src/common/uri.cpp

   1 /////////////////////////////////////////////////////////////////////////////
   2 // Name:        uri.cpp
   3 // Purpose:     Implementation of a uri parser
   4 // Author:      Ryan Norton
   5 // Created:     10/26/04
   6 // RCS-ID:      $Id$
   7 // Copyright:   (c) 2004 Ryan Norton
   8 // Licence:     wxWindows
   9 /////////////////////////////////////////////////////////////////////////////
  10
  11 // ===========================================================================
  12 // declarations
  13 // ===========================================================================
  14
  15 // ---------------------------------------------------------------------------
  16 // headers
  17 // ---------------------------------------------------------------------------
  18
  19 // For compilers that support precompilation, includes "wx.h".
  20 #include "wx/wxprec.h"
  21
  22 #ifdef __BORLANDC__
  23     #pragma hdrstop
  24 #endif
  25
  26 #include "wx/uri.h"
  27
  28 // ---------------------------------------------------------------------------
  29 // definitions
  30 // ---------------------------------------------------------------------------
  31
  32 IMPLEMENT_CLASS(wxURI, wxObject)
  33
  34 // ===========================================================================
  35 // implementation
  36 // ===========================================================================
  37
  38 // ---------------------------------------------------------------------------
  39 // utilities
  40 // ---------------------------------------------------------------------------
  41
  42 // ---------------------------------------------------------------------------
  43 //
  44 //                        wxURI
  45 //
  46 // ---------------------------------------------------------------------------
  47
  48 // ---------------------------------------------------------------------------
  49 //  Constructors
  50 // ---------------------------------------------------------------------------
  51
  52 wxURI::wxURI() : m_hostType(wxURI_REGNAME), m_fields(0)
  53 {
  54 }
  55
  56 wxURI::wxURI(const wxString& uri) : m_hostType(wxURI_REGNAME), m_fields(0)
  57 {
  58     Create(uri);
  59 }
  60
  61 wxURI::wxURI(const wxURI& uri)  : wxObject(), m_hostType(wxURI_REGNAME), m_fields(0)
  62 {
  63     Assign(uri);
  64 }
  65
  66 // ---------------------------------------------------------------------------
  67 // Destructor and cleanup
  68 // ---------------------------------------------------------------------------
  69
  70 wxURI::~wxURI()
  71 {
  72     Clear();
  73 }
  74
  75 void wxURI::Clear()
  76 {
  77     m_scheme = m_userinfo = m_server = m_port = m_path =
  78     m_query = m_fragment = wxEmptyString;
  79
  80     m_hostType = wxURI_REGNAME;
  81
  82     m_fields = 0;
  83 }
  84
  85 // ---------------------------------------------------------------------------
  86 // Create
  87 //
  88 // This creates the URI - all we do here is call the main parsing method
  89 // ---------------------------------------------------------------------------
  90
  91 const wxChar* wxURI::Create(const wxString& uri)
  92 {
  93     if (m_fields)
  94         Clear();
  95
  96     return Parse(uri);
  97 }
  98
  99 // ---------------------------------------------------------------------------
 100 // Escape Methods
 101 //
 102 // TranslateEscape unencodes a 3 character URL escape sequence
 103 //
 104 // Escape encodes an invalid URI character into a 3 character sequence
 105 //
 106 // IsEscape determines if the input string contains an escape sequence,
 107 // if it does, then it moves the input string past the escape sequence
 108 //
 109 // Unescape unencodes all 3 character URL escape sequences in a wxString
 110 // ---------------------------------------------------------------------------
 111
 112 wxChar wxURI::TranslateEscape(const wxChar* s)
 113 {
 114     wxASSERT_MSG( IsHex(s[0]) && IsHex(s[1]), wxT("Invalid escape sequence!"));
 115
 116     return (wxChar)( CharToHex(s[0]) << 4 ) | CharToHex(s[1]);
 117 }
 118
 119 wxString wxURI::Unescape(const wxString& uri)
 120 {
 121     wxString new_uri;
 122
 123     for(size_t i = 0; i < uri.length(); ++i)
 124     {
 125         if (uri[i] == wxT('%'))
 126         {
 127             new_uri += wxURI::TranslateEscape( &(uri.c_str()[i+1]) );
 128             i += 2;
 129         }
 130         else
 131             new_uri += uri[i];
 132     }
 133
 134     return new_uri;
 135 }
 136
 137 void wxURI::Escape(wxString& s, const wxChar& c)
 138 {
 139     const wxChar* hdig = wxT("0123456789abcdef");
 140     s += wxT('%');
 141     s += hdig[(c >> 4) & 15];
 142     s += hdig[c & 15];
 143 }
 144
 145 bool wxURI::IsEscape(const wxChar*& uri)
 146 {
 147     // pct-encoded   = "%" HEXDIG HEXDIG
 148     if(*uri == wxT('%') && IsHex(*(uri+1)) && IsHex(*(uri+2)))
 149         return true;
 150     else
 151         return false;
 152 }
 153
 154 // ---------------------------------------------------------------------------
 155 // GetUser
 156 // GetPassword
 157 //
 158 // Gets the username and password via the old URL method.
 159 // ---------------------------------------------------------------------------
 160 wxString wxURI::GetUser() const
 161 {
 162       size_t dwPasswordPos = m_userinfo.find(':');
 163
 164       if (dwPasswordPos == wxString::npos)
 165           dwPasswordPos = 0;
 166
 167       return m_userinfo(0, dwPasswordPos);
 168 }
 169
 170 wxString wxURI::GetPassword() const
 171 {
 172       size_t dwPasswordPos = m_userinfo.find(':');
 173
 174       if (dwPasswordPos == wxString::npos)
 175           return wxT("");
 176       else
 177           return m_userinfo(dwPasswordPos+1, m_userinfo.length() + 1);
 178 }
 179
 180 // ---------------------------------------------------------------------------
 181 // BuildURI
 182 //
 183 // BuildURI() builds the entire URI into a useable
 184 // representation, including proper identification characters such as slashes
 185 //
 186 // BuildUnescapedURI() does the same thing as BuildURI(), only it unescapes
 187 // the components that accept escape sequences
 188 // ---------------------------------------------------------------------------
 189
 190 wxString wxURI::BuildURI() const
 191 {
 192     wxString ret;
 193
 194     if (HasScheme())
 195         ret = ret + m_scheme + wxT(":");
 196
 197     if (HasServer())
 198     {
 199         ret += wxT("//");
 200
 201         if (HasUserInfo())
 202             ret = ret + m_userinfo + wxT("@");
 203
 204         ret += m_server;
 205
 206         if (HasPort())
 207             ret = ret + wxT(":") + m_port;
 208     }
 209
 210     ret += m_path;
 211
 212     if (HasQuery())
 213         ret = ret + wxT("?") + m_query;
 214
 215     if (HasFragment())
 216         ret = ret + wxT("#") + m_fragment;
 217
 218     return ret;
 219 }
 220
 221 wxString wxURI::BuildUnescapedURI() const
 222 {
 223     wxString ret;
 224
 225     if (HasScheme())
 226         ret = ret + m_scheme + wxT(":");
 227
 228     if (HasServer())
 229     {
 230         ret += wxT("//");
 231
 232         if (HasUserInfo())
 233             ret = ret + wxURI::Unescape(m_userinfo) + wxT("@");
 234
 235         if (m_hostType == wxURI_REGNAME)
 236             ret += wxURI::Unescape(m_server);
 237         else
 238             ret += m_server;
 239
 240         if (HasPort())
 241             ret = ret + wxT(":") + m_port;
 242     }
 243
 244     ret += wxURI::Unescape(m_path);
 245
 246     if (HasQuery())
 247         ret = ret + wxT("?") + wxURI::Unescape(m_query);
 248
 249     if (HasFragment())
 250         ret = ret + wxT("#") + wxURI::Unescape(m_fragment);
 251
 252     return ret;
 253 }
 254
 255 // ---------------------------------------------------------------------------
 256 // Assignment
 257 // ---------------------------------------------------------------------------
 258
 259 wxURI& wxURI::Assign(const wxURI& uri)
 260 {
 261     //assign fields
 262     m_fields = uri.m_fields;
 263
 264     //ref over components
 265     m_scheme = uri.m_scheme;
 266     m_userinfo = uri.m_userinfo;
 267     m_server = uri.m_server;
 268     m_hostType = uri.m_hostType;
 269     m_port = uri.m_port;
 270     m_path = uri.m_path;
 271     m_query = uri.m_query;
 272     m_fragment = uri.m_fragment;
 273
 274     return *this;
 275 }
 276
 277 wxURI& wxURI::operator = (const wxURI& uri)
 278 {
 279     return Assign(uri);
 280 }
 281
 282 wxURI& wxURI::operator = (const wxString& string)
 283 {
 284     Create(string);
 285     return *this;
 286 }
 287
 288 // ---------------------------------------------------------------------------
 289 // Comparison
 290 // ---------------------------------------------------------------------------
 291
 292 bool wxURI::operator == (const wxURI& uri) const
 293 {
 294     if (HasScheme())
 295     {
 296         if(m_scheme != uri.m_scheme)
 297             return false;
 298     }
 299     else if (uri.HasScheme())
 300         return false;
 301
 302
 303     if (HasServer())
 304     {
 305         if (HasUserInfo())
 306         {
 307             if (m_userinfo != uri.m_userinfo)
 308                 return false;
 309         }
 310         else if (uri.HasUserInfo())
 311             return false;
 312
 313         if (m_server != uri.m_server ||
 314             m_hostType != uri.m_hostType)
 315             return false;
 316
 317         if (HasPort())
 318         {
 319             if(m_port != uri.m_port)
 320                 return false;
 321         }
 322         else if (uri.HasPort())
 323             return false;
 324     }
 325     else if (uri.HasServer())
 326         return false;
 327
 328
 329     if (HasPath())
 330     {
 331         if(m_path != uri.m_path)
 332             return false;
 333     }
 334     else if (uri.HasPath())
 335         return false;
 336
 337     if (HasQuery())
 338     {
 339         if (m_query != uri.m_query)
 340             return false;
 341     }
 342     else if (uri.HasQuery())
 343         return false;
 344
 345     if (HasFragment())
 346     {
 347         if (m_fragment != uri.m_fragment)
 348             return false;
 349     }
 350     else if (uri.HasFragment())
 351         return false;
 352
 353     return true;
 354 }
 355
 356 // ---------------------------------------------------------------------------
 357 // IsReference
 358 //
 359 // if there is no authority or scheme, it is a reference
 360 // ---------------------------------------------------------------------------
 361
 362 bool wxURI::IsReference() const
 363 {   return !HasScheme() || !HasServer();  }
 364
 365 // ---------------------------------------------------------------------------
 366 // Parse
 367 //
 368 // Master URI parsing method.  Just calls the individual parsing methods
 369 //
 370 // URI = scheme ":" hier-part [ "?" query ] [ "#" fragment ]
 371 // URI-reference = URI / relative
 372 // ---------------------------------------------------------------------------
 373
 374 const wxChar* wxURI::Parse(const wxChar* uri)
 375 {
 376     uri = ParseScheme(uri);
 377     uri = ParseAuthority(uri);
 378     uri = ParsePath(uri);
 379     uri = ParseQuery(uri);
 380     return ParseFragment(uri);
 381 }
 382
 383 // ---------------------------------------------------------------------------
 384 // ParseXXX
 385 //
 386 // Individual parsers for each URI component
 387 // ---------------------------------------------------------------------------
 388
 389 const wxChar* wxURI::ParseScheme(const wxChar* uri)
 390 {
 391     wxASSERT(uri != NULL);
 392
 393     //copy of the uri - used for figuring out
 394     //length of each component
 395     const wxChar* uricopy = uri;
 396
 397     //Does the uri have a scheme (first character alpha)?
 398     if (IsAlpha(*uri))
 399     {
 400         m_scheme += *uri++;
 401
 402         //scheme        = ALPHA *( ALPHA / DIGIT / "+" / "-" / "." )
 403         while (IsAlpha(*uri) || IsDigit(*uri) ||
 404                *uri == wxT('+')   ||
 405                *uri == wxT('-')   ||
 406                *uri == wxT('.'))
 407         {
 408             m_scheme += *uri++;
 409         }
 410
 411         //valid scheme?
 412         if (*uri == wxT(':'))
 413         {
 414             //mark the scheme as valid
 415             m_fields |= wxURI_SCHEME;
 416
 417             //move reference point up to input buffer
 418             uricopy = ++uri;
 419         }
 420         else
 421             //relative uri with relative path reference
 422             m_scheme = wxEmptyString;
 423     }
 424 //    else
 425         //relative uri with _possible_ relative path reference
 426
 427     return uricopy;
 428 }
 429
 430 const wxChar* wxURI::ParseAuthority(const wxChar* uri)
 431 {
 432     // authority     = [ userinfo "@" ] host [ ":" port ]
 433     if (*uri == wxT('/') && *(uri+1) == wxT('/'))
 434     {
 435         uri += 2;
 436
 437         uri = ParseUserInfo(uri);
 438         uri = ParseServer(uri);
 439         return ParsePort(uri);
 440     }
 441
 442     return uri;
 443 }
 444
 445 const wxChar* wxURI::ParseUserInfo(const wxChar* uri)
 446 {
 447     wxASSERT(uri != NULL);
 448
 449     //copy of the uri - used for figuring out
 450     //length of each component
 451     const wxChar* uricopy = uri;
 452
 453     // userinfo      = *( unreserved / pct-encoded / sub-delims / ":" )
 454     while(*uri && *uri != wxT('@') && *uri != wxT('/') && *uri != wxT('#') && *uri != wxT('?'))
 455     {
 456         if(IsUnreserved(*uri) ||
 457            IsSubDelim(*uri) || *uri == wxT(':'))
 458             m_userinfo += *uri++;
 459         else if (IsEscape(uri))
 460         {
 461             m_userinfo += *uri++;
 462             m_userinfo += *uri++;
 463             m_userinfo += *uri++;
 464         }
 465         else
 466             Escape(m_userinfo, *uri++);
 467     }
 468
 469     if(*uri == wxT('@'))
 470     {
 471         //valid userinfo
 472         m_fields |= wxURI_USERINFO;
 473
 474         uricopy = ++uri;
 475     }
 476     else
 477         m_userinfo = wxEmptyString;
 478
 479     return uricopy;
 480 }
 481
 482 const wxChar* wxURI::ParseServer(const wxChar* uri)
 483 {
 484     wxASSERT(uri != NULL);
 485
 486     //copy of the uri - used for figuring out
 487     //length of each component
 488     const wxChar* uricopy = uri;
 489
 490     // host          = IP-literal / IPv4address / reg-name
 491     // IP-literal    = "[" ( IPv6address / IPvFuture  ) "]"
 492     if (*uri == wxT('['))
 493     {
 494         ++uri; //some compilers don't support *&ing a ++*
 495         if (ParseIPv6address(uri) && *uri == wxT(']'))
 496         {
 497             ++uri;
 498             m_hostType = wxURI_IPV6ADDRESS;
 499
 500             wxStringBufferLength theBuffer(m_server, uri - uricopy);
 501             wxTmemcpy(theBuffer, uricopy, uri-uricopy);
 502             theBuffer.SetLength(uri-uricopy);
 503         }
 504         else
 505         {
 506             uri = uricopy;
 507
 508             ++uri; //some compilers don't support *&ing a ++*
 509             if (ParseIPvFuture(uri) && *uri == wxT(']'))
 510             {
 511                 ++uri;
 512                 m_hostType = wxURI_IPVFUTURE;
 513
 514                 wxStringBufferLength theBuffer(m_server, uri - uricopy);
 515                 wxTmemcpy(theBuffer, uricopy, uri-uricopy);
 516                 theBuffer.SetLength(uri-uricopy);
 517             }
 518             else
 519                 uri = uricopy;
 520         }
 521     }
 522     else
 523     {
 524         if (ParseIPv4address(uri))
 525         {
 526             m_hostType = wxURI_IPV4ADDRESS;
 527
 528             wxStringBufferLength theBuffer(m_server, uri - uricopy);
 529             wxTmemcpy(theBuffer, uricopy, uri-uricopy);
 530             theBuffer.SetLength(uri-uricopy);
 531         }
 532         else
 533             uri = uricopy;
 534     }
 535
 536     if(m_hostType == wxURI_REGNAME)
 537     {
 538         uri = uricopy;
 539         // reg-name      = *( unreserved / pct-encoded / sub-delims )
 540         while(*uri && *uri != wxT('/') && *uri != wxT(':') && *uri != wxT('#') && *uri != wxT('?'))
 541         {
 542             if(IsUnreserved(*uri) ||  IsSubDelim(*uri))
 543                 m_server += *uri++;
 544             else if (IsEscape(uri))
 545             {
 546                 m_server += *uri++;
 547                 m_server += *uri++;
 548                 m_server += *uri++;
 549             }
 550             else
 551                 Escape(m_server, *uri++);
 552         }
 553     }
 554
 555     //mark the server as valid
 556     m_fields |= wxURI_SERVER;
 557
 558     return uri;
 559 }
 560
 561
 562 const wxChar* wxURI::ParsePort(const wxChar* uri)
 563 {
 564     wxASSERT(uri != NULL);
 565
 566     // port          = *DIGIT
 567     if(*uri == wxT(':'))
 568     {
 569         ++uri;
 570         while(IsDigit(*uri))
 571         {
 572             m_port += *uri++;
 573         }
 574
 575         //mark the port as valid
 576         m_fields |= wxURI_PORT;
 577     }
 578
 579     return uri;
 580 }
 581
 582 const wxChar* wxURI::ParsePath(const wxChar* uri, bool bReference, bool bNormalize)
 583 {
 584     wxASSERT(uri != NULL);
 585
 586     //copy of the uri - used for figuring out
 587     //length of each component
 588     const wxChar* uricopy = uri;
 589
 590     /// hier-part     = "//" authority path-abempty
 591     ///               / path-absolute
 592     ///               / path-rootless
 593     ///               / path-empty
 594     ///
 595     /// relative-part = "//" authority path-abempty
 596     ///               / path-absolute
 597     ///               / path-noscheme
 598     ///               / path-empty
 599     ///
 600     /// path-abempty  = *( "/" segment )
 601     /// path-absolute = "/" [ segment-nz *( "/" segment ) ]
 602     /// path-noscheme = segment-nz-nc *( "/" segment )
 603     /// path-rootless = segment-nz *( "/" segment )
 604     /// path-empty    = 0<pchar>
 605     ///
 606     /// segment       = *pchar
 607     /// segment-nz    = 1*pchar
 608     /// segment-nz-nc = 1*( unreserved / pct-encoded / sub-delims / "@" )
 609     ///               ; non-zero-length segment without any colon ":"
 610     ///
 611     /// pchar         = unreserved / pct-encoded / sub-delims / ":" / "@"
 612     if (*uri == wxT('/'))
 613     {
 614         m_path += *uri++;
 615
 616         while(*uri && *uri != wxT('#') && *uri != wxT('?'))
 617         {
 618             if( IsUnreserved(*uri) || IsSubDelim(*uri) ||
 619                 *uri == wxT(':') || *uri == wxT('@') || *uri == wxT('/'))
 620                 m_path += *uri++;
 621             else if (IsEscape(uri))
 622             {
 623                 m_path += *uri++;
 624                 m_path += *uri++;
 625                 m_path += *uri++;
 626             }
 627             else
 628                 Escape(m_path, *uri++);
 629         }
 630
 631         if (bNormalize)
 632         {
 633             wxStringBufferLength theBuffer(m_path, m_path.length() + 1);
 634 #if wxUSE_STL
 635             wxTmemcpy(theBuffer, m_path.c_str(), m_path.length()+1);
 636 #endif
 637             Normalize(theBuffer, true);
 638             theBuffer.SetLength(wxStrlen(theBuffer));
 639         }
 640         //mark the path as valid
 641         m_fields |= wxURI_PATH;
 642     }
 643     else if(*uri) //Relative path
 644     {
 645         if (bReference)
 646         {
 647             //no colon allowed
 648             while(*uri && *uri != wxT('#') && *uri != wxT('?'))
 649             {
 650                 if(IsUnreserved(*uri) || IsSubDelim(*uri) ||
 651                   *uri == wxT('@') || *uri == wxT('/'))
 652                     m_path += *uri++;
 653                 else if (IsEscape(uri))
 654                 {
 655                     m_path += *uri++;
 656                     m_path += *uri++;
 657                     m_path += *uri++;
 658                 }
 659                 else
 660                     Escape(m_path, *uri++);
 661             }
 662         }
 663         else
 664         {
 665             while(*uri && *uri != wxT('#') && *uri != wxT('?'))
 666             {
 667                 if(IsUnreserved(*uri) || IsSubDelim(*uri) ||
 668                    *uri == wxT(':') || *uri == wxT('@') || *uri == wxT('/'))
 669                     m_path += *uri++;
 670                 else if (IsEscape(uri))
 671                 {
 672                     m_path += *uri++;
 673                     m_path += *uri++;
 674                     m_path += *uri++;
 675                 }
 676                 else
 677                     Escape(m_path, *uri++);
 678             }
 679         }
 680
 681         if (uri != uricopy)
 682         {
 683             if (bNormalize)
 684             {
 685                 wxStringBufferLength theBuffer(m_path, m_path.length() + 1);
 686 #if wxUSE_STL
 687                 wxTmemcpy(theBuffer, m_path.c_str(), m_path.length()+1);
 688 #endif
 689                 Normalize(theBuffer);
 690                 theBuffer.SetLength(wxStrlen(theBuffer));
 691             }
 692
 693             //mark the path as valid
 694             m_fields |= wxURI_PATH;
 695         }
 696     }
 697
 698     return uri;
 699 }
 700
 701
 702 const wxChar* wxURI::ParseQuery(const wxChar* uri)
 703 {
 704     wxASSERT(uri != NULL);
 705
 706     // query         = *( pchar / "/" / "?" )
 707     if (*uri == wxT('?'))
 708     {
 709         ++uri;
 710         while(*uri && *uri != wxT('#'))
 711         {
 712             if (IsUnreserved(*uri) || IsSubDelim(*uri) ||
 713                 *uri == wxT(':') || *uri == wxT('@') || *uri == wxT('/') || *uri == wxT('?'))
 714                   m_query += *uri++;
 715             else if (IsEscape(uri))
 716             {
 717                   m_query += *uri++;
 718                   m_query += *uri++;
 719                   m_query += *uri++;
 720             }
 721             else
 722                   Escape(m_query, *uri++);
 723         }
 724
 725         //mark the server as valid
 726         m_fields |= wxURI_QUERY;
 727     }
 728
 729     return uri;
 730 }
 731
 732
 733 const wxChar* wxURI::ParseFragment(const wxChar* uri)
 734 {
 735     wxASSERT(uri != NULL);
 736
 737     // fragment      = *( pchar / "/" / "?" )
 738     if (*uri == wxT('#'))
 739     {
 740         ++uri;
 741         while(*uri)
 742         {
 743             if (IsUnreserved(*uri) || IsSubDelim(*uri) ||
 744                 *uri == wxT(':') || *uri == wxT('@') || *uri == wxT('/') || *uri == wxT('?'))
 745                   m_fragment += *uri++;
 746             else if (IsEscape(uri))
 747             {
 748                   m_fragment += *uri++;
 749                   m_fragment += *uri++;
 750                   m_fragment += *uri++;
 751             }
 752             else
 753                   Escape(m_fragment, *uri++);
 754         }
 755
 756         //mark the server as valid
 757         m_fields |= wxURI_FRAGMENT;
 758     }
 759
 760     return uri;
 761 }
 762
 763 // ---------------------------------------------------------------------------
 764 // Resolve
 765 //
 766 // Builds missing components of this uri from a base uri
 767 //
 768 // A version of the algorithm outlined in the RFC is used here
 769 // (it is shown in comments)
 770 //
 771 // Note that an empty URI inherits all components
 772 // ---------------------------------------------------------------------------
 773
 774 void wxURI::Resolve(const wxURI& base, int flags)
 775 {
 776     wxASSERT_MSG(!base.IsReference(),
 777                 wxT("wxURI to inherit from must not be a reference!"));
 778
 779     // If we arn't being strict, enable the older (pre-RFC2396)
 780     // loophole that allows this uri to inherit other
 781     // properties from the base uri - even if the scheme
 782     // is defined
 783     if ( !(flags & wxURI_STRICT) &&
 784             HasScheme() && base.HasScheme() &&
 785                 m_scheme == base.m_scheme )
 786     {
 787         m_fields -= wxURI_SCHEME;
 788     }
 789
 790
 791     // Do nothing if this is an absolute wxURI
 792     //    if defined(R.scheme) then
 793     //       T.scheme    = R.scheme;
 794     //       T.authority = R.authority;
 795     //       T.path      = remove_dot_segments(R.path);
 796     //       T.query     = R.query;
 797     if (HasScheme())
 798     {
 799         return;
 800     }
 801
 802     //No scheme - inherit
 803     m_scheme = base.m_scheme;
 804     m_fields |= wxURI_SCHEME;
 805
 806     // All we need to do for relative URIs with an
 807     // authority component is just inherit the scheme
 808     //       if defined(R.authority) then
 809     //          T.authority = R.authority;
 810     //          T.path      = remove_dot_segments(R.path);
 811     //          T.query     = R.query;
 812     if (HasServer())
 813     {
 814         return;
 815     }
 816
 817     //No authority - inherit
 818     if (base.HasUserInfo())
 819     {
 820         m_userinfo = base.m_userinfo;
 821         m_fields |= wxURI_USERINFO;
 822     }
 823
 824     m_server = base.m_server;
 825     m_hostType = base.m_hostType;
 826     m_fields |= wxURI_SERVER;
 827
 828     if (base.HasPort())
 829     {
 830         m_port = base.m_port;
 831         m_fields |= wxURI_PORT;
 832     }
 833
 834
 835     // Simple path inheritance from base
 836     if (!HasPath())
 837     {
 838         //             T.path = Base.path;
 839         m_path = base.m_path;
 840         m_fields |= wxURI_PATH;
 841
 842
 843         //             if defined(R.query) then
 844         //                T.query = R.query;
 845         //             else
 846         //                T.query = Base.query;
 847         //             endif;
 848         if (!HasQuery())
 849         {
 850             m_query = base.m_query;
 851             m_fields |= wxURI_QUERY;
 852         }
 853     }
 854     else
 855     {
 856         //             if (R.path starts-with "/") then
 857         //                T.path = remove_dot_segments(R.path);
 858         //             else
 859         //                T.path = merge(Base.path, R.path);
 860         //                T.path = remove_dot_segments(T.path);
 861         //             endif;
 862         //             T.query = R.query;
 863         if (m_path[0u] != wxT('/'))
 864         {
 865             //Merge paths
 866             const wxChar* op = m_path.c_str();
 867             const wxChar* bp = base.m_path.c_str() + base.m_path.Length();
 868
 869             //not a ending directory?  move up
 870             if (base.m_path[0] && *(bp-1) != wxT('/'))
 871                 UpTree(base.m_path, bp);
 872
 873             //normalize directories
 874             while(*op == wxT('.') && *(op+1) == wxT('.') &&
 875                        (*(op+2) == '\0' || *(op+2) == wxT('/')) )
 876             {
 877                 UpTree(base.m_path, bp);
 878
 879                 if (*(op+2) == '\0')
 880                     op += 2;
 881                 else
 882                     op += 3;
 883             }
 884
 885             m_path = base.m_path.substr(0, bp - base.m_path.c_str()) +
 886                     m_path.substr((op - m_path.c_str()), m_path.Length());
 887         }
 888     }
 889
 890     //T.fragment = R.fragment;
 891 }
 892
 893 // ---------------------------------------------------------------------------
 894 // UpTree
 895 //
 896 // Moves a URI path up a directory
 897 // ---------------------------------------------------------------------------
 898
 899 //static
 900 void wxURI::UpTree(const wxChar* uristart, const wxChar*& uri)
 901 {
 902     if (uri != uristart && *(uri-1) == wxT('/'))
 903     {
 904         uri -= 2;
 905     }
 906
 907     for(;uri != uristart; --uri)
 908     {
 909         if (*uri == wxT('/'))
 910         {
 911             ++uri;
 912             break;
 913         }
 914     }
 915
 916     //!!!TODO:HACK!!!//
 917     if (uri == uristart && *uri == wxT('/'))
 918         ++uri;
 919     //!!!//
 920 }
 921
 922 // ---------------------------------------------------------------------------
 923 // Normalize
 924 //
 925 // Normalizes directories in-place
 926 //
 927 // I.E. ./ and . are ignored
 928 //
 929 // ../ and .. are removed if a directory is before it, along
 930 // with that directory (leading .. and ../ are kept)
 931 // ---------------------------------------------------------------------------
 932
 933 //static
 934 void wxURI::Normalize(wxChar* s, bool bIgnoreLeads)
 935 {
 936     wxChar* cp = s;
 937     wxChar* bp = s;
 938
 939     if(s[0] == wxT('/'))
 940         ++bp;
 941
 942     while(*cp)
 943     {
 944         if (*cp == wxT('.') && (*(cp+1) == wxT('/') || *(cp+1) == '\0')
 945             && (bp == cp || *(cp-1) == wxT('/')))
 946         {
 947             //. _or_ ./  - ignore
 948             if (*(cp+1) == '\0')
 949                 cp += 1;
 950             else
 951                 cp += 2;
 952         }
 953         else if (*cp == wxT('.') && *(cp+1) == wxT('.') &&
 954                 (*(cp+2) == wxT('/') || *(cp+2) == '\0')
 955                 && (bp == cp || *(cp-1) == wxT('/')))
 956         {
 957             //.. _or_ ../ - go up the tree
 958             if (s != bp)
 959             {
 960                 UpTree((const wxChar*)bp, (const wxChar*&)s);
 961
 962                 if (*(cp+2) == '\0')
 963                     cp += 2;
 964                 else
 965                     cp += 3;
 966             }
 967             else if (!bIgnoreLeads)
 968
 969             {
 970                 *bp++ = *cp++;
 971                 *bp++ = *cp++;
 972                 if (*cp)
 973                     *bp++ = *cp++;
 974
 975                 s = bp;
 976             }
 977             else
 978             {
 979                 if (*(cp+2) == '\0')
 980                     cp += 2;
 981                 else
 982                     cp += 3;
 983             }
 984         }
 985         else
 986             *s++ = *cp++;
 987     }
 988
 989     *s = '\0';
 990 }
 991
 992 // ---------------------------------------------------------------------------
 993 // ParseH16
 994 //
 995 // Parses 1 to 4 hex values.  Returns true if the first character of the input
 996 // string is a valid hex character.  It is the caller's responsability to move
 997 // the input string back to its original position on failure.
 998 // ---------------------------------------------------------------------------
 999
1000 bool wxURI::ParseH16(const wxChar*& uri)
1001 {
1002     // h16           = 1*4HEXDIG
1003     if(!IsHex(*++uri))
1004         return false;
1005
1006     if(IsHex(*++uri) && IsHex(*++uri) && IsHex(*++uri))
1007         ++uri;
1008
1009     return true;
1010 }
1011
1012 // ---------------------------------------------------------------------------
1013 // ParseIPXXX
1014 //
1015 // Parses a certain version of an IP address and moves the input string past
1016 // it.  Returns true if the input  string contains the proper version of an ip
1017 // address.  It is the caller's responsability to move the input string back
1018 // to its original position on failure.
1019 // ---------------------------------------------------------------------------
1020
1021 bool wxURI::ParseIPv4address(const wxChar*& uri)
1022 {
1023     //IPv4address   = dec-octet "." dec-octet "." dec-octet "." dec-octet
1024     //
1025     //dec-octet     =      DIGIT                    ; 0-9
1026     //                / %x31-39 DIGIT               ; 10-99
1027     //                / "1" 2DIGIT                  ; 100-199
1028     //                / "2" %x30-34 DIGIT           ; 200-249
1029     //                / "25" %x30-35                ; 250-255
1030     size_t iIPv4 = 0;
1031     if (IsDigit(*uri))
1032     {
1033         ++iIPv4;
1034
1035
1036         //each ip part must be between 0-255 (dupe of version in for loop)
1037         if( IsDigit(*++uri) && IsDigit(*++uri) &&
1038            //100 or less  (note !)
1039            !( (*(uri-2) < wxT('2')) ||
1040            //240 or less
1041              (*(uri-2) == wxT('2') &&
1042                (*(uri-1) < wxT('5') || (*(uri-1) == wxT('5') && *uri <= wxT('5')))
1043              )
1044             )
1045           )
1046         {
1047             return false;
1048         }
1049
1050         if(IsDigit(*uri))++uri;
1051
1052         //compilers should unroll this loop
1053         for(; iIPv4 < 4; ++iIPv4)
1054         {
1055             if (*uri != wxT('.') || !IsDigit(*++uri))
1056                 break;
1057
1058             //each ip part must be between 0-255
1059             if( IsDigit(*++uri) && IsDigit(*++uri) &&
1060                //100 or less  (note !)
1061                !( (*(uri-2) < wxT('2')) ||
1062                //240 or less
1063                  (*(uri-2) == wxT('2') &&
1064                    (*(uri-1) < wxT('5') || (*(uri-1) == wxT('5') && *uri <= wxT('5')))
1065                  )
1066                 )
1067               )
1068             {
1069                 return false;
1070             }
1071             if(IsDigit(*uri))++uri;
1072         }
1073     }
1074     return iIPv4 == 4;
1075 }
1076
1077 bool wxURI::ParseIPv6address(const wxChar*& uri)
1078 {
1079     // IPv6address   =                            6( h16 ":" ) ls32
1080     //               /                       "::" 5( h16 ":" ) ls32
1081     //               / [               h16 ] "::" 4( h16 ":" ) ls32
1082     //               / [ *1( h16 ":" ) h16 ] "::" 3( h16 ":" ) ls32
1083     //               / [ *2( h16 ":" ) h16 ] "::" 2( h16 ":" ) ls32
1084     //               / [ *3( h16 ":" ) h16 ] "::"    h16 ":"   ls32
1085     //               / [ *4( h16 ":" ) h16 ] "::"              ls32
1086     //               / [ *5( h16 ":" ) h16 ] "::"              h16
1087     //               / [ *6( h16 ":" ) h16 ] "::"
1088
1089     size_t numPrefix = 0,
1090               maxPostfix;
1091
1092     bool bEndHex = false;
1093
1094     for( ; numPrefix < 6; ++numPrefix)
1095     {
1096         if(!ParseH16(uri))
1097         {
1098             --uri;
1099             bEndHex = true;
1100             break;
1101         }
1102
1103         if(*uri != wxT(':'))
1104         {
1105             break;
1106         }
1107     }
1108
1109     if(!bEndHex && !ParseH16(uri))
1110     {
1111         --uri;
1112
1113         if (numPrefix)
1114             return false;
1115
1116         if (*uri == wxT(':'))
1117         {
1118             if (*++uri != wxT(':'))
1119                 return false;
1120
1121             maxPostfix = 5;
1122         }
1123         else
1124             maxPostfix = 6;
1125     }
1126     else
1127     {
1128         if (*uri != wxT(':') || *(uri+1) != wxT(':'))
1129         {
1130             if (numPrefix != 6)
1131                 return false;
1132
1133             while (*--uri != wxT(':')) {}
1134             ++uri;
1135
1136             const wxChar* uristart = uri;
1137             //parse ls32
1138             // ls32          = ( h16 ":" h16 ) / IPv4address
1139             if (ParseH16(uri) && *uri == wxT(':') && ParseH16(uri))
1140                 return true;
1141
1142             uri = uristart;
1143
1144             if (ParseIPv4address(uri))
1145                 return true;
1146             else
1147                 return false;
1148         }
1149         else
1150         {
1151             uri += 2;
1152
1153             if (numPrefix > 3)
1154                 maxPostfix = 0;
1155             else
1156                 maxPostfix = 4 - numPrefix;
1157         }
1158     }
1159
1160     bool bAllowAltEnding = maxPostfix == 0;
1161
1162     for(; maxPostfix != 0; --maxPostfix)
1163     {
1164         if(!ParseH16(uri) || *uri != wxT(':'))
1165             return false;
1166     }
1167
1168     if(numPrefix <= 4)
1169     {
1170         const wxChar* uristart = uri;
1171         //parse ls32
1172         // ls32          = ( h16 ":" h16 ) / IPv4address
1173         if (ParseH16(uri) && *uri == wxT(':') && ParseH16(uri))
1174             return true;
1175
1176         uri = uristart;
1177
1178         if (ParseIPv4address(uri))
1179             return true;
1180
1181         uri = uristart;
1182
1183         if (!bAllowAltEnding)
1184             return false;
1185     }
1186
1187     if(numPrefix <= 5 && ParseH16(uri))
1188         return true;
1189
1190     return true;
1191 }
1192
1193 bool wxURI::ParseIPvFuture(const wxChar*& uri)
1194 {
1195     // IPvFuture     = "v" 1*HEXDIG "." 1*( unreserved / sub-delims / ":" )
1196     if (*++uri != wxT('v') || !IsHex(*++uri))
1197         return false;
1198
1199     while (IsHex(*++uri)) {}
1200
1201     if (*uri != wxT('.') || !(IsUnreserved(*++uri) || IsSubDelim(*uri) || *uri == wxT(':')))
1202         return false;
1203
1204     while(IsUnreserved(*++uri) || IsSubDelim(*uri) || *uri == wxT(':')) {}
1205
1206     return true;
1207 }
1208
1209
1210 // ---------------------------------------------------------------------------
1211 // CharToHex
1212 //
1213 // Converts a character into a numeric hexidecimal value, or 0 if the
1214 // passed in character is not a valid hex character
1215 // ---------------------------------------------------------------------------
1216
1217 //static
1218 wxChar wxURI::CharToHex(const wxChar& c)
1219 {
1220     if ((c >= wxT('A')) && (c <= wxT('Z'))) return wxChar(c - wxT('A') + 0x0A);
1221     if ((c >= wxT('a')) && (c <= wxT('z'))) return wxChar(c - wxT('a') + 0x0a);
1222     if ((c >= wxT('0')) && (c <= wxT('9'))) return wxChar(c - wxT('0') + 0x00);
1223
1224     return 0;
1225 }
1226
1227 // ---------------------------------------------------------------------------
1228 // IsXXX
1229 //
1230 // Returns true if the passed in character meets the criteria of the method
1231 // ---------------------------------------------------------------------------
1232
1233 //! unreserved    = ALPHA / DIGIT / "-" / "." / "_" / "~"
1234 bool wxURI::IsUnreserved (const wxChar& c)
1235 {   return IsAlpha(c) || IsDigit(c) ||
1236            c == wxT('-') ||
1237            c == wxT('.') ||
1238            c == wxT('_') ||
1239            c == wxT('~') //tilde
1240            ;
1241 }
1242
1243 bool wxURI::IsReserved (const wxChar& c)
1244 {
1245     return IsGenDelim(c) || IsSubDelim(c);
1246 }
1247
1248 //! gen-delims    = ":" / "/" / "?" / "#" / "[" / "]" / "@"
1249 bool wxURI::IsGenDelim (const wxChar& c)
1250 {
1251     return c == wxT(':') ||
1252            c == wxT('/') ||
1253            c == wxT('?') ||
1254            c == wxT('#') ||
1255            c == wxT('[') ||
1256            c == wxT(']') ||
1257            c == wxT('@');
1258 }
1259
1260 //! sub-delims    = "!" / "$" / "&" / "'" / "(" / ")"
1261 //!               / "*" / "+" / "," / ";" / "="
1262 bool wxURI::IsSubDelim (const wxChar& c)
1263 {
1264     return c == wxT('!') ||
1265            c == wxT('$') ||
1266            c == wxT('&') ||
1267            c == wxT('\'') ||
1268            c == wxT('(') ||
1269            c == wxT(')') ||
1270            c == wxT('*') ||
1271            c == wxT('+') ||
1272            c == wxT(',') ||
1273            c == wxT(';') ||
1274            c == wxT('=')
1275            ;
1276 }
1277
1278 bool wxURI::IsHex(const wxChar& c)
1279 {   return IsDigit(c) || (c >= wxT('a') && c <= wxT('f')) || (c >= wxT('A') && c <= wxT('F')); }
1280
1281 bool wxURI::IsAlpha(const wxChar& c)
1282 {   return (c >= wxT('a') && c <= wxT('z')) || (c >= wxT('A') && c <= wxT('Z'));  }
1283
1284 bool wxURI::IsDigit(const wxChar& c)
1285 {   return c >= wxT('0') && c <= wxT('9');        }
1286
1287
1288 //end of uri.cpp
1289
1290
1291