/**

 * uri.c: set of generic URI related routines

 *

 * Reference: RFCs 3986, 2732 and 2373

 *

 * See Copyright for the status of this software.

 *

 * daniel@veillard.com

 */

#define IN_LIBXML

#include "libxml.h"

#include <string.h>

#include <libxml/xmlmemory.h>

#include <libxml/uri.h>

#include <libxml/globals.h>

#include <libxml/xmlerror.h>

/**

 * MAX_URI_LENGTH:

 *

 * The definition of the URI regexp in the above RFC has no size limit

 * In practice they are usually relativey short except for the

 * data URI scheme as defined in RFC 2397. Even for data URI the usual

 * maximum size before hitting random practical limits is around 64 KB

 * and 4KB is usually a maximum admitted limit for proper operations.

 * The value below is more a security limit than anything else and

 * really should never be hit by 'normal' operations

 * Set to 1 MByte in 2012, this is only enforced on output

 */

#define MAX_URI_LENGTH 1024 * 1024

static void

xmlURIErrMemory(const char *extra)

{

    if (extra)

        __xmlRaiseError(NULL, NULL, NULL,

                        NULL, NULL, XML_FROM_URI,

                        XML_ERR_NO_MEMORY, XML_ERR_FATAL, NULL, 0,

                        extra, NULL, NULL, 0, 0,

                        "Memory allocation failed : %s\n", extra);

    else

        __xmlRaiseError(NULL, NULL, NULL,

                        NULL, NULL, XML_FROM_URI,

                        XML_ERR_NO_MEMORY, XML_ERR_FATAL, NULL, 0,

                        NULL, NULL, NULL, 0, 0,

                        "Memory allocation failed\n");

}

static void xmlCleanURI(xmlURIPtr uri);

/*

 * Old rule from 2396 used in legacy handling code

 * alpha    = lowalpha | upalpha

 */

#define IS_ALPHA(x) (IS_LOWALPHA(x) || IS_UPALPHA(x))

/*

 * lowalpha = "a" | "b" | "c" | "d" | "e" | "f" | "g" | "h" | "i" | "j" |

 *            "k" | "l" | "m" | "n" | "o" | "p" | "q" | "r" | "s" | "t" |

 *            "u" | "v" | "w" | "x" | "y" | "z"

 */

#define IS_LOWALPHA(x) (((x) >= 'a') && ((x) <= 'z'))

/*

 * upalpha = "A" | "B" | "C" | "D" | "E" | "F" | "G" | "H" | "I" | "J" |

 *           "K" | "L" | "M" | "N" | "O" | "P" | "Q" | "R" | "S" | "T" |

 *           "U" | "V" | "W" | "X" | "Y" | "Z"

 */

#define IS_UPALPHA(x) (((x) >= 'A') && ((x) <= 'Z'))

#ifdef IS_DIGIT

#undef IS_DIGIT

#endif

/*

 * digit = "0" | "1" | "2" | "3" | "4" | "5" | "6" | "7" | "8" | "9"

 */

#define IS_DIGIT(x) (((x) >= '0') && ((x) <= '9'))

/*

 * alphanum = alpha | digit

 */

#define IS_ALPHANUM(x) (IS_ALPHA(x) || IS_DIGIT(x))

/*

 * mark = "-" | "_" | "." | "!" | "~" | "*" | "'" | "(" | ")"

 */

#define IS_MARK(x) (((x) == '-') || ((x) == '_') || ((x) == '.') ||     \

    ((x) == '!') || ((x) == '~') || ((x) == '*') || ((x) == '\'') ||    \

    ((x) == '(') || ((x) == ')'))

/*

 * unwise = "{" | "}" | "|" | "\" | "^" | "`"

 */

#define IS_UNWISE(p)                                                    \

      (((*(p) == '{')) || ((*(p) == '}')) || ((*(p) == '|')) ||         \

       ((*(p) == '\\')) || ((*(p) == '^')) || ((*(p) == '[')) ||        \

       ((*(p) == ']')) || ((*(p) == '`')))

/*

 * reserved = ";" | "/" | "?" | ":" | "@" | "&" | "=" | "+" | "$" | "," |

 *            "[" | "]"

 */

#define IS_RESERVED(x) (((x) == ';') || ((x) == '/') || ((x) == '?') || \

        ((x) == ':') || ((x) == '@') || ((x) == '&') || ((x) == '=') || \

        ((x) == '+') || ((x) == '$') || ((x) == ',') || ((x) == '[') || \

        ((x) == ']'))

/*

 * unreserved = alphanum | mark

 */

#define IS_UNRESERVED(x) (IS_ALPHANUM(x) || IS_MARK(x))

/*

 * Skip to next pointer char, handle escaped sequences

 */

#define NEXT(p) ((*p == '%')? p += 3 : p++)

/*

 * Productions from the spec.

 *

 *    authority     = server | reg_name

 *    reg_name      = 1*( unreserved | escaped | "$" | "," |

 *                        ";" | ":" | "@" | "&" | "=" | "+" )

 *

 * path          = [ abs_path | opaque_part ]

 */

#define STRNDUP(s, n) (char *) xmlStrndup((const xmlChar *)(s), (n))

/************************************************************************

 *									*

 *                         RFC 3986 parser				*

 *									*

 ************************************************************************/

#define ISA_DIGIT(p) ((*(p) >= '0') && (*(p) <= '9'))

#define ISA_ALPHA(p) (((*(p) >= 'a') && (*(p) <= 'z')) ||		\

                      ((*(p) >= 'A') && (*(p) <= 'Z')))

#define ISA_HEXDIG(p)							\

       (ISA_DIGIT(p) || ((*(p) >= 'a') && (*(p) <= 'f')) ||		\

        ((*(p) >= 'A') && (*(p) <= 'F')))

/*

 *    sub-delims    = "!" / "$" / "&" / "'" / "(" / ")"

 *                     / "*" / "+" / "," / ";" / "="

 */

#define ISA_SUB_DELIM(p)						\

      (((*(p) == '!')) || ((*(p) == '$')) || ((*(p) == '&')) ||		\

       ((*(p) == '(')) || ((*(p) == ')')) || ((*(p) == '*')) ||		\

       ((*(p) == '+')) || ((*(p) == ',')) || ((*(p) == ';')) ||		\

       ((*(p) == '=')) || ((*(p) == '\'')))

/*

 *    gen-delims    = ":" / "/" / "?" / "#" / "[" / "]" / "@"

 */

#define ISA_GEN_DELIM(p)						\

      (((*(p) == ':')) || ((*(p) == '/')) || ((*(p) == '?')) ||         \

       ((*(p) == '#')) || ((*(p) == '[')) || ((*(p) == ']')) ||         \

       ((*(p) == '@')))

/*

 *    reserved      = gen-delims / sub-delims

 */

#define ISA_RESERVED(p) (ISA_GEN_DELIM(p) || (ISA_SUB_DELIM(p)))

/*

 *    unreserved    = ALPHA / DIGIT / "-" / "." / "_" / "~"

 */

#define ISA_UNRESERVED(p)						\

      ((ISA_ALPHA(p)) || (ISA_DIGIT(p)) || ((*(p) == '-')) ||		\

       ((*(p) == '.')) || ((*(p) == '_')) || ((*(p) == '~')))

/*

 *    pct-encoded   = "%" HEXDIG HEXDIG

 */

#define ISA_PCT_ENCODED(p)						\

     ((*(p) == '%') && (ISA_HEXDIG(p + 1)) && (ISA_HEXDIG(p + 2)))

/*

 *    pchar         = unreserved / pct-encoded / sub-delims / ":" / "@"

 */

#define ISA_PCHAR(p)							\

     (ISA_UNRESERVED(p) || ISA_PCT_ENCODED(p) || ISA_SUB_DELIM(p) ||	\

      ((*(p) == ':')) || ((*(p) == '@')))

/**

 * xmlParse3986Scheme:

 * @uri:  pointer to an URI structure

 * @str:  pointer to the string to analyze

 *

 * Parse an URI scheme

 *

 * ALPHA *( ALPHA / DIGIT / "+" / "-" / "." )

 *

 * Returns 0 or the error code

 */

static int

xmlParse3986Scheme(xmlURIPtr uri, const char **str) {

    const char *cur;

    if (str == NULL)

	return(-1);

    cur = *str;

    if (!ISA_ALPHA(cur))

	return(2);

    cur++;

    while (ISA_ALPHA(cur) || ISA_DIGIT(cur) ||

           (*cur == '+') || (*cur == '-') || (*cur == '.')) cur++;

    if (uri != NULL) {

	if (uri->scheme != NULL) xmlFree(uri->scheme);

	uri->scheme = STRNDUP(*str, cur - *str);

    }

    *str = cur;

    return(0);

}

/**

 * xmlParse3986Fragment:

 * @uri:  pointer to an URI structure

 * @str:  pointer to the string to analyze

 *

 * Parse the query part of an URI

 *

 * fragment      = *( pchar / "/" / "?" )

 * NOTE: the strict syntax as defined by 3986 does not allow '[' and ']'

 *       in the fragment identifier but this is used very broadly for

 *       xpointer scheme selection, so we are allowing it here to not break

 *       for example all the DocBook processing chains.

 *

 * Returns 0 or the error code

 */

static int

xmlParse3986Fragment(xmlURIPtr uri, const char **str)

{

    const char *cur;

    if (str == NULL)

        return (-1);

    cur = *str;

    while ((ISA_PCHAR(cur)) || (*cur == '/') || (*cur == '?') ||

           (*cur == '[') || (*cur == ']') ||

           ((uri != NULL) && (uri->cleanup & 1) && (IS_UNWISE(cur))))

        NEXT(cur);

    if (uri != NULL) {

        if (uri->fragment != NULL)

            xmlFree(uri->fragment);

	if (uri->cleanup & 2)

	    uri->fragment = STRNDUP(*str, cur - *str);

	else

	    uri->fragment = xmlURIUnescapeString(*str, cur - *str, NULL);

    }

    *str = cur;

    return (0);

}

/**

 * xmlParse3986Query:

 * @uri:  pointer to an URI structure

 * @str:  pointer to the string to analyze

 *

 * Parse the query part of an URI

 *

 * query = *uric

 *

 * Returns 0 or the error code

 */

static int

xmlParse3986Query(xmlURIPtr uri, const char **str)

{

    const char *cur;

    if (str == NULL)

        return (-1);

    cur = *str;

    while ((ISA_PCHAR(cur)) || (*cur == '/') || (*cur == '?') ||

           ((uri != NULL) && (uri->cleanup & 1) && (IS_UNWISE(cur))))

        NEXT(cur);

    if (uri != NULL) {

        if (uri->query != NULL)

            xmlFree(uri->query);

	if (uri->cleanup & 2)

	    uri->query = STRNDUP(*str, cur - *str);

	else

	    uri->query = xmlURIUnescapeString(*str, cur - *str, NULL);

	/* Save the raw bytes of the query as well.

	 * See: http://mail.gnome.org/archives/xml/2007-April/thread.html#00114

	 */

	if (uri->query_raw != NULL)

	    xmlFree (uri->query_raw);

	uri->query_raw = STRNDUP (*str, cur - *str);

    }

    *str = cur;

    return (0);

}

/**

 * xmlParse3986Port:

 * @uri:  pointer to an URI structure

 * @str:  the string to analyze

 *

 * Parse a port part and fills in the appropriate fields

 * of the @uri structure

 *

 * port          = *DIGIT

 *

 * Returns 0 or the error code

 */

static int

xmlParse3986Port(xmlURIPtr uri, const char **str)

{

    const char *cur = *str;

    unsigned port = 0; /* unsigned for defined overflow behavior */

    if (ISA_DIGIT(cur)) {

	while (ISA_DIGIT(cur)) {

	    port = port * 10 + (*cur - '0');

	    cur++;

	}

	if (uri != NULL)

	    uri->port = port & INT_MAX; /* port value modulo INT_MAX+1 */

	*str = cur;

	return(0);

    }

    return(1);

}

/**

 * xmlParse3986Userinfo:

 * @uri:  pointer to an URI structure

 * @str:  the string to analyze

 *

 * Parse an user informations part and fills in the appropriate fields

 * of the @uri structure

 *

 * userinfo      = *( unreserved / pct-encoded / sub-delims / ":" )

 *

 * Returns 0 or the error code

 */

static int

xmlParse3986Userinfo(xmlURIPtr uri, const char **str)

{

    const char *cur;

    cur = *str;

    while (ISA_UNRESERVED(cur) || ISA_PCT_ENCODED(cur) ||

           ISA_SUB_DELIM(cur) || (*cur == ':'))

	NEXT(cur);

    if (*cur == '@') {

	if (uri != NULL) {

	    if (uri->user != NULL) xmlFree(uri->user);

	    if (uri->cleanup & 2)

		uri->user = STRNDUP(*str, cur - *str);

	    else

		uri->user = xmlURIUnescapeString(*str, cur - *str, NULL);

	}

	*str = cur;

	return(0);

    }

    return(1);

}

/**

 * xmlParse3986DecOctet:

 * @str:  the string to analyze

 *

 *    dec-octet     = DIGIT                 ; 0-9

 *                  / %x31-39 DIGIT         ; 10-99

 *                  / "1" 2DIGIT            ; 100-199

 *                  / "2" %x30-34 DIGIT     ; 200-249

 *                  / "25" %x30-35          ; 250-255

 *

 * Skip a dec-octet.

 *

 * Returns 0 if found and skipped, 1 otherwise

 */

static int

xmlParse3986DecOctet(const char **str) {

    const char *cur = *str;

    if (!(ISA_DIGIT(cur)))

        return(1);

    if (!ISA_DIGIT(cur+1))

	cur++;

    else if ((*cur != '0') && (ISA_DIGIT(cur + 1)) && (!ISA_DIGIT(cur+2)))

	cur += 2;

    else if ((*cur == '1') && (ISA_DIGIT(cur + 1)) && (ISA_DIGIT(cur + 2)))

	cur += 3;

    else if ((*cur == '2') && (*(cur + 1) >= '0') &&

	     (*(cur + 1) <= '4') && (ISA_DIGIT(cur + 2)))

	cur += 3;

    else if ((*cur == '2') && (*(cur + 1) == '5') &&

	     (*(cur + 2) >= '0') && (*(cur + 1) <= '5'))

	cur += 3;

    else

        return(1);

    *str = cur;

    return(0);

}

/**

 * xmlParse3986Host:

 * @uri:  pointer to an URI structure

 * @str:  the string to analyze

 *

 * Parse an host part and fills in the appropriate fields

 * of the @uri structure

 *

 * host          = IP-literal / IPv4address / reg-name

 * IP-literal    = "[" ( IPv6address / IPvFuture  ) "]"

 * IPv4address   = dec-octet "." dec-octet "." dec-octet "." dec-octet

 * reg-name      = *( unreserved / pct-encoded / sub-delims )

 *

 * Returns 0 or the error code

 */

static int

xmlParse3986Host(xmlURIPtr uri, const char **str)

{

    const char *cur = *str;

    const char *host;

    host = cur;

    /*

     * IPv6 and future adressing scheme are enclosed between brackets

     */

    if (*cur == '[') {

        cur++;

	while ((*cur != ']') && (*cur != 0))

	    cur++;

	if (*cur != ']')

	    return(1);

	cur++;

	goto found;

    }

    /*

     * try to parse an IPv4

     */

    if (ISA_DIGIT(cur)) {

        if (xmlParse3986DecOctet(&cur) != 0)

	    goto not_ipv4;

	if (*cur != '.')

	    goto not_ipv4;

	cur++;

        if (xmlParse3986DecOctet(&cur) != 0)

	    goto not_ipv4;

	if (*cur != '.')

	    goto not_ipv4;

        if (xmlParse3986DecOctet(&cur) != 0)

	    goto not_ipv4;

	if (*cur != '.')

	    goto not_ipv4;

        if (xmlParse3986DecOctet(&cur) != 0)

	    goto not_ipv4;

	goto found;

not_ipv4:

        cur = *str;

    }

    /*

     * then this should be a hostname which can be empty

     */

    while (ISA_UNRESERVED(cur) || ISA_PCT_ENCODED(cur) || ISA_SUB_DELIM(cur))

        NEXT(cur);

found:

    if (uri != NULL) {

	if (uri->authority != NULL) xmlFree(uri->authority);

	uri->authority = NULL;

	if (uri->server != NULL) xmlFree(uri->server);

	if (cur != host) {

	    if (uri->cleanup & 2)

		uri->server = STRNDUP(host, cur - host);

	    else

		uri->server = xmlURIUnescapeString(host, cur - host, NULL);

	} else

	    uri->server = NULL;

    }

    *str = cur;

    return(0);

}

/**

 * xmlParse3986Authority:

 * @uri:  pointer to an URI structure

 * @str:  the string to analyze

 *

 * Parse an authority part and fills in the appropriate fields

 * of the @uri structure

 *

 * authority     = [ userinfo "@" ] host [ ":" port ]

 *

 * Returns 0 or the error code

 */

static int

xmlParse3986Authority(xmlURIPtr uri, const char **str)

{

    const char *cur;

    int ret;

    cur = *str;

    /*

     * try to parse an userinfo and check for the trailing @

     */

    ret = xmlParse3986Userinfo(uri, &cur);

    if ((ret != 0) || (*cur != '@'))

        cur = *str;

    else

        cur++;

    ret = xmlParse3986Host(uri, &cur);

    if (ret != 0) return(ret);

    if (*cur == ':') {

        cur++;

        ret = xmlParse3986Port(uri, &cur);

	if (ret != 0) return(ret);

    }

    *str = cur;

    return(0);

}

/**

 * xmlParse3986Segment:

 * @str:  the string to analyze

 * @forbid: an optional forbidden character

 * @empty: allow an empty segment

 *

 * Parse a segment and fills in the appropriate fields

 * of the @uri structure

 *

 * segment       = *pchar

 * segment-nz    = 1*pchar

 * segment-nz-nc = 1*( unreserved / pct-encoded / sub-delims / "@" )

 *               ; non-zero-length segment without any colon ":"

 *

 * Returns 0 or the error code

 */

static int

xmlParse3986Segment(const char **str, char forbid, int empty)

{

    const char *cur;

    cur = *str;

    if (!ISA_PCHAR(cur)) {

        if (empty)

	    return(0);

	return(1);

    }

    while (ISA_PCHAR(cur) && (*cur != forbid))

        NEXT(cur);

    *str = cur;

    return (0);

}

/**

 * xmlParse3986PathAbEmpty:

 * @uri:  pointer to an URI structure

 * @str:  the string to analyze

 *

 * Parse an path absolute or empty and fills in the appropriate fields

 * of the @uri structure

 *

 * path-abempty  = *( "/" segment )

 *

 * Returns 0 or the error code

 */

static int

xmlParse3986PathAbEmpty(xmlURIPtr uri, const char **str)

{

    const char *cur;

    int ret;

    cur = *str;

    while (*cur == '/') {

        cur++;

	ret = xmlParse3986Segment(&cur, 0, 1);

	if (ret != 0) return(ret);

    }

    if (uri != NULL) {

	if (uri->path != NULL) xmlFree(uri->path);

        if (*str != cur) {

            if (uri->cleanup & 2)

                uri->path = STRNDUP(*str, cur - *str);

            else

                uri->path = xmlURIUnescapeString(*str, cur - *str, NULL);

        } else {

            uri->path = NULL;

        }

    }

    *str = cur;

    return (0);

}

/**

 * xmlParse3986PathAbsolute:

 * @uri:  pointer to an URI structure

 * @str:  the string to analyze

 *

 * Parse an path absolute and fills in the appropriate fields

 * of the @uri structure

 *

 * path-absolute = "/" [ segment-nz *( "/" segment ) ]

 *

 * Returns 0 or the error code

 */

static int

xmlParse3986PathAbsolute(xmlURIPtr uri, const char **str)

{

    const char *cur;

    int ret;

    cur = *str;

    if (*cur != '/')

        return(1);

    cur++;

    ret = xmlParse3986Segment(&cur, 0, 0);

    if (ret == 0) {

	while (*cur == '/') {

	    cur++;

	    ret = xmlParse3986Segment(&cur, 0, 1);

	    if (ret != 0) return(ret);

	}

    }

    if (uri != NULL) {

	if (uri->path != NULL) xmlFree(uri->path);

        if (cur != *str) {

            if (uri->cleanup & 2)

                uri->path = STRNDUP(*str, cur - *str);

            else

                uri->path = xmlURIUnescapeString(*str, cur - *str, NULL);

        } else {

            uri->path = NULL;

        }

    }

    *str = cur;

    return (0);

}

/**

 * xmlParse3986PathRootless:

 * @uri:  pointer to an URI structure

 * @str:  the string to analyze

 *

 * Parse an path without root and fills in the appropriate fields

 * of the @uri structure

 *

 * path-rootless = segment-nz *( "/" segment )

 *

 * Returns 0 or the error code

 */

static int

xmlParse3986PathRootless(xmlURIPtr uri, const char **str)

{

    const char *cur;

    int ret;

    cur = *str;

    ret = xmlParse3986Segment(&cur, 0, 0);

    if (ret != 0) return(ret);

    while (*cur == '/') {

        cur++;

	ret = xmlParse3986Segment(&cur, 0, 1);

	if (ret != 0) return(ret);

    }

    if (uri != NULL) {

	if (uri->path != NULL) xmlFree(uri->path);

        if (cur != *str) {

            if (uri->cleanup & 2)

                uri->path = STRNDUP(*str, cur - *str);

            else

                uri->path = xmlURIUnescapeString(*str, cur - *str, NULL);

        } else {

            uri->path = NULL;

        }

    }

    *str = cur;

    return (0);

}

/**

 * xmlParse3986PathNoScheme:

 * @uri:  pointer to an URI structure

 * @str:  the string to analyze

 *

 * Parse an path which is not a scheme and fills in the appropriate fields

 * of the @uri structure

 *

 * path-noscheme = segment-nz-nc *( "/" segment )

 *

 * Returns 0 or the error code

 */

static int

xmlParse3986PathNoScheme(xmlURIPtr uri, const char **str)

{

    const char *cur;

    int ret;

    cur = *str;

    ret = xmlParse3986Segment(&cur, ':', 0);

    if (ret != 0) return(ret);

    while (*cur == '/') {

        cur++;

	ret = xmlParse3986Segment(&cur, 0, 1);

	if (ret != 0) return(ret);

    }

    if (uri != NULL) {

	if (uri->path != NULL) xmlFree(uri->path);

        if (cur != *str) {

            if (uri->cleanup & 2)

                uri->path = STRNDUP(*str, cur - *str);

            else

                uri->path = xmlURIUnescapeString(*str, cur - *str, NULL);

        } else {

            uri->path = NULL;

        }

    }

    *str = cur;

    return (0);

}

/**

 * xmlParse3986HierPart:

 * @uri:  pointer to an URI structure

 * @str:  the string to analyze

 *

 * Parse an hierarchical part and fills in the appropriate fields

 * of the @uri structure

 *

 * hier-part     = "//" authority path-abempty

 *                / path-absolute

 *                / path-rootless

 *                / path-empty

 *

 * Returns 0 or the error code

 */

static int

xmlParse3986HierPart(xmlURIPtr uri, const char **str)

{

    const char *cur;

    int ret;

    cur = *str;

    if ((*cur == '/') && (*(cur + 1) == '/')) {

        cur += 2;

	ret = xmlParse3986Authority(uri, &cur);

	if (ret != 0) return(ret);

	if (uri->server == NULL)

	    uri->port = -1;

	ret = xmlParse3986PathAbEmpty(uri, &cur);

	if (ret != 0) return(ret);

	*str = cur;

	return(0);

    } else if (*cur == '/') {

        ret = xmlParse3986PathAbsolute(uri, &cur);

	if (ret != 0) return(ret);

    } else if (ISA_PCHAR(cur)) {

        ret = xmlParse3986PathRootless(uri, &cur);

	if (ret != 0) return(ret);

    } else {

	/* path-empty is effectively empty */

	if (uri != NULL) {

	    if (uri->path != NULL) xmlFree(uri->path);

	    uri->path = NULL;

	}

    }

    *str = cur;

    return (0);

}

/**

 * xmlParse3986RelativeRef:

 * @uri:  pointer to an URI structure

 * @str:  the string to analyze

 *

 * Parse an URI string and fills in the appropriate fields

 * of the @uri structure

 *

 * relative-ref  = relative-part [ "?" query ] [ "#" fragment ]

 * relative-part = "//" authority path-abempty

 *               / path-absolute

 *               / path-noscheme

 *               / path-empty

 *

 * Returns 0 or the error code

 */

static int

xmlParse3986RelativeRef(xmlURIPtr uri, const char *str) {

    int ret;

    if ((*str == '/') && (*(str + 1) == '/')) {

        str += 2;

	ret = xmlParse3986Authority(uri, &str);

	if (ret != 0) return(ret);

	ret = xmlParse3986PathAbEmpty(uri, &str);

	if (ret != 0) return(ret);

    } else if (*str == '/') {

	ret = xmlParse3986PathAbsolute(uri, &str);

	if (ret != 0) return(ret);

    } else if (ISA_PCHAR(str)) {

        ret = xmlParse3986PathNoScheme(uri, &str);

	if (ret != 0) return(ret);

    } else {

	/* path-empty is effectively empty */

	if (uri != NULL) {

	    if (uri->path != NULL) xmlFree(uri->path);

	    uri->path = NULL;

	}

    }

    if (*str == '?') {

	str++;

	ret = xmlParse3986Query(uri, &str);

	if (ret != 0) return(ret);

    }

    if (*str == '#') {

	str++;

	ret = xmlParse3986Fragment(uri, &str);

	if (ret != 0) return(ret);

    }

    if (*str != 0) {

	xmlCleanURI(uri);

	return(1);

    }

    return(0);

}

/**

 * xmlParse3986URI:

 * @uri:  pointer to an URI structure

 * @str:  the string to analyze

 *

 * Parse an URI string and fills in the appropriate fields

 * of the @uri structure

 *

 * scheme ":" hier-part [ "?" query ] [ "#" fragment ]

 *

 * Returns 0 or the error code

 */

static int

xmlParse3986URI(xmlURIPtr uri, const char *str) {

    int ret;

    ret = xmlParse3986Scheme(uri, &str);

    if (ret != 0) return(ret);

    if (*str != ':') {

	return(1);

    }

    str++;

    ret = xmlParse3986HierPart(uri, &str);

    if (ret != 0) return(ret);

    if (*str == '?') {

	str++;

	ret = xmlParse3986Query(uri, &str);

	if (ret != 0) return(ret);

    }

    if (*str == '#') {

	str++;

	ret = xmlParse3986Fragment(uri, &str);

	if (ret != 0) return(ret);

    }

    if (*str != 0) {

	xmlCleanURI(uri);

	return(1);