Class: String

Inherits:
Object
  • Object
show all
Defined in:
(unknown)

Instance Method Summary collapse

Instance Method Details

#bit_and(other) ⇒ Object



926
927
928
929
930
931
932
933
934
935
936
937
# File 'ext/string_bits/string_bits.c', line 926

static VALUE
rb_str_bit_and(VALUE self, VALUE other)
{
    check_binary_op_lengths(self, other);
    ssize_t len = RSTRING_LEN(self);
    VALUE result = alloc_result(self);
    const unsigned char *a = (const unsigned char *)RSTRING_PTR(self);
    const unsigned char *b = (const unsigned char *)RSTRING_PTR(other);
    unsigned char *dst = (unsigned char *)RSTRING_PTR(result);
    for (ssize_t i = 0; i < len; i++) dst[i] = a[i] & b[i];
    return result;
}

#bit_and!(other) ⇒ Object



939
940
941
942
943
944
945
946
947
948
949
# File 'ext/string_bits/string_bits.c', line 939

static VALUE
rb_str_bit_and_bang(VALUE self, VALUE other)
{
    check_binary_op_lengths(self, other);
    rb_str_modify(self);
    ssize_t len = RSTRING_LEN(self);
    unsigned char *a = (unsigned char *)RSTRING_PTR(self);
    const unsigned char *b = (const unsigned char *)RSTRING_PTR(other);
    for (ssize_t i = 0; i < len; i++) a[i] &= b[i];
    return self;
}

#bit_at(*args) ⇒ Object

String#bit_at(n, lsb_first: true) -> true or false

bit_at uses flat/Arrow convention: byte_index = n/8 from start, bit = n%8 from LSB e.g. “xAAxCC”: bit 0..7 live in byte=0xAA, bit 8..15 live in byte=0xCC

str = "\xFF\xAA" # 11111111 10101010
str.bit_at(0) # => true (1st bit is set)
str.bit_at(7) # => true (8th bit is set)
str.bit_at(8) # => false (9th bit is clear)
str.bit_at(9) # => true (10th bit is set)
str.bit_at(16) # => nil


351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
# File 'ext/string_bits/string_bits.c', line 351

static VALUE
rb_str_bit_at(int argc, VALUE *argv, VALUE self)
{
    VALUE n, opts;
    rb_scan_args(argc, argv, "1:", &n, &opts);
    validate_option_hash(opts, SB_KW_LSB_FIRST);

    if (!rb_integer_type_p(n)) {
        rb_raise(rb_eTypeError, "bit index must be an integer");
    }
    ssize_t idx = integer_to_bit_idx(n);
    if (idx < 0) {
        rb_raise(rb_eArgError, "bit index must be non-negative");
    }
    ssize_t size = RSTRING_LEN(self) * 8;
    if (size <= idx) {
        return Qnil;
    }

    int lsb_first = parse_lsb_first_opt(opts);

    if (!lsb_first) {
        idx = (idx & ~7L) | (7 - (idx & 7L));
    }

    if (test_bit(RSTRING_PTR(self), idx)) {
        return Qtrue;
    } else {
        return Qfalse;
    }
}

#bit_countObject



383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
# File 'ext/string_bits/string_bits.c', line 383

static VALUE
rb_str_bit_count(VALUE self)
{
    ssize_t count = 0;
    ssize_t len = RSTRING_LEN(self);
    const char *str = RSTRING_PTR(self);
    ssize_t off = 0;
    ssize_t unrolled_end = len & ~31L;
    ssize_t aligned_end  = len & ~7L;

    /* Use memcpy to avoid unaligned loads (SIGBUS on SPARC, MIPS, etc.)
     * and strict-aliasing violations. Modern compilers fold 8-byte memcpy
     * into a single load on platforms that allow unaligned access. */
    for (; off < unrolled_end; off += 32) {
        uint64_t w0, w1, w2, w3;
        memcpy(&w0, str + off,      8);
        memcpy(&w1, str + off + 8,  8);
        memcpy(&w2, str + off + 16, 8);
        memcpy(&w3, str + off + 24, 8);
        count += sb_popcount64(w0);
        count += sb_popcount64(w1);
        count += sb_popcount64(w2);
        count += sb_popcount64(w3);
    }

    for (; off < aligned_end; off += 8) {
        uint64_t w;
        memcpy(&w, str + off, 8);
        count += sb_popcount64(w);
    }

    ssize_t remainder = len - aligned_end;
    if (remainder > 0) {
        uint64_t last = 0;
        const unsigned char *tail = (const unsigned char *)(str + aligned_end);
        for (ssize_t i = 0; i < remainder; i++) {
            last |= (uint64_t)tail[i] << (i * 8);
        }
        count += sb_popcount64(last);
    }

    return SSIZET2NUM(count);
}

#bit_fields(*args) ⇒ Object

String#bit_fields(*bitlens, lsb_first: true) -> Array String#bit_fields(*bitlens, lsb_first: true) { |*fields| } -> self

Non-iterator complement of each_bit_field. Without a block, returns an Array of all extracted records. With a single bitlen the array is flat (matching each_bit_field(n).to_a); with multiple bitlens each record is itself an Array (matching each_bit_field(a, b, …).to_a).

With a block, behaves identically to each_bit_field without with: — yielding one Integer per field and returning self.

Porting to Ruby Core:

1. Move alongside each_bit_field in string.c.
2. Share extract_uint64 and the bitlen validation logic.
3. Register with rb_define_method in Init_String().


1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
# File 'ext/string_bits/string_bits.c', line 1144

static VALUE
rb_str_bit_fields(int argc, VALUE *argv, VALUE self)
{
    VALUE rest, opts;
    rb_scan_args(argc, argv, "*:", &rest, &opts);
    validate_option_hash(opts, SB_KW_LSB_FIRST);

    ssize_t num_fields = RARRAY_LEN(rest);
    if (num_fields == 0) {
        rb_raise(rb_eArgError, "wrong number of arguments (given 0, expected 1+)");
    }

    ssize_t *bitlens = ALLOCA_N(ssize_t, num_fields);
    ssize_t step = 0;
    for (ssize_t f = 0; f < num_fields; f++) {
        VALUE v = RARRAY_AREF(rest, f);
        if (!rb_integer_type_p(v)) {
            rb_raise(rb_eTypeError, "bitlen must be an integer");
        }
        ssize_t bl = NUM2SSIZET(v);
        if (bl <= 0) {
            rb_raise(rb_eArgError, "bitlen must be positive");
        }
        if (bl > 64) {
            rb_raise(rb_eArgError, "bitlen must be <= 64 (got %ld)", bl);
        }
        bitlens[f] = bl;
        step += bl;
    }

    int lsb_first = parse_lsb_first_opt(opts);

    ssize_t src_len = RSTRING_LEN(self);
    ssize_t total_bits = src_len * 8;
    ssize_t iterations = total_bits / step;

    int have_block = rb_block_given_p();
    VALUE result = have_block ? Qnil : rb_ary_new_capa(iterations);

    VALUE *field_vals = ALLOCA_N(VALUE, num_fields);

    for (ssize_t iter = 0; iter < iterations; iter++) {
        ssize_t base_bit = iter * step;
        const unsigned char *src = (const unsigned char *)RSTRING_PTR(self);
        ssize_t field_bit = base_bit;
        for (ssize_t f = 0; f < num_fields; f++) {
            uint64_t val = extract_uint64(src, src_len, field_bit, bitlens[f], lsb_first);
            field_vals[f] = ULL2NUM(val);
            field_bit += bitlens[f];
        }
        if (have_block) {
            rb_yield_values2((int)num_fields, field_vals);
        } else if (num_fields == 1) {
            rb_ary_push(result, field_vals[0]);
        } else {
            rb_ary_push(result, rb_ary_new_from_values(num_fields, field_vals));
        }
    }

    return have_block ? self : result;
}

#bit_notObject



905
906
907
908
909
910
911
912
913
914
# File 'ext/string_bits/string_bits.c', line 905

static VALUE
rb_str_bit_not(VALUE self)
{
    ssize_t len = RSTRING_LEN(self);
    VALUE result = alloc_result(self);
    const unsigned char *src = (const unsigned char *)RSTRING_PTR(self);
    unsigned char *dst = (unsigned char *)RSTRING_PTR(result);
    for (ssize_t i = 0; i < len; i++) dst[i] = ~src[i];
    return result;
}

#bit_not!Object



916
917
918
919
920
921
922
923
924
# File 'ext/string_bits/string_bits.c', line 916

static VALUE
rb_str_bit_not_bang(VALUE self)
{
    rb_str_modify(self);
    ssize_t len = RSTRING_LEN(self);
    unsigned char *ptr = (unsigned char *)RSTRING_PTR(self);
    for (ssize_t i = 0; i < len; i++) ptr[i] = ~ptr[i];
    return self;
}

#bit_or(other) ⇒ Object



951
952
953
954
955
956
957
958
959
960
961
962
# File 'ext/string_bits/string_bits.c', line 951

static VALUE
rb_str_bit_or(VALUE self, VALUE other)
{
    check_binary_op_lengths(self, other);
    ssize_t len = RSTRING_LEN(self);
    VALUE result = alloc_result(self);
    const unsigned char *a = (const unsigned char *)RSTRING_PTR(self);
    const unsigned char *b = (const unsigned char *)RSTRING_PTR(other);
    unsigned char *dst = (unsigned char *)RSTRING_PTR(result);
    for (ssize_t i = 0; i < len; i++) dst[i] = a[i] | b[i];
    return result;
}

#bit_or!(other) ⇒ Object



964
965
966
967
968
969
970
971
972
973
974
# File 'ext/string_bits/string_bits.c', line 964

static VALUE
rb_str_bit_or_bang(VALUE self, VALUE other)
{
    check_binary_op_lengths(self, other);
    rb_str_modify(self);
    ssize_t len = RSTRING_LEN(self);
    unsigned char *a = (unsigned char *)RSTRING_PTR(self);
    const unsigned char *b = (const unsigned char *)RSTRING_PTR(other);
    for (ssize_t i = 0; i < len; i++) a[i] |= b[i];
    return self;
}

#bit_run_count(*args) ⇒ Object

String#bit_run_count(pos, bit) -> Integer | nil

Returns the length of the consecutive run of ‘bit` starting at flat position `pos`. Returns nil when `pos` is out of range or the bit at `pos` does not equal `bit`.

‘bit` accepts 0, 1, false, or true (false/true are aliases for 0/1, matching the values yielded by each_bit_run).

Counts forward from ‘pos` toward higher bit indices.

Inspired by Gauche Scheme’s (bitvector-count-run bit bvec i).

Uses the same flat LSB-first addressing as bit_at: byte bit pos%8.

Porting to Ruby Core:

1. Move to string.c; register in Init_String().
2. Reuse integer_to_bit_idx for consistent Bignum handling.


1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
# File 'ext/string_bits/string_bits.c', line 1294

static VALUE
rb_str_bit_run_count(int argc, VALUE *argv, VALUE self)
{
    VALUE pos_val, bit_val, opts;
    rb_scan_args(argc, argv, "20:", &pos_val, &bit_val, &opts);
    validate_option_hash(opts, SB_KW_LSB_FIRST);
    int lsb_first = parse_lsb_first_opt(opts);

    if (!rb_integer_type_p(pos_val)) {
        rb_raise(rb_eTypeError, "position must be an integer");
    }
    int target;
    if (bit_val == Qtrue || bit_val == INT2FIX(1)) {
        target = 1;
    } else if (bit_val == Qfalse || bit_val == INT2FIX(0)) {
        target = 0;
    } else {
        rb_raise(rb_eArgError, "bit must be 0, 1, false, or true");
    }
    ssize_t pos     = integer_to_bit_idx(pos_val);
    ssize_t src_len = RSTRING_LEN(self);
    if (pos < 0 || pos >= src_len * 8) return Qnil;

    const unsigned char *src = (const unsigned char *)RSTRING_PTR(self);
    if (lsb_first) {
        if (((src[pos >> 3] >> (pos & 7)) & 1) != target) return Qnil;
        return SSIZET2NUM(count_run_lsb(src, src_len, pos, target));
    }

    if (logical_get_bit(src, pos, 0) != target) return Qnil;

    ssize_t run = 1;
    ssize_t total_bits = src_len * 8;
    while (pos + run < total_bits && logical_get_bit(src, pos + run, 0) == target) {
        run++;
    }
    return SSIZET2NUM(run);
}

#bit_runs(*args) ⇒ Object

String#bit_runs(lsb_first: true) -> Array String#bit_runs(lsb_first: true) { |bit, len| } -> self

Non-iterator complement of each_bit_run. Without a block, collects all (bit, run_length) pairs into an Array and returns it. With a block, yields each pair and returns self.

Follows the same pattern as String#bytes vs String#each_byte.

Porting to Ruby Core:

1. Move to string.c alongside each_bit_run; register in Init_String().


1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
# File 'ext/string_bits/string_bits.c', line 1401

static VALUE
rb_str_bit_runs(int argc, VALUE *argv, VALUE self)
{
    int lsb_first = parse_lsb_first(argc, argv);
    ssize_t src_len   = RSTRING_LEN(self);
    int have_block = rb_block_given_p();

    if (src_len == 0) return have_block ? self : rb_ary_new();

    ssize_t total_bits = src_len * 8;
    VALUE result    = have_block ? Qnil : rb_ary_new();

    if (lsb_first) {
        ssize_t pos = 0;
        while (pos < total_bits) {
            const unsigned char *src = (const unsigned char *)RSTRING_PTR(self);
            int bit  = (src[pos >> 3] >> (pos & 7)) & 1;
            ssize_t run = count_run_lsb(src, src_len, pos, bit);
            VALUE bval = bit ? Qtrue : Qfalse;
            VALUE lval = SSIZET2NUM(run);
            have_block ? rb_yield_values(2, bval, lval)
                       : rb_ary_push(result, rb_assoc_new(bval, lval));
            pos += run;
        }
    } else {
        ssize_t pos = 0;
        while (pos < total_bits) {
            const unsigned char *src = (const unsigned char *)RSTRING_PTR(self);
            int bit = logical_get_bit(src, pos, 0);
            ssize_t run = 1;
            while (pos + run < total_bits && logical_get_bit(src, pos + run, 0) == bit) {
                run++;
            }
            VALUE bval = bit ? Qtrue : Qfalse;
            VALUE lval = SSIZET2NUM(run);
            have_block ? rb_yield_values(2, bval, lval)
                       : rb_ary_push(result, rb_assoc_new(bval, lval));
            pos += run;
        }
    }

    return have_block ? self : result;
}

#bit_slice(*args) ⇒ Object

String#bit_slice(bit_offset, bit_length) -> String String#bit_slice(range) -> String

str = "\xFF\x00" # 11111111 00000000
str.bit_slice(4, 8) # => "\xF0" (11110000)


721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
# File 'ext/string_bits/string_bits.c', line 721

static VALUE
rb_str_bit_slice(int argc, VALUE *argv, VALUE self)
{
    ssize_t src_len = RSTRING_LEN(self);
    ssize_t total_bits = src_len * 8;
    ssize_t offset, length;
    VALUE v0, v1, opts;
    int n_pos = rb_scan_args(argc, argv, "11:", &v0, &v1, &opts);
    validate_option_hash(opts, SB_KW_LSB_FIRST);
    int lsb_first = parse_lsb_first_opt(opts);

    if (n_pos == 1 && rb_obj_is_kind_of(v0, rb_cRange)) {
        ssize_t beg, len;
        if (!RTEST(sb_range_beg_len(v0, &beg, &len, total_bits, 0))) {
            return Qnil;
        }
        offset = beg;
        length = len;
    }
    else if (n_pos == 2) {
        if (!rb_integer_type_p(v0) || !rb_integer_type_p(v1)) {
            return Qnil;
        }

        offset = integer_to_bit_idx(v0);
        length = integer_to_bit_idx(v1);

        if (offset < 0 || length < 0) return Qnil;
    }
    else if (n_pos == 1) {
        return Qnil;
    }
    else {
        rb_raise(rb_eArgError,
                 "wrong number of arguments (given %d, expected 1 or 2)", n_pos);
    }

    if (offset > total_bits) return Qnil;
    ssize_t available = total_bits - offset;
    if (length > available) length = available;

    if (length == 0) return rb_str_new("", 0);

    ssize_t out_bytes = (length + 7) / 8;
    VALUE result = rb_str_buf_new(out_bytes);
    rb_str_resize(result, out_bytes);
    rb_enc_associate(result, rb_enc_get(self));
    unsigned char *dst = (unsigned char *)RSTRING_PTR(result);
    const unsigned char *src = (const unsigned char *)RSTRING_PTR(self);

    memset(dst, 0, out_bytes);

    if (lsb_first) {
        bit_copy_core(dst, 0, src, src_len, offset, length);
    } else {
        ssize_t dst_bit = 0;
        ssize_t start_byte = offset >> 3;
        ssize_t end_byte = (offset + length - 1) >> 3;

        for (ssize_t b = start_byte; b <= end_byte; b++) {
            ssize_t b_start_l = b << 3;
            ssize_t b_end_l = b_start_l + 7;
            ssize_t l_min = (offset > b_start_l) ? offset : b_start_l;
            ssize_t l_max = ((offset + length - 1) < b_end_l) ? (offset + length - 1) : b_end_l;

            ssize_t p_min = b_start_l + (7 - (l_max & 7L));
            ssize_t p_max = b_start_l + (7 - (l_min & 7L));
            ssize_t chunk_len = p_max - p_min + 1;

            bit_copy_core(dst, dst_bit, src, src_len, p_min, chunk_len);
            dst_bit += chunk_len;
        }
    }
    return result;
}

#bit_splice(*args) ⇒ Object

String#bit_splice(bit_index, bit_length, str) -> self String#bit_splice(bit_index, bit_length, str, str_bit_index, str_bit_length) -> self String#bit_splice(range, str) -> self String#bit_splice(range, str, str_range) -> self

Writes bits from str into self at bit-level granularity. The inverse of bit_slice: where bit_slice reads a sub-sequence of bits, bit_splice writes one.

The destination and source bit lengths must be equal; bit_splice does not resize self (sub-byte resize is undefined). This mirrors the constraint that bytesplice imposes when the replacement has the same byte length.

Negative indices count backward from the end, exactly as in bytesplice. Returns self.

Porting to Ruby Core:

1. Move to string.c; register in Init_String().
2. Use rb_str_modify_expand if resize support is ever added.
3. bit_copy_core moves with it; share ebs_extract with bit_slice.


1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
# File 'ext/string_bits/string_bits.c', line 1465

static VALUE
rb_str_bit_splice(int argc, VALUE *argv, VALUE self)
{
    ssize_t dst_bit_off, dst_bit_len;
    ssize_t src_bit_off, src_bit_len;
    VALUE str;
    ssize_t dst_total = RSTRING_LEN(self) * 8;
    VALUE v0, v1, v2, v3, v4, opts;

    int n_pos = rb_scan_args(argc, argv, "23:", &v0, &v1, &v2, &v3, &v4, &opts);
    validate_option_hash(opts, SB_KW_LSB_FIRST);
    int lsb_first = parse_lsb_first_opt(opts);

    if (n_pos == 2 && rb_obj_is_kind_of(v0, rb_cRange)) {
        /* bit_splice(range, str) */
        ssize_t beg, len;
        sb_range_beg_len(v0, &beg, &len, dst_total, 1);
        dst_bit_off = beg;
        dst_bit_len = len;
        str = v1;
        Check_Type(str, T_STRING);
        src_bit_off = 0;
        src_bit_len = dst_bit_len;
    }
    else if (n_pos == 3 && rb_obj_is_kind_of(v0, rb_cRange)) {
        /* bit_splice(range, str, str_range) */
        ssize_t beg, len;
        sb_range_beg_len(v0, &beg, &len, dst_total, 1);
        dst_bit_off = beg;
        dst_bit_len = len;
        str = v1;
        Check_Type(str, T_STRING);
        if (!rb_obj_is_kind_of(v2, rb_cRange)) {
            rb_raise(rb_eTypeError, "third argument must be a Range");
        }
        ssize_t src_total = RSTRING_LEN(str) * 8;
        sb_range_beg_len(v2, &beg, &len, src_total, 1);
        src_bit_off = beg;
        src_bit_len = len;
    }
    else if (n_pos == 3) {
        /* bit_splice(bit_index, bit_length, str) */
        if (!rb_integer_type_p(v0) || !rb_integer_type_p(v1)) {
            rb_raise(rb_eTypeError, "bit index and length must be integers");
        }
        dst_bit_off = integer_to_bit_idx(v0);
        dst_bit_len = integer_to_bit_idx(v1);
        if (dst_bit_off < 0) dst_bit_off += dst_total;

        /*
         * Integer source support was prototyped here, but it is intentionally
         * disabled in the current proposal to keep the public API limited to
         * String-to-String splicing.
         */
        if (rb_integer_type_p(v2)) {
            rb_raise(rb_eArgError,
                     "bit_splice source must be a String in the current proposal");
        }

        str = v2;
        Check_Type(str, T_STRING);
        src_bit_off = 0;
        src_bit_len = dst_bit_len;
    }
    else if (n_pos == 5) {
        /* bit_splice(bit_index, bit_length, str, str_bit_index, str_bit_length) */
        if (!rb_integer_type_p(v0) || !rb_integer_type_p(v1) ||
            !rb_integer_type_p(v3) || !rb_integer_type_p(v4)) {
            rb_raise(rb_eTypeError, "bit indices and lengths must be integers");
        }
        dst_bit_off = integer_to_bit_idx(v0);
        dst_bit_len = integer_to_bit_idx(v1);
        if (dst_bit_off < 0) dst_bit_off += dst_total;
        str = v2;
        Check_Type(str, T_STRING);
        ssize_t src_total = RSTRING_LEN(str) * 8;
        src_bit_off = integer_to_bit_idx(v3);
        src_bit_len = integer_to_bit_idx(v4);
        if (src_bit_off < 0) src_bit_off += src_total;
    }
    else {
        rb_raise(rb_eArgError,
                 "wrong number of arguments (given %d, expected 2, 3, or 5)", n_pos);
    }

    if (dst_bit_off < 0 || dst_bit_len < 0 || dst_bit_off + dst_bit_len > dst_total) {
        rb_raise(rb_eIndexError,
                 "bit_splice: destination range [%ld, %ld] out of bounds (total %ld bits)",
                 dst_bit_off, dst_bit_len, dst_total);
    }

    ssize_t src_total_bits = RSTRING_LEN(str) * 8;
    if (src_bit_off < 0 || src_bit_len < 0 || src_bit_off + src_bit_len > src_total_bits) {
        rb_raise(rb_eIndexError,
                 "bit_splice: source range [%ld, %ld] out of bounds (total %ld bits)",
                 src_bit_off, src_bit_len, src_total_bits);
    }

    if (dst_bit_len != src_bit_len) {
        rb_raise(rb_eArgError,
                 "bit_splice: destination length (%ld) must equal source length (%ld)",
                 dst_bit_len, src_bit_len);
    }

    if (dst_bit_len == 0) return self;

    /* Guard against self-aliasing: duplicate src before modifying self */
    VALUE src_str = (str == self) ? rb_str_dup(str) : str;

    rb_str_modify(self);

    unsigned char       *dst          = (unsigned char *)RSTRING_PTR(self);
    const unsigned char *src          = (const unsigned char *)RSTRING_PTR(src_str);
    ssize_t                 src_len_bytes = RSTRING_LEN(src_str);

    if (lsb_first) {
        bit_copy_core(dst, dst_bit_off, src, src_len_bytes, src_bit_off, dst_bit_len);
    } else {
        ssize_t current_src_bit = src_bit_off;
        ssize_t start_byte = dst_bit_off >> 3;
        ssize_t end_byte = (dst_bit_off + dst_bit_len - 1) >> 3;

        for (ssize_t b = start_byte; b <= end_byte; b++) {
            ssize_t b_start_l = b << 3;
            ssize_t b_end_l = b_start_l + 7;
            ssize_t l_min = (dst_bit_off > b_start_l) ? dst_bit_off : b_start_l;
            ssize_t l_max = ((dst_bit_off + dst_bit_len - 1) < b_end_l) ? (dst_bit_off + dst_bit_len - 1) : b_end_l;

            ssize_t p_min = b_start_l + (7 - (l_max & 7L));
            ssize_t p_max = b_start_l + (7 - (l_min & 7L));
            ssize_t chunk_len = p_max - p_min + 1;

            bit_copy_core(dst, p_min, src, src_len_bytes, current_src_bit, chunk_len);
            current_src_bit += chunk_len;
        }
    }

    RB_GC_GUARD(src_str);
    return self;
}

#bit_xor(other) ⇒ Object



976
977
978
979
980
981
982
983
984
985
986
987
# File 'ext/string_bits/string_bits.c', line 976

static VALUE
rb_str_bit_xor(VALUE self, VALUE other)
{
    check_binary_op_lengths(self, other);
    ssize_t len = RSTRING_LEN(self);
    VALUE result = alloc_result(self);
    const unsigned char *a = (const unsigned char *)RSTRING_PTR(self);
    const unsigned char *b = (const unsigned char *)RSTRING_PTR(other);
    unsigned char *dst = (unsigned char *)RSTRING_PTR(result);
    for (ssize_t i = 0; i < len; i++) dst[i] = a[i] ^ b[i];
    return result;
}

#bit_xor!(other) ⇒ Object



989
990
991
992
993
994
995
996
997
998
999
# File 'ext/string_bits/string_bits.c', line 989

static VALUE
rb_str_bit_xor_bang(VALUE self, VALUE other)
{
    check_binary_op_lengths(self, other);
    rb_str_modify(self);
    ssize_t len = RSTRING_LEN(self);
    unsigned char *a = (unsigned char *)RSTRING_PTR(self);
    const unsigned char *b = (const unsigned char *)RSTRING_PTR(other);
    for (ssize_t i = 0; i < len; i++) a[i] ^= b[i];
    return self;
}

#bits(*args) ⇒ Object



454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
# File 'ext/string_bits/string_bits.c', line 454

static VALUE
rb_str_bits(int argc, VALUE *argv, VALUE self)
{
    int lsb_first = parse_lsb_first(argc, argv);
    ssize_t len = RSTRING_LEN(self);
    const unsigned char *str = (const unsigned char *)RSTRING_PTR(self);
    ssize_t total_bits = len * 8;
    int have_block = rb_block_given_p();

    VALUE ary = have_block ? Qnil : rb_ary_new_capa(total_bits);

    for (ssize_t i = 0; i < len; i++) {
        unsigned char b = str[i];
        if (lsb_first) {
            for (int j = 0; j < 8; j++) {
                VALUE bit = (b >> j) & 1 ? Qtrue : Qfalse;
                have_block ? rb_yield(bit) : rb_ary_push(ary, bit);
            }
        } else {
            for (int j = 7; j >= 0; j--) {
                VALUE bit = (b >> j) & 1 ? Qtrue : Qfalse;
                have_block ? rb_yield(bit) : rb_ary_push(ary, bit);
            }
        }
    }

    return have_block ? self : ary;
}

#clear_bit(*args) ⇒ Object



872
873
874
875
876
# File 'ext/string_bits/string_bits.c', line 872

static VALUE
rb_str_clear_bit(int argc, VALUE *argv, VALUE self)
{
    return rb_str_mutate_bits(argc, argv, self, SB_MUT_CLEAR);
}

#each_bit(*args) ⇒ Object

iterate bits ————————————————————



429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
# File 'ext/string_bits/string_bits.c', line 429

static VALUE
rb_str_each_bit(int argc, VALUE *argv, VALUE self)
{
    RETURN_ENUMERATOR(self, argc, argv);

    int lsb_first = parse_lsb_first(argc, argv);
    ssize_t len = RSTRING_LEN(self);
    const unsigned char *str = (const unsigned char *)RSTRING_PTR(self);

    for (ssize_t i = 0; i < len; i++) {
        unsigned char b = str[i];
        if (lsb_first) {
            for (int j = 0; j < 8; j++) {
                rb_yield((b >> j) & 1 ? Qtrue : Qfalse);
            }
        } else {
            for (int j = 7; j >= 0; j--) {
                rb_yield((b >> j) & 1 ? Qtrue : Qfalse);
            }
        }
    }

    return self;
}

#each_bit_field(*args) ⇒ Object

String#each_bit_field(*bitlens, lsb_first: true) -> self String#each_bit_field(*bitlens, lsb_first: true) -> Enumerator

Iterates over the string as a sequence of packed bit-field records. Each positional argument specifies the width (in bits) of one field in the record. On each iteration, one Integer per field is yielded (LSB-first bit layout). Each bitlen must be in the range 1..64.

lsb_first: true (default) – intra-byte field extraction uses bit 0..7. lsb_first: false – intra-byte field extraction uses bit 7..0.

Incomplete trailing bits (when bytesize*8 is not a multiple of sum(bitlens)) are silently dropped, matching the behavior of Enumerable#each_slice.

Porting to Ruby Core:

1. Move extract_uint64 and this function into string.c.
2. Register with rb_define_method in Init_String().
3. Replace ALLOCA_N with stack arrays for small field counts and heap otherwise.


1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
# File 'ext/string_bits/string_bits.c', line 1073

static VALUE
rb_str_each_bit_field(int argc, VALUE *argv, VALUE self)
{
    RETURN_ENUMERATOR(self, argc, argv);

    VALUE rest, opts;
    rb_scan_args(argc, argv, "*:", &rest, &opts);
    validate_option_hash(opts, SB_KW_LSB_FIRST);

    ssize_t num_fields = RARRAY_LEN(rest);
    if (num_fields == 0) {
        rb_raise(rb_eArgError, "wrong number of arguments (given 0, expected 1+)");
    }

    ssize_t *bitlens = ALLOCA_N(ssize_t, num_fields);
    ssize_t step = 0;
    for (ssize_t f = 0; f < num_fields; f++) {
        VALUE v = RARRAY_AREF(rest, f);
        if (!rb_integer_type_p(v)) {
            rb_raise(rb_eTypeError, "bitlen must be an integer");
        }
        ssize_t bl = NUM2SSIZET(v);
        if (bl <= 0) {
            rb_raise(rb_eArgError, "bitlen must be positive");
        }
        if (bl > 64) {
            rb_raise(rb_eArgError, "bitlen must be <= 64 (got %ld)", bl);
        }
        bitlens[f] = bl;
        step += bl;
    }

    int lsb_first = parse_lsb_first_opt(opts);

    ssize_t src_len = RSTRING_LEN(self);
    ssize_t total_bits = src_len * 8;
    ssize_t iterations = total_bits / step;

    VALUE *field_vals = ALLOCA_N(VALUE, num_fields);

    for (ssize_t iter = 0; iter < iterations; iter++) {
        ssize_t base_bit = iter * step;
        const unsigned char *src = (const unsigned char *)RSTRING_PTR(self);
        ssize_t field_bit = base_bit;
        for (ssize_t f = 0; f < num_fields; f++) {
            uint64_t val = extract_uint64(src, src_len, field_bit, bitlens[f], lsb_first);
            field_vals[f] = ULL2NUM(val);
            field_bit += bitlens[f];
        }
        rb_yield_values2((int)num_fields, field_vals);
    }

    return self;
}

#each_bit_run(*args) ⇒ Object

String#each_bit_run(lsb_first: true) { |bit, len| } -> self String#each_bit_run(lsb_first: true) -> Enumerator

Yields (bit, run_length) pairs for each consecutive run of identical bits. Run-length boundary detection and counting happen entirely in C, replacing the Ruby-level current/count state machine required when using each_bit.

For random data (~50% density) each_bit_run yields ~half as many times as each_bit. For structured data (sparse validity bitmaps, sensor bursts) the ratio is proportional to the average run length.

lsb_first: true (default) iterates bit 0..7 within each byte. lsb_first: false iterates bit 7..0 within each byte.

Porting to Ruby Core:

1. Move to string.c; register in Init_String().
2. count_run_lsb / count_run_msb move with it.


1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
# File 'ext/string_bits/string_bits.c', line 1351

static VALUE
rb_str_each_bit_run(int argc, VALUE *argv, VALUE self)
{
    RETURN_ENUMERATOR(self, argc, argv);

    int lsb_first = parse_lsb_first(argc, argv);
    ssize_t src_len  = RSTRING_LEN(self);
    if (src_len == 0) return self;

    ssize_t total_bits = src_len * 8;

    if (lsb_first) {
        ssize_t pos = 0;
        while (pos < total_bits) {
            const unsigned char *src = (const unsigned char *)RSTRING_PTR(self);
            int bit  = (src[pos >> 3] >> (pos & 7)) & 1;
            ssize_t run = count_run_lsb(src, src_len, pos, bit);
            rb_yield_values(2, bit ? Qtrue : Qfalse, SSIZET2NUM(run));
            pos += run;
        }
    }
    else {
        ssize_t pos = 0;
        while (pos < total_bits) {
            const unsigned char *src = (const unsigned char *)RSTRING_PTR(self);
            int bit = logical_get_bit(src, pos, 0);
            ssize_t run = 1;
            while (pos + run < total_bits && logical_get_bit(src, pos + run, 0) == bit) {
                run++;
            }
            rb_yield_values(2, bit ? Qtrue : Qfalse, SSIZET2NUM(run));
            pos += run;
        }
    }

    return self;
}

#each_set_bit_offset(*args) ⇒ Object

iterate set-bit positions ———————————————–



485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
# File 'ext/string_bits/string_bits.c', line 485

static VALUE
rb_str_each_set_bit_offset(int argc, VALUE *argv, VALUE self)
{
    RETURN_ENUMERATOR(self, argc, argv);

    int lsb_first = parse_lsb_first(argc, argv);
    ssize_t len = RSTRING_LEN(self);
    const unsigned char *str = (const unsigned char *)RSTRING_PTR(self);
    if (lsb_first) {
        /* LSB-first: ascending positions 0, 1, 2, ...
         * On little-endian, loading 8 bytes as a uint64_t preserves the flat
         * LSB-first bit numbering: word bit 0 == position 0, bit 63 == 63.
         * memcpy avoids unaligned-load SIGBUS on strict-alignment platforms. */
#if SB_LITTLE_ENDIAN
        ssize_t n_words = len >> 3;
        for (ssize_t wi = 0; wi < n_words; wi++) {
            uint64_t w;
            memcpy(&w, str + wi * 8, 8);
            while (w != 0) {
                int bit = sb_ctzll(w);
                rb_yield(SSIZET2NUM(wi * 64 + bit));
                w &= w - 1;
            }
        }
        for (ssize_t bi = n_words << 3; bi < len; bi++) {
            unsigned int b = str[bi];
            while (b != 0) {
                int bit = sb_ctz8(b);
                rb_yield(SSIZET2NUM(bi * 8 + bit));
                b &= b - 1;
            }
        }
#else
        for (ssize_t bi = 0; bi < len; bi++) {
            unsigned int b = str[bi];
            while (b != 0) {
                int bit = sb_ctz8(b);
                rb_yield(SSIZET2NUM(bi * 8 + bit));
                b &= b - 1;
            }
        }
#endif
    }
    else {
        /* lsb_first: false => byte order preserved, bits 7..0 map to logical 0..7 */
        for (ssize_t bi = 0; bi < len; bi++) {
            unsigned int b = str[bi];
            while (b != 0) {
                int bit = sb_highest_bit8(b);
                ssize_t physical = bi * 8 + bit;
                rb_yield(SSIZET2NUM(physical_to_count_from(physical, 0)));
                b ^= (1u << bit);  /* clear highest set bit */
            }
        }
    }

    return self;
}

#flip_bit(*args) ⇒ Object



878
879
880
881
882
# File 'ext/string_bits/string_bits.c', line 878

static VALUE
rb_str_flip_bit(int argc, VALUE *argv, VALUE self)
{
    return rb_str_mutate_bits(argc, argv, self, SB_MUT_FLIP);
}

#set_bit(*args) ⇒ Object



866
867
868
869
870
# File 'ext/string_bits/string_bits.c', line 866

static VALUE
rb_str_set_bit(int argc, VALUE *argv, VALUE self)
{
    return rb_str_mutate_bits(argc, argv, self, SB_MUT_SET);
}

#set_bit_offsets(*args) ⇒ Object



544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
# File 'ext/string_bits/string_bits.c', line 544

static VALUE
rb_str_set_bit_offsets(int argc, VALUE *argv, VALUE self)
{
    int lsb_first = parse_lsb_first(argc, argv);
    ssize_t len = RSTRING_LEN(self);
    const unsigned char *str = (const unsigned char *)RSTRING_PTR(self);
    int have_block = rb_block_given_p();

    VALUE ary;
    if (have_block) {
        ary = Qnil;
    }
    else {
        /* Pre-size the Array with popcount to avoid repeated reallocation.
         * memcpy avoids unaligned-load issues on strict-alignment platforms. */
        ssize_t count = 0;
        ssize_t nw = len >> 3;
        for (ssize_t wi = 0; wi < nw; wi++) {
            uint64_t w;
            memcpy(&w, str + wi * 8, 8);
            count += sb_popcount64(w);
        }
        for (ssize_t bi = nw << 3; bi < len; bi++)
            count += sb_popcount64((uint64_t)(unsigned char)str[bi]);
        ary = rb_ary_new_capa(count);
    }

    if (lsb_first) {
#if SB_LITTLE_ENDIAN
        ssize_t n_words = len >> 3;
        for (ssize_t wi = 0; wi < n_words; wi++) {
            uint64_t w;
            memcpy(&w, str + wi * 8, 8);
            while (w != 0) {
                int bit = sb_ctzll(w);
                VALUE pos = SSIZET2NUM(wi * 64 + bit);
                have_block ? rb_yield(pos) : rb_ary_push(ary, pos);
                w &= w - 1;
            }
        }
        for (ssize_t bi = n_words << 3; bi < len; bi++) {
            unsigned int b = str[bi];
            while (b != 0) {
                int bit = sb_ctz8(b);
                VALUE pos = SSIZET2NUM(bi * 8 + bit);
                have_block ? rb_yield(pos) : rb_ary_push(ary, pos);
                b &= b - 1;
            }
        }
#else
        for (ssize_t bi = 0; bi < len; bi++) {
            unsigned int b = str[bi];
            while (b != 0) {
                int bit = sb_ctz8(b);
                VALUE pos = SSIZET2NUM(bi * 8 + bit);
                have_block ? rb_yield(pos) : rb_ary_push(ary, pos);
                b &= b - 1;
            }
        }
#endif
    }
    else {
        for (ssize_t bi = 0; bi < len; bi++) {
            unsigned int b = str[bi];
            while (b != 0) {
                int bit = sb_highest_bit8(b);
                ssize_t physical = bi * 8 + bit;
                VALUE pos = SSIZET2NUM(physical_to_count_from(physical, 0));
                have_block ? rb_yield(pos) : rb_ary_push(ary, pos);
                b ^= (1u << bit);
            }
        }
    }

    return have_block ? self : ary;
}