curl/lib/urlapi.c

/***************************************************************************
 *                                  _   _ ____  _
 *  Project                     ___| | | |  _ \| |
 *                             / __| | | | |_) | |
 *                            | (__| |_| |  _ <| |___
 *                             \___|\___/|_| \_\_____|
 *
 * Copyright (C) 1998 - 2020, Daniel Stenberg, <daniel@haxx.se>, et al.
 *
 * This software is licensed as described in the file COPYING, which
 * you should have received as part of this distribution. The terms
 * are also available at https://curl.haxx.se/docs/copyright.html.
 *
 * You may opt to use, copy, modify, merge, publish, distribute and/or sell
 * copies of the Software, and permit persons to whom the Software is
 * furnished to do so, under the terms of the COPYING file.
 *
 * This software is distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY
 * KIND, either express or implied.
 *
 ***************************************************************************/

#include "curl_setup.h"

#include "urldata.h"
#include "urlapi-int.h"
#include "strcase.h"
#include "dotdot.h"
#include "url.h"
#include "escape.h"
#include "curl_ctype.h"
#include "inet_pton.h"

/* The last 3 #include files should be in this order */
#include "curl_printf.h"
#include "curl_memory.h"
#include "memdebug.h"

  /* MSDOS/Windows style drive prefix, eg c: in c:foo */
#define STARTS_WITH_DRIVE_PREFIX(str) \
  ((('a' <= str[0] && str[0] <= 'z') || \
    ('A' <= str[0] && str[0] <= 'Z')) && \
   (str[1] == ':'))

  /* MSDOS/Windows style drive prefix, optionally with
   * a '|' instead of ':', followed by a slash or NUL */
#define STARTS_WITH_URL_DRIVE_PREFIX(str) \
  ((('a' <= (str)[0] && (str)[0] <= 'z') || \
    ('A' <= (str)[0] && (str)[0] <= 'Z')) && \
   ((str)[1] == ':' || (str)[1] == '|') && \
   ((str)[2] == '/' || (str)[2] == '\\' || (str)[2] == 0))

/* Internal representation of CURLU. Point to URL-encoded strings. */
struct Curl_URL {
  char *scheme;
  char *user;
  char *password;
  char *options; /* IMAP only? */
  char *host;
  char *zoneid; /* for numerical IPv6 addresses */
  char *port;
  char *path;
  char *query;
  char *fragment;

  char *scratch; /* temporary scratch area */
  char *temppath; /* temporary path pointer */
  long portnum; /* the numerical version */
};

#define DEFAULT_SCHEME "https"

static void free_urlhandle(struct Curl_URL *u)
{
  free(u->scheme);
  free(u->user);
  free(u->password);
  free(u->options);
  free(u->host);
  free(u->zoneid);
  free(u->port);
  free(u->path);
  free(u->query);
  free(u->fragment);
  free(u->scratch);
  free(u->temppath);
}

/* move the full contents of one handle onto another and
   free the original */
static void mv_urlhandle(struct Curl_URL *from,
                         struct Curl_URL *to)
{
  free_urlhandle(to);
  *to = *from;
  free(from);
}

/*
 * Find the separator at the end of the host name, or the '?' in cases like
 * http://www.url.com?id=2380
 */
static const char *find_host_sep(const char *url)
{
  const char *sep;
  const char *query;

  /* Find the start of the hostname */
  sep = strstr(url, "//");
  if(!sep)
    sep = url;
  else
    sep += 2;

  query = strchr(sep, '?');
  sep = strchr(sep, '/');

  if(!sep)
    sep = url + strlen(url);

  if(!query)
    query = url + strlen(url);

  return sep < query ? sep : query;
}

/*
 * Decide in an encoding-independent manner whether a character in an
 * URL must be escaped. The same criterion must be used in strlen_url()
 * and strcpy_url().
 */
static bool urlchar_needs_escaping(int c)
{
    return !(ISCNTRL(c) || ISSPACE(c) || ISGRAPH(c));
}

/*
 * strlen_url() returns the length of the given URL if the spaces within the
 * URL were properly URL encoded.
 * URL encoding should be skipped for host names, otherwise IDN resolution
 * will fail.
 */
static size_t strlen_url(const char *url, bool relative)
{
  const unsigned char *ptr;
  size_t newlen = 0;
  bool left = TRUE; /* left side of the ? */
  const unsigned char *host_sep = (const unsigned char *) url;

  if(!relative)
    host_sep = (const unsigned char *) find_host_sep(url);

  for(ptr = (unsigned char *)url; *ptr; ptr++) {

    if(ptr < host_sep) {
      ++newlen;
      continue;
    }

    switch(*ptr) {
    case '?':
      left = FALSE;
      /* FALLTHROUGH */
    default:
      if(urlchar_needs_escaping(*ptr))
        newlen += 2;
      newlen++;
      break;
    case ' ':
      if(left)
        newlen += 3;
      else
        newlen++;
      break;
    }
  }
  return newlen;
}

/* strcpy_url() copies a url to a output buffer and URL-encodes the spaces in
 * the source URL accordingly.
 * URL encoding should be skipped for host names, otherwise IDN resolution
 * will fail.
 */
static void strcpy_url(char *output, const char *url, bool relative)
{
  /* we must add this with whitespace-replacing */
  bool left = TRUE;
  const unsigned char *iptr;
  char *optr = output;
  const unsigned char *host_sep = (const unsigned char *) url;

  if(!relative)
    host_sep = (const unsigned char *) find_host_sep(url);

  for(iptr = (unsigned char *)url;    /* read from here */
      *iptr;         /* until zero byte */
      iptr++) {

    if(iptr < host_sep) {
      *optr++ = *iptr;
      continue;
    }

    switch(*iptr) {
    case '?':
      left = FALSE;
      /* FALLTHROUGH */
    default:
      if(urlchar_needs_escaping(*iptr)) {
        msnprintf(optr, 4, "%%%02x", *iptr);
        optr += 3;
      }
      else
        *optr++=*iptr;
      break;
    case ' ':
      if(left) {
        *optr++='%'; /* add a '%' */
        *optr++='2'; /* add a '2' */
        *optr++='0'; /* add a '0' */
      }
      else
        *optr++='+'; /* add a '+' here */
      break;
    }
  }
  *optr = 0; /* null-terminate output buffer */

}

/*
 * Returns true if the given URL is absolute (as opposed to relative) within
 * the buffer size. Returns the scheme in the buffer if TRUE and 'buf' is
 * non-NULL.
 */
bool Curl_is_absolute_url(const char *url, char *buf, size_t buflen)
{
  size_t i;
#ifdef WIN32
  if(STARTS_WITH_DRIVE_PREFIX(url))
    return FALSE;
#endif
  for(i = 0; i < buflen && url[i]; ++i) {
    char s = url[i];
    if((s == ':') && (url[i + 1] == '/')) {
      if(buf)
        buf[i] = 0;
      return TRUE;
    }
    /* RFC 3986 3.1 explains:
      scheme      = ALPHA *( ALPHA / DIGIT / "+" / "-" / "." )
    */
    else if(ISALNUM(s) || (s == '+') || (s == '-') || (s == '.') ) {
      if(buf)
        buf[i] = (char)TOLOWER(s);
    }
    else
      break;
  }
  return FALSE;
}

/*
 * Concatenate a relative URL to a base URL making it absolute.
 * URL-encodes any spaces.
 * The returned pointer must be freed by the caller unless NULL
 * (returns NULL on out of memory).
 */
static char *concat_url(const char *base, const char *relurl)
{
  /***
   TRY to append this new path to the old URL
   to the right of the host part. Oh crap, this is doomed to cause
   problems in the future...
  */
  char *newest;
  char *protsep;
  char *pathsep;
  size_t newlen;
  bool host_changed = FALSE;

  const char *useurl = relurl;
  size_t urllen;

  /* we must make our own copy of the URL to play with, as it may
     point to read-only data */
  char *url_clone = strdup(base);

  if(!url_clone)
    return NULL; /* skip out of this NOW */

  /* protsep points to the start of the host name */
  protsep = strstr(url_clone, "//");
  if(!protsep)
    protsep = url_clone;
  else
    protsep += 2; /* pass the slashes */

  if('/' != relurl[0]) {
    int level = 0;

    /* First we need to find out if there's a ?-letter in the URL,
       and cut it and the right-side of that off */
    pathsep = strchr(protsep, '?');
    if(pathsep)
      *pathsep = 0;

    /* we have a relative path to append to the last slash if there's one
       available, or if the new URL is just a query string (starts with a
       '?')  we append the new one at the end of the entire currently worked
       out URL */
    if(useurl[0] != '?') {
      pathsep = strrchr(protsep, '/');
      if(pathsep)
        *pathsep = 0;
    }

    /* Check if there's any slash after the host name, and if so, remember
       that position instead */
    pathsep = strchr(protsep, '/');
    if(pathsep)
      protsep = pathsep + 1;
    else
      protsep = NULL;

    /* now deal with one "./" or any amount of "../" in the newurl
       and act accordingly */

    if((useurl[0] == '.') && (useurl[1] == '/'))
      useurl += 2; /* just skip the "./" */

    while((useurl[0] == '.') &&
          (useurl[1] == '.') &&
          (useurl[2] == '/')) {
      level++;
      useurl += 3; /* pass the "../" */
    }

    if(protsep) {
      while(level--) {
        /* cut off one more level from the right of the original URL */
        pathsep = strrchr(protsep, '/');
        if(pathsep)
          *pathsep = 0;
        else {
          *protsep = 0;
          break;
        }
      }
    }
  }
  else {
    /* We got a new absolute path for this server */

    if(relurl[1] == '/') {
      /* the new URL starts with //, just keep the protocol part from the
         original one */
      *protsep = 0;
      useurl = &relurl[2]; /* we keep the slashes from the original, so we
                              skip the new ones */
      host_changed = TRUE;
    }
    else {
      /* cut off the original URL from the first slash, or deal with URLs
         without slash */
      pathsep = strchr(protsep, '/');
      if(pathsep) {
        /* When people use badly formatted URLs, such as
           "http://www.url.com?dir=/home/daniel" we must not use the first
           slash, if there's a ?-letter before it! */
        char *sep = strchr(protsep, '?');
        if(sep && (sep < pathsep))
          pathsep = sep;
        *pathsep = 0;
      }
      else {
        /* There was no slash. Now, since we might be operating on a badly
           formatted URL, such as "http://www.url.com?id=2380" which doesn't
           use a slash separator as it is supposed to, we need to check for a
           ?-letter as well! */
        pathsep = strchr(protsep, '?');
        if(pathsep)
          *pathsep = 0;
      }
    }
  }

  /* If the new part contains a space, this is a mighty stupid redirect
     but we still make an effort to do "right". To the left of a '?'
     letter we replace each space with %20 while it is replaced with '+'
     on the right side of the '?' letter.
  */
  newlen = strlen_url(useurl, !host_changed);

  urllen = strlen(url_clone);

  newest = malloc(urllen + 1 + /* possible slash */
                  newlen + 1 /* zero byte */);

  if(!newest) {
    free(url_clone); /* don't leak this */
    return NULL;
  }

  /* copy over the root url part */
  memcpy(newest, url_clone, urllen);

  /* check if we need to append a slash */
  if(('/' == useurl[0]) || (protsep && !*protsep) || ('?' == useurl[0]))
    ;
  else
    newest[urllen++]='/';

  /* then append the new piece on the right side */
  strcpy_url(&newest[urllen], useurl, !host_changed);

  free(url_clone);

  return newest;
}

/*
 * parse_hostname_login()
 *
 * Parse the login details (user name, password and options) from the URL and
 * strip them out of the host name
 *
 */
static CURLUcode parse_hostname_login(struct Curl_URL *u,
                                      char **hostname,
                                      unsigned int flags)
{
  CURLUcode result = CURLUE_OK;
  CURLcode ccode;
  char *userp = NULL;
  char *passwdp = NULL;
  char *optionsp = NULL;
  const struct Curl_handler *h = NULL;

  /* At this point, we're hoping all the other special cases have
   * been taken care of, so conn->host.name is at most
   *    [user[:password][;options]]@]hostname
   *
   * We need somewhere to put the embedded details, so do that first.
   */

  char *ptr = strchr(*hostname, '@');
  char *login = *hostname;

  if(!ptr)
    goto out;

  /* We will now try to extract the
   * possible login information in a string like:
   * ftp://user:password@ftp.my.site:8021/README */
  *hostname = ++ptr;

  /* if this is a known scheme, get some details */
  if(u->scheme)
    h = Curl_builtin_scheme(u->scheme);

  /* We could use the login information in the URL so extract it. Only parse
     options if the handler says we should. Note that 'h' might be NULL! */
  ccode = Curl_parse_login_details(login, ptr - login - 1,
                                   &userp, &passwdp,
                                   (h && (h->flags & PROTOPT_URLOPTIONS)) ?
                                   &optionsp:NULL);
  if(ccode) {
    result = CURLUE_MALFORMED_INPUT;
    goto out;
  }

  if(userp) {
    if(flags & CURLU_DISALLOW_USER) {
      /* Option DISALLOW_USER is set and url contains username. */
      result = CURLUE_USER_NOT_ALLOWED;
      goto out;
    }

    u->user = userp;
  }

  if(passwdp)
    u->password = passwdp;

  if(optionsp)
    u->options = optionsp;

  return CURLUE_OK;
  out:

  free(userp);
  free(passwdp);
  free(optionsp);

  return result;
}

UNITTEST CURLUcode Curl_parse_port(struct Curl_URL *u, char *hostname)
{
  char *portptr = NULL;
  char endbracket;
  int len;

  /*
   * Find the end of an IPv6 address, either on the ']' ending bracket or
   * a percent-encoded zone index.
   */
  if(1 == sscanf(hostname, "[%*45[0123456789abcdefABCDEF:.]%c%n",
                 &endbracket, &len)) {
    if(']' == endbracket)
      portptr = &hostname[len];
    else if('%' == endbracket) {
      int zonelen = len;
      if(1 == sscanf(hostname + zonelen, "%*[^]]%c%n", &endbracket, &len)) {
        if(']' != endbracket)
          return CURLUE_MALFORMED_INPUT;
        portptr = &hostname[--zonelen + len + 1];
      }
      else
        return CURLUE_MALFORMED_INPUT;
    }
    else
      return CURLUE_MALFORMED_INPUT;

    /* this is a RFC2732-style specified IP-address */
    if(portptr && *portptr) {
      if(*portptr != ':')
        return CURLUE_MALFORMED_INPUT;
    }
    else
      portptr = NULL;
  }
  else
    portptr = strchr(hostname, ':');

  if(portptr) {
    char *rest;
    long port;
    char portbuf[7];

    /* Browser behavior adaptation. If there's a colon with no digits after,
       just cut off the name there which makes us ignore the colon and just
       use the default port. Firefox, Chrome and Safari all do that. */
    if(!portptr[1]) {
      *portptr = '\0';
      return CURLUE_OK;
    }

    if(!ISDIGIT(portptr[1]))
      return CURLUE_BAD_PORT_NUMBER;

    port = strtol(portptr + 1, &rest, 10);  /* Port number must be decimal */

    if((port <= 0) || (port > 0xffff))
      /* Single unix standard says port numbers are 16 bits long, but we don't
         treat port zero as OK. */
      return CURLUE_BAD_PORT_NUMBER;

    if(rest[0])
      return CURLUE_BAD_PORT_NUMBER;

    *portptr++ = '\0'; /* cut off the name there */
    *rest = 0;
    /* generate a new port number string to get rid of leading zeroes etc */
    msnprintf(portbuf, sizeof(portbuf), "%ld", port);
    u->portnum = port;
    u->port = strdup(portbuf);
    if(!u->port)
      return CURLUE_OUT_OF_MEMORY;
  }

  return CURLUE_OK;
}

/* scan for byte values < 31 or 127 */
static CURLUcode junkscan(const char *part)
{
  if(part) {
    static const char badbytes[]={
      /* */ 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
      0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
      0x10, 0x11, 0x12, 0x13, 0x14, 0x15, 0x16, 0x17,
      0x18, 0x19, 0x1a, 0x1b, 0x1c, 0x1d, 0x1e, 0x1f,
      0x7f,
      0x00 /* null-terminate */
    };
    size_t n = strlen(part);
    size_t nfine = strcspn(part, badbytes);
    if(nfine != n)
      /* since we don't know which part is scanned, return a generic error
         code */
      return CURLUE_MALFORMED_INPUT;
  }
  return CURLUE_OK;
}

static CURLUcode hostname_check(struct Curl_URL *u, char *hostname)
{
  size_t len;
  size_t hlen = strlen(hostname);

  if(hostname[0] == '[') {
#ifdef ENABLE_IPV6
    char dest[16]; /* fits a binary IPv6 address */
#endif
    const char *l = "0123456789abcdefABCDEF:.";
    if(hlen < 4) /* '[::]' is the shortest possible valid string */
      return CURLUE_MALFORMED_INPUT;
    hostname++;
    hlen -= 2;

    if(hostname[hlen] != ']')
      return CURLUE_MALFORMED_INPUT;

    /* only valid letters are ok */
    len = strspn(hostname, l);
    if(hlen != len) {
      hlen = len;
      if(hostname[len] == '%') {
        /* this could now be '%[zone id]' */
        char zoneid[16];
        int i = 0;
        char *h = &hostname[len + 1];
        /* pass '25' if present and is a url encoded percent sign */
        if(!strncmp(h, "25", 2) && h[2] && (h[2] != ']'))
          h += 2;
        while(*h && (*h != ']') && (i < 15))
          zoneid[i++] = *h++;
        if(!i || (']' != *h))
          return CURLUE_MALFORMED_INPUT;
        zoneid[i] = 0;
        u->zoneid = strdup(zoneid);
        if(!u->zoneid)
          return CURLUE_OUT_OF_MEMORY;
        hostname[len] = ']'; /* insert end bracket */
        hostname[len + 1] = 0; /* terminate the hostname */
      }
      else
        return CURLUE_MALFORMED_INPUT;
      /* hostname is fine */
    }
#ifdef ENABLE_IPV6
    hostname[hlen] = 0; /* end the address there */
    if(1 != Curl_inet_pton(AF_INET6, hostname, dest))
      return CURLUE_MALFORMED_INPUT;
    hostname[hlen] = ']'; /* restore ending bracket */
#endif
  }
  else {
    /* letters from the second string is not ok */
    len = strcspn(hostname, " ");
    if(hlen != len)
      /* hostname with bad content */
      return CURLUE_MALFORMED_INPUT;
  }
  if(!hostname[0])
    return CURLUE_NO_HOST;
  return CURLUE_OK;
}

#define HOSTNAME_END(x) (((x) == '/') || ((x) == '?') || ((x) == '#'))

static CURLUcode seturl(const char *url, CURLU *u, unsigned int flags)
{
  char *path;
  bool path_alloced = FALSE;
  char *hostname;
  char *query = NULL;
  char *fragment = NULL;
  CURLUcode result;
  bool url_has_scheme = FALSE;
  char schemebuf[MAX_SCHEME_LEN + 1];
  const char *schemep = NULL;
  size_t schemelen = 0;
  size_t urllen;

  if(!url)
    return CURLUE_MALFORMED_INPUT;

  /*************************************************************
   * Parse the URL.
   ************************************************************/
  /* allocate scratch area */
  urllen = strlen(url);
  if(urllen > CURL_MAX_INPUT_LENGTH)
    /* excessive input length */
    return CURLUE_MALFORMED_INPUT;

  path = u->scratch = malloc(urllen * 2 + 2);
  if(!path)
    return CURLUE_OUT_OF_MEMORY;

  hostname = &path[urllen + 1];
  hostname[0] = 0;

  if(Curl_is_absolute_url(url, schemebuf, sizeof(schemebuf))) {
    url_has_scheme = TRUE;
    schemelen = strlen(schemebuf);
  }

  /* handle the file: scheme */
  if(url_has_scheme && strcasecompare(schemebuf, "file")) {
    /* path has been allocated large enough to hold this */
    strcpy(path, &url[5]);

    hostname = NULL; /* no host for file: URLs */
    u->scheme = strdup("file");
    if(!u->scheme)
      return CURLUE_OUT_OF_MEMORY;

    /* Extra handling URLs with an authority component (i.e. that start with
     * "file://")
     *
     * We allow omitted hostname (e.g. file:/<path>) -- valid according to
     * RFC 8089, but not the (current) WHAT-WG URL spec.
     */
    if(path[0] == '/' && path[1] == '/') {
      /* swallow the two slashes */
      char *ptr = &path[2];

      /*
       * According to RFC 8089, a file: URL can be reliably dereferenced if:
       *
       *  o it has no/blank hostname, or
       *
       *  o the hostname matches "localhost" (case-insensitively), or
       *
       *  o the hostname is a FQDN that resolves to this machine.
       *
       * For brevity, we only consider URLs with empty, "localhost", or
       * "127.0.0.1" hostnames as local.
       *
       * Additionally, there is an exception for URLs with a Windows drive
       * letter in the authority (which was accidentally omitted from RFC 8089
       * Appendix E, but believe me, it was meant to be there. --MK)
       */
      if(ptr[0] != '/' && !STARTS_WITH_URL_DRIVE_PREFIX(ptr)) {
        /* the URL includes a host name, it must match "localhost" or
           "127.0.0.1" to be valid */
        if(!checkprefix("localhost/", ptr) &&
           !checkprefix("127.0.0.1/", ptr)) {
          /* Invalid file://hostname/, expected localhost or 127.0.0.1 or
             none */
          return CURLUE_MALFORMED_INPUT;
        }
        ptr += 9; /* now points to the slash after the host */
      }

      path = ptr;
    }

#if !defined(MSDOS) && !defined(WIN32) && !defined(__CYGWIN__)
    /* Don't allow Windows drive letters when not in Windows.
     * This catches both "file:/c:" and "file:c:" */
    if(('/' == path[0] && STARTS_WITH_URL_DRIVE_PREFIX(&path[1])) ||
       STARTS_WITH_URL_DRIVE_PREFIX(path)) {
      /* File drive letters are only accepted in MSDOS/Windows */
      return CURLUE_MALFORMED_INPUT;
    }
#else
    /* If the path starts with a slash and a drive letter, ditch the slash */
    if('/' == path[0] && STARTS_WITH_URL_DRIVE_PREFIX(&path[1])) {
      /* This cannot be done with strcpy, as the memory chunks overlap! */
      memmove(path, &path[1], strlen(&path[1]) + 1);
    }
#endif

  }
  else {
    /* clear path */
    const char *p;
    const char *hostp;
    size_t len;
    path[0] = 0;

    if(url_has_scheme) {
      int i = 0;
      p = &url[schemelen + 1];
      while(p && (*p == '/') && (i < 4)) {
        p++;
        i++;
      }
      if((i < 1) || (i>3))
        /* less than one or more than three slashes */
        return CURLUE_MALFORMED_INPUT;

      schemep = schemebuf;
      if(!Curl_builtin_scheme(schemep) &&
         !(flags & CURLU_NON_SUPPORT_SCHEME))
        return CURLUE_UNSUPPORTED_SCHEME;

      if(junkscan(schemep))
        return CURLUE_MALFORMED_INPUT;

    }
    else {
      /* no scheme! */

      if(!(flags & (CURLU_DEFAULT_SCHEME|CURLU_GUESS_SCHEME)))
        return CURLUE_MALFORMED_INPUT;
      if(flags & CURLU_DEFAULT_SCHEME)
        schemep = DEFAULT_SCHEME;

      /*
       * The URL was badly formatted, let's try without scheme specified.
       */
      p = url;
    }
    hostp = p; /* host name starts here */

    while(*p && !HOSTNAME_END(*p)) /* find end of host name */
      p++;

    len = p - hostp;
    if(len) {
      memcpy(hostname, hostp, len);
      hostname[len] = 0;
    }
    else {
      if(!(flags & CURLU_NO_AUTHORITY))
        return CURLUE_MALFORMED_INPUT;
    }

    len = strlen(p);
    memcpy(path, p, len);
    path[len] = 0;

    if(schemep) {
      u->scheme = strdup(schemep);
      if(!u->scheme)
        return CURLUE_OUT_OF_MEMORY;
    }
  }

  if(junkscan(path))
    return CURLUE_MALFORMED_INPUT;

  if((flags & CURLU_URLENCODE) && path[0]) {
    /* worst case output length is 3x the original! */
    char *newp = malloc(strlen(path) * 3);
    if(!newp)
      return CURLUE_OUT_OF_MEMORY;
    path_alloced = TRUE;
    strcpy_url(newp, path, TRUE); /* consider it relative */
    u->temppath = path = newp;
  }

  fragment = strchr(path, '#');
  if(fragment) {
    *fragment++ = 0;
    if(fragment[0]) {
      u->fragment = strdup(fragment);
      if(!u->fragment)
        return CURLUE_OUT_OF_MEMORY;
    }
  }

  query = strchr(path, '?');
  if(query) {
    *query++ = 0;
    /* done even if the query part is a blank string */
    u->query = strdup(query);
    if(!u->query)
      return CURLUE_OUT_OF_MEMORY;
  }

  if(!path[0])
    /* if there's no path left set, unset */
    path = NULL;
  else {
    if(!(flags & CURLU_PATH_AS_IS)) {
      /* remove ../ and ./ sequences according to RFC3986 */
      char *newp = Curl_dedotdotify(path);
      if(!newp)
        return CURLUE_OUT_OF_MEMORY;

      if(strcmp(newp, path)) {
        /* if we got a new version */
        if(path_alloced)
          Curl_safefree(u->temppath);
        u->temppath = path = newp;
        path_alloced = TRUE;
      }
      else
        free(newp);
    }

    u->path = path_alloced?path:strdup(path);
    if(!u->path)
      return CURLUE_OUT_OF_MEMORY;
    u->temppath = NULL; /* used now */
  }

  if(hostname) {
    /*
     * Parse the login details and strip them out of the host name.
     */
    if(junkscan(hostname))
      return CURLUE_MALFORMED_INPUT;

    result = parse_hostname_login(u, &hostname, flags);
    if(result)
      return result;

    result = Curl_parse_port(u, hostname);
    if(result)
      return result;

    if(0 == strlen(hostname) && (flags & CURLU_NO_AUTHORITY)) {
      /* Skip hostname check, it's allowed to be empty. */
    }
    else {
      result = hostname_check(u, hostname);
      if(result)
        return result;
    }

    u->host = strdup(hostname);
    if(!u->host)
      return CURLUE_OUT_OF_MEMORY;

    if((flags & CURLU_GUESS_SCHEME) && !schemep) {
      /* legacy curl-style guess based on host name */
      if(checkprefix("ftp.", hostname))
        schemep = "ftp";
      else if(checkprefix("dict.", hostname))
        schemep = "dict";
      else if(checkprefix("ldap.", hostname))
        schemep = "ldap";
      else if(checkprefix("imap.", hostname))
        schemep = "imap";
      else if(checkprefix("smtp.", hostname))
        schemep = "smtp";
      else if(checkprefix("pop3.", hostname))
        schemep = "pop3";
      else
        schemep = "http";

      u->scheme = strdup(schemep);
      if(!u->scheme)
        return CURLUE_OUT_OF_MEMORY;
    }
  }

  Curl_safefree(u->scratch);
  Curl_safefree(u->temppath);

  return CURLUE_OK;
}

/*
 * Parse the URL and set the relevant members of the Curl_URL struct.
 */
static CURLUcode parseurl(const char *url, CURLU *u, unsigned int flags)
{
  CURLUcode result = seturl(url, u, flags);
  if(result) {
    free_urlhandle(u);
    memset(u, 0, sizeof(struct Curl_URL));
  }
  return result;
}

/*
 */
CURLU *curl_url(void)
{
  return calloc(sizeof(struct Curl_URL), 1);
}

void curl_url_cleanup(CURLU *u)
{
  if(u) {
    free_urlhandle(u);
    free(u);
  }
}

#define DUP(dest, src, name)         \
  if(src->name) {                    \
    dest->name = strdup(src->name);  \
    if(!dest->name)                  \
      goto fail;                     \
  }

CURLU *curl_url_dup(CURLU *in)
{
  struct Curl_URL *u = calloc(sizeof(struct Curl_URL), 1);
  if(u) {
    DUP(u, in, scheme);
    DUP(u, in, user);
    DUP(u, in, password);
    DUP(u, in, options);
    DUP(u, in, host);
    DUP(u, in, port);
    DUP(u, in, path);
    DUP(u, in, query);
    DUP(u, in, fragment);
    u->portnum = in->portnum;
  }
  return u;
  fail:
  curl_url_cleanup(u);
  return NULL;
}

CURLUcode curl_url_get(CURLU *u, CURLUPart what,
                       char **part, unsigned int flags)
{
  char *ptr;
  CURLUcode ifmissing = CURLUE_UNKNOWN_PART;
  char portbuf[7];
  bool urldecode = (flags & CURLU_URLDECODE)?1:0;
  bool plusdecode = FALSE;
  (void)flags;
  if(!u)
    return CURLUE_BAD_HANDLE;
  if(!part)
    return CURLUE_BAD_PARTPOINTER;
  *part = NULL;

  switch(what) {
  case CURLUPART_SCHEME:
    ptr = u->scheme;
    ifmissing = CURLUE_NO_SCHEME;
    urldecode = FALSE; /* never for schemes */
    break;
  case CURLUPART_USER:
    ptr = u->user;
    ifmissing = CURLUE_NO_USER;
    break;
  case CURLUPART_PASSWORD:
    ptr = u->password;
    ifmissing = CURLUE_NO_PASSWORD;
    break;
  case CURLUPART_OPTIONS:
    ptr = u->options;
    ifmissing = CURLUE_NO_OPTIONS;
    break;
  case CURLUPART_HOST:
    ptr = u->host;
    ifmissing = CURLUE_NO_HOST;
    break;
  case CURLUPART_ZONEID:
    ptr = u->zoneid;
    break;
  case CURLUPART_PORT:
    ptr = u->port;
    ifmissing = CURLUE_NO_PORT;
    urldecode = FALSE; /* never for port */
    if(!ptr && (flags & CURLU_DEFAULT_PORT) && u->scheme) {
      /* there's no stored port number, but asked to deliver
         a default one for the scheme */
      const struct Curl_handler *h =
        Curl_builtin_scheme(u->scheme);
      if(h) {
        msnprintf(portbuf, sizeof(portbuf), "%ld", h->defport);
        ptr = portbuf;
      }
    }
    else if(ptr && u->scheme) {
      /* there is a stored port number, but ask to inhibit if
         it matches the default one for the scheme */
      const struct Curl_handler *h =
        Curl_builtin_scheme(u->scheme);
      if(h && (h->defport == u->portnum) &&
         (flags & CURLU_NO_DEFAULT_PORT))
        ptr = NULL;
    }
    break;
  case CURLUPART_PATH:
    ptr = u->path;
    if(!ptr) {
      ptr = u->path = strdup("/");
      if(!u->path)
        return CURLUE_OUT_OF_MEMORY;
    }
    break;
  case CURLUPART_QUERY:
    ptr = u->query;
    ifmissing = CURLUE_NO_QUERY;
    plusdecode = urldecode;
    break;
  case CURLUPART_FRAGMENT:
    ptr = u->fragment;
    ifmissing = CURLUE_NO_FRAGMENT;
    break;
  case CURLUPART_URL: {
    char *url;
    char *scheme;
    char *options = u->options;
    char *port = u->port;
    char *allochost = NULL;
    if(u->scheme && strcasecompare("file", u->scheme)) {
      url = aprintf("file://%s%s%s",
                    u->path,
                    u->fragment? "#": "",
                    u->fragment? u->fragment : "");
    }
    else if(!u->host)
      return CURLUE_NO_HOST;
    else {
      const struct Curl_handler *h = NULL;
      if(u->scheme)
        scheme = u->scheme;
      else if(flags & CURLU_DEFAULT_SCHEME)
        scheme = (char *) DEFAULT_SCHEME;
      else
        return CURLUE_NO_SCHEME;

      h = Curl_builtin_scheme(scheme);
      if(!port && (flags & CURLU_DEFAULT_PORT)) {
        /* there's no stored port number, but asked to deliver
           a default one for the scheme */
        if(h) {
          msnprintf(portbuf, sizeof(portbuf), "%ld", h->defport);
          port = portbuf;
        }
      }
      else if(port) {
        /* there is a stored port number, but asked to inhibit if it matches
           the default one for the scheme */
        if(h && (h->defport == u->portnum) &&
           (flags & CURLU_NO_DEFAULT_PORT))
          port = NULL;
      }

      if(h && !(h->flags & PROTOPT_URLOPTIONS))
        options = NULL;

      if((u->host[0] == '[') && u->zoneid) {
        /* make it '[ host %25 zoneid ]' */
        size_t hostlen = strlen(u->host);
        size_t alen = hostlen + 3 + strlen(u->zoneid) + 1;
        allochost = malloc(alen);
        if(!allochost)
          return CURLUE_OUT_OF_MEMORY;
        memcpy(allochost, u->host, hostlen - 1);
        msnprintf(&allochost[hostlen - 1], alen - hostlen + 1,
                  "%%25%s]", u->zoneid);
      }

      url = aprintf("%s://%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s",
                    scheme,
                    u->user ? u->user : "",
                    u->password ? ":": "",
                    u->password ? u->password : "",
                    options ? ";" : "",
                    options ? options : "",
                    (u->user || u->password || options) ? "@": "",
                    allochost ? allochost : u->host,
                    port ? ":": "",
                    port ? port : "",
                    (u->path && (u->path[0] != '/')) ? "/": "",
                    u->path ? u->path : "/",
                    (u->query && u->query[0]) ? "?": "",
                    (u->query && u->query[0]) ? u->query : "",
                    u->fragment? "#": "",
                    u->fragment? u->fragment : "");
      free(allochost);
    }
    if(!url)
      return CURLUE_OUT_OF_MEMORY;
    *part = url;
    return CURLUE_OK;
  }
  default:
    ptr = NULL;
    break;
  }
  if(ptr) {
    *part = strdup(ptr);
    if(!*part)
      return CURLUE_OUT_OF_MEMORY;
    if(plusdecode) {
      /* convert + to space */
      char *plus;
      for(plus = *part; *plus; ++plus) {
        if(*plus == '+')
          *plus = ' ';
      }
    }
    if(urldecode) {
      char *decoded;
      size_t dlen;
      /* this unconditional rejection of control bytes is documented
         API behavior */
      CURLcode res = Curl_urldecode(NULL, *part, 0, &decoded, &dlen,
                                    REJECT_CTRL);
      free(*part);
      if(res) {
        *part = NULL;
        return CURLUE_URLDECODE;
      }
      *part = decoded;
    }
    return CURLUE_OK;
  }
  else
    return ifmissing;
}

CURLUcode curl_url_set(CURLU *u, CURLUPart what,
                       const char *part, unsigned int flags)
{
  char **storep = NULL;
  long port = 0;
  bool urlencode = (flags & CURLU_URLENCODE)? 1 : 0;
  bool plusencode = FALSE;
  bool urlskipslash = FALSE;
  bool appendquery = FALSE;
  bool equalsencode = FALSE;

  if(!u)
    return CURLUE_BAD_HANDLE;
  if(!part) {
    /* setting a part to NULL clears it */
    switch(what) {
    case CURLUPART_URL:
      break;
    case CURLUPART_SCHEME:
      storep = &u->scheme;
      break;
    case CURLUPART_USER:
      storep = &u->user;
      break;
    case CURLUPART_PASSWORD:
      storep = &u->password;
      break;
    case CURLUPART_OPTIONS:
      storep = &u->options;
      break;
    case CURLUPART_HOST:
      storep = &u->host;
      break;
    case CURLUPART_ZONEID:
      storep = &u->zoneid;
      break;
    case CURLUPART_PORT:
      u->portnum = 0;
      storep = &u->port;
      break;
    case CURLUPART_PATH:
      storep = &u->path;
      break;
    case CURLUPART_QUERY:
      storep = &u->query;
      break;
    case CURLUPART_FRAGMENT:
      storep = &u->fragment;
      break;
    default:
      return CURLUE_UNKNOWN_PART;
    }
    if(storep && *storep) {
      free(*storep);
      *storep = NULL;
    }
    return CURLUE_OK;
  }

  switch(what) {
  case CURLUPART_SCHEME:
    if(strlen(part) > MAX_SCHEME_LEN)
      /* too long */
      return CURLUE_MALFORMED_INPUT;
    if(!(flags & CURLU_NON_SUPPORT_SCHEME) &&
       /* verify that it is a fine scheme */
       !Curl_builtin_scheme(part))
      return CURLUE_UNSUPPORTED_SCHEME;
    storep = &u->scheme;
    urlencode = FALSE; /* never */
    break;
  case CURLUPART_USER:
    storep = &u->user;
    break;
  case CURLUPART_PASSWORD:
    storep = &u->password;
    break;
  case CURLUPART_OPTIONS:
    storep = &u->options;
    break;
  case CURLUPART_HOST:
    storep = &u->host;
    free(u->zoneid);
    u->zoneid = NULL;
    break;
  case CURLUPART_ZONEID:
    storep = &u->zoneid;
    break;
  case CURLUPART_PORT:
  {
    char *endp;
    urlencode = FALSE; /* never */
    port = strtol(part, &endp, 10);  /* Port number must be decimal */
    if((port <= 0) || (port > 0xffff))
      return CURLUE_BAD_PORT_NUMBER;
    if(*endp)
      /* weirdly provided number, not good! */
      return CURLUE_MALFORMED_INPUT;
    storep = &u->port;
  }
  break;
  case CURLUPART_PATH:
    urlskipslash = TRUE;
    storep = &u->path;
    break;
  case CURLUPART_QUERY:
    plusencode = urlencode;
    appendquery = (flags & CURLU_APPENDQUERY)?1:0;
    equalsencode = appendquery;
    storep = &u->query;
    break;
  case CURLUPART_FRAGMENT:
    storep = &u->fragment;
    break;
  case CURLUPART_URL: {
    /*
     * Allow a new URL to replace the existing (if any) contents.
     *
     * If the existing contents is enough for a URL, allow a relative URL to
     * replace it.
     */
    CURLUcode result;
    char *oldurl;
    char *redired_url;
    CURLU *handle2;

    if(Curl_is_absolute_url(part, NULL, MAX_SCHEME_LEN + 1)) {
      handle2 = curl_url();
      if(!handle2)
        return CURLUE_OUT_OF_MEMORY;
      result = parseurl(part, handle2, flags);
      if(!result)
        mv_urlhandle(handle2, u);
      else
        curl_url_cleanup(handle2);
      return result;
    }
    /* extract the full "old" URL to do the redirect on */
    result = curl_url_get(u, CURLUPART_URL, &oldurl, flags);
    if(result) {
      /* couldn't get the old URL, just use the new! */
      handle2 = curl_url();
      if(!handle2)
        return CURLUE_OUT_OF_MEMORY;
      result = parseurl(part, handle2, flags);
      if(!result)
        mv_urlhandle(handle2, u);
      else
        curl_url_cleanup(handle2);
      return result;
    }

    /* apply the relative part to create a new URL */
    redired_url = concat_url(oldurl, part);
    free(oldurl);
    if(!redired_url)
      return CURLUE_OUT_OF_MEMORY;

    /* now parse the new URL */
    handle2 = curl_url();
    if(!handle2) {
      free(redired_url);
      return CURLUE_OUT_OF_MEMORY;
    }
    result = parseurl(redired_url, handle2, flags);
    free(redired_url);
    if(!result)
      mv_urlhandle(handle2, u);
    else
      curl_url_cleanup(handle2);
    return result;
  }
  default:
    return CURLUE_UNKNOWN_PART;
  }
  DEBUGASSERT(storep);
  {
    const char *newp = part;
    size_t nalloc = strlen(part);

    if(nalloc > CURL_MAX_INPUT_LENGTH)
      /* excessive input length */
      return CURLUE_MALFORMED_INPUT;

    if(urlencode) {
      const unsigned char *i;
      char *o;
      bool free_part = FALSE;
      char *enc = malloc(nalloc * 3 + 1); /* for worst case! */
      if(!enc)
        return CURLUE_OUT_OF_MEMORY;
      if(plusencode) {
        /* space to plus */
        i = (const unsigned char *)part;
        for(o = enc; *i; ++o, ++i)
          *o = (*i == ' ') ? '+' : *i;
        *o = 0; /* null-terminate */
        part = strdup(enc);
        if(!part) {
          free(enc);
          return CURLUE_OUT_OF_MEMORY;
        }
        free_part = TRUE;
      }
      for(i = (const unsigned char *)part, o = enc; *i; i++) {
        if(Curl_isunreserved(*i) ||
           ((*i == '/') && urlskipslash) ||
           ((*i == '=') && equalsencode) ||
           ((*i == '+') && plusencode)) {
          if((*i == '=') && equalsencode)
            /* only skip the first equals sign */
            equalsencode = FALSE;
          *o = *i;
          o++;
        }
        else {
          msnprintf(o, 4, "%%%02x", *i);
          o += 3;
        }
      }
      *o = 0; /* null-terminate */
      newp = enc;
      if(free_part)
        free((char *)part);
    }
    else {
      char *p;
      newp = strdup(part);
      if(!newp)
        return CURLUE_OUT_OF_MEMORY;
      p = (char *)newp;
      while(*p) {
        /* make sure percent encoded are lower case */
        if((*p == '%') && ISXDIGIT(p[1]) && ISXDIGIT(p[2]) &&
           (ISUPPER(p[1]) || ISUPPER(p[2]))) {
          p[1] = (char)TOLOWER(p[1]);
          p[2] = (char)TOLOWER(p[2]);
          p += 3;
        }
        else
          p++;
      }
    }

    if(appendquery) {
      /* Append the string onto the old query. Add a '&' separator if none is
         present at the end of the exsting query already */
      size_t querylen = u->query ? strlen(u->query) : 0;
      bool addamperand = querylen && (u->query[querylen -1] != '&');
      if(querylen) {
        size_t newplen = strlen(newp);
        char *p = malloc(querylen + addamperand + newplen + 1);
        if(!p) {
          free((char *)newp);
          return CURLUE_OUT_OF_MEMORY;
        }
        strcpy(p, u->query); /* original query */
        if(addamperand)
          p[querylen] = '&'; /* ampersand */
        strcpy(&p[querylen + addamperand], newp); /* new suffix */
        free((char *)newp);
        free(*storep);
        *storep = p;
        return CURLUE_OK;
      }
    }

    if(what == CURLUPART_HOST) {
      if(0 == strlen(newp) && (flags & CURLU_NO_AUTHORITY)) {
        /* Skip hostname check, it's allowed to be empty. */
      }
      else {
        if(hostname_check(u, (char *)newp)) {
          free((char *)newp);
          return CURLUE_MALFORMED_INPUT;
        }
      }
    }

    free(*storep);
    *storep = (char *)newp;
  }
  /* set after the string, to make it not assigned if the allocation above
     fails */
  if(port)
    u->portnum = port;
  return CURLUE_OK;
}