This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Integrate mainline
[perl5.git] / ext / Encode / encengine.c
CommitLineData
017e2add
NIS
1/*
2Data structures for encoding transformations.
3
4Perl works internally in either a native 'byte' encoding or
5in UTF-8 encoded Unicode. We have no immediate need for a "wchar_t"
6representation. When we do we can use utf8_to_uv().
7
8Most character encodings are either simple byte mappings or
9variable length multi-byte encodings. UTF-8 can be viewed as a
10rather extreme case of the latter.
11
12So to solve an important part of perl's encode needs we need to solve the
13"multi-byte -> multi-byte" case. The simple byte forms are then just degenerate
14case. (Where one of multi-bytes will usually be UTF-8.)
15
16The other type of encoding is a shift encoding where a prefix sequence
17determines what subsequent bytes mean. Such encodings have state.
18
19We also need to handle case where a character in one encoding has to be
20represented as multiple characters in the other. e.g. letter+diacritic.
21
22The process can be considered as pseudo perl:
23
24my $dst = '';
25while (length($src))
26 {
27 my $size = $count($src);
28 my $in_seq = substr($src,0,$size,'');
29 my $out_seq = $s2d_hash{$in_seq};
30 if (defined $out_seq)
31 {
32 $dst .= $out_seq;
33 }
34 else
35 {
36 # an error condition
37 }
38 }
39return $dst;
40
41That has the following components:
42 &src_count - a "rule" for how many bytes make up the next character in the
43 source.
44 %s2d_hash - a mapping from input sequences to output sequences
45
46The problem with that scheme is that it does not allow the output
47character repertoire to affect the characters considered from the
48input.
49
50So we use a "trie" representation which can also be considered
51a state machine:
52
53my $dst = '';
54my $seq = \@s2d_seq;
55my $next = \@s2d_next;
56while (length($src))
57 {
58 my $byte = $substr($src,0,1,'');
59 my $out_seq = $seq->[$byte];
60 if (defined $out_seq)
61 {
62 $dst .= $out_seq;
63 }
64 else
65 {
66 # an error condition
67 }
68 ($next,$seq) = @$next->[$byte] if $next;
69 }
70return $dst;
71
72There is now a pair of data structures to represent everything.
73It is valid for output sequence at a particular point to
74be defined but zero length, that just means "don't know yet".
75For the single byte case there is no 'next' so new tables will be the same as
76the original tables. For a multi-byte case a prefix byte will flip to the tables
77for the next page (adding nothing to the output), then the tables for the page
78will provide the actual output and set tables back to original base page.
79
80This scheme can also handle shift encodings.
81
82A slight enhancement to the scheme also allows for look-ahead - if
83we add a flag to re-add the removed byte to the source we could handle
84 a" -> รค
85 ab -> a (and take b back please)
86
87*/
88
89#include <EXTERN.h>
90#include <perl.h>
91#define U8 U8
92#include "encode.h"
93
94STRLEN
95translate(encpage_t *enc, const U8 *src, STRLEN slen, U8 *dst, STRLEN dlen)
96{
97 const U8 *send = src+slen;
98 U8 *dend = dst+dlen;
99 U8 *dptr = dst;
100 while (src < send)
101 {
102 encpage_t *e = enc;
103 U8 byte = *src++;
104 while (byte > e->max)
105 e++;
106 if (byte >= e->min)
107 {
108 STRLEN n = e->dlen;
109 if (n)
110 {
111 const U8 *out = e->seq+n*(byte - e->min);
112 STRLEN n = *out++;
113 if (dptr+n <= dend)
114 {
115 if (dst)
116 Copy(out,dptr,n,U8);
117 dptr += n;
118 }
119 else
120 {
121 /* No room */
122 }
123 }
124 enc = e->next;
125 }
126 else
127 {
128 /* Cannot represent */
129 }
130 }
131 return dptr-dst;
132}
133
134