path: root/netwerk/base/nsURLParsers.cpp

/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
/* This Source Code Form is subject to the terms of the Mozilla Public
 * License, v. 2.0. If a copy of the MPL was not distributed with this
 * file, You can obtain one at http://mozilla.org/MPL/2.0/. */

#include <string.h>

#include "mozilla/RangedPtr.h"
#include "mozilla/TextUtils.h"

#include "nsCRTGlue.h"
#include "nsURLParsers.h"
#include "nsURLHelper.h"
#include "nsString.h"

using namespace mozilla;

//----------------------------------------------------------------------------

static uint32_t CountConsecutiveSlashes(const char* str, int32_t len) {
  RangedPtr<const char> p(str, len);
  uint32_t count = 0;
  while (len-- && *p++ == '/') ++count;
  return count;
}

//----------------------------------------------------------------------------
// nsBaseURLParser implementation
//----------------------------------------------------------------------------

NS_IMPL_ISUPPORTS(nsAuthURLParser, nsIURLParser)
NS_IMPL_ISUPPORTS(nsNoAuthURLParser, nsIURLParser)

#define SET_RESULT(component, pos, len)                \
  PR_BEGIN_MACRO                                       \
  if (component##Pos) *component##Pos = uint32_t(pos); \
  if (component##Len) *component##Len = int32_t(len);  \
  PR_END_MACRO

#define OFFSET_RESULT(component, offset)         \
  PR_BEGIN_MACRO                                 \
  if (component##Pos) *component##Pos += offset; \
  PR_END_MACRO

NS_IMETHODIMP
nsBaseURLParser::ParseURL(const char* spec, int32_t specLen,
                          uint32_t* schemePos, int32_t* schemeLen,
                          uint32_t* authorityPos, int32_t* authorityLen,
                          uint32_t* pathPos, int32_t* pathLen) {
  if (NS_WARN_IF(!spec)) {
    return NS_ERROR_INVALID_POINTER;
  }

  if (specLen < 0) specLen = strlen(spec);

  const char* stop = nullptr;
  const char* colon = nullptr;
  const char* slash = nullptr;
  const char* p = spec;
  uint32_t offset = 0;
  int32_t len = specLen;

  // skip leading whitespace
  while (*p == ' ' || *p == '\n' || *p == '\r' || *p == '\t') {
    spec++;
    specLen--;
    offset++;

    p++;
    len--;
  }

  for (; len && *p && !colon && !slash; ++p, --len) {
    switch (*p) {
      case ':':
        if (!colon) colon = p;
        break;
      case '/':  // start of filepath
      case '?':  // start of query
      case '#':  // start of ref
        if (!slash) slash = p;
        break;
      case '@':  // username@hostname
      case '[':  // start of IPv6 address literal
        if (!stop) stop = p;
        break;
    }
  }
  // disregard the first colon if it follows an '@' or a '['
  if (colon && stop && colon > stop) colon = nullptr;

  // if the spec only contained whitespace ...
  if (specLen == 0) {
    SET_RESULT(scheme, 0, -1);
    SET_RESULT(authority, 0, 0);
    SET_RESULT(path, 0, 0);
    return NS_OK;
  }

  // ignore trailing whitespace and control characters
  for (p = spec + specLen - 1; ((unsigned char)*p <= ' ') && (p != spec); --p)
    ;

  specLen = p - spec + 1;

  if (colon && (colon < slash || !slash)) {
    //
    // spec = <scheme>:/<the-rest>
    //
    // or
    //
    // spec = <scheme>:<authority>
    // spec = <scheme>:<path-no-slashes>
    //
    if (!net_IsValidScheme(nsDependentCSubstring(spec, colon - spec)) ||
        (*(colon + 1) == ':')) {
      return NS_ERROR_MALFORMED_URI;
    }
    SET_RESULT(scheme, offset, colon - spec);
    if (authorityLen || pathLen) {
      uint32_t schemeLen = colon + 1 - spec;
      offset += schemeLen;
      ParseAfterScheme(colon + 1, specLen - schemeLen, authorityPos,
                       authorityLen, pathPos, pathLen);
      OFFSET_RESULT(authority, offset);
      OFFSET_RESULT(path, offset);
    }
  } else {
    //
    // spec = <authority-no-port-or-password>/<path>
    // spec = <path>
    //
    // or
    //
    // spec = <authority-no-port-or-password>/<path-with-colon>
    // spec = <path-with-colon>
    //
    // or
    //
    // spec = <authority-no-port-or-password>
    // spec = <path-no-slashes-or-colon>
    //
    SET_RESULT(scheme, 0, -1);
    if (authorityLen || pathLen) {
      ParseAfterScheme(spec, specLen, authorityPos, authorityLen, pathPos,
                       pathLen);
      OFFSET_RESULT(authority, offset);
      OFFSET_RESULT(path, offset);
    }
  }
  return NS_OK;
}

NS_IMETHODIMP
nsBaseURLParser::ParseAuthority(const char* auth, int32_t authLen,
                                uint32_t* usernamePos, int32_t* usernameLen,
                                uint32_t* passwordPos, int32_t* passwordLen,
                                uint32_t* hostnamePos, int32_t* hostnameLen,
                                int32_t* port) {
  if (NS_WARN_IF(!auth)) {
    return NS_ERROR_INVALID_POINTER;
  }

  if (authLen < 0) authLen = strlen(auth);

  SET_RESULT(username, 0, -1);
  SET_RESULT(password, 0, -1);
  SET_RESULT(hostname, 0, authLen);
  if (port) *port = -1;
  return NS_OK;
}

NS_IMETHODIMP
nsBaseURLParser::ParseUserInfo(const char* userinfo, int32_t userinfoLen,
                               uint32_t* usernamePos, int32_t* usernameLen,
                               uint32_t* passwordPos, int32_t* passwordLen) {
  SET_RESULT(username, 0, -1);
  SET_RESULT(password, 0, -1);
  return NS_OK;
}

NS_IMETHODIMP
nsBaseURLParser::ParseServerInfo(const char* serverinfo, int32_t serverinfoLen,
                                 uint32_t* hostnamePos, int32_t* hostnameLen,
                                 int32_t* port) {
  SET_RESULT(hostname, 0, -1);
  if (port) *port = -1;
  return NS_OK;
}

NS_IMETHODIMP
nsBaseURLParser::ParsePath(const char* path, int32_t pathLen,
                           uint32_t* filepathPos, int32_t* filepathLen,
                           uint32_t* queryPos, int32_t* queryLen,
                           uint32_t* refPos, int32_t* refLen) {
  if (NS_WARN_IF(!path)) {
    return NS_ERROR_INVALID_POINTER;
  }

  if (pathLen < 0) pathLen = strlen(path);

  // path = [/]<segment1>/<segment2>/<...>/<segmentN>?<query>#<ref>

  // XXX PL_strnpbrk would be nice, but it's buggy

  // search for first occurrence of either ? or #
  const char *query_beg = nullptr, *query_end = nullptr;
  const char* ref_beg = nullptr;
  const char* p = nullptr;
  for (p = path; p < path + pathLen; ++p) {
    // only match the query string if it precedes the reference fragment
    if (!ref_beg && !query_beg && *p == '?')
      query_beg = p + 1;
    else if (*p == '#') {
      ref_beg = p + 1;
      if (query_beg) query_end = p;
      break;
    }
  }

  if (query_beg) {
    if (query_end)
      SET_RESULT(query, query_beg - path, query_end - query_beg);
    else
      SET_RESULT(query, query_beg - path, pathLen - (query_beg - path));
  } else
    SET_RESULT(query, 0, -1);

  if (ref_beg)
    SET_RESULT(ref, ref_beg - path, pathLen - (ref_beg - path));
  else
    SET_RESULT(ref, 0, -1);

  const char* end;
  if (query_beg)
    end = query_beg - 1;
  else if (ref_beg)
    end = ref_beg - 1;
  else
    end = path + pathLen;

  // an empty file path is no file path
  if (end != path)
    SET_RESULT(filepath, 0, end - path);
  else
    SET_RESULT(filepath, 0, -1);
  return NS_OK;
}

NS_IMETHODIMP
nsBaseURLParser::ParseFilePath(const char* filepath, int32_t filepathLen,
                               uint32_t* directoryPos, int32_t* directoryLen,
                               uint32_t* basenamePos, int32_t* basenameLen,
                               uint32_t* extensionPos, int32_t* extensionLen) {
  if (NS_WARN_IF(!filepath)) {
    return NS_ERROR_INVALID_POINTER;
  }

  if (filepathLen < 0) filepathLen = strlen(filepath);

  if (filepathLen == 0) {
    SET_RESULT(directory, 0, -1);
    SET_RESULT(basename, 0, 0);  // assume a zero length file basename
    SET_RESULT(extension, 0, -1);
    return NS_OK;
  }

  const char* p;
  const char* end = filepath + filepathLen;

  // search backwards for filename
  for (p = end - 1; *p != '/' && p > filepath; --p)
    ;
  if (*p == '/') {
    // catch /.. and /.
    if ((p + 1 < end && *(p + 1) == '.') &&
        (p + 2 == end || (*(p + 2) == '.' && p + 3 == end)))
      p = end - 1;
    // filepath = <directory><filename>.<extension>
    SET_RESULT(directory, 0, p - filepath + 1);
    ParseFileName(p + 1, end - (p + 1), basenamePos, basenameLen, extensionPos,
                  extensionLen);
    OFFSET_RESULT(basename, p + 1 - filepath);
    OFFSET_RESULT(extension, p + 1 - filepath);
  } else {
    // filepath = <filename>.<extension>
    SET_RESULT(directory, 0, -1);
    ParseFileName(filepath, filepathLen, basenamePos, basenameLen, extensionPos,
                  extensionLen);
  }
  return NS_OK;
}

nsresult nsBaseURLParser::ParseFileName(
    const char* filename, int32_t filenameLen, uint32_t* basenamePos,
    int32_t* basenameLen, uint32_t* extensionPos, int32_t* extensionLen) {
  if (NS_WARN_IF(!filename)) {
    return NS_ERROR_INVALID_POINTER;
  }

  if (filenameLen < 0) filenameLen = strlen(filename);

  // no extension if filename ends with a '.'
  if (filename[filenameLen - 1] != '.') {
    // ignore '.' at the beginning
    for (const char* p = filename + filenameLen - 1; p > filename; --p) {
      if (*p == '.') {
        // filename = <basename.extension>
        SET_RESULT(basename, 0, p - filename);
        SET_RESULT(extension, p + 1 - filename,
                   filenameLen - (p - filename + 1));
        return NS_OK;
      }
    }
  }
  // filename = <basename>
  SET_RESULT(basename, 0, filenameLen);
  SET_RESULT(extension, 0, -1);
  return NS_OK;
}

//----------------------------------------------------------------------------
// nsNoAuthURLParser implementation
//----------------------------------------------------------------------------

NS_IMETHODIMP
nsNoAuthURLParser::ParseAuthority(const char* auth, int32_t authLen,
                                  uint32_t* usernamePos, int32_t* usernameLen,
                                  uint32_t* passwordPos, int32_t* passwordLen,
                                  uint32_t* hostnamePos, int32_t* hostnameLen,
                                  int32_t* port) {
  MOZ_ASSERT_UNREACHABLE("Shouldn't parse auth in a NoAuthURL!");
  return NS_ERROR_UNEXPECTED;
}

void nsNoAuthURLParser::ParseAfterScheme(const char* spec, int32_t specLen,
                                         uint32_t* authPos, int32_t* authLen,
                                         uint32_t* pathPos, int32_t* pathLen) {
  MOZ_ASSERT(specLen >= 0, "unexpected");

  // everything is the path
  uint32_t pos = 0;
  switch (CountConsecutiveSlashes(spec, specLen)) {
    case 0:
    case 1:
      break;
    case 2: {
      const char* p = nullptr;
      if (specLen > 2) {
        // looks like there is an authority section

        // if the authority looks like a drive number then we
        // really want to treat it as part of the path
        // [a-zA-Z][:|]{/\}
        // i.e one of:   c:   c:\foo  c:/foo  c|  c|\foo  c|/foo
        if ((specLen > 3) && (spec[3] == ':' || spec[3] == '|') &&
            IsAsciiAlpha(spec[2]) &&
            ((specLen == 4) || (spec[4] == '/') || (spec[4] == '\\'))) {
          pos = 1;
          break;
        }
        // Ignore apparent authority; path is everything after it
        for (p = spec + 2; p < spec + specLen; ++p) {
          if (*p == '/' || *p == '?' || *p == '#') break;
        }
      }
      SET_RESULT(auth, 0, -1);
      if (p && p != spec + specLen)
        SET_RESULT(path, p - spec, specLen - (p - spec));
      else
        SET_RESULT(path, 0, -1);
      return;
    }
    default:
      pos = 2;
      break;
  }
  SET_RESULT(auth, pos, 0);
  SET_RESULT(path, pos, specLen - pos);
}

#if defined(XP_WIN)
NS_IMETHODIMP
nsNoAuthURLParser::ParseFilePath(const char* filepath, int32_t filepathLen,
                                 uint32_t* directoryPos, int32_t* directoryLen,
                                 uint32_t* basenamePos, int32_t* basenameLen,
                                 uint32_t* extensionPos,
                                 int32_t* extensionLen) {
  if (NS_WARN_IF(!filepath)) {
    return NS_ERROR_INVALID_POINTER;
  }

  if (filepathLen < 0) filepathLen = strlen(filepath);

  // look for a filepath consisting of only a drive number, which may or
  // may not have a leading slash.
  if (filepathLen > 1 && filepathLen < 4) {
    const char* end = filepath + filepathLen;
    const char* p = filepath;
    if (*p == '/') p++;
    if ((end - p == 2) && (p[1] == ':' || p[1] == '|') && IsAsciiAlpha(*p)) {
      // filepath = <drive-number>:
      SET_RESULT(directory, 0, filepathLen);
      SET_RESULT(basename, 0, -1);
      SET_RESULT(extension, 0, -1);
      return NS_OK;
    }
  }

  // otherwise fallback on common implementation
  return nsBaseURLParser::ParseFilePath(filepath, filepathLen, directoryPos,
                                        directoryLen, basenamePos, basenameLen,
                                        extensionPos, extensionLen);
}
#endif

//----------------------------------------------------------------------------
// nsAuthURLParser implementation
//----------------------------------------------------------------------------

NS_IMETHODIMP
nsAuthURLParser::ParseAuthority(const char* auth, int32_t authLen,
                                uint32_t* usernamePos, int32_t* usernameLen,
                                uint32_t* passwordPos, int32_t* passwordLen,
                                uint32_t* hostnamePos, int32_t* hostnameLen,
                                int32_t* port) {
  nsresult rv;

  if (NS_WARN_IF(!auth)) {
    return NS_ERROR_INVALID_POINTER;
  }

  if (authLen < 0) authLen = strlen(auth);

  if (authLen == 0) {
    SET_RESULT(username, 0, -1);
    SET_RESULT(password, 0, -1);
    SET_RESULT(hostname, 0, 0);
    if (port) *port = -1;
    return NS_OK;
  }

  // search backwards for @
  const char* p = auth + authLen - 1;
  for (; (*p != '@') && (p > auth); --p) {
  }
  if (*p == '@') {
    // auth = <user-info@server-info>
    rv = ParseUserInfo(auth, p - auth, usernamePos, usernameLen, passwordPos,
                       passwordLen);
    if (NS_FAILED(rv)) return rv;
    rv = ParseServerInfo(p + 1, authLen - (p - auth + 1), hostnamePos,
                         hostnameLen, port);
    if (NS_FAILED(rv)) return rv;
    OFFSET_RESULT(hostname, p + 1 - auth);

    // malformed if has a username or password
    // but no host info, such as: http://u:p@/
    if ((usernamePos || passwordPos) && (!hostnamePos || !*hostnameLen)) {
      return NS_ERROR_MALFORMED_URI;
    }
  } else {
    // auth = <server-info>
    SET_RESULT(username, 0, -1);
    SET_RESULT(password, 0, -1);
    rv = ParseServerInfo(auth, authLen, hostnamePos, hostnameLen, port);
    if (NS_FAILED(rv)) return rv;
  }
  return NS_OK;
}

NS_IMETHODIMP
nsAuthURLParser::ParseUserInfo(const char* userinfo, int32_t userinfoLen,
                               uint32_t* usernamePos, int32_t* usernameLen,
                               uint32_t* passwordPos, int32_t* passwordLen) {
  if (NS_WARN_IF(!userinfo)) {
    return NS_ERROR_INVALID_POINTER;
  }

  if (userinfoLen < 0) userinfoLen = strlen(userinfo);

  if (userinfoLen == 0) {
    SET_RESULT(username, 0, -1);
    SET_RESULT(password, 0, -1);
    return NS_OK;
  }

  const char* p = (const char*)memchr(userinfo, ':', userinfoLen);
  if (p) {
    // userinfo = <username:password>
    SET_RESULT(username, 0, p - userinfo);
    SET_RESULT(password, p - userinfo + 1, userinfoLen - (p - userinfo + 1));
  } else {
    // userinfo = <username>
    SET_RESULT(username, 0, userinfoLen);
    SET_RESULT(password, 0, -1);
  }
  return NS_OK;
}

NS_IMETHODIMP
nsAuthURLParser::ParseServerInfo(const char* serverinfo, int32_t serverinfoLen,
                                 uint32_t* hostnamePos, int32_t* hostnameLen,
                                 int32_t* port) {
  if (NS_WARN_IF(!serverinfo)) {
    return NS_ERROR_INVALID_POINTER;
  }

  if (serverinfoLen < 0) serverinfoLen = strlen(serverinfo);

  if (serverinfoLen == 0) {
    SET_RESULT(hostname, 0, 0);
    if (port) *port = -1;
    return NS_OK;
  }

  // search backwards for a ':' but stop on ']' (IPv6 address literal
  // delimiter).  check for illegal characters in the hostname.
  const char* p = serverinfo + serverinfoLen - 1;
  const char *colon = nullptr, *bracket = nullptr;
  for (; p > serverinfo; --p) {
    switch (*p) {
      case ']':
        bracket = p;
        break;
      case ':':
        if (bracket == nullptr) colon = p;
        break;
      case ' ':
        // hostname must not contain a space
        return NS_ERROR_MALFORMED_URI;
    }
  }

  if (colon) {
    // serverinfo = <hostname:port>
    SET_RESULT(hostname, 0, colon - serverinfo);
    if (port) {
      // XXX unfortunately ToInteger is not defined for substrings
      nsAutoCString buf(colon + 1, serverinfoLen - (colon + 1 - serverinfo));
      if (buf.Length() == 0) {
        *port = -1;
      } else {
        const char* nondigit = NS_strspnp("0123456789", buf.get());
        if (nondigit && *nondigit) return NS_ERROR_MALFORMED_URI;

        nsresult err;
        *port = buf.ToInteger(&err);
        if (NS_FAILED(err) || *port < 0 ||
            *port > std::numeric_limits<uint16_t>::max())
          return NS_ERROR_MALFORMED_URI;
      }
    }
  } else {
    // serverinfo = <hostname>
    SET_RESULT(hostname, 0, serverinfoLen);
    if (port) *port = -1;
  }

  // In case of IPv6 address check its validity
  if (*hostnameLen > 1 && *(serverinfo + *hostnamePos) == '[' &&
      *(serverinfo + *hostnamePos + *hostnameLen - 1) == ']' &&
      !net_IsValidIPv6Addr(
          Substring(serverinfo + *hostnamePos + 1, *hostnameLen - 2)))
    return NS_ERROR_MALFORMED_URI;

  return NS_OK;
}

void nsAuthURLParser::ParseAfterScheme(const char* spec, int32_t specLen,
                                       uint32_t* authPos, int32_t* authLen,
                                       uint32_t* pathPos, int32_t* pathLen) {
  MOZ_ASSERT(specLen >= 0, "unexpected");

  uint32_t nslash = CountConsecutiveSlashes(spec, specLen);

  // search for the end of the authority section
  const char* end = spec + specLen;
  const char* p;
  for (p = spec + nslash; p < end; ++p) {
    if (*p == '/' || *p == '?' || *p == '#') break;
  }
  if (p < end) {
    // spec = [/]<auth><path>
    SET_RESULT(auth, nslash, p - (spec + nslash));
    SET_RESULT(path, p - spec, specLen - (p - spec));
  } else {
    // spec = [/]<auth>
    SET_RESULT(auth, nslash, specLen - nslash);
    SET_RESULT(path, 0, -1);
  }
}

//----------------------------------------------------------------------------
// nsStdURLParser implementation
//----------------------------------------------------------------------------

void nsStdURLParser::ParseAfterScheme(const char* spec, int32_t specLen,
                                      uint32_t* authPos, int32_t* authLen,
                                      uint32_t* pathPos, int32_t* pathLen) {
  MOZ_ASSERT(specLen >= 0, "unexpected");

  uint32_t nslash = CountConsecutiveSlashes(spec, specLen);

  // search for the end of the authority section
  const char* end = spec + specLen;
  const char* p;
  for (p = spec + nslash; p < end; ++p) {
    if (strchr("/?#;", *p)) break;
  }
  switch (nslash) {
    case 0:
    case 2:
      if (p < end) {
        // spec = (//)<auth><path>
        SET_RESULT(auth, nslash, p - (spec + nslash));
        SET_RESULT(path, p - spec, specLen - (p - spec));
      } else {
        // spec = (//)<auth>
        SET_RESULT(auth, nslash, specLen - nslash);
        SET_RESULT(path, 0, -1);
      }
      break;
    case 1:
      // spec = /<path>
      SET_RESULT(auth, 0, -1);
      SET_RESULT(path, 0, specLen);
      break;
    default:
      // spec = ///[/]<path>
      SET_RESULT(auth, 2, 0);
      SET_RESULT(path, 2, specLen - 2);
  }
}