
    }
    else {
	str2 = rb_str_new5(str, RSTRING(str)->ptr+beg, len);
    }
    OBJ_INFECT(str2, str);

    return str2;
}

看到这两句

if (len < 0) return Qnil;
if (beg > RSTRING(str)->len) return Qnil;

意思是说要截取的字串长度如果为负数,返回nil
开始的位置超出了原字串的长度,也返回nil

往下看,因为s.lenght刚好等于RSTRING(str)->len,其实会执行到

str2 = rb_str_new5(str, RSTRING(str)->ptr+beg, len);

rb_str_new5是这样的

VALUE
rb_str_new5(obj, ptr, len)
    VALUE obj;
    const char *ptr;
    long len;
{
    return str_new(rb_obj_class(obj), ptr, len);
}

主要是调用了str_new

static VALUE
str_new(klass, ptr, len)
    VALUE klass;
    const char *ptr;
    long len;
{
    VALUE str;

    if (len < 0) {
	rb_raise(rb_eArgError, "negative string size (or size too big)");
    }

    str = str_alloc(klass);
    RSTRING(str)->len = len;
    RSTRING(str)->aux.capa = len;
    RSTRING(str)->ptr = ALLOC_N(char,len+1);
    if (ptr) {
	memcpy(RSTRING(str)->ptr, ptr, len);
    }
    RSTRING(str)->ptr[len] = '\0';
    return str;
}

到了

  if (ptr) {
	memcpy(RSTRING(str)->ptr, ptr, len);
    }

ptr其实已经指向了"hello"的最后一个字符,也就是'\0'了,执行完memcpy后,str成了"",再最后执行

RSTRING(str)->ptr[len] = '\0';

这一句加""加上'\0',是c语言的空字符串.

最后的结果就是""了.

路线是这样的
s[s.length,1]
==> rb_str_substr("hello", 5, 1);
==> rb_str_new5("hello",ptr=5,1);
==> str_new("hello",ptr=5,1)
==>memcpy("hello",ptr=5,1)
==>""