mbfilter.c

php-4.4.7学习linux时下载的源代码

	if (pc->output >= pc->stop) {
		return -1;
	}

	if (pc->output >= pc->start) {
		(*pc->next_filter->filter_function)(c, pc->next_filter);
	}

	pc->output++;

	return c;
}

mbfl_string *
mbfl_substr(
    mbfl_string *string,
    mbfl_string *result,
    int from,
    int length)
{
	const mbfl_encoding *encoding;
	int n, m, k, len, start, end;
	unsigned char *p, *w;
	const unsigned char *mbtab;

	encoding = mbfl_no2encoding(string->no_encoding);
	if (encoding == NULL || string == NULL || result == NULL) {
		return NULL;
	}
	mbfl_string_init(result);
	result->no_language = string->no_language;
	result->no_encoding = string->no_encoding;

	if ((encoding->flag & (MBFL_ENCTYPE_SBCS | MBFL_ENCTYPE_WCS2BE | MBFL_ENCTYPE_WCS2LE | MBFL_ENCTYPE_WCS4BE | MBFL_ENCTYPE_WCS4LE)) ||
	   encoding->mblen_table != NULL) {
		len = string->len;
		start = from;
		end = from + length;
		if (encoding->flag & (MBFL_ENCTYPE_WCS2BE | MBFL_ENCTYPE_MWC2LE)) {
			start *= 2;
			end = start + length*2;
		} else if (encoding->flag & (MBFL_ENCTYPE_WCS4BE | MBFL_ENCTYPE_MWC4LE)) {
			start *= 4;
			end = start + length*4;
		} else if (encoding->mblen_table != NULL) {
			mbtab = encoding->mblen_table;
			start = 0;
			end = 0;
			n = 0;
			k = 0;
			p = string->val;
			if (p != NULL) {
				/* search start position */
				while (k <= from) {
					start = n;
					if (n >= len) {
						break;
					}
					m = mbtab[*p];
					n += m;
					p += m;
					k++;
				}
				/* detect end position */
				k = 0;
				end = start;
				while (k < length) {
					end = n;
					if (n >= len) {
						break;
					}
					m = mbtab[*p];
					n += m;
					p += m;
					k++;
				}
			}
		}

		if (start > len) {
			start = len;
		}
		if (start < 0) {
			start = 0;
		}
		if (end > len) {
			end = len;
		}
		if (end < 0) {
			end = 0;
		}
		if (start > end) {
			start = end;
		}

		/* allocate memory and copy */
		n = end - start;
		result->len = 0;
		result->val = w = (unsigned char*)mbfl_malloc((n + 8)*sizeof(unsigned char));
		if (w != NULL) {
			p = string->val;
			if (p != NULL) {
				p += start;
				result->len = n;
				while (n > 0) {
					*w++ = *p++;
					n--;
				}
			}
			*w++ = '\0';
			*w++ = '\0';
			*w++ = '\0';
			*w = '\0';
		} else {
			result = NULL;
		}
	} else {
		mbfl_memory_device device;
		struct collector_substr_data pc;
		mbfl_convert_filter *decoder;
		mbfl_convert_filter *encoder;

		mbfl_memory_device_init(&device, length + 1, 0);
		mbfl_string_init(result);
		result->no_language = string->no_language;
		result->no_encoding = string->no_encoding;
		/* output code filter */
		decoder = mbfl_convert_filter_new(
		    mbfl_no_encoding_wchar,
		    string->no_encoding,
		    mbfl_memory_device_output, 0, &device);
		/* wchar filter */
		encoder = mbfl_convert_filter_new(
		    string->no_encoding,
		    mbfl_no_encoding_wchar,
		    collector_substr, 0, &pc);
		if (decoder == NULL || encoder == NULL) {
			mbfl_convert_filter_delete(encoder);
			mbfl_convert_filter_delete(decoder);
			return NULL;
		}
		pc.next_filter = decoder;
		pc.start = from;
		pc.stop = from + length;
		pc.output = 0;

		/* feed data */
		p = string->val;
		n = string->len;
		if (p != NULL) {
			while (n > 0) {
				if ((*encoder->filter_function)(*p++, encoder) < 0) {
					break;
				}
				n--;
			}
		}

		mbfl_convert_filter_flush(encoder);
		mbfl_convert_filter_flush(decoder);
		result = mbfl_memory_device_result(&device, result);
		mbfl_convert_filter_delete(encoder);
		mbfl_convert_filter_delete(decoder);
	}

	return result;
}


/*
 *  strcut
 */
mbfl_string *
mbfl_strcut(
    mbfl_string *string,
    mbfl_string *result,
    int from,
    int length)
{
	const mbfl_encoding *encoding;
	int n, m, k, len, start, end;
	unsigned char *p, *w;
	const unsigned char *mbtab;
	mbfl_memory_device device;
	mbfl_convert_filter *encoder, *encoder_tmp, *decoder, *decoder_tmp;

	encoding = mbfl_no2encoding(string->no_encoding);
	if (encoding == NULL || string == NULL || result == NULL) {
		return NULL;
	}
	mbfl_string_init(result);
	result->no_language = string->no_language;
	result->no_encoding = string->no_encoding;

	if ((encoding->flag & (MBFL_ENCTYPE_SBCS | MBFL_ENCTYPE_WCS2BE | MBFL_ENCTYPE_WCS2LE | MBFL_ENCTYPE_WCS4BE | MBFL_ENCTYPE_WCS4LE)) ||
	   encoding->mblen_table != NULL) {
		len = string->len;
		start = from;
		end = from + length;
		if (encoding->flag & (MBFL_ENCTYPE_WCS2BE | MBFL_ENCTYPE_WCS2LE)) {
			start /= 2;
			start *= 2;
			end = length/2;
			end *= 2;
			end += start;
		} else if (encoding->flag & (MBFL_ENCTYPE_WCS4BE | MBFL_ENCTYPE_WCS4LE)) {
			start /= 4;
			start *= 4;
			end = length/4;
			end *= 4;
			end += start;
		} else if (encoding->mblen_table != NULL) {
			mbtab = encoding->mblen_table;
			start = 0;
			end = 0;
			n = 0;
			p = string->val;
			if (p != NULL) {
				/* search start position */
				for (;;) {
					m = mbtab[*p];
					n += m;
					p += m;
					if (n > from) {
						break;
					}
					start = n;
				}
				/* search end position */
				k = start + length;
				if (k >= (int)string->len) {
					end = string->len;
				} else {
					end = start;
					while (n <= k) {
						end = n;
						m = mbtab[*p];
						n += m;
						p += m;
					}
				}
			}
		}

		if (start > len) {
			start = len;
		}
		if (start < 0) {
			start = 0;
		}
		if (end > len) {
			end = len;
		}
		if (end < 0) {
			end = 0;
		}
		if (start > end) {
			start = end;
		}
		/* allocate memory and copy string */
		n = end - start;
		result->len = 0;
		result->val = w = (unsigned char*)mbfl_malloc((n + 8)*sizeof(unsigned char));
		if (w != NULL) {
			result->len = n;
			p = &(string->val[start]);
			while (n > 0) {
				*w++ = *p++;
				n--;
			}
			*w++ = '\0';
			*w++ = '\0';
			*w++ = '\0';
			*w = '\0';
		} else {
			result = NULL;
		}
	} else {
		/* wchar filter */
		encoder = mbfl_convert_filter_new(
		  string->no_encoding,
		  mbfl_no_encoding_wchar,
		  mbfl_filter_output_null, 0, 0);
		encoder_tmp = mbfl_convert_filter_new(
		  string->no_encoding,
		  mbfl_no_encoding_wchar,
		  mbfl_filter_output_null, 0, 0);
		/* output code filter */
		decoder = mbfl_convert_filter_new(
		  mbfl_no_encoding_wchar,
		  string->no_encoding,
		  mbfl_memory_device_output, 0, &device);
		decoder_tmp = mbfl_convert_filter_new(
		  mbfl_no_encoding_wchar,
		  string->no_encoding,
		  mbfl_memory_device_output, 0, &device);
		if (encoder == NULL || encoder_tmp == NULL || decoder == NULL || decoder_tmp == NULL) {
			mbfl_convert_filter_delete(encoder);
			mbfl_convert_filter_delete(encoder_tmp);
			mbfl_convert_filter_delete(decoder);
			mbfl_convert_filter_delete(decoder_tmp);
			return NULL;
		}
		mbfl_memory_device_init(&device, length + 8, 0);
		k = 0;
		n = 0;
		p = string->val;
		if (p != NULL) {
			/* seartch start position */
			while (n < from) {
				(*encoder->filter_function)(*p++, encoder);
				n++;
			}
			/* output a little shorter than "length" */
			encoder->output_function = mbfl_filter_output_pipe;
			encoder->data = decoder;
			k = length - 20;
			len = string->len;
			while (n < len && device.pos < k) {
				(*encoder->filter_function)(*p++, encoder);
				n++;
			}
			/* detect end position */
			for (;;) {
				/* backup current state */
				k = device.pos;
				mbfl_convert_filter_copy(encoder, encoder_tmp);
				mbfl_convert_filter_copy(decoder, decoder_tmp);
				if (n >= len) {
					break;
				}
				/* feed 1byte and flush */
				(*encoder->filter_function)(*p, encoder);
				(*encoder->filter_flush)(encoder);
				(*decoder->filter_flush)(decoder);
				if (device.pos > length) {
					break;
				}
				/* restore filter and re-feed data */
				device.pos = k;
				mbfl_convert_filter_copy(encoder_tmp, encoder);
				mbfl_convert_filter_copy(decoder_tmp, decoder);
				(*encoder->filter_function)(*p, encoder);
				p++;
				n++;
			}
			device.pos = k;
			mbfl_convert_filter_copy(encoder_tmp, encoder);
			mbfl_convert_filter_copy(decoder_tmp, decoder);
			mbfl_convert_filter_flush(encoder);
			mbfl_convert_filter_flush(decoder);
		}
		result = mbfl_memory_device_result(&device, result);
		mbfl_convert_filter_delete(encoder);
		mbfl_convert_filter_delete(encoder_tmp);
		mbfl_convert_filter_delete(decoder);
		mbfl_convert_filter_delete(decoder_tmp);
	}

	return result;
}


/*
 *  strwidth
 */
static int is_fullwidth(int c)
{
	int i;

	if (c < mbfl_eaw_table[0].begin) {
		return 0;
	}

	for (i = 0; i < sizeof(mbfl_eaw_table) / sizeof(mbfl_eaw_table[0]); i++) {
		if (mbfl_eaw_table[i].begin <= c && c <= mbfl_eaw_table[i].end) {
			return 1;
		}
	}

	return 0;
}

static int
filter_count_width(int c, void* data)
{
	(*(int *)data) += (is_fullwidth(c) ? 2: 1);
	return c;
}

int
mbfl_strwidth(mbfl_string *string)
{
	int len, n;
	unsigned char *p;
	mbfl_convert_filter *filter;

	len = 0;
	if (string->len > 0 && string->val != NULL) {
		/* wchar filter */
		filter = mbfl_convert_filter_new(
		    string->no_encoding,
		    mbfl_no_encoding_wchar,
		    filter_count_width, 0, &len);
		if (filter == NULL) {
			mbfl_convert_filter_delete(filter);
			return -1;
		}

		/* feed data */
		p = string->val;
		n = string->len;
		while (n > 0) {
			(*filter->filter_function)(*p++, filter);
			n--;
		}

		mbfl_convert_filter_flush(filter);
		mbfl_convert_filter_delete(filter);
	}

	return len;
}


/*
 *  strimwidth
 */
struct collector_strimwidth_data {
	mbfl_convert_filter *decoder;
	mbfl_convert_filter *decoder_backup;
	mbfl_memory_device device;
	int from;
	int width;
	int outwidth;
	int outchar;
	int status;
	int endpos;
};

static int
collector_strimwidth(int c, void* data)
{
	struct collector_strimwidth_data *pc = (struct collector_strimwidth_data*)data;

	switch (pc->status) {
	case 10:
		(*pc->decoder->filter_function)(c, pc->decoder);
		break;
	default:
		if (pc->outchar >= pc->from) {
			pc->outwidth += (is_fullwidth(c) ? 2: 1);

			if (pc->outwidth > pc->width) {
				if (pc->status == 0) {
					pc->endpos = pc->device.pos;
					mbfl_convert_filter_copy(pc->decoder, pc->decoder_backup);
				}
				pc->status++;
				(*pc->decoder->filter_function)(c, pc->decoder);
				c = -1;
			} else {
				(*pc->decoder->filter_function)(c, pc->decoder);
			}
		}
		pc->outchar++;
		break;
	}

	return c;
}

mbfl_string *
mbfl_strimwidth(
    mbfl_string *string,
    mbfl_string *marker,
    mbfl_string *result,
    int from,
    int width)
{
	struct collector_strimwidth_data pc;
	mbfl_convert_filter *encoder;
	int n, mkwidth;
	unsigned char *p;

	if (string == NULL || result == NULL) {
		return NULL;
	}
	mbfl_string_init(result);
	result->no_language = string->no_language;
	result->no_encoding = string->no_encoding;
	mbfl_memory_device_init(&pc.device, width, 0);

	/* output code filter */
	pc.decoder = mbfl_convert_filter_new(
	    mbfl_no_encoding_wchar,
	    string->no_encoding,
	    mbfl_memory_device_output, 0, &pc.device);
	pc.decoder_backup = mbfl_convert_filter_new(
	    mbfl_no_encoding_wchar,
	    string->no_encoding,
	    mbfl_memory_device_output, 0, &pc.device);
	/* wchar filter */
	encoder = mbfl_convert_filter_new(
	    string->no_encoding,
	    mbfl_no_encoding_wchar,
	    collector_strimwidth, 0, &pc);
	if (pc.decoder == NULL || pc.decoder_backup == NULL || encoder == NULL) {
		mbfl_convert_filter_delete(encoder);
		mbfl_convert_filter_delete(pc.decoder);
		mbfl_convert_filter_delete(pc.decoder_backup);
		return NULL;
	}
	mkwidth = 0;
	if (marker) {
		mkwidth = mbfl_strwidth(marker);