CMake/lib/urlapi.c

/***************************************************************************
 *                                  _   _ ____  _
 *  Project                     ___| | | |  _ \| |
 *                             / __| | | | |_) | |
 *                            | (__| |_| |  _ <| |___
 *                             \___|\___/|_| \_\_____|
 *
 * Copyright (C) Daniel Stenberg, <[email protected]>, 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.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.
 *
 * SPDX-License-Identifier: curl
 *
 ***************************************************************************/
#include "curl_setup.h"

#include "urldata.h"
#include "urlapi-int.h"
#include "strcase.h"
#include "url.h"
#include "escape.h"
#include "curlx/inet_pton.h"
#include "curlx/inet_ntop.h"
#include "strdup.h"
#include "idn.h"
#include "curlx/strparse.h"
#include "curl_memrchr.h"

#ifdef _WIN32
/* MS-DOS/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] == ':'))
#endif

/* MS-DOS/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))

/* scheme is not URL encoded, the longest libcurl supported ones are... */
#define MAX_SCHEME_LEN 40

/*
 * If USE_IPV6 is disabled, we still want to parse IPv6 addresses, so make
 * sure we have _some_ value for AF_INET6 without polluting our fake value
 * everywhere.
 */
#if !defined(USE_IPV6) && !defined(AF_INET6)
#define AF_INET6 (AF_INET + 1)
#endif

/* 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;
  unsigned short portnum; /* the numerical version (if 'port' is set) */
  BIT(query_present);    /* to support blank */
  BIT(fragment_present); /* to support blank */
  BIT(guessed_scheme);   /* when a URL without scheme is parsed */
};

#define DEFAULT_SCHEME "https"

static CURLUcode parseurl_and_replace(const char *url, CURLU *u,
                                      unsigned int flags);

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

/*
 * Find the separator at the end of the hostname, or the '?' in cases like
 * http://www.example.com?id=2380
 */
static const char *find_host_sep(const char *url)
{
  /* Find the start of the hostname */
  const char *sep = strstr(url, "//");
  if(!sep)
    sep = url;
  else
    sep += 2;

  /* Find first / or ? */
  while(*sep && *sep != '/' && *sep != '?')
    sep++;

  return sep;
}

/* convert CURLcode to CURLUcode */
#define cc2cu(x) ((x) == CURLE_TOO_LARGE ? CURLUE_TOO_LARGE :   \
                  CURLUE_OUT_OF_MEMORY)

/* urlencode_str() writes data into an output dynbuf and URL-encodes the
 * spaces in the source URL accordingly.
 *
 * URL encoding should be skipped for hostnames, otherwise IDN resolution
 * will fail.
 */
static CURLUcode urlencode_str(struct dynbuf *o, const char *url,
                               size_t len, bool relative,
                               bool query)
{
  /* we must add this with whitespace-replacing */
  bool left = !query;
  const unsigned char *iptr;
  const unsigned char *host_sep = (const unsigned char *)url;
  CURLcode result = CURLE_OK;

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

    /* output the first piece as-is */
    n = (const char *)host_sep - url;
    result = curlx_dyn_addn(o, url, n);
    len -= n;
  }

  for(iptr = host_sep; len && !result; iptr++, len--) {
    if(*iptr == ' ') {
      if(left)
        result = curlx_dyn_addn(o, "%20", 3);
      else
        result = curlx_dyn_addn(o, "+", 1);
    }
    else if((*iptr < ' ') || (*iptr >= 0x7f)) {
      unsigned char out[3] = { '%' };
      Curl_hexbyte(&out[1], *iptr);
      result = curlx_dyn_addn(o, out, 3);
    }
    else {
      result = curlx_dyn_addn(o, iptr, 1);
      if(*iptr == '?')
        left = FALSE;
    }
  }

  if(result)
    return cc2cu(result);
  return CURLUE_OK;
}

/*
 * Returns the length of the scheme if the given URL is absolute (as opposed
 * to relative). Stores the scheme in the buffer if TRUE and 'buf' is
 * non-NULL. The buflen must be larger than MAX_SCHEME_LEN if buf is set.
 *
 * If 'guess_scheme' is TRUE, it means the URL might be provided without
 * scheme.
 */
size_t Curl_is_absolute_url(const char *url, char *buf, size_t buflen,
                            bool guess_scheme)
{
  size_t i = 0;
  DEBUGASSERT(!buf || (buflen > MAX_SCHEME_LEN));
  (void)buflen; /* only used in debug-builds */
  if(buf)
    buf[0] = 0; /* always leave a defined value in buf */
#ifdef _WIN32
  if(guess_scheme && STARTS_WITH_DRIVE_PREFIX(url))
    return 0;
#endif
  if(ISALPHA(url[0]))
    for(i = 1; i < MAX_SCHEME_LEN; ++i) {
      char s = url[i];
      if(s && (ISALNUM(s) || (s == '+') || (s == '-') || (s == '.'))) {
        /* RFC 3986 3.1 explains:
           scheme      = ALPHA *( ALPHA / DIGIT / "+" / "-" / "." )
        */
      }
      else {
        break;
      }
    }
  if(i && (url[i] == ':') && ((url[i + 1] == '/') || !guess_scheme)) {
    /* If this does not guess scheme, the scheme always ends with the colon so
       that this also detects data: URLs etc. In guessing mode, data: could
       be the hostname "data" with a specified port number. */

    /* the length of the scheme is the name part only */
    size_t len = i;
    if(buf) {
      Curl_strntolower(buf, url, i);
      buf[i] = 0;
    }
    return len;
  }
  return 0;
}

/*
 * Concatenate a relative URL onto a base URL making it absolute.
 */
static CURLUcode redirect_url(const char *base, const char *relurl,
                              CURLU *u, unsigned int flags)
{
  struct dynbuf urlbuf;
  bool host_changed = FALSE;
  const char *useurl = relurl;
  const char *cutoff = NULL;
  size_t prelen;
  CURLUcode uc;

  /* protsep points to the start of the hostname, after [scheme]:// */
  const char *protsep = base + strlen(u->scheme) + 3;
  DEBUGASSERT(base && relurl && u); /* all set here */
  if(!base)
    return CURLUE_MALFORMED_INPUT; /* should never happen */

  /* handle different relative URL types */
  switch(relurl[0]) {
  case '/':
    if(relurl[1] == '/') {
      /* protocol-relative URL: //example.com/path */
      cutoff = protsep;
      useurl = &relurl[2];
      host_changed = TRUE;
    }
    else
      /* absolute /path */
      cutoff = strchr(protsep, '/');
    break;

  case '#':
    /* fragment-only change */
    if(u->fragment)
      cutoff = strchr(protsep, '#');
    break;

  default:
    /* path or query-only change */
    if(u->query && u->query[0])
      /* remove existing query */
      cutoff = strchr(protsep, '?');
    else if(u->fragment && u->fragment[0])
      /* Remove existing fragment */
      cutoff = strchr(protsep, '#');

    if(relurl[0] != '?') {
      /* append a relative path after the last slash */
      cutoff = memrchr(protsep, '/',
                       cutoff ? (size_t)(cutoff - protsep) : strlen(protsep));
      if(cutoff)
        cutoff++; /* truncate after last slash */
    }
    break;
  }

  prelen = cutoff ? (size_t)(cutoff - base) : strlen(base);

  /* build new URL */
  curlx_dyn_init(&urlbuf, CURL_MAX_INPUT_LENGTH);

  if(!curlx_dyn_addn(&urlbuf, base, prelen) &&
     !urlencode_str(&urlbuf, useurl, strlen(useurl), !host_changed, FALSE)) {
    uc = parseurl_and_replace(curlx_dyn_ptr(&urlbuf), u,
                              flags & ~CURLU_PATH_AS_IS);
  }
  else
    uc = CURLUE_OUT_OF_MEMORY;

  curlx_dyn_free(&urlbuf);
  return uc;
}

/* scan for byte values <= 31, 127 and sometimes space */
CURLUcode Curl_junkscan(const char *url, size_t *urllen, bool allowspace)
{
  size_t n = strlen(url);
  size_t i;
  unsigned char control;
  const unsigned char *p = (const unsigned char *)url;
  if(n > CURL_MAX_INPUT_LENGTH)
    return CURLUE_MALFORMED_INPUT;

  control = allowspace ? 0x1f : 0x20;
  for(i = 0; i < n; i++) {
    if(p[i] <= control || p[i] == 127)
      return CURLUE_MALFORMED_INPUT;
  }
  *urllen = n;
  return CURLUE_OK;
}

/*
 * parse_hostname_login()
 *
 * Parse the login details (username, password and options) from the URL and
 * strip them out of the hostname
 *
 */
static CURLUcode parse_hostname_login(struct Curl_URL *u,
                                      const char *login,
                                      size_t len,
                                      unsigned int flags,
                                      size_t *offset) /* to the hostname */
{
  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 assume all the other special cases have been taken
   * care of, so the host is at most
   *
   *   [user[:password][;options]]@]hostname
   *
   * We need somewhere to put the embedded details, so do that first.
   */
  char *ptr;

  DEBUGASSERT(login);

  *offset = 0;
  ptr = memchr(login, '@', len);
  if(!ptr)
    goto out;

  /* We will now try to extract the
   * possible login information in a string like:
   * ftp://user:[email protected]:8021/README */
  ptr++;

  /* if this is a known scheme, get some details */
  if(u->scheme)
    h = Curl_get_scheme_handler(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) {
    /* the only possible error from Curl_parse_login_details is out of
       memory: */
    result = CURLUE_OUT_OF_MEMORY;
    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;
    }
    curlx_free(u->user);
    u->user = userp;
  }

  if(passwdp) {
    curlx_free(u->password);
    u->password = passwdp;
  }

  if(optionsp) {
    curlx_free(u->options);
    u->options = optionsp;
  }

  /* the hostname starts at this offset */
  *offset = ptr - login;
  return CURLUE_OK;

out:

  curlx_free(userp);
  curlx_free(passwdp);
  curlx_free(optionsp);
  u->user = NULL;
  u->password = NULL;
  u->options = NULL;

  return result;
}

UNITTEST CURLUcode Curl_parse_port(struct Curl_URL *u, struct dynbuf *host,
                                   bool has_scheme)
{
  const char *portptr;
  char *hostname = curlx_dyn_ptr(host);
  /*
   * Find the end of an IPv6 address on the ']' ending bracket.
   */
  if(hostname[0] == '[') {
    portptr = strchr(hostname, ']');
    if(!portptr)
      return CURLUE_BAD_IPV6;
    portptr++;
    /* this is a RFC2732-style specified IP-address */
    if(*portptr) {
      if(*portptr != ':')
        return CURLUE_BAD_PORT_NUMBER;
    }
    else
      portptr = NULL;
  }
  else
    portptr = strchr(hostname, ':');

  if(portptr) {
    curl_off_t port;
    size_t keep = portptr - hostname;

    /* Browser behavior adaptation. If there is 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.

       Do not do it if the URL has no scheme, to make something that looks like
       a scheme not work!
    */
    curlx_dyn_setlen(host, keep);
    portptr++;
    if(!*portptr)
      return has_scheme ? CURLUE_OK : CURLUE_BAD_PORT_NUMBER;

    if(curlx_str_number(&portptr, &port, 0xffff) || *portptr)
      return CURLUE_BAD_PORT_NUMBER;

    u->portnum = (unsigned short)port;
    /* generate a new port number string to get rid of leading zeroes etc */
    curlx_free(u->port);
    u->port = curl_maprintf("%" CURL_FORMAT_CURL_OFF_T, port);
    if(!u->port)
      return CURLUE_OUT_OF_MEMORY;
  }

  return CURLUE_OK;
}

/* this assumes 'hostname' now starts with [ */
static CURLUcode ipv6_parse(struct Curl_URL *u, char *hostname,
                            size_t hlen) /* length of hostname */
{
  size_t len;
  DEBUGASSERT(*hostname == '[');
  if(hlen < 4) /* '[::]' is the shortest possible valid string */
    return CURLUE_BAD_IPV6;
  hostname++;
  hlen -= 2;

  /* only valid IPv6 letters are ok */
  len = strspn(hostname, "0123456789abcdefABCDEF:.");

  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_BAD_IPV6;
      zoneid[i] = 0;
      u->zoneid = curlx_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_BAD_IPV6;
    /* hostname is fine */
  }

  /* Normalize the IPv6 address */
  {
    char dest[16]; /* fits a binary IPv6 address */
    hostname[hlen] = 0; /* end the address there */
    if(curlx_inet_pton(AF_INET6, hostname, dest) != 1)
      return CURLUE_BAD_IPV6;
    if(curlx_inet_ntop(AF_INET6, dest, hostname, hlen + 1)) {
      hlen = strlen(hostname); /* might be shorter now */
      hostname[hlen + 1] = 0;
    }
    hostname[hlen] = ']'; /* restore ending bracket */
  }
  return CURLUE_OK;
}

static CURLUcode hostname_check(struct Curl_URL *u, char *hostname,
                                size_t hlen) /* length of hostname */
{
  size_t len;
  DEBUGASSERT(hostname);

  if(!hlen)
    return CURLUE_NO_HOST;
  else if(hostname[0] == '[')
    return ipv6_parse(u, hostname, hlen);
  else {
    /* letters from the second string are not ok */
    len = strcspn(hostname, " \r\n\t/:#?!@{}[]\\$\'\"^`*<>=;,+&()%");
    if(hlen != len)
      /* hostname with bad content */
      return CURLUE_BAD_HOSTNAME;
  }
  return CURLUE_OK;
}

/*
 * Handle partial IPv4 numerical addresses and different bases, like
 * '16843009', '0x7f', '0x7f.1' '0177.1.1.1' etc.
 *
 * If the given input string is syntactically wrong IPv4 or any part for
 * example is too big, this function returns HOST_NAME.
 *
 * Output the "normalized" version of that input string in plain quad decimal
 * integers.
 *
 * Returns the host type.
 */

#define HOST_ERROR   -1 /* out of memory */

#define HOST_NAME    1
#define HOST_IPV4    2
#define HOST_IPV6    3

static int ipv4_normalize(struct dynbuf *host)
{
  bool done = FALSE;
  int n = 0;
  const char *c = curlx_dyn_ptr(host);
  unsigned int parts[4] = { 0, 0, 0, 0 };
  CURLcode result = CURLE_OK;

  if(*c == '[')
    return HOST_IPV6;

  while(!done) {
    int rc;
    curl_off_t l;
    if(*c == '0') {
      if(c[1] == 'x') {
        c += 2; /* skip the prefix */
        rc = curlx_str_hex(&c, &l, UINT_MAX);
      }
      else
        rc = curlx_str_octal(&c, &l, UINT_MAX);
    }
    else
      rc = curlx_str_number(&c, &l, UINT_MAX);

    if(rc)
      return HOST_NAME;

    parts[n] = (unsigned int)l;

    switch(*c) {
    case '.':
      if(n == 3)
        return HOST_NAME;
      n++;
      c++;
      break;

    case '\0':
      done = TRUE;
      break;

    default:
      return HOST_NAME;
    }
  }

  switch(n) {
  case 0: /* a -- 32 bits */
    curlx_dyn_reset(host);

    result = curlx_dyn_addf(host, "%u.%u.%u.%u",
                            (parts[0] >> 24),
                            ((parts[0] >> 16) & 0xff),
                            ((parts[0] >> 8) & 0xff),
                            (parts[0] & 0xff));
    break;
  case 1: /* a.b -- 8.24 bits */
    if((parts[0] > 0xff) || (parts[1] > 0xffffff))
      return HOST_NAME;
    curlx_dyn_reset(host);
    result = curlx_dyn_addf(host, "%u.%u.%u.%u",
                            (parts[0]),
                            ((parts[1] >> 16) & 0xff),
                            ((parts[1] >> 8) & 0xff),
                            (parts[1] & 0xff));
    break;
  case 2: /* a.b.c -- 8.8.16 bits */
    if((parts[0] > 0xff) || (parts[1] > 0xff) || (parts[2] > 0xffff))
      return HOST_NAME;
    curlx_dyn_reset(host);
    result = curlx_dyn_addf(host, "%u.%u.%u.%u",
                            (parts[0]),
                            (parts[1]),
                            ((parts[2] >> 8) & 0xff),
                            (parts[2] & 0xff));
    break;
  case 3: /* a.b.c.d -- 8.8.8.8 bits */
    if((parts[0] > 0xff) || (parts[1] > 0xff) || (parts[2] > 0xff) ||
       (parts[3] > 0xff))
      return HOST_NAME;
    curlx_dyn_reset(host);
    result = curlx_dyn_addf(host, "%u.%u.%u.%u",
                            (parts[0]),
                            (parts[1]),
                            (parts[2]),
                            (parts[3]));
    break;
  }
  if(result)
    return HOST_ERROR;
  return HOST_IPV4;
}

/* if necessary, replace the host content with a URL decoded version */
static CURLUcode urldecode_host(struct dynbuf *host)
{
  char *per = NULL;
  const char *hostname = curlx_dyn_ptr(host);
  per = strchr(hostname, '%');
  if(!per)
    /* nothing to decode */
    return CURLUE_OK;
  else {
    /* encoded */
    size_t dlen;
    char *decoded;
    CURLcode result = Curl_urldecode(hostname, 0, &decoded, &dlen,
                                     REJECT_CTRL);
    if(result)
      return CURLUE_BAD_HOSTNAME;
    curlx_dyn_reset(host);
    result = curlx_dyn_addn(host, decoded, dlen);
    curlx_free(decoded);
    if(result)
      return cc2cu(result);
  }

  return CURLUE_OK;
}

static CURLUcode parse_authority(struct Curl_URL *u,
                                 const char *auth, size_t authlen,
                                 unsigned int flags,
                                 struct dynbuf *host,
                                 bool has_scheme)
{
  size_t offset;
  CURLUcode uc;
  CURLcode result;

  /*
   * Parse the login details and strip them out of the hostname.
   */
  uc = parse_hostname_login(u, auth, authlen, flags, &offset);
  if(uc)
    goto out;

  result = curlx_dyn_addn(host, auth + offset, authlen - offset);
  if(result) {
    uc = cc2cu(result);
    goto out;
  }

  uc = Curl_parse_port(u, host, has_scheme);
  if(uc)
    goto out;

  if(!curlx_dyn_len(host))
    return CURLUE_NO_HOST;

  switch(ipv4_normalize(host)) {
  case HOST_IPV4:
    break;
  case HOST_IPV6:
    uc = ipv6_parse(u, curlx_dyn_ptr(host), curlx_dyn_len(host));
    break;
  case HOST_NAME:
    uc = urldecode_host(host);
    if(!uc)
      uc = hostname_check(u, curlx_dyn_ptr(host), curlx_dyn_len(host));
    break;
  case HOST_ERROR:
    uc = CURLUE_OUT_OF_MEMORY;
    break;
  default:
    uc = CURLUE_BAD_HOSTNAME; /* Bad IPv4 address even */
    break;
  }

out:
  return uc;
}

/* used for HTTP/2 server push */
CURLUcode Curl_url_set_authority(CURLU *u, const char *authority)
{
  CURLUcode result;
  struct dynbuf host;

  DEBUGASSERT(authority);
  curlx_dyn_init(&host, CURL_MAX_INPUT_LENGTH);

  result = parse_authority(u, authority, strlen(authority),
                           CURLU_DISALLOW_USER, &host, !!u->scheme);
  if(result)
    curlx_dyn_free(&host);
  else {
    curlx_free(u->host);
    u->host = curlx_dyn_ptr(&host);
  }
  return result;
}

/*
 * "Remove Dot Segments"
 * https://datatracker.ietf.org/doc/html/rfc3986#section-5.2.4
 */

static bool is_dot(const char **str, size_t *clen)
{
  const char *p = *str;
  if(*p == '.') {
    (*str)++;
    (*clen)--;
    return TRUE;
  }
  else if((*clen >= 3) &&
          (p[0] == '%') && (p[1] == '2') && ((p[2] | 0x20) == 'e')) {
    *str += 3;
    *clen -= 3;
    return TRUE;
  }
  return FALSE;
}

#define ISSLASH(x) ((x) == '/')

/*
 * dedotdotify()
 * @unittest: 1395
 *
 * This function gets a null-terminated path with dot and dotdot sequences
 * passed in and strips them off according to the rules in RFC 3986 section
 * 5.2.4.
 *
 * The function handles a path. It should not contain the query nor fragment.
 *
 * RETURNS
 *
 * Zero for success and 'out' set to an allocated dedotdotified string.
 */
UNITTEST int dedotdotify(const char *input, size_t clen, char **outp);
UNITTEST int dedotdotify(const char *input, size_t clen, char **outp)
{
  struct dynbuf out;
  CURLcode result = CURLE_OK;

  *outp = NULL;
  /* the path always starts with a slash, and a slash has not dot */
  if(clen < 2)
    return 0;

  curlx_dyn_init(&out, clen + 1);

  /*  A. If the input buffer begins with a prefix of "../" or "./", then
      remove that prefix from the input buffer; otherwise, */
  if(is_dot(&input, &clen)) {
    const char *p = input;
    size_t blen = clen;

    if(!clen)
      /* . [end] */
      goto end;
    else if(ISSLASH(*p)) {
      /* one dot followed by a slash */
      input = p + 1;
      clen--;
    }

    /*  D. if the input buffer consists only of "." or "..", then remove
        that from the input buffer; otherwise, */
    else if(is_dot(&p, &blen)) {
      if(!blen)
        /* .. [end] */
        goto end;
      else if(ISSLASH(*p)) {
        /* ../ */
        input = p + 1;
        clen = blen - 1;
      }
    }
  }

  while(clen && !result) { /* until end of path content */
    if(ISSLASH(*input)) {
      const char *p = &input[1];
      size_t blen = clen - 1;
      /*  B. if the input buffer begins with a prefix of "/./" or "/.", where
          "."  is a complete path segment, then replace that prefix with "/" in
          the input buffer; otherwise, */
      if(is_dot(&p, &blen)) {
        if(!blen) { /* /. */
          result = curlx_dyn_addn(&out, "/", 1);
          break;
        }
        else if(ISSLASH(*p)) { /* /./ */
          input = p;
          clen = blen;
          continue;
        }

        /*  C. if the input buffer begins with a prefix of "/../" or "/..",
            where ".." is a complete path segment, then replace that prefix
            with "/" in the input buffer and remove the last segment and its
            preceding "/" (if any) from the output buffer; otherwise, */
        else if(is_dot(&p, &blen) && (ISSLASH(*p) || !blen)) {
          /* remove the last segment from the output buffer */
          size_t len = curlx_dyn_len(&out);
          if(len) {
            char *ptr = curlx_dyn_ptr(&out);
            char *last = memrchr(ptr, '/', len);
            if(last)
              /* trim the output at the slash */
              curlx_dyn_setlen(&out, last - ptr);
          }

          if(blen) { /* /../ */
            input = p;
            clen = blen;
            continue;
          }
          result = curlx_dyn_addn(&out, "/", 1);
          break;
        }
      }
    }

    /*  E. move the first path segment in the input buffer to the end of
        the output buffer, including the initial "/" character (if any) and
        any subsequent characters up to, but not including, the next "/"
        character or the end of the input buffer. */

    result = curlx_dyn_addn(&out, input, 1);
    input++;
    clen--;
  }
end:
  if(!result) {
    if(curlx_dyn_len(&out))
      *outp = curlx_dyn_ptr(&out);
    else {
      *outp = curlx_strdup("");
      if(!*outp)
        return 1;
    }
  }
  return result ? 1 : 0; /* success */
}

static CURLUcode parseurl(const char *url, CURLU *u, unsigned int flags)
{
  const char *path;
  size_t pathlen;
  char *query = NULL;
  char *fragment = NULL;
  char schemebuf[MAX_SCHEME_LEN + 1];
  size_t schemelen = 0;
  size_t urllen;
  CURLUcode result = CURLUE_OK;
  size_t fraglen = 0;
  struct dynbuf host;

  DEBUGASSERT(url);

  curlx_dyn_init(&host, CURL_MAX_INPUT_LENGTH);

  result = Curl_junkscan(url, &urllen, !!(flags & CURLU_ALLOW_SPACE));
  if(result)
    goto fail;

  schemelen = Curl_is_absolute_url(url, schemebuf, sizeof(schemebuf),
                                   flags & (CURLU_GUESS_SCHEME |
                                            CURLU_DEFAULT_SCHEME));

  /* handle the file: scheme */
  if(schemelen && !strcmp(schemebuf, "file")) {
    bool uncpath = FALSE;
    if(urllen <= 6) {
      /* file:/ is not enough to actually be a complete file: URL */
      result = CURLUE_BAD_FILE_URL;
      goto fail;
    }

    /* path has been allocated large enough to hold this */
    path = &url[5];
    pathlen = urllen - 5;

    u->scheme = curlx_strdup("file");
    if(!u->scheme) {
      result = CURLUE_OUT_OF_MEMORY;
      goto fail;
    }

    /* 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 */
      const 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, or
       *
       *  o it is an UNC String transformed to an URI (Windows only, RFC 8089
       *    Appendix E.3).
       *
       * For brevity, we only consider URLs with empty, "localhost", or
       * "127.0.0.1" hostnames as local, otherwise as an UNC String.
       *
       * 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 hostname, it must match "localhost" or
           "127.0.0.1" to be valid */
        if(checkprefix("localhost/", ptr) ||
           checkprefix("127.0.0.1/", ptr)) {
          ptr += 9; /* now points to the slash after the host */
        }
        else {
#ifdef _WIN32
          size_t len;

          /* the hostname, NetBIOS computer name, can not contain disallowed
             chars, and the delimiting slash character must be appended to the
             hostname */
          path = strpbrk(ptr, "/\\:*?\"<>|");
          if(!path || *path != '/') {
            result = CURLUE_BAD_FILE_URL;
            goto fail;
          }

          len = path - ptr;
          if(len) {
            CURLcode code = curlx_dyn_addn(&host, ptr, len);
            if(code) {
              result = cc2cu(code);
              goto fail;
            }
            uncpath = TRUE;
          }

          ptr -= 2; /* now points to the // before the host in UNC */
#else
          /* Invalid file://hostname/, expected localhost or 127.0.0.1 or
             none */
          result = CURLUE_BAD_FILE_URL;
          goto fail;
#endif
        }
      }

      path = ptr;
      pathlen = urllen - (ptr - url);
    }

    if(!uncpath)
      /* no host for file: URLs by default */
      curlx_dyn_reset(&host);

#if !defined(_WIN32) && !defined(MSDOS) && !defined(__CYGWIN__)
    /* Do not 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 MS-DOS/Windows */
      result = CURLUE_BAD_FILE_URL;
      goto fail;
    }
#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! */
      path++;
      pathlen--;
    }
#endif
  }
  else {
    /* clear path */
    const char *schemep = NULL;
    const char *hostp;
    size_t hostlen;

    if(schemelen) {
      int i = 0;
      const char *p = &url[schemelen + 1];
      while((*p == '/') && (i < 4)) {
        p++;
        i++;
      }

      schemep = schemebuf;
      if(!Curl_get_scheme_handler(schemep) &&
         !(flags & CURLU_NON_SUPPORT_SCHEME)) {
        result = CURLUE_UNSUPPORTED_SCHEME;
        goto fail;
      }

      if((i < 1) || (i > 3)) {
        /* less than one or more than three slashes */
        result = CURLUE_BAD_SLASHES;
        goto fail;
      }
      hostp = p; /* hostname starts here */
    }
    else {
      /* no scheme! */

      if(!(flags & (CURLU_DEFAULT_SCHEME | CURLU_GUESS_SCHEME))) {
        result = CURLUE_BAD_SCHEME;
        goto fail;
      }
      if(flags & CURLU_DEFAULT_SCHEME)
        schemep = DEFAULT_SCHEME;

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

    if(schemep) {
      u->scheme = curlx_strdup(schemep);
      if(!u->scheme) {
        result = CURLUE_OUT_OF_MEMORY;
        goto fail;
      }
    }

    /* find the end of the hostname + port number */
    hostlen = strcspn(hostp, "/?#");
    path = &hostp[hostlen];

    /* this pathlen also contains the query and the fragment */
    pathlen = urllen - (path - url);
    if(hostlen) {

      result = parse_authority(u, hostp, hostlen, flags, &host, schemelen);
      if(result)
        goto fail;

      if((flags & CURLU_GUESS_SCHEME) && !schemep) {
        const char *hostname = curlx_dyn_ptr(&host);
        /* legacy curl-style guess based on hostname */
        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 = curlx_strdup(schemep);
        if(!u->scheme) {
          result = CURLUE_OUT_OF_MEMORY;
          goto fail;
        }
        u->guessed_scheme = TRUE;
      }
    }
    else if(flags & CURLU_NO_AUTHORITY) {
      /* allowed to be empty. */
      if(curlx_dyn_add(&host, "")) {
        result = CURLUE_OUT_OF_MEMORY;
        goto fail;
      }
    }
    else {
      result = CURLUE_NO_HOST;
      goto fail;
    }
  }

  fragment = strchr(path, '#');
  if(fragment) {
    fraglen = pathlen - (fragment - path);
    u->fragment_present = TRUE;
    if(fraglen > 1) {
      /* skip the leading '#' in the copy but include the terminating null */
      if(flags & CURLU_URLENCODE) {
        struct dynbuf enc;
        curlx_dyn_init(&enc, CURL_MAX_INPUT_LENGTH);
        result = urlencode_str(&enc, fragment + 1, fraglen - 1, TRUE, FALSE);
        if(result)
          goto fail;
        u->fragment = curlx_dyn_ptr(&enc);
      }
      else {
        u->fragment = Curl_memdup0(fragment + 1, fraglen - 1);
        if(!u->fragment) {
          result = CURLUE_OUT_OF_MEMORY;
          goto fail;
        }
      }
    }
    /* after this, pathlen still contains the query */
    pathlen -= fraglen;
  }

  query = memchr(path, '?', pathlen);
  if(query) {
    size_t qlen = fragment ? (size_t)(fragment - query) :
      pathlen - (query - path);
    pathlen -= qlen;
    u->query_present = TRUE;
    if(qlen > 1) {
      if(flags & CURLU_URLENCODE) {
        struct dynbuf enc;
        curlx_dyn_init(&enc, CURL_MAX_INPUT_LENGTH);
        /* skip the leading question mark */
        result = urlencode_str(&enc, query + 1, qlen - 1, TRUE, TRUE);
        if(result)
          goto fail;
        u->query = curlx_dyn_ptr(&enc);
      }
      else {
        u->query = Curl_memdup0(query + 1, qlen - 1);
        if(!u->query) {
          result = CURLUE_OUT_OF_MEMORY;
          goto fail;
        }
      }
    }
    else {
      /* single byte query */
      u->query = curlx_strdup("");
      if(!u->query) {
        result = CURLUE_OUT_OF_MEMORY;
        goto fail;
      }
    }
  }

  if(pathlen && (flags & CURLU_URLENCODE)) {
    struct dynbuf enc;
    curlx_dyn_init(&enc, CURL_MAX_INPUT_LENGTH);
    result = urlencode_str(&enc, path, pathlen, TRUE, FALSE);
    if(result)
      goto fail;
    pathlen = curlx_dyn_len(&enc);
    path = u->path = curlx_dyn_ptr(&enc);
  }

  if(pathlen <= 1) {
    /* there is no path left or just the slash, unset */
    path = NULL;
  }
  else {
    if(!u->path) {
      u->path = Curl_memdup0(path, pathlen);
      if(!u->path) {
        result = CURLUE_OUT_OF_MEMORY;
        goto fail;
      }
      path = u->path;
    }
    else if(flags & CURLU_URLENCODE)
      /* it might have encoded more than just the path so cut it */
      u->path[pathlen] = 0;

    if(!(flags & CURLU_PATH_AS_IS)) {
      /* remove ../ and ./ sequences according to RFC3986 */
      char *dedot;
      int err = dedotdotify(path, pathlen, &dedot);
      if(err) {
        result = CURLUE_OUT_OF_MEMORY;
        goto fail;
      }
      if(dedot) {
        curlx_free(u->path);
        u->path = dedot;
      }
    }
  }

  u->host = curlx_dyn_ptr(&host);

  return result;
fail:
  curlx_dyn_free(&host);
  free_urlhandle(u);
  return result;
}

/*
 * Parse the URL and, if successful, replace everything in the Curl_URL struct.
 */
static CURLUcode parseurl_and_replace(const char *url, CURLU *u,
                                      unsigned int flags)
{
  CURLUcode result;
  CURLU tmpurl;
  memset(&tmpurl, 0, sizeof(tmpurl));
  result = parseurl(url, &tmpurl, flags);
  if(!result) {
    free_urlhandle(u);
    *u = tmpurl;
  }
  return result;
}

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

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

#define DUP(dest, src, name)                    \
  do {                                          \
    if(src->name) {                             \
      dest->name = curlx_strdup(src->name);     \
      if(!dest->name)                           \
        goto fail;                              \
    }                                           \
  } while(0)

CURLU *curl_url_dup(const CURLU *in)
{
  struct Curl_URL *u = curlx_calloc(1, sizeof(struct Curl_URL));
  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);
    DUP(u, in, zoneid);
    u->portnum = in->portnum;
    u->fragment_present = in->fragment_present;
    u->query_present = in->query_present;
  }
  return u;
fail:
  curl_url_cleanup(u);
  return NULL;
}

#ifndef USE_IDN
#define host_decode(x, y) CURLUE_LACKS_IDN
#define host_encode(x, y) CURLUE_LACKS_IDN
#else
static CURLUcode host_decode(const char *host, char **allochost)
{
  CURLcode result = Curl_idn_decode(host, allochost);
  if(result)
    return (result == CURLE_OUT_OF_MEMORY) ?
      CURLUE_OUT_OF_MEMORY : CURLUE_BAD_HOSTNAME;
  return CURLUE_OK;
}

static CURLUcode host_encode(const char *host, char **allochost)
{
  CURLcode result = Curl_idn_encode(host, allochost);
  if(result)
    return (result == CURLE_OUT_OF_MEMORY) ?
      CURLUE_OUT_OF_MEMORY : CURLUE_BAD_HOSTNAME;
  return CURLUE_OK;
}
#endif

static CURLUcode urlget_format(const CURLU *u, CURLUPart what,
                               const char *ptr, char **partp,
                               bool plusdecode, unsigned int flags)
{
  CURLUcode uc = CURLUE_OK;
  size_t partlen = strlen(ptr);
  bool urldecode = (flags & CURLU_URLDECODE) ? 1 : 0;
  bool urlencode = (flags & CURLU_URLENCODE) ? 1 : 0;
  bool punycode = (flags & CURLU_PUNYCODE) && (what == CURLUPART_HOST);
  bool depunyfy = (flags & CURLU_PUNY2IDN) && (what == CURLUPART_HOST);
  char *part = Curl_memdup0(ptr, partlen);
  *partp = NULL;
  if(!part)
    return CURLUE_OUT_OF_MEMORY;
  if(plusdecode) {
    /* convert + to space */
    char *plus = part;
    size_t i = 0;
    for(i = 0; i < partlen; ++plus, i++) {
      if(*plus == '+')
        *plus = ' ';
    }
  }
  if(urldecode) {
    char *decoded;
    size_t dlen;
    /* this unconditional rejection of control bytes is documented
       API behavior */
    CURLcode res = Curl_urldecode(part, partlen, &decoded, &dlen, REJECT_CTRL);
    curlx_free(part);
    if(res)
      return CURLUE_URLDECODE;
    part = decoded;
    partlen = dlen;
  }
  if(urlencode) {
    struct dynbuf enc;
    curlx_dyn_init(&enc, CURL_MAX_INPUT_LENGTH);
    uc = urlencode_str(&enc, part, partlen, TRUE, what == CURLUPART_QUERY);
    curlx_free(part);
    if(uc)
      return uc;
    part = curlx_dyn_ptr(&enc);
  }
  else if(punycode) {
    if(!Curl_is_ASCII_name(u->host)) {
      char *punyversion = NULL;
      uc = host_decode(part, &punyversion);
      curlx_free(part);
      if(uc)
        return uc;
      part = punyversion;
    }
  }
  else if(depunyfy) {
    if(Curl_is_ASCII_name(u->host)) {
      char *unpunified = NULL;
      uc = host_encode(part, &unpunified);
      curlx_free(part);
      if(uc)
        return uc;
      part = unpunified;
    }
  }
  *partp = part;
  return CURLUE_OK;
}

static CURLUcode urlget_url(const CURLU *u, char **part, unsigned int flags)
{
  char *url;
  const char *scheme;
  char *options = u->options;
  char *port = u->port;
  char *allochost = NULL;
  bool show_fragment =
    u->fragment || (u->fragment_present && flags & CURLU_GET_EMPTY);
  bool show_query = (u->query && u->query[0]) ||
    (u->query_present && flags & CURLU_GET_EMPTY);
  bool punycode = (flags & CURLU_PUNYCODE) ? 1 : 0;
  bool depunyfy = (flags & CURLU_PUNY2IDN) ? 1 : 0;
  bool urlencode = (flags & CURLU_URLENCODE) ? 1 : 0;
  char portbuf[7];
  if(u->scheme && curl_strequal("file", u->scheme)) {
    url = curl_maprintf("file://%s%s%s%s%s",
                        u->path,
                        show_query ? "?": "",
                        u->query ? u->query : "",
                        show_fragment ? "#": "",
                        u->fragment ? u->fragment : "");
  }
  else if(!u->host)
    return CURLUE_NO_HOST;
  else {
    const struct Curl_handler *h = NULL;
    char schemebuf[MAX_SCHEME_LEN + 5];
    if(u->scheme)
      scheme = u->scheme;
    else if(flags & CURLU_DEFAULT_SCHEME)
      scheme = DEFAULT_SCHEME;
    else
      return CURLUE_NO_SCHEME;

    h = Curl_get_scheme_handler(scheme);
    if(!port && (flags & CURLU_DEFAULT_PORT)) {
      /* there is no stored port number, but asked to deliver
         a default one for the scheme */
      if(h) {
        curl_msnprintf(portbuf, sizeof(portbuf), "%u", 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] == '[') {
      if(u->zoneid) {
        /* make it '[ host %25 zoneid ]' */
        struct dynbuf enc;
        size_t hostlen = strlen(u->host);
        curlx_dyn_init(&enc, CURL_MAX_INPUT_LENGTH);
        if(curlx_dyn_addf(&enc, "%.*s%%25%s]", (int)hostlen - 1, u->host,
                          u->zoneid))
          return CURLUE_OUT_OF_MEMORY;
        allochost = curlx_dyn_ptr(&enc);
      }
    }
    else if(urlencode) {
      allochost = curl_easy_escape(NULL, u->host, 0);
      if(!allochost)
        return CURLUE_OUT_OF_MEMORY;
    }
    else if(punycode) {
      if(!Curl_is_ASCII_name(u->host)) {
        CURLUcode ret = host_decode(u->host, &allochost);
        if(ret)
          return ret;
      }
    }
    else if(depunyfy) {
      if(Curl_is_ASCII_name(u->host)) {
        CURLUcode ret = host_encode(u->host, &allochost);
        if(ret)
          return ret;
      }
    }

    if(!(flags & CURLU_NO_GUESS_SCHEME) || !u->guessed_scheme)
      curl_msnprintf(schemebuf, sizeof(schemebuf), "%s://", scheme);
    else
      schemebuf[0] = 0;

    url = curl_maprintf("%s%s%s%s%s%s%s%s%s%s%s%s%s%s%s",
                        schemebuf,
                        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 : "/",
                        show_query ? "?": "",
                        u->query ? u->query : "",
                        show_fragment ? "#": "",
                        u->fragment ? u->fragment : "");
    curlx_free(allochost);
  }
  if(!url)
    return CURLUE_OUT_OF_MEMORY;
  *part = url;
  return CURLUE_OK;
}

CURLUcode curl_url_get(const CURLU *u, CURLUPart what,
                       char **part, unsigned int flags)
{
  const char *ptr;
  CURLUcode ifmissing = CURLUE_UNKNOWN_PART;
  char portbuf[7];
  bool plusdecode = FALSE;
  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;
    flags &= ~CURLU_URLDECODE; /* never for schemes */
    if((flags & CURLU_NO_GUESS_SCHEME) && u->guessed_scheme)
      return CURLUE_NO_SCHEME;
    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;
    ifmissing = CURLUE_NO_ZONEID;
    break;
  case CURLUPART_PORT:
    ptr = u->port;
    ifmissing = CURLUE_NO_PORT;
    flags &= ~CURLU_URLDECODE; /* never for port */
    if(!ptr && (flags & CURLU_DEFAULT_PORT) && u->scheme) {
      /* there is no stored port number, but asked to deliver
         a default one for the scheme */
      const struct Curl_handler *h = Curl_get_scheme_handler(u->scheme);
      if(h) {
        curl_msnprintf(portbuf, sizeof(portbuf), "%u", 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_get_scheme_handler(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 = "/";
    break;
  case CURLUPART_QUERY:
    ptr = u->query;
    ifmissing = CURLUE_NO_QUERY;
    plusdecode = flags & CURLU_URLDECODE;
    if(ptr && !ptr[0] && !(flags & CURLU_GET_EMPTY))
      /* there was a blank query and the user do not ask for it */
      ptr = NULL;
    break;
  case CURLUPART_FRAGMENT:
    ptr = u->fragment;
    ifmissing = CURLUE_NO_FRAGMENT;
    if(!ptr && u->fragment_present && flags & CURLU_GET_EMPTY)
      /* there was a blank fragment and the user asks for it */
      ptr = "";
    break;
  case CURLUPART_URL:
    return urlget_url(u, part, flags);
  default:
    ptr = NULL;
    break;
  }
  if(ptr)
    return urlget_format(u, what, ptr, part, plusdecode, flags);

  return ifmissing;
}

static CURLUcode set_url_scheme(CURLU *u, const char *scheme,
                                unsigned int flags)
{
  size_t plen = strlen(scheme);
  const struct Curl_handler *h = NULL;
  if((plen > MAX_SCHEME_LEN) || (plen < 1))
    /* too long or too short */
    return CURLUE_BAD_SCHEME;
  /* verify that it is a fine scheme */
  h = Curl_get_scheme_handler(scheme);
  if(!h) {
    const char *s = scheme;
    if(!(flags & CURLU_NON_SUPPORT_SCHEME))
      return CURLUE_UNSUPPORTED_SCHEME;
    if(ISALPHA(*s)) {
      /* ALPHA *( ALPHA / DIGIT / "+" / "-" / "." ) */
      while(--plen) {
        if(ISALNUM(*s) || (*s == '+') || (*s == '-') || (*s == '.'))
          s++; /* fine */
        else
          return CURLUE_BAD_SCHEME;
      }
    }
    else
      return CURLUE_BAD_SCHEME;
  }
  u->guessed_scheme = FALSE;
  return CURLUE_OK;
}

static CURLUcode set_url_port(CURLU *u, const char *provided_port)
{
  char *tmp;
  curl_off_t port;
  if(!ISDIGIT(provided_port[0]))
    /* not a number */
    return CURLUE_BAD_PORT_NUMBER;
  if(curlx_str_number(&provided_port, &port, 0xffff) || *provided_port)
    /* weirdly provided number, not good! */
    return CURLUE_BAD_PORT_NUMBER;
  tmp = curl_maprintf("%" CURL_FORMAT_CURL_OFF_T, port);
  if(!tmp)
    return CURLUE_OUT_OF_MEMORY;
  curlx_free(u->port);
  u->port = tmp;
  u->portnum = (unsigned short)port;
  return CURLUE_OK;
}

static CURLUcode set_url(CURLU *u, const char *url, size_t part_size,
                         unsigned int flags)
{
  /*
   * 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 uc;
  char *oldurl = NULL;

  if(!part_size) {
    /* a blank URL is not a valid URL unless we already have a complete one
       and this is a redirect */
    uc = curl_url_get(u, CURLUPART_URL, &oldurl, flags);
    if(!uc) {
      /* success, meaning the "" is a fine relative URL, but nothing
         changes */
      curlx_free(oldurl);
      return CURLUE_OK;
    }
    if(uc == CURLUE_OUT_OF_MEMORY)
      return uc;
    return CURLUE_MALFORMED_INPUT;
  }

  /* if the new URL is absolute replace the existing with the new. */
  if(Curl_is_absolute_url(url, NULL, 0,
                          flags & (CURLU_GUESS_SCHEME | CURLU_DEFAULT_SCHEME)))
    return parseurl_and_replace(url, u, flags);

  /* if the old URL is incomplete (we cannot get an absolute URL in
     'oldurl'), replace the existing with the new */
  uc = curl_url_get(u, CURLUPART_URL, &oldurl, flags);
  if(uc == CURLUE_OUT_OF_MEMORY)
    return uc;
  else if(uc)
    return parseurl_and_replace(url, u, flags);

  DEBUGASSERT(oldurl); /* it is set here */
  /* apply the relative part to create a new URL */
  uc = redirect_url(oldurl, url, u, flags);
  curlx_free(oldurl);
  return uc;
}

static CURLUcode urlset_clear(CURLU *u, CURLUPart what)
{
  switch(what) {
  case CURLUPART_URL:
    free_urlhandle(u);
    memset(u, 0, sizeof(struct Curl_URL));
    break;
  case CURLUPART_SCHEME:
    Curl_safefree(u->scheme);
    u->guessed_scheme = FALSE;
    break;
  case CURLUPART_USER:
    Curl_safefree(u->user);
    break;
  case CURLUPART_PASSWORD:
    Curl_safefree(u->password);
    break;
  case CURLUPART_OPTIONS:
    Curl_safefree(u->options);
    break;
  case CURLUPART_HOST:
    Curl_safefree(u->host);
    break;
  case CURLUPART_ZONEID:
    Curl_safefree(u->zoneid);
    break;
  case CURLUPART_PORT:
    u->portnum = 0;
    Curl_safefree(u->port);
    break;
  case CURLUPART_PATH:
    Curl_safefree(u->path);
    break;
  case CURLUPART_QUERY:
    Curl_safefree(u->query);
    u->query_present = FALSE;
    break;
  case CURLUPART_FRAGMENT:
    Curl_safefree(u->fragment);
    u->fragment_present = FALSE;
    break;
  default:
    return CURLUE_UNKNOWN_PART;
  }
  return CURLUE_OK;
}

static bool allowed_in_path(unsigned char x)
{
  switch(x) {
  case '!':
  case '$':
  case '&':
  case '\'':
  case '(':
  case ')':
  case '{':
  case '}':
  case '[':
  case ']':
  case '*':
  case '+':
  case ',':
  case ';':
  case '=':
  case ':':
  case '@':
  case '/':
    return TRUE;
  }
  return FALSE;
}

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

  if(!u)
    return CURLUE_BAD_HANDLE;
  if(!part)
    /* setting a part to NULL clears it */
    return urlset_clear(u, what);

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

  switch(what) {
  case CURLUPART_SCHEME: {
    CURLUcode status = set_url_scheme(u, part, flags);
    if(status)
      return status;
    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;
    Curl_safefree(u->zoneid);
    break;
  case CURLUPART_ZONEID:
    storep = &u->zoneid;
    break;
  case CURLUPART_PORT:
    return set_url_port(u, part);
  case CURLUPART_PATH:
    pathmode = TRUE;
    leadingslash = TRUE; /* enforce */
    storep = &u->path;
    break;
  case CURLUPART_QUERY:
    plusencode = urlencode;
    appendquery = (flags & CURLU_APPENDQUERY) ? 1 : 0;
    equalsencode = appendquery;
    storep = &u->query;
    u->query_present = TRUE;
    break;
  case CURLUPART_FRAGMENT:
    storep = &u->fragment;
    u->fragment_present = TRUE;
    break;
  case CURLUPART_URL:
    return set_url(u, part, nalloc, flags);
  default:
    return CURLUE_UNKNOWN_PART;
  }
  DEBUGASSERT(storep);
  {
    const char *newp;
    struct dynbuf enc;
    curlx_dyn_init(&enc, nalloc * 3 + 1 + leadingslash);

    if(leadingslash && (part[0] != '/')) {
      CURLcode result = curlx_dyn_addn(&enc, "/", 1);
      if(result)
        return cc2cu(result);
    }
    if(urlencode) {
      const unsigned char *i;

      for(i = (const unsigned char *)part; *i; i++) {
        CURLcode result;
        if((*i == ' ') && plusencode) {
          result = curlx_dyn_addn(&enc, "+", 1);
          if(result)
            return CURLUE_OUT_OF_MEMORY;
        }
        else if(ISUNRESERVED(*i) ||
                (pathmode && allowed_in_path(*i)) ||
                ((*i == '=') && equalsencode)) {
          if((*i == '=') && equalsencode)
            /* only skip the first equals sign */
            equalsencode = FALSE;
          result = curlx_dyn_addn(&enc, i, 1);
          if(result)
            return cc2cu(result);
        }
        else {
          unsigned char out[3] = { '%' };
          Curl_hexbyte(&out[1], *i);
          result = curlx_dyn_addn(&enc, out, 3);
          if(result)
            return cc2cu(result);
        }
      }
    }
    else {
      char *p;
      CURLcode result = curlx_dyn_add(&enc, part);
      if(result)
        return cc2cu(result);
      p = curlx_dyn_ptr(&enc);
      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] = Curl_raw_tolower(p[1]);
          p[2] = Curl_raw_tolower(p[2]);
          p += 3;
        }
        else
          p++;
      }
    }
    newp = curlx_dyn_ptr(&enc);

    if(appendquery && newp) {
      /* Append the 'newp' string onto the old query. Add a '&' separator if
         none is present at the end of the existing query already */

      size_t querylen = u->query ? strlen(u->query) : 0;
      bool addamperand = querylen && (u->query[querylen - 1] != '&');
      if(querylen) {
        struct dynbuf qbuf;
        curlx_dyn_init(&qbuf, CURL_MAX_INPUT_LENGTH);

        if(curlx_dyn_addn(&qbuf, u->query, querylen)) /* add original query */
          goto nomem;

        if(addamperand) {
          if(curlx_dyn_addn(&qbuf, "&", 1))
            goto nomem;
        }
        if(curlx_dyn_add(&qbuf, newp))
          goto nomem;
        curlx_dyn_free(&enc);
        curlx_free(*storep);
        *storep = curlx_dyn_ptr(&qbuf);
        return CURLUE_OK;
nomem:
        curlx_dyn_free(&enc);
        return CURLUE_OUT_OF_MEMORY;
      }
    }

    else if(what == CURLUPART_HOST) {
      size_t n = curlx_dyn_len(&enc);
      if(!n && (flags & CURLU_NO_AUTHORITY)) {
        /* Skip hostname check, it is allowed to be empty. */
      }
      else {
        bool bad = FALSE;
        if(!n)
          bad = TRUE; /* empty hostname is not okay */
        else if(!urlencode) {
          /* if the hostname part was not URL encoded here, it was set ready
             URL encoded so we need to decode it to check */
          size_t dlen;
          char *decoded = NULL;
          CURLcode result =
            Curl_urldecode(newp, n, &decoded, &dlen, REJECT_CTRL);
          if(result || hostname_check(u, decoded, dlen))
            bad = TRUE;
          curlx_free(decoded);
        }
        else if(hostname_check(u, (char *)CURL_UNCONST(newp), n))
          bad = TRUE;
        if(bad) {
          curlx_dyn_free(&enc);
          return CURLUE_BAD_HOSTNAME;
        }
      }
    }

    curlx_free(*storep);
    *storep = (char *)CURL_UNCONST(newp);
  }
  return CURLUE_OK;
}