utf8itor.cc

Go to the documentation of this file.

 
/* Copyright (C) 2006,2007,2010,2013,2015,2019,2024 Olly Betts
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, see
 * <https://www.gnu.org/licenses/>.
 */
 
#include <config.h>
 
#include <xapian/unicode.h>
 
using namespace std;
 
static inline bool bad_cont(unsigned char ch) {
    return static_cast<signed char>(ch) >= static_cast<signed char>(0xc0);
}
 
namespace Xapian {
 
namespace Unicode {
 
// buf should be at least 4 bytes.
unsigned
nonascii_to_utf8(unsigned ch, char* buf)
{
    if (ch < 0x800) {
        buf[0] = char(0xc0 | (ch >> 6));
        buf[1] = char(0x80 | (ch & 0x3f));
        return 2;
    }
    if (ch < 0x10000) {
        buf[0] = char(0xe0 | (ch >> 12));
        buf[1] = char(0x80 | ((ch >> 6) & 0x3f));
        buf[2] = char(0x80 | (ch & 0x3f));
        return 3;
    }
    if (ch < 0x200000) {
        buf[0] = char(0xf0 | (ch >> 18));
        buf[1] = char(0x80 | ((ch >> 12) & 0x3f));
        buf[2] = char(0x80 | ((ch >> 6) & 0x3f));
        buf[3] = char(0x80 | (ch & 0x3f));
        return 4;
    }
    // Unicode doesn't specify any characters above 0x10ffff.
    // Should we be presented with such a numeric character
    // entity or similar, we just replace it with nothing.
    return 0;
}
 
}
 
bool
Utf8Iterator::calculate_sequence_length() const noexcept
{
    // Handle invalid UTF-8, overlong sequences, surrogate pair halves, and
    // truncated sequences as if the text was actually in ISO-8859-1 since we
    // need to do something with it, and this seems the most likely reason why
    // we'd have invalid UTF-8.
 
    unsigned char ch = *p;
 
    seqlen = 1;
    // Single byte encoding (0x00-0x7f) or invalid (0x80-0xbf) or overlong
    // sequence (0xc0-0xc1).
    //
    // (0xc0 and 0xc1 would start 2 byte sequences for characters which are
    // representable in a single byte, and we should not decode these.)
    if (ch < 0xc2) return (ch < 0x80);
 
    if (ch < 0xe0) {
        if (p + 1 == end || // Not enough bytes
            bad_cont(p[1])) // Invalid
            return false;
        seqlen = 2;
        return true;
    }
    if (ch < 0xf0) {
        if (end - p < 3 || // Not enough bytes
            bad_cont(p[1]) || bad_cont(p[2]) || // Invalid
            (p[0] == 0xe0 && p[1] < 0xa0) || // Overlong encoding
            (p[0] == 0xed && p[1] >= 0xa0)) // Surrogate pair half
            return false;
        seqlen = 3;
        return true;
    }
    if (ch >= 0xf5 || // Code value above Unicode
        end - p < 4 || // Not enough bytes
        bad_cont(p[1]) || bad_cont(p[2]) || bad_cont(p[3]) || // Invalid
        (p[0] == 0xf0 && p[1] < 0x90) || // Overlong encoding
        (p[0] == 0xf4 && p[1] >= 0x90)) // Code value above Unicode
        return false;
    seqlen = 4;
    return true;
}
 
unsigned Utf8Iterator::operator*() const noexcept {
    if (p == NULL) return unsigned(-1);
    if (seqlen == 0) calculate_sequence_length();
    unsigned char ch = *p;
    if (seqlen == 1) return ch;
    if (seqlen == 2) return ((ch & 0x1f) << 6) | (p[1] & 0x3f);
    if (seqlen == 3)
        return ((ch & 0x0f) << 12) | ((p[1] & 0x3f) << 6) | (p[2] & 0x3f);
    return ((ch & 0x07) << 18) | ((p[1] & 0x3f) << 12) |
            ((p[2] & 0x3f) << 6) | (p[3] & 0x3f);
}
 
unsigned
Utf8Iterator::strict_deref() const noexcept
{
    if (p == NULL) return unsigned(-1);
    if (seqlen == 0) {
        if (!calculate_sequence_length())
            return unsigned(*p) | 0x80000000;
    }
    unsigned char ch = *p;
    if (seqlen == 1) return ch;
    if (seqlen == 2) return ((ch & 0x1f) << 6) | (p[1] & 0x3f);
    if (seqlen == 3)
        return ((ch & 0x0f) << 12) | ((p[1] & 0x3f) << 6) | (p[2] & 0x3f);
    return ((ch & 0x07) << 18) | ((p[1] & 0x3f) << 12) |
            ((p[2] & 0x3f) << 6) | (p[3] & 0x3f);
}
 
}