diff options
author | David Mitchell <davem@iabyn.com> | 2018-01-15 15:29:27 +0000 |
---|---|---|
committer | David Mitchell <davem@iabyn.com> | 2018-01-19 13:45:20 +0000 |
commit | 0b9a13c37566388c6489a7e2d45b4e92ed36819d (patch) | |
tree | 3abb6e9e5a0301e2aba27ef046bee5e41c0f9e7b /op.c | |
parent | 6d63cc8e88a2b96ed80956f16c0978d790bf4411 (diff) | |
download | perl-0b9a13c37566388c6489a7e2d45b4e92ed36819d.tar.gz |
tr///; simplify $utf8 =~ tr/nonutf8/nonutf8/
The run-time code to handle a non-utf8 tr/// against a utf8 string
is complex, with many variants of similar code repeated depending on the
presence of the /s and /c flags.
Simplify them all into a single code block by changing how the translation
table is stored. Formerly, the tr struct contained possibly two tables:
the basic 0-255 slot one, plus in the presence of /c, a second one
to map the implicit search range (\x{100}...) against any residual
replacement chars not consumed by the first table.
This commit merges the two tables into a single unified whole. For example
tr/\x00-\xfe/abcd/c
is equivalent to
tr/xff-\x{7fffffff}/abcd/
which generates a 259-entry translation table consisting of:
0x00 => -1
0x01 => -1
...
0xfe => -1
0xff => a
0x100 => b
0x101 => c
0x102 => d
In addition we store:
1) the size of the translation table (0x103 in the example above);
2) an extra 'wildcard' entry stored 1 slot beyond the main table,
which specifies the action for any codepoints outside the range of
the table (i.e. chars 0x103..0x7fffffff). This can be either:
a) a character, when the last replacement char is repeated;
b) -1 when /c isn't in effect;
c) -2 when /d is in effect;
c) -3 identity: when the replacement list is empty but not /d.
In the example above, this would be
0x103 => d
The addition of -3 as a valid slot value is new.
This makes the main runtime code for the utf8 string with non-utf8 tr//
case look like, at its core:
size = tbl->size;
mapped_ch = tbl->map[ch >= size ? size : ch];
which then processes mapped_ch based on whether its >=0, or -1/-2/-3.
This is a lot simpler than the old scheme, and should generally be faster
too.
Diffstat (limited to 'op.c')
-rw-r--r-- | op.c | 100 |
1 files changed, 44 insertions, 56 deletions
@@ -6344,8 +6344,9 @@ S_pmtrans(pTHX_ OP *o, OP *expr, OP *repl) I32 j; I32 grows = 0; OPtrans_map *tbl; + SSize_t struct_size; /* malloced size of table struct */ - const I32 complement = o->op_private & OPpTRANS_COMPLEMENT; + const bool complement = cBOOL(o->op_private & OPpTRANS_COMPLEMENT); const I32 squash = o->op_private & OPpTRANS_SQUASH; I32 del = o->op_private & OPpTRANS_DELETE; SV* swash; @@ -6611,35 +6612,43 @@ S_pmtrans(pTHX_ OP *o, OP *expr, OP *repl) goto warnins; } - /* Non-utf8 case: set o->op_pv to point to a simple 256-entry lookup + /* Non-utf8 case: set o->op_pv to point to a simple 256+ entry lookup * table. Entries with the value -1 indicate chars not to be * translated, while -2 indicates a search char without a * corresponding replacement char under /d. * - * With /c, an extra length arg is stored at the end of the table to - * indicate the number of chars in the replacement string, plus any - * excess replacement chars not paired with search chars. The extra - * chars are needed for utf8 strings. For example, - * tr/\x00-\xfd/abcd/c is logically equivalent to - * tr/\xfe\xff\x{100}\x{101}.../abcdddd.../, so the c,d chars need to - * be kept even though they aren't paired with any chars in the table - * (which represents chars \x00-\xff). Even without excess chars, the - * last replacement char needs to be kept. + * Normally, the table has 256 slots. However, in the presence of + * /c, the search charlist has an implicit \x{100}-\x{7fffffff} + * added, and if there are enough replacement chars to start pairing + * with the \x{100},... search chars, then a larger (> 256) table + * is allocated. + * + * In addition, regardless of whether under /c, an extra slot at the + * end is used to store the final repeating char, or -3 under an empty + * replacement list, or -2 under /d; which makes the runtime code + * easier. * * The toker will have already expanded char ranges in t and r. */ - tbl = (OPtrans_map*)PerlMemShared_calloc( - complement ? sizeof(OPtrans_map_ex) : sizeof(OPtrans_map), - sizeof(char)); + /* Initially allocate 257-slot table: 256 for basic (non /c) usage, + * plus final slot for repeat/-2/-3. Later we realloc if excess > * 0. + * The OPtrans_map struct already contains one slot; hence the -1. + */ + struct_size = sizeof(OPtrans_map) + (256 - 1 + 1)*sizeof(short); + tbl = (OPtrans_map*)PerlMemShared_calloc(struct_size, 1); + tbl->size = 256; cPVOPo->op_pv = (char*)tbl; if (complement) { + SSize_t excess; + /* in this branch, j is a count of 'consumed' (i.e. paired off * with a search char) replacement chars (so j <= rlen always) */ for (i = 0; i < (I32)tlen; i++) tbl->map[t[i]] = -1; + for (i = 0, j = 0; i < 256; i++) { if (!tbl->map[i]) { if (j == (I32)rlen) { @@ -6659,52 +6668,29 @@ S_pmtrans(pTHX_ OP *o, OP *expr, OP *repl) } assert(j <= (I32)rlen); + excess = rlen - (SSize_t)j; - /* populate extended portion of table */ - - { - /* the repeat char: it may be used to fill the 0x100+ - * range. For example, - * tr/\x00-AE-\xff/bcd/c - * is equivalent to - * tr/BCD\x{100}-\x{7fffffff}/bcd/ - * which is equivalent to - * tr/BCD\x{100}-\x{7fffffff}/bcddddddddd..../ - * So remember the 'd'. - */ - short repeat_char; - SSize_t excess = rlen - (SSize_t)j; - OPtrans_map_ex *extbl = (OPtrans_map_ex*)tbl; + if (excess) { + /* More replacement chars than search chars: + * store excess replacement chars at end of main table. + */ - if (excess) { - /* More replacement chars than search chars: - * store excess replacement chars at end of main table. - */ + struct_size += excess; + tbl = (OPtrans_map*)PerlMemShared_realloc(tbl, + struct_size + excess * sizeof(short)); + tbl->size += excess; + cPVOPo->op_pv = (char*)tbl; - extbl = (OPtrans_map_ex *) PerlMemShared_realloc(extbl, - sizeof(OPtrans_map_ex) + excess * sizeof(short)); - cPVOPo->op_pv = (char*)extbl; - for (i = 0; i < (I32)excess; i++) - extbl->map_ex[i] = r[j+i]; - repeat_char = r[rlen-1]; - } - else { - /* no more replacement chars than search chars */ + for (i = 0; i < (I32)excess; i++) + tbl->map[i + 256] = r[j+i]; + } + else { + /* no more replacement chars than search chars */ + if (!rlen && !del && !squash) + o->op_private |= OPpTRANS_IDENTICAL; + } - if (rlen) - repeat_char = r[rlen - 1]; - else { - /* empty replacement list */ - repeat_char = 0; /* this value isn't used at runtime */ - /* -1 excess count indicates empty replacement charlist */ - excess = -1; - if (!(squash | del)) - o->op_private |= OPpTRANS_IDENTICAL; - } - } - extbl->excess_len = excess; /* excess char count */ - extbl->repeat_char = (short)repeat_char; /* repeated replace char */ - } + tbl->map[tbl->size] = del ? -2 : rlen ? r[rlen - 1] : -3; } else { if (!rlen && !del) { @@ -6715,6 +6701,7 @@ S_pmtrans(pTHX_ OP *o, OP *expr, OP *repl) else if (!squash && rlen == tlen && memEQ((char*)t, (char*)r, tlen)) { o->op_private |= OPpTRANS_IDENTICAL; } + for (i = 0; i < 256; i++) tbl->map[i] = -1; for (i = 0, j = 0; i < (I32)tlen; i++,j++) { @@ -6733,6 +6720,7 @@ S_pmtrans(pTHX_ OP *o, OP *expr, OP *repl) tbl->map[t[i]] = r[j]; } } + tbl->map[tbl->size] = del ? -1 : rlen ? -1 : -3; } /* both non-utf8 and utf8 code paths end up here */ |