StringScanner provides for lexical scanning operations on a String. Here is an example of its usage:
s = StringScanner.new('This is an example string') s.eos? # -> false p s.scan(/\w+/) # -> "This" p s.scan(/\w+/) # -> nil p s.scan(/\s+/) # -> " " p s.scan(/\s+/) # -> nil p s.scan(/\w+/) # -> "is" s.eos? # -> false p s.scan(/\s+/) # -> " " p s.scan(/\w+/) # -> "an" p s.scan(/\s+/) # -> " " p s.scan(/\w+/) # -> "example" p s.scan(/\s+/) # -> " " p s.scan(/\w+/) # -> "string" s.eos? # -> true p s.scan(/\s+/) # -> nil p s.scan(/\w+/) # -> nil
Scanning a string means remembering the position of a scan pointer, which is just an index. The point of scanning is to move forward a bit at a time, so matches are sought after the scan pointer; usually immediately after it.
Given the string “test string”, here are the pertinent scan pointer positions:
t e s t s t r i n g
0 1 2 ... 1
0
When you scan for a pattern (a regular expression), the match must occur at the character after the scan pointer. If you use scan_until, then the match can occur anywhere after the scan pointer. In both cases, the scan pointer moves just beyond the last character of the match, ready to scan again from the next character onwards. This is demonstrated by the example above.
Method Categories
There are other methods besides the plain scanners. You can look ahead in the string without actually scanning. You can access the most recent match. You can modify the string being scanned, reset or terminate the scanner, find out or change the position of the scan pointer, skip ahead, and so on.
Advancing the Scan Pointer
Looking Ahead
Finding Where we Are
-
beginning_of_line?(bol?)
Setting Where we Are
Match Data
Miscellaneous
There are aliases to several of the methods.
static VALUE
strscan_s_mustc(VALUE self)
{
return self;
}
This method is defined for backward compatibility.
static VALUE
strscan_initialize(int argc, VALUE *argv, VALUE self)
{
struct strscanner *p;
VALUE str, options;
p = check_strscan(self);
rb_scan_args(argc, argv, "11", &str, &options);
options = rb_check_hash_type(options);
if (!NIL_P(options)) {
VALUE fixed_anchor;
ID keyword_ids[1];
keyword_ids[0] = rb_intern("fixed_anchor");
rb_get_kwargs(options, keyword_ids, 0, 1, &fixed_anchor);
if (fixed_anchor == Qundef) {
p->fixed_anchor_p = false;
}
else {
p->fixed_anchor_p = RTEST(fixed_anchor);
}
}
else {
p->fixed_anchor_p = false;
}
StringValue(str);
p->str = str;
return self;
}
Creates a new StringScanner object to scan over the given string.
If fixed_anchor is true, \A always matches the beginning of the string. Otherwise, \A always matches the current position.
dup argument is obsolete and not used now.
static VALUE
strscan_aref(VALUE self, VALUE idx)
{
const char *name;
struct strscanner *p;
long i;
GET_SCANNER(self, p);
if (! MATCHED_P(p)) return Qnil;
switch (TYPE(idx)) {
case T_SYMBOL:
idx = rb_sym2str(idx);
/* fall through */
case T_STRING:
if (!p->regex) return Qnil;
RSTRING_GETMEM(idx, name, i);
i = name_to_backref_number(&(p->regs), p->regex, name, name + i, rb_enc_get(idx));
break;
default:
i = NUM2LONG(idx);
}
if (i < 0)
i += p->regs.num_regs;
if (i < 0) return Qnil;
if (i >= p->regs.num_regs) return Qnil;
if (p->regs.beg[i] == -1) return Qnil;
return extract_range(p,
adjust_register_position(p, p->regs.beg[i]),
adjust_register_position(p, p->regs.end[i]));
}
Returns the n-th subgroup in the most recent match.
s = StringScanner.new("Fri Dec 12 1975 14:39") s.scan(/(\w+) (\w+) (\d+) /) # -> "Fri Dec 12 " s[0] # -> "Fri Dec 12 " s[1] # -> "Fri" s[2] # -> "Dec" s[3] # -> "12" s.post_match # -> "1975 14:39" s.pre_match # -> "" s.reset s.scan(/(?<wday>\w+) (?<month>\w+) (?<day>\d+) /) # -> "Fri Dec 12 " s[0] # -> "Fri Dec 12 " s[1] # -> "Fri" s[2] # -> "Dec" s[3] # -> "12" s[:wday] # -> "Fri" s[:month] # -> "Dec" s[:day] # -> "12" s.post_match # -> "1975 14:39" s.pre_match # -> ""
static VALUE
strscan_bol_p(VALUE self)
{
struct strscanner *p;
GET_SCANNER(self, p);
if (CURPTR(p) > S_PEND(p)) return Qnil;
if (p->curr == 0) return Qtrue;
return (*(CURPTR(p) - 1) == '\n') ? Qtrue : Qfalse;
}
Returns true iff the scan pointer is at the beginning of the line.
s = StringScanner.new("test\ntest\n") s.bol? # => true s.scan(/te/) s.bol? # => false s.scan(/st\n/) s.bol? # => true s.terminate s.bol? # => true
static VALUE
strscan_captures(VALUE self)
{
struct strscanner *p;
int i, num_regs;
VALUE new_ary;
GET_SCANNER(self, p);
if (! MATCHED_P(p)) return Qnil;
num_regs = p->regs.num_regs;
new_ary = rb_ary_new2(num_regs);
for (i = 1; i < num_regs; i++) {
VALUE str = extract_range(p,
adjust_register_position(p, p->regs.beg[i]),
adjust_register_position(p, p->regs.end[i]));
rb_ary_push(new_ary, str);
}
return new_ary;
}
Returns the subgroups in the most recent match (not including the full match). If nothing was priorly matched, it returns nil.
s = StringScanner.new("Fri Dec 12 1975 14:39") s.scan(/(\w+) (\w+) (\d+) /) # -> "Fri Dec 12 " s.captures # -> ["Fri", "Dec", "12"] s.scan(/(\w+) (\w+) (\d+) /) # -> nil s.captures # -> nil
static VALUE
strscan_get_charpos(VALUE self)
{
struct strscanner *p;
VALUE substr;
GET_SCANNER(self, p);
substr = rb_funcall(p->str, id_byteslice, 2, INT2FIX(0), INT2NUM(p->curr));
return rb_str_length(substr);
}
Returns the character position of the scan pointer. In the ‘reset’ position, this value is zero. In the ‘terminated’ position (i.e. the string is exhausted), this value is the size of the string.
In short, it’s a 0-based index into the string.
s = StringScanner.new("abcädeföghi") s.charpos # -> 0 s.scan_until(/ä/) # -> "abcä" s.pos # -> 5 s.charpos # -> 4
static VALUE
strscan_check(VALUE self, VALUE re)
{
return strscan_do_scan(self, re, 0, 1, 1);
}
This returns the value that scan would return, without advancing the scan pointer. The match register is affected, though.
s = StringScanner.new("Fri Dec 12 1975 14:39") s.check /Fri/ # -> "Fri" s.pos # -> 0 s.matched # -> "Fri" s.check /12/ # -> nil s.matched # -> nil
Mnemonic: it “checks” to see whether a scan will return a value.
static VALUE
strscan_check_until(VALUE self, VALUE re)
{
return strscan_do_scan(self, re, 0, 1, 0);
}
This returns the value that scan_until would return, without advancing the scan pointer. The match register is affected, though.
s = StringScanner.new("Fri Dec 12 1975 14:39") s.check_until /12/ # -> "Fri Dec 12" s.pos # -> 0 s.matched # -> 12
Mnemonic: it “checks” to see whether a scan_until will return a value.
static VALUE
strscan_clear(VALUE self)
{
rb_warning("StringScanner#clear is obsolete; use #terminate instead");
return strscan_terminate(self);
}
static VALUE
strscan_concat(VALUE self, VALUE str)
{
struct strscanner *p;
GET_SCANNER(self, p);
StringValue(str);
rb_str_append(p->str, str);
return self;
}
Appends str to the string being scanned. This method does not affect scan pointer.
s = StringScanner.new("Fri Dec 12 1975 14:39") s.scan(/Fri /) s << " +1000 GMT" s.string # -> "Fri Dec 12 1975 14:39 +1000 GMT" s.scan(/Dec/) # -> "Dec"
static VALUE
strscan_empty_p(VALUE self)
{
rb_warning("StringScanner#empty? is obsolete; use #eos? instead");
return strscan_eos_p(self);
}
static VALUE
strscan_eos_p(VALUE self)
{
struct strscanner *p;
GET_SCANNER(self, p);
return EOS_P(p) ? Qtrue : Qfalse;
}
Returns true if the scan pointer is at the end of the string.
s = StringScanner.new('test string') p s.eos? # => false s.scan(/test/) p s.eos? # => false s.terminate p s.eos? # => true
static VALUE
strscan_exist_p(VALUE self, VALUE re)
{
return strscan_do_scan(self, re, 0, 0, 0);
}
Looks ahead to see if the pattern exists anywhere in the string, without advancing the scan pointer. This predicates whether a scan_until will return a value.
s = StringScanner.new('test string') s.exist? /s/ # -> 3 s.scan /test/ # -> "test" s.exist? /s/ # -> 2 s.exist? /e/ # -> nil
static VALUE
strscan_fixed_anchor_p(VALUE self)
{
struct strscanner *p;
p = check_strscan(self);
return p->fixed_anchor_p ? Qtrue : Qfalse;
}
Whether scanner uses fixed anchor mode or not.
If fixed anchor mode is used, \A always matches the beginning of the string. Otherwise, \A always matches the current position.
static VALUE
strscan_get_byte(VALUE self)
{
struct strscanner *p;
GET_SCANNER(self, p);
CLEAR_MATCH_STATUS(p);
if (EOS_P(p))
return Qnil;
p->prev = p->curr;
p->curr++;
MATCHED(p);
adjust_registers_to_matched(p);
return extract_range(p,
adjust_register_position(p, p->regs.beg[0]),
adjust_register_position(p, p->regs.end[0]));
}
Scans one byte and returns it. This method is not multibyte character sensitive. See also: getch.
s = StringScanner.new('ab') s.get_byte # => "a" s.get_byte # => "b" s.get_byte # => nil $KCODE = 'EUC' s = StringScanner.new("\244\242") s.get_byte # => "\244" s.get_byte # => "\242" s.get_byte # => nil
static VALUE
strscan_getbyte(VALUE self)
{
rb_warning("StringScanner#getbyte is obsolete; use #get_byte instead");
return strscan_get_byte(self);
}
static VALUE
strscan_getch(VALUE self)
{
struct strscanner *p;
long len;
GET_SCANNER(self, p);
CLEAR_MATCH_STATUS(p);
if (EOS_P(p))
return Qnil;
len = rb_enc_mbclen(CURPTR(p), S_PEND(p), rb_enc_get(p->str));
len = minl(len, S_RESTLEN(p));
p->prev = p->curr;
p->curr += len;
MATCHED(p);
adjust_registers_to_matched(p);
return extract_range(p,
adjust_register_position(p, p->regs.beg[0]),
adjust_register_position(p, p->regs.end[0]));
}
Scans one character and returns it. This method is multibyte character sensitive.
s = StringScanner.new("ab") s.getch # => "a" s.getch # => "b" s.getch # => nil $KCODE = 'EUC' s = StringScanner.new("\244\242") s.getch # => "\244\242" # Japanese hira-kana "A" in EUC-JP s.getch # => nil
static VALUE
strscan_init_copy(VALUE vself, VALUE vorig)
{
struct strscanner *self, *orig;
self = check_strscan(vself);
orig = check_strscan(vorig);
if (self != orig) {
self->flags = orig->flags;
self->str = orig->str;
self->prev = orig->prev;
self->curr = orig->curr;
if (rb_reg_region_copy(&self->regs, &orig->regs))
rb_memerror();
RB_GC_GUARD(vorig);
}
return vself;
}
Duplicates a StringScanner object.
static VALUE
strscan_inspect(VALUE self)
{
struct strscanner *p;
VALUE a, b;
p = check_strscan(self);
if (NIL_P(p->str)) {
a = rb_sprintf("#<%"PRIsVALUE" (uninitialized)>", rb_obj_class(self));
return a;
}
if (EOS_P(p)) {
a = rb_sprintf("#<%"PRIsVALUE" fin>", rb_obj_class(self));
return a;
}
if (p->curr == 0) {
b = inspect2(p);
a = rb_sprintf("#<%"PRIsVALUE" %ld/%ld @ %"PRIsVALUE">",
rb_obj_class(self),
p->curr, S_LEN(p),
b);
return a;
}
a = inspect1(p);
b = inspect2(p);
a = rb_sprintf("#<%"PRIsVALUE" %ld/%ld %"PRIsVALUE" @ %"PRIsVALUE">",
rb_obj_class(self),
p->curr, S_LEN(p),
a, b);
return a;
}
Returns a string that represents the StringScanner object, showing:
-
the current position
-
the size of the string
-
the characters surrounding the scan pointer
s =
StringScanner.new(“Fri Dec 12 1975 14:39”) s.inspect # -> ‘#<StringScanner 0/21 @ “Fri D…”>’ s.scan_until /12/ # -> “Fri Dec 12” s.inspect # -> ‘#<StringScanner 10/21 “…ec 12” @ “ 1975…”>’
static VALUE
strscan_match_p(VALUE self, VALUE re)
{
return strscan_do_scan(self, re, 0, 0, 1);
}
Tests whether the given pattern is matched from the current scan pointer. Returns the length of the match, or nil. The scan pointer is not advanced.
s = StringScanner.new('test string') p s.match?(/\w+/) # -> 4 p s.match?(/\w+/) # -> 4 p s.match?("test") # -> 4 p s.match?(/\s+/) # -> nil
static VALUE
strscan_matched(VALUE self)
{
struct strscanner *p;
GET_SCANNER(self, p);
if (! MATCHED_P(p)) return Qnil;
return extract_range(p,
adjust_register_position(p, p->regs.beg[0]),
adjust_register_position(p, p->regs.end[0]));
}
Returns the last matched string.
s = StringScanner.new('test string') s.match?(/\w+/) # -> 4 s.matched # -> "test"
static VALUE
strscan_matched_p(VALUE self)
{
struct strscanner *p;
GET_SCANNER(self, p);
return MATCHED_P(p) ? Qtrue : Qfalse;
}
Returns true iff the last match was successful.
s = StringScanner.new('test string') s.match?(/\w+/) # => 4 s.matched? # => true s.match?(/\d+/) # => nil s.matched? # => false
static VALUE
strscan_matched_size(VALUE self)
{
struct strscanner *p;
GET_SCANNER(self, p);
if (! MATCHED_P(p)) return Qnil;
return INT2NUM(p->regs.end[0] - p->regs.beg[0]);
}
Returns the size of the most recent match (see matched), or nil if there was no recent match.
s = StringScanner.new('test string') s.check /\w+/ # -> "test" s.matched_size # -> 4 s.check /\d+/ # -> nil s.matched_size # -> nil
static VALUE
strscan_peek(VALUE self, VALUE vlen)
{
struct strscanner *p;
long len;
GET_SCANNER(self, p);
len = NUM2LONG(vlen);
if (EOS_P(p))
return str_new(p, "", 0);
len = minl(len, S_RESTLEN(p));
return extract_beg_len(p, p->curr, len);
}
Extracts a string corresponding to string[pos,len], without advancing the scan pointer.
s = StringScanner.new('test string') s.peek(7) # => "test st" s.peek(7) # => "test st"
static VALUE
strscan_peep(VALUE self, VALUE vlen)
{
rb_warning("StringScanner#peep is obsolete; use #peek instead");
return strscan_peek(self, vlen);
}
static VALUE
strscan_get_pos(VALUE self)
{
struct strscanner *p;
GET_SCANNER(self, p);
return INT2FIX(p->curr);
}
Returns the byte position of the scan pointer. In the ‘reset’ position, this value is zero. In the ‘terminated’ position (i.e. the string is exhausted), this value is the bytesize of the string.
In short, it’s a 0-based index into bytes of the string.
s = StringScanner.new('test string') s.pos # -> 0 s.scan_until /str/ # -> "test str" s.pos # -> 8 s.terminate # -> #<StringScanner fin> s.pos # -> 11
static VALUE
strscan_set_pos(VALUE self, VALUE v)
{
struct strscanner *p;
long i;
GET_SCANNER(self, p);
i = NUM2INT(v);
if (i < 0) i += S_LEN(p);
if (i < 0) rb_raise(rb_eRangeError, "index out of range");
if (i > S_LEN(p)) rb_raise(rb_eRangeError, "index out of range");
p->curr = i;
return INT2NUM(i);
}
Sets the byte position of the scan pointer.
s = StringScanner.new('test string') s.pos = 7 # -> 7 s.rest # -> "ring"
static VALUE
strscan_post_match(VALUE self)
{
struct strscanner *p;
GET_SCANNER(self, p);
if (! MATCHED_P(p)) return Qnil;
return extract_range(p,
adjust_register_position(p, p->regs.end[0]),
S_LEN(p));
}
Returns the post-match
(in the regular expression sense) of the last scan.s = StringScanner.new('test string') s.scan(/\w+/) # -> "test" s.scan(/\s+/) # -> " " s.pre_match # -> "test" s.post_match # -> "string"
static VALUE
strscan_pre_match(VALUE self)
{
struct strscanner *p;
GET_SCANNER(self, p);
if (! MATCHED_P(p)) return Qnil;
return extract_range(p,
0,
adjust_register_position(p, p->regs.beg[0]));
}
Returns the pre-match
(in the regular expression sense) of the last scan.s = StringScanner.new('test string') s.scan(/\w+/) # -> "test" s.scan(/\s+/) # -> " " s.pre_match # -> "test" s.post_match # -> "string"
static VALUE
strscan_reset(VALUE self)
{
struct strscanner *p;
GET_SCANNER(self, p);
p->curr = 0;
CLEAR_MATCH_STATUS(p);
return self;
}
Reset the scan pointer (index 0) and clear matching data.
static VALUE
strscan_rest(VALUE self)
{
struct strscanner *p;
GET_SCANNER(self, p);
if (EOS_P(p)) {
return str_new(p, "", 0);
}
return extract_range(p, p->curr, S_LEN(p));
}
Returns the “rest” of the string (i.e. everything after the scan pointer). If there is no more data (eos? = true), it returns "".
static VALUE
strscan_rest_p(VALUE self)
{
struct strscanner *p;
GET_SCANNER(self, p);
return EOS_P(p) ? Qfalse : Qtrue;
}
static VALUE
strscan_rest_size(VALUE self)
{
struct strscanner *p;
long i;
GET_SCANNER(self, p);
if (EOS_P(p)) {
return INT2FIX(0);
}
i = S_RESTLEN(p);
return INT2FIX(i);
}
s.rest_size is equivalent to s.rest.size.
static VALUE
strscan_restsize(VALUE self)
{
rb_warning("StringScanner#restsize is obsolete; use #rest_size instead");
return strscan_rest_size(self);
}
s.restsize is equivalent to s.rest_size. This method is obsolete; use rest_size instead.
static VALUE
strscan_scan(VALUE self, VALUE re)
{
return strscan_do_scan(self, re, 1, 1, 1);
}
Tries to match with pattern at the current position. If there’s a match, the scanner advances the “scan pointer” and returns the matched string. Otherwise, the scanner returns nil.
s = StringScanner.new('test string') p s.scan(/\w+/) # -> "test" p s.scan(/\w+/) # -> nil p s.scan(/\s+/) # -> " " p s.scan("str") # -> "str" p s.scan(/\w+/) # -> "ing" p s.scan(/./) # -> nil
static VALUE
strscan_scan_full(VALUE self, VALUE re, VALUE s, VALUE f)
{
return strscan_do_scan(self, re, RTEST(s), RTEST(f), 1);
}
Tests whether the given pattern is matched from the current scan pointer. Advances the scan pointer if advance_pointer_p is true. Returns the matched string if return_string_p is true. The match register is affected.
“full” means “#scan with full parameters”.
static VALUE
strscan_scan_until(VALUE self, VALUE re)
{
return strscan_do_scan(self, re, 1, 1, 0);
}
Scans the string until the pattern is matched. Returns the substring up to and including the end of the match, advancing the scan pointer to that location. If there is no match, nil is returned.
s = StringScanner.new("Fri Dec 12 1975 14:39") s.scan_until(/1/) # -> "Fri Dec 1" s.pre_match # -> "Fri Dec " s.scan_until(/XYZ/) # -> nil
static VALUE
strscan_search_full(VALUE self, VALUE re, VALUE s, VALUE f)
{
return strscan_do_scan(self, re, RTEST(s), RTEST(f), 0);
}
Scans the string until the pattern is matched. Advances the scan pointer if advance_pointer_p, otherwise not. Returns the matched string if return_string_p is true, otherwise returns the number of bytes advanced. This method does affect the match register.
static VALUE
strscan_size(VALUE self)
{
struct strscanner *p;
GET_SCANNER(self, p);
if (! MATCHED_P(p)) return Qnil;
return INT2FIX(p->regs.num_regs);
}
Returns the amount of subgroups in the most recent match. The full match counts as a subgroup.
s = StringScanner.new("Fri Dec 12 1975 14:39") s.scan(/(\w+) (\w+) (\d+) /) # -> "Fri Dec 12 " s.size # -> 4
static VALUE
strscan_skip(VALUE self, VALUE re)
{
return strscan_do_scan(self, re, 1, 0, 1);
}
Attempts to skip over the given pattern beginning with the scan pointer. If it matches, the scan pointer is advanced to the end of the match, and the length of the match is returned. Otherwise, nil is returned.
It’s similar to scan, but without returning the matched string.
s = StringScanner.new('test string') p s.skip(/\w+/) # -> 4 p s.skip(/\w+/) # -> nil p s.skip(/\s+/) # -> 1 p s.skip("st") # -> 2 p s.skip(/\w+/) # -> 4 p s.skip(/./) # -> nil
static VALUE
strscan_skip_until(VALUE self, VALUE re)
{
return strscan_do_scan(self, re, 1, 0, 0);
}
Advances the scan pointer until pattern is matched and consumed. Returns the number of bytes advanced, or nil if no match was found.
Look ahead to match pattern, and advance the scan pointer to the end of the match. Return the number of characters advanced, or nil if the match was unsuccessful.
It’s similar to scan_until, but without returning the intervening string.
s = StringScanner.new("Fri Dec 12 1975 14:39") s.skip_until /12/ # -> 10 s #
static VALUE
strscan_get_string(VALUE self)
{
struct strscanner *p;
GET_SCANNER(self, p);
return p->str;
}
Returns the string being scanned.
static VALUE
strscan_set_string(VALUE self, VALUE str)
{
struct strscanner *p = check_strscan(self);
StringValue(str);
p->str = str;
p->curr = 0;
CLEAR_MATCH_STATUS(p);
return str;
}
Changes the string being scanned to str and resets the scanner. Returns str.
static VALUE
strscan_terminate(VALUE self)
{
struct strscanner *p;
GET_SCANNER(self, p);
p->curr = S_LEN(p);
CLEAR_MATCH_STATUS(p);
return self;
}
Sets the scan pointer to the end of the string and clear matching data.
static VALUE
strscan_unscan(VALUE self)
{
struct strscanner *p;
GET_SCANNER(self, p);
if (! MATCHED_P(p))
rb_raise(ScanError, "unscan failed: previous match record not exist");
p->curr = p->prev;
CLEAR_MATCH_STATUS(p);
return self;
}
Sets the scan pointer to the previous position. Only one previous position is remembered, and it changes with each scanning operation.
s = StringScanner.new('test string') s.scan(/\w+/) # => "test" s.unscan s.scan(/../) # => "te" s.scan(/\d/) # => nil s.unscan # ScanError: unscan failed: previous match record not exist
static VALUE
strscan_values_at(int argc, VALUE *argv, VALUE self)
{
struct strscanner *p;
long i;
VALUE new_ary;
GET_SCANNER(self, p);
if (! MATCHED_P(p)) return Qnil;
new_ary = rb_ary_new2(argc);
for (i = 0; i<argc; i++) {
rb_ary_push(new_ary, strscan_aref(self, argv[i]));
}
return new_ary;
}
Returns the subgroups in the most recent match at the given indices. If nothing was priorly matched, it returns nil.
s = StringScanner.new("Fri Dec 12 1975 14:39") s.scan(/(\w+) (\w+) (\d+) /) # -> "Fri Dec 12 " s.values_at 0, -1, 5, 2 # -> ["Fri Dec 12 ", "12", nil, "Dec"] s.scan(/(\w+) (\w+) (\d+) /) # -> nil s.values_at 0, -1, 5, 2 # -> nil