[libyaml.git] / src / reader.c


#if HAVE_CONFIG_H
#include <config.h>
#endif

#include <yaml/yaml.h>

#include <assert.h>

/* Check for the UTF-16-BE BOM. */
#define IS_UTF16BE_BOM(pointer) ((pointer)[0] == 0xFE && (pointer)[1] == 0xFF)

/* Check for the UTF-16-LE BOM. */
#define IS_UTF16LE_BOM(pointer) ((pointer)[0] == 0xFF && (pointer)[1] == 0xFE)

/* Get a UTF-16-BE character. */
#define UTF16BE_CHAR(pointer)   ((pointer)[0] << 8 + (pointer)[1])

/* Get a UTF-16-LE character. */
#define UTF16LE_CHAR(pointer)   ((pointer)[0] + (pointer)[1] << 8)

/*
 * From http://www.ietf.org/rfc/rfc3629.txt:
 *
 *    Char. number range  |        UTF-8 octet sequence
 *      (hexadecimal)    |              (binary)
 *   --------------------+---------------------------------------------
 *   0000 0000-0000 007F | 0xxxxxxx
 *   0000 0080-0000 07FF | 110xxxxx 10xxxxxx
 *   0000 0800-0000 FFFF | 1110xxxx 10xxxxxx 10xxxxxx
 *   0001 0000-0010 FFFF | 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx
 */

/* Get the length of a UTF-8 character (0 on error). */
#define UTF8_LENGTH(pointer)    \
    ((pointer)[0] < 0x80 ? 1 :  \
    (pointer)[0] < 0xC0 ? 0 :   \
    (pointer)[0] < 0xE0 ? 2 :   \
    (pointer)[0] < 0xF0 ? 3 :   \
    (pointer)[0] < 0xF8 ? 4 : 0)

/* Get the value of the first byte of a UTF-8 sequence (0xFF on error). */
#define UTF8_FIRST_CHUNK(pointer)   \
    ((pointer)[0] < 0x80 ? (pointer)[0] & 0x7F :    \
    (pointer)[0] < 0xC0 ? 0xFF :    \
    (pointer)[0] < 0xE0 ? (pointer)[0] & 0x1F : \
    (pointer)[0] < 0xF0 ? (pointer)[0] & 0x0F : \
    (pointer)[0] < 0xF8 ? (pointer)[0] & 0x07 : 0xFF)

/* Get the value of a non-first byte of a UTF-8 sequence (0xFF on error). */
#define UTF8_NEXT_CHUNK(pointer)    \
    ((pointer)[0] >= 0x80 && (pointer)[0] < 0xC0 ? (pointer)[0] & 0x3F : 0xFF)

/* Determine the length of a UTF-8 character. */

/*
 * Ensure that the buffer contains at least length characters.
 * Return 1 on success, 0 on failure.
 *
 * The length is supposed to be significantly less that the buffer size.
 */

int
yaml_parser_update_buffer(yaml_parser_t *parser, size_t length)
{
    /* If the EOF flag is set, do nothing. */

    if (parser->eof)
        return 1;

    /* Return if the buffer contains enough characters. */

    if (parser->unread >= length)
        return 1;

    /* Determine the input encoding if it is not known yet. */

    if (!parser->encoding) {
        if (!yaml_parser_determine_encoding(parser))
            return 0;
    }

    /* Move the unread characters to the beginning of the buffer. */

    if (parser->buffer < parser->pointer
            && parser->pointer < parser->buffer_end) {
        size_t size = parser->buffer_end - parser->pointer;
        memmove(parser->buffer, parser->pointer, size);
        parser->pointer = parser->buffer;
        parser->buffer_end -= size;
    }
    else if (parser->pointer == parser->buffer_end) {
        parser->pointer = parser->buffer;
        parser->buffer_end = parser->buffer;
    }

    /* Fill the buffer until it has enough characters. */

    while (parser->unread < length)
    {
        /* Fill the raw buffer. */

        if (!yaml_parser_update_raw_buffer(parser)) return 0;

        /* Decode the raw buffer. */

        while (parser->raw_unread)
        {
            unsigned int ch;
            int incomplete = 0;

            /* Decode the next character. */

            switch (parser->encoding)
            {
                case YAML_UTF8_ENCODING:

                    unsigned int utf8_length = UTF8_LENGTH(parser->raw_pointer);
                    unsigned int utf8_chunk;

                    /* Check if the raw buffer contains an incomplete character. */

                    if (utf8_length > parser->raw_unread) {
                        if (parser->eof) {
                            parser->error = YAML_READER_ERROR;
                            return 0;
                        }
                        incomplete = 1;
                    }

                    /* Get the character checking it for validity. */

                    utf8_chunk = UTF8_FIRST_CHUNK(parser->raw_pointer ++);
                    if (utf8_chunk == 0xFF) {
                        parser->error = YAML_READER_ERROR;
                        return 0;
                    }
                    ch = utf8_chunk;
                    parser->raw_unread --;
                    while (-- utf8_length) {
                        utf8_chunk = UTF8_NEXT_CHUNK(parser->raw_pointer ++);
                        if (utf8_chunk == 0xFF) {
                            parser->error = YAML_READER_ERROR;
                            return 0;
                        }
                        ch = ch << 6 + utf8_chunk;
                        parser->raw_unread --;
                    }

                    break;
                
                case YAML_UTF16LE_ENCODING:

                    /* Check if the raw buffer contains an incomplete character. */

                    if (parser->raw_unread < 2) {
                        if (parser->eof) {
                            parser->error = YAML_READER_ERROR;
                            return 0;
                        }
                        incomplete = 1;
                    }

                    /* Get the current character. */

                    ch = UTF16LE_CHAR(parser->raw_pointer);
                    parser->raw_pointer += 2;
                    parser->raw_unread -= 2;

                    break;

                case YAML_UTF16BE_ENCODING:

                    /* Check if the raw buffer contains an incomplete character. */

                    if (parser->raw_unread < 2) {
                        if (parser->eof) {
                            parser->error = YAML_READER_ERROR;
                            return 0;
                        }
                        incomplete = 1;
                    }

                    /* Get the current character. */

                    ch = UTF16BE_CHAR(parser->raw_pointer);
                    parser->raw_pointer += 2;
                    parser->raw_unread -= 2;

                    break;
            }

            /*
             * Check if the character is in the allowed range:
             *      #x9 | #xA | #xD | [#x20-#x7E]               (8 bit)
             *      | #x85 | [#xA0-#xD7FF] | [#xE000-#xFFFD]    (16 bit)
             *      | [#x10000-#x10FFFF]                        (32 bit)
             */

            if (! (ch == 0x09 || ch == 0x0A || ch == 0x0D
                        || (ch >= 0x20 && ch <= 0x7E)
                        || (ch == 0x85) || (ch >= 0xA0 && ch <= 0xD7FF)
                        || (ch >= 0xE000 && ch <= 0xFFFD)
                        || (ch >= 0x10000 && ch <= 0x10FFFF))) {
                parser->error = YAML_READER_ERROR;
                return 0;
            }

            /* Finally put the character into the buffer. */

            /* 0000 0000-0000 007F -> 0xxxxxxx */
            if (ch <= 0x7F) {
                *(parser->buffer_end++) = ch;
            }
            /* 0000 0080-0000 07FF -> 110xxxxx 10xxxxxx */
            else if (ch <= 0x7FF) {
                *(parser->buffer_end++) = 0xC0 + (ch >> 6) & 0x1F;
                *(parser->buffer_end++) = 0x80 + ch & 0x3F;
            }
            /* 0000 0800-0000 FFFF -> 1110xxxx 10xxxxxx 10xxxxxx */
            else if (ch <= 0xFFFF) {
                *(parser->buffer_end++) = 0x80 + ch & 0x3F;
                *(parser->buffer_end++) = 0xC0 + (ch >> 6) & 0x1F;
                
            }
            /* 0001 0000-0010 FFFF -> 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx */
            else {
            }
        }
        
    }

}

/*
 * Determine the input stream encoding by checking the BOM symbol. If no BOM is
 * found, the UTF-8 encoding is assumed. Return 1 on success, 0 on failure.
 */

int
yaml_parser_determine_encoding(yaml_parser_t *parser)
{
    /* Ensure that we had enough bytes in the raw buffer. */

    while (!parser->eof && parser->raw_unread < 2) {
        if (!yaml_parser_update_raw_buffer(parser)) {
            return 0;
        }
    }

    /* Determine the encoding. */

    if (parser->raw_unread >= 2 && IS_UTF16BE_BOM(parser->raw_pointer)) {
        parser->encoding = YAML_UTF16BE_ENCODING;
    }
    else if (parser->raw_unread >= 2 && IS_UTF16LE_BOM(parser->raw_pointer)) {
        parser->encoding = YAML_UTF16LE_ENCODING;
    }
    else {
        parser->encoding = YAML_UTF8_ENCODING;
    }
}

/*
 * Update the raw buffer.
 */

int
yaml_parser_update_raw_buffer(yaml_parser_t *parser)
{
    size_t size_read = 0;

    /* Return if the raw buffer is full. */

    if (parser->raw_unread == YAML_RAW_BUFFER_SIZE) return 1;

    /* Return on EOF. */

    if (parser->eof) return 1;

    /* Move the remaining bytes in the raw buffer to the beginning. */

    if (parser->raw_unread && parser->raw_buffer < parser->raw_pointer) {
        memmove(parser->raw_buffer, parser->raw_pointer, parser->raw_unread);
    }
    parser->raw_pointer = parser->raw_buffer;

    /* Call the read handler to fill the buffer. */

    if (!parser->read_handler(parser->read_handler_data,
                parser->raw_buffer + parser->raw_unread,
                YAML_RAW_BUFFER_SIZE - parser->raw_unread,
                &size_read)) {
        parser->error = YAML_READER_ERROR;
        return 0;
    }
    parser->raw_unread += size_read;
    if (!size_read) {
        parser->eof = 1;
    }

    return 1;
}
Commit	Line	Data
95b98ba9	1
6eb1ded4 KS	2	#if HAVE_CONFIG_H
	3	#include <config.h>
	4	#endif
	5
	6	#include <yaml/yaml.h>
	7
	8	#include <assert.h>
	9
	10	/* Check for the UTF-16-BE BOM. */
	11	#define IS_UTF16BE_BOM(pointer) ((pointer)[0] == 0xFE && (pointer)[1] == 0xFF)
	12
	13	/* Check for the UTF-16-LE BOM. */
	14	#define IS_UTF16LE_BOM(pointer) ((pointer)[0] == 0xFF && (pointer)[1] == 0xFE)
	15
	16	/* Get a UTF-16-BE character. */
	17	#define UTF16BE_CHAR(pointer) ((pointer)[0] << 8 + (pointer)[1])
	18
	19	/* Get a UTF-16-LE character. */
	20	#define UTF16LE_CHAR(pointer) ((pointer)[0] + (pointer)[1] << 8)
	21
	22	/*
	23	* From http://www.ietf.org/rfc/rfc3629.txt:
	24	*
	25	* Char. number range \| UTF-8 octet sequence
	26	* (hexadecimal) \| (binary)
	27	* --------------------+---------------------------------------------
	28	* 0000 0000-0000 007F \| 0xxxxxxx
	29	* 0000 0080-0000 07FF \| 110xxxxx 10xxxxxx
	30	* 0000 0800-0000 FFFF \| 1110xxxx 10xxxxxx 10xxxxxx
	31	* 0001 0000-0010 FFFF \| 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx
	32	*/
	33
	34	/* Get the length of a UTF-8 character (0 on error). */
	35	#define UTF8_LENGTH(pointer) \
	36	((pointer)[0] < 0x80 ? 1 : \
	37	(pointer)[0] < 0xC0 ? 0 : \
	38	(pointer)[0] < 0xE0 ? 2 : \
	39	(pointer)[0] < 0xF0 ? 3 : \
	40	(pointer)[0] < 0xF8 ? 4 : 0)
	41
	42	/* Get the value of the first byte of a UTF-8 sequence (0xFF on error). */
	43	#define UTF8_FIRST_CHUNK(pointer) \
	44	((pointer)[0] < 0x80 ? (pointer)[0] & 0x7F : \
	45	(pointer)[0] < 0xC0 ? 0xFF : \
	46	(pointer)[0] < 0xE0 ? (pointer)[0] & 0x1F : \
	47	(pointer)[0] < 0xF0 ? (pointer)[0] & 0x0F : \
	48	(pointer)[0] < 0xF8 ? (pointer)[0] & 0x07 : 0xFF)
	49
	50	/* Get the value of a non-first byte of a UTF-8 sequence (0xFF on error). */
	51	#define UTF8_NEXT_CHUNK(pointer) \
	52	((pointer)[0] >= 0x80 && (pointer)[0] < 0xC0 ? (pointer)[0] & 0x3F : 0xFF)
	53
	54	/* Determine the length of a UTF-8 character. */
95b98ba9 KS	55
	56	/*
	57	* Ensure that the buffer contains at least length characters.
	58	* Return 1 on success, 0 on failure.
6eb1ded4 KS	59	*
6eb1ded4 KS	60	* The length is supposed to be significantly less that the buffer size.
95b98ba9 KS	61	*/
	62
	63	int
6eb1ded4	64	yaml_parser_update_buffer(yaml_parser_t *parser, size_t length)
95b98ba9 KS	65	{
	66	/* If the EOF flag is set, do nothing. */
	67
	68	if (parser->eof)
	69	return 1;
	70
6eb1ded4 KS	71	/* Return if the buffer contains enough characters. */
	72
	73	if (parser->unread >= length)
	74	return 1;
	75
	76	/* Determine the input encoding if it is not known yet. */
95b98ba9	77
6eb1ded4 KS	78	if (!parser->encoding) {
6eb1ded4 KS	79	if (!yaml_parser_determine_encoding(parser))
95b98ba9	80	return 0;
95b98ba9 KS	81	}
95b98ba9 KS	82
6eb1ded4 KS	83	/* Move the unread characters to the beginning of the buffer. */
	84
	85	if (parser->buffer < parser->pointer
	86	&& parser->pointer < parser->buffer_end) {
	87	size_t size = parser->buffer_end - parser->pointer;
	88	memmove(parser->buffer, parser->pointer, size);
	89	parser->pointer = parser->buffer;
	90	parser->buffer_end -= size;
	91	}
	92	else if (parser->pointer == parser->buffer_end) {
	93	parser->pointer = parser->buffer;
	94	parser->buffer_end = parser->buffer;
	95	}
	96
	97	/* Fill the buffer until it has enough characters. */
	98
	99	while (parser->unread < length)
	100	{
	101	/* Fill the raw buffer. */
	102
	103	if (!yaml_parser_update_raw_buffer(parser)) return 0;
	104
	105	/* Decode the raw buffer. */
	106
	107	while (parser->raw_unread)
	108	{
	109	unsigned int ch;
	110	int incomplete = 0;
	111
	112	/* Decode the next character. */
	113
	114	switch (parser->encoding)
	115	{
	116	case YAML_UTF8_ENCODING:
	117
	118	unsigned int utf8_length = UTF8_LENGTH(parser->raw_pointer);
	119	unsigned int utf8_chunk;
	120
	121	/* Check if the raw buffer contains an incomplete character. */
	122
	123	if (utf8_length > parser->raw_unread) {
	124	if (parser->eof) {
	125	parser->error = YAML_READER_ERROR;
	126	return 0;
	127	}
	128	incomplete = 1;
	129	}
	130
	131	/* Get the character checking it for validity. */
	132
	133	utf8_chunk = UTF8_FIRST_CHUNK(parser->raw_pointer ++);
	134	if (utf8_chunk == 0xFF) {
	135	parser->error = YAML_READER_ERROR;
	136	return 0;
	137	}
	138	ch = utf8_chunk;
	139	parser->raw_unread --;
	140	while (-- utf8_length) {
	141	utf8_chunk = UTF8_NEXT_CHUNK(parser->raw_pointer ++);
	142	if (utf8_chunk == 0xFF) {
	143	parser->error = YAML_READER_ERROR;
	144	return 0;
	145	}
	146	ch = ch << 6 + utf8_chunk;
147	parser->raw_unread --;
148	}
149
150	break;
151
152	case YAML_UTF16LE_ENCODING:
153
154	/* Check if the raw buffer contains an incomplete character. */
155
156	if (parser->raw_unread < 2) {
157	if (parser->eof) {
158	parser->error = YAML_READER_ERROR;
159	return 0;
160	}
161	incomplete = 1;
162	}
163
164	/* Get the current character. */
165
166	ch = UTF16LE_CHAR(parser->raw_pointer);
167	parser->raw_pointer += 2;
168	parser->raw_unread -= 2;
169
170	break;
171
172	case YAML_UTF16BE_ENCODING:
173
174	/* Check if the raw buffer contains an incomplete character. */
175
176	if (parser->raw_unread < 2) {
177	if (parser->eof) {
178	parser->error = YAML_READER_ERROR;
179	return 0;
180	}
181	incomplete = 1;
182	}
183
184	/* Get the current character. */
185
186	ch = UTF16BE_CHAR(parser->raw_pointer);
187	parser->raw_pointer += 2;
188	parser->raw_unread -= 2;
189
190	break;
191	}
192
193	/*
194	* Check if the character is in the allowed range:
195	* #x9 \| #xA \| #xD \| [#x20-#x7E] (8 bit)
196	* \| #x85 \| [#xA0-#xD7FF] \| [#xE000-#xFFFD] (16 bit)
197	* \| [#x10000-#x10FFFF] (32 bit)
198	*/
199
200	if (! (ch == 0x09 \|\| ch == 0x0A \|\| ch == 0x0D
201	\|\| (ch >= 0x20 && ch <= 0x7E)
202	\|\| (ch == 0x85) \|\| (ch >= 0xA0 && ch <= 0xD7FF)
203	\|\| (ch >= 0xE000 && ch <= 0xFFFD)
204	\|\| (ch >= 0x10000 && ch <= 0x10FFFF))) {
205	parser->error = YAML_READER_ERROR;
206	return 0;
207	}
208
209	/* Finally put the character into the buffer. */
210
211	/* 0000 0000-0000 007F -> 0xxxxxxx */
212	if (ch <= 0x7F) {
213	*(parser->buffer_end++) = ch;
214	}
215	/* 0000 0080-0000 07FF -> 110xxxxx 10xxxxxx */
216	else if (ch <= 0x7FF) {
217	*(parser->buffer_end++) = 0xC0 + (ch >> 6) & 0x1F;
218	*(parser->buffer_end++) = 0x80 + ch & 0x3F;
219	}
220	/* 0000 0800-0000 FFFF -> 1110xxxx 10xxxxxx 10xxxxxx */
221	else if (ch <= 0xFFFF) {
222	*(parser->buffer_end++) = 0x80 + ch & 0x3F;
223	*(parser->buffer_end++) = 0xC0 + (ch >> 6) & 0x1F;
224
225	}
226	/* 0001 0000-0010 FFFF -> 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx */
227	else {
228	}
95b98ba9	229	}
6eb1ded4	230
95b98ba9 KS	231	}
95b98ba9 KS	232
6eb1ded4	233	}
95b98ba9	234
6eb1ded4 KS	235	/*
	236	* Determine the input stream encoding by checking the BOM symbol. If no BOM is
	237	* found, the UTF-8 encoding is assumed. Return 1 on success, 0 on failure.
	238	*/
	239
	240	int
	241	yaml_parser_determine_encoding(yaml_parser_t *parser)
	242	{
	243	/* Ensure that we had enough bytes in the raw buffer. */
	244
	245	while (!parser->eof && parser->raw_unread < 2) {
	246	if (!yaml_parser_update_raw_buffer(parser)) {
95b98ba9	247	return 0;
6eb1ded4	248	}
95b98ba9 KS	249	}
95b98ba9 KS	250
6eb1ded4	251	/* Determine the encoding. */
95b98ba9	252
6eb1ded4 KS	253	if (parser->raw_unread >= 2 && IS_UTF16BE_BOM(parser->raw_pointer)) {
	254	parser->encoding = YAML_UTF16BE_ENCODING;
	255	}
	256	else if (parser->raw_unread >= 2 && IS_UTF16LE_BOM(parser->raw_pointer)) {
	257	parser->encoding = YAML_UTF16LE_ENCODING;
	258	}
	259	else {
	260	parser->encoding = YAML_UTF8_ENCODING;
	261	}
95b98ba9 KS	262	}
95b98ba9 KS	263
6eb1ded4 KS	264	/*
	265	* Update the raw buffer.
	266	*/
	267
	268	int
	269	yaml_parser_update_raw_buffer(yaml_parser_t *parser)
	270	{
	271	size_t size_read = 0;
	272
	273	/* Return if the raw buffer is full. */
	274
	275	if (parser->raw_unread == YAML_RAW_BUFFER_SIZE) return 1;
95b98ba9	276
6eb1ded4 KS	277	/* Return on EOF. */
	278
	279	if (parser->eof) return 1;
	280
	281	/* Move the remaining bytes in the raw buffer to the beginning. */
	282
	283	if (parser->raw_unread && parser->raw_buffer < parser->raw_pointer) {
	284	memmove(parser->raw_buffer, parser->raw_pointer, parser->raw_unread);
	285	}
	286	parser->raw_pointer = parser->raw_buffer;
	287
	288	/* Call the read handler to fill the buffer. */
	289
	290	if (!parser->read_handler(parser->read_handler_data,
	291	parser->raw_buffer + parser->raw_unread,
	292	YAML_RAW_BUFFER_SIZE - parser->raw_unread,
	293	&size_read)) {
	294	parser->error = YAML_READER_ERROR;
	295	return 0;
	296	}
	297	parser->raw_unread += size_read;
	298	if (!size_read) {
	299	parser->eof = 1;
	300	}
	301
	302	return 1;
	303	}
95b98ba9	304