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 #if defined(__GNUG__) && !defined(NO_GCC_PRAGMA)
  20     #pragma implementation "uri.h"
  21 #endif
  22
  23 // For compilers that support precompilation, includes "wx.h".
  24 #include "wx/wxprec.h"
  25
  26 #ifdef __BORLANDC__
  27     #pragma hdrstop
  28 #endif
  29
  30 #include "wx/uri.h"
  31
  32 // ---------------------------------------------------------------------------
  33 // definitions
  34 // ---------------------------------------------------------------------------
  35
  36 IMPLEMENT_CLASS(wxURI, wxObject);
  37
  38 // ===========================================================================
  39 // implementation
  40 // ===========================================================================
  41
  42 // ---------------------------------------------------------------------------
  43 // utilities
  44 // ---------------------------------------------------------------------------
  45
  46 // ---------------------------------------------------------------------------
  47 //
  48 //                        wxURI
  49 //
  50 // ---------------------------------------------------------------------------
  51
  52 // ---------------------------------------------------------------------------
  53 //  Constructors
  54 // ---------------------------------------------------------------------------
  55
  56 wxURI::wxURI() : m_hostType(wxURI_REGNAME), m_fields(0)
  57 {
  58 }
  59
  60 wxURI::wxURI(const wxString& uri) : m_hostType(wxURI_REGNAME), m_fields(0)
  61 {
  62     Create(uri);
  63 }
  64
  65 wxURI::wxURI(const wxURI& uri)  : wxObject(), m_hostType(wxURI_REGNAME), m_fields(0)
  66 {
  67     Assign(uri);
  68 }
  69
  70 // ---------------------------------------------------------------------------
  71 // Destructor and cleanup
  72 // ---------------------------------------------------------------------------
  73
  74 wxURI::~wxURI()
  75 {
  76     Clear();
  77 }
  78
  79 void wxURI::Clear()
  80 {
  81     m_scheme = m_userinfo = m_server = m_port = m_path =
  82     m_query = m_fragment = wxEmptyString;
  83
  84     m_hostType = wxURI_REGNAME;
  85
  86     m_fields = 0;
  87 }
  88
  89 // ---------------------------------------------------------------------------
  90 // Create
  91 //
  92 // This creates the URI - all we do here is call the main parsing method
  93 // ---------------------------------------------------------------------------
  94
  95 const wxChar* wxURI::Create(const wxString& uri)
  96 {
  97     if (m_fields)
  98         Clear();
  99
 100     return Parse(uri);
 101 }
 102
 103 // ---------------------------------------------------------------------------
 104 // Escape Methods
 105 //
 106 // TranslateEscape unencodes a 3 character URL escape sequence
 107 //
 108 // Escape encodes an invalid URI character into a 3 character sequence
 109 //
 110 // IsEscape determines if the input string contains an escape sequence,
 111 // if it does, then it moves the input string past the escape sequence
 112 //
 113 // Unescape unencodes all 3 character URL escape sequences in a wxString
 114 // ---------------------------------------------------------------------------
 115
 116 wxChar wxURI::TranslateEscape(const wxChar* s)
 117 {
 118     wxASSERT_MSG( IsHex(s[0]) && IsHex(s[1]), wxT("Invalid escape sequence!"));
 119
 120     return (wxChar)( CharToHex(s[0]) << 4 ) | CharToHex(s[1]);
 121 }
 122
 123 wxString wxURI::Unescape(const wxString& uri)
 124 {
 125     wxString new_uri;
 126
 127     for(size_t i = 0; i < uri.length(); ++i)
 128     {
 129         if (uri[i] == wxT('%'))
 130         {
 131             new_uri += wxURI::TranslateEscape( &(uri.c_str()[i+1]) );
 132             i += 2;
 133         }
 134         else
 135             new_uri += uri[i];
 136     }
 137
 138     return new_uri;
 139 }
 140
 141 void wxURI::Escape(wxString& s, const wxChar& c)
 142 {
 143     const wxChar* hdig = wxT("0123456789abcdef");
 144     s += wxT('%');
 145     s += hdig[(c >> 4) & 15];
 146     s += hdig[c & 15];
 147 }
 148
 149 bool wxURI::IsEscape(const wxChar*& uri)
 150 {
 151     // pct-encoded   = "%" HEXDIG HEXDIG
 152     if(*uri == wxT('%') && IsHex(*(uri+1)) && IsHex(*(uri+2)))
 153     {
 154         uri += 3;
 155         return true;
 156     }
 157     else
 158         return false;
 159 }
 160
 161 // ---------------------------------------------------------------------------
 162 // GetUser
 163 // GetPassword
 164 //
 165 // Gets the username and password via the old URL method.
 166 // ---------------------------------------------------------------------------
 167 wxString wxURI::GetUser() const
 168 {
 169       size_t dwPasswordPos = m_userinfo.find(':');
 170
 171       if (dwPasswordPos == wxString::npos)
 172           dwPasswordPos = 0;
 173
 174       return m_userinfo(0, dwPasswordPos);
 175 }
 176
 177 wxString wxURI::GetPassword() const
 178 {
 179       size_t dwPasswordPos = m_userinfo.find(':');
 180
 181       if (dwPasswordPos == wxString::npos)
 182           return wxT("");
 183       else
 184           return m_userinfo(dwPasswordPos+1, m_userinfo.length() + 1);
 185 }
 186
 187 // ---------------------------------------------------------------------------
 188 // BuildURI
 189 //
 190 // BuildURI() builds the entire URI into a useable
 191 // representation, including proper identification characters such as slashes
 192 //
 193 // BuildUnescapedURI() does the same thing as BuildURI(), only it unescapes
 194 // the components that accept escape sequences
 195 // ---------------------------------------------------------------------------
 196
 197 wxString wxURI::BuildURI() const
 198 {
 199     wxString ret;
 200
 201     if (HasScheme())
 202         ret = ret + m_scheme + wxT(":");
 203
 204     if (HasServer())
 205     {
 206         ret += wxT("//");
 207
 208         if (HasUserInfo())
 209             ret = ret + m_userinfo + wxT("@");
 210
 211         ret += m_server;
 212
 213         if (HasPort())
 214             ret = ret + wxT(":") + m_port;
 215     }
 216
 217     ret += m_path;
 218
 219     if (HasQuery())
 220         ret = ret + wxT("?") + m_query;
 221
 222     if (HasFragment())
 223         ret = ret + wxT("#") + m_fragment;
 224
 225     return ret;
 226 }
 227
 228 wxString wxURI::BuildUnescapedURI() const
 229 {
 230     wxString ret;
 231
 232     if (HasScheme())
 233         ret = ret + m_scheme + wxT(":");
 234
 235     if (HasServer())
 236     {
 237         ret += wxT("//");
 238
 239         if (HasUserInfo())
 240             ret = ret + wxURI::Unescape(m_userinfo) + wxT("@");
 241
 242         if (m_hostType == wxURI_REGNAME)
 243             ret += wxURI::Unescape(m_server);
 244         else
 245             ret += m_server;
 246
 247         if (HasPort())
 248             ret = ret + wxT(":") + m_port;
 249     }
 250
 251     ret += wxURI::Unescape(m_path);
 252
 253     if (HasQuery())
 254         ret = ret + wxT("?") + wxURI::Unescape(m_query);
 255
 256     if (HasFragment())
 257         ret = ret + wxT("#") + wxURI::Unescape(m_fragment);
 258
 259     return ret;
 260 }
 261
 262 // ---------------------------------------------------------------------------
 263 // Assignment
 264 // ---------------------------------------------------------------------------
 265
 266 wxURI& wxURI::Assign(const wxURI& uri)
 267 {
 268     //assign fields
 269     m_fields = uri.m_fields;
 270
 271     //ref over components
 272     m_scheme = uri.m_scheme;
 273     m_userinfo = uri.m_userinfo;
 274     m_server = uri.m_server;
 275     m_hostType = uri.m_hostType;
 276     m_port = uri.m_port;
 277     m_path = uri.m_path;
 278     m_query = uri.m_query;
 279     m_fragment = uri.m_fragment;
 280
 281     return *this;
 282 }
 283
 284 wxURI& wxURI::operator = (const wxURI& uri)
 285 {
 286     return Assign(uri);
 287 }
 288
 289 wxURI& wxURI::operator = (const wxString& string)
 290 {
 291     Create(string);
 292     return *this;
 293 }
 294
 295 // ---------------------------------------------------------------------------
 296 // Comparison
 297 // ---------------------------------------------------------------------------
 298
 299 bool wxURI::operator == (const wxURI& uri) const
 300 {
 301     if (HasScheme())
 302     {
 303         if(m_scheme != uri.m_scheme)
 304             return false;
 305     }
 306     else if (uri.HasScheme())
 307         return false;
 308
 309
 310     if (HasServer())
 311     {
 312         if (HasUserInfo())
 313         {
 314             if (m_userinfo != uri.m_userinfo)
 315                 return false;
 316         }
 317         else if (uri.HasUserInfo())
 318             return false;
 319
 320         if (m_server != uri.m_server ||
 321             m_hostType != uri.m_hostType)
 322             return false;
 323
 324         if (HasPort())
 325         {
 326             if(m_port != uri.m_port)
 327                 return false;
 328         }
 329         else if (uri.HasPort())
 330             return false;
 331     }
 332     else if (uri.HasServer())
 333         return false;
 334
 335
 336     if (HasPath())
 337     {
 338         if(m_path != uri.m_path)
 339             return false;
 340     }
 341     else if (uri.HasPath())
 342         return false;
 343
 344     if (HasQuery())
 345     {
 346         if (m_query != uri.m_query)
 347             return false;
 348     }
 349     else if (uri.HasQuery())
 350         return false;
 351
 352     if (HasFragment())
 353     {
 354         if (m_fragment != uri.m_fragment)
 355             return false;
 356     }
 357     else if (uri.HasFragment())
 358         return false;
 359
 360     return true;
 361 }
 362
 363 // ---------------------------------------------------------------------------
 364 // IsReference
 365 //
 366 // if there is no authority or scheme, it is a reference
 367 // ---------------------------------------------------------------------------
 368
 369 bool wxURI::IsReference() const
 370 {   return !HasScheme() || !HasServer();  }
 371
 372 // ---------------------------------------------------------------------------
 373 // Parse
 374 //
 375 // Master URI parsing method.  Just calls the individual parsing methods
 376 //
 377 // URI = scheme ":" hier-part [ "?" query ] [ "#" fragment ]
 378 // URI-reference = URI / relative-URITestCase
 379 // ---------------------------------------------------------------------------
 380
 381 const wxChar* wxURI::Parse(const wxChar* uri)
 382 {
 383     uri = ParseScheme(uri);
 384     uri = ParseAuthority(uri);
 385     uri = ParsePath(uri);
 386     uri = ParseQuery(uri);
 387     return ParseFragment(uri);
 388 }
 389
 390 // ---------------------------------------------------------------------------
 391 // ParseXXX
 392 //
 393 // Individual parsers for each URI component
 394 // ---------------------------------------------------------------------------
 395
 396 const wxChar* wxURI::ParseScheme(const wxChar* uri)
 397 {
 398     wxASSERT(uri != NULL);
 399
 400     //copy of the uri - used for figuring out
 401     //length of each component
 402     const wxChar* uricopy = uri;
 403
 404     //Does the uri have a scheme (first character alpha)?
 405     if (IsAlpha(*uri))
 406     {
 407         m_scheme += *uri++;
 408
 409         //scheme        = ALPHA *( ALPHA / DIGIT / "+" / "-" / "." )
 410         while (IsAlpha(*uri) || IsDigit(*uri) ||
 411                *uri == wxT('+')   ||
 412                *uri == wxT('-')   ||
 413                *uri == wxT('.'))
 414         {
 415             m_scheme += *uri++;
 416         }
 417
 418         //valid scheme?
 419         if (*uri == wxT(':'))
 420         {
 421             //mark the scheme as valid
 422             m_fields |= wxURI_SCHEME;
 423
 424             //move reference point up to input buffer
 425             uricopy = ++uri;
 426         }
 427         else
 428             //relative uri with relative path reference
 429             m_scheme = wxEmptyString;
 430     }
 431 //    else
 432         //relative uri with _possible_ relative path reference
 433
 434     return uricopy;
 435 }
 436
 437 const wxChar* wxURI::ParseAuthority(const wxChar* uri)
 438 {
 439     // authority     = [ userinfo "@" ] host [ ":" port ]
 440     if (*uri == wxT('/') && *(uri+1) == wxT('/'))
 441     {
 442         uri += 2;
 443
 444         uri = ParseUserInfo(uri);
 445         uri = ParseServer(uri);
 446         return ParsePort(uri);
 447     }
 448
 449     return uri;
 450 }
 451
 452 const wxChar* wxURI::ParseUserInfo(const wxChar* uri)
 453 {
 454     wxASSERT(uri != NULL);
 455
 456     //copy of the uri - used for figuring out
 457     //length of each component
 458     const wxChar* uricopy = uri;
 459
 460     // userinfo      = *( unreserved / pct-encoded / sub-delims / ":" )
 461     while(*uri && *uri != wxT('@') && *uri != wxT('/') && *uri != wxT('#') && *uri != wxT('?'))
 462     {
 463         if(IsUnreserved(*uri) || IsEscape(uri) ||
 464            IsSubDelim(*uri) || *uri == wxT(':'))
 465             m_userinfo += *uri++;
 466         else
 467             Escape(m_userinfo, *uri++);
 468     }
 469
 470     if(*uri == wxT('@'))
 471     {
 472         //valid userinfo
 473         m_fields |= wxURI_USERINFO;
 474
 475         uricopy = ++uri;
 476     }
 477     else
 478         m_userinfo = wxEmptyString;
 479
 480     return uricopy;
 481 }
 482
 483 const wxChar* wxURI::ParseServer(const wxChar* uri)
 484 {
 485     wxASSERT(uri != NULL);
 486
 487     //copy of the uri - used for figuring out
 488     //length of each component
 489     const wxChar* uricopy = uri;
 490
 491     // host          = IP-literal / IPv4address / reg-name
 492     // IP-literal    = "[" ( IPv6address / IPvFuture  ) "]"
 493     if (*uri == wxT('['))
 494     {
 495         ++uri; //some compilers don't support *&ing a ++*
 496         if (ParseIPv6address(uri) && *uri == wxT(']'))
 497         {
 498             ++uri;
 499             m_hostType = wxURI_IPV6ADDRESS;
 500
 501             wxStringBufferLength theBuffer(m_server, uri - uricopy);
 502             wxTmemcpy(theBuffer, uricopy, uri-uricopy);
 503             theBuffer.SetLength(uri-uricopy);
 504         }
 505         else
 506         {
 507             uri = uricopy;
 508
 509             ++uri; //some compilers don't support *&ing a ++*
 510             if (ParseIPvFuture(uri) && *uri == wxT(']'))
 511             {
 512                 ++uri;
 513                 m_hostType = wxURI_IPVFUTURE;
 514
 515                 wxStringBufferLength theBuffer(m_server, uri - uricopy);
 516                 wxTmemcpy(theBuffer, uricopy, uri-uricopy);
 517                 theBuffer.SetLength(uri-uricopy);
 518             }
 519             else
 520                 uri = uricopy;
 521         }
 522     }
 523     else
 524     {
 525         if (ParseIPv4address(uri))
 526         {
 527             m_hostType = wxURI_IPV4ADDRESS;
 528
 529             wxStringBufferLength theBuffer(m_server, uri - uricopy);
 530             wxTmemcpy(theBuffer, uricopy, uri-uricopy);
 531             theBuffer.SetLength(uri-uricopy);
 532         }
 533         else
 534             uri = uricopy;
 535     }
 536
 537     if(m_hostType == wxURI_REGNAME)
 538     {
 539         uri = uricopy;
 540         // reg-name      = *( unreserved / pct-encoded / sub-delims )
 541         while(*uri && *uri != wxT('/') && *uri != wxT(':') && *uri != wxT('#') && *uri != wxT('?'))
 542         {
 543             if(IsUnreserved(*uri) || IsEscape(uri) ||  IsSubDelim(*uri))
 544                 m_server += *uri++;
 545             else
 546                 Escape(m_server, *uri++);
 547         }
 548     }
 549
 550     //mark the server as valid
 551     m_fields |= wxURI_SERVER;
 552
 553     return uri;
 554 }
 555
 556
 557 const wxChar* wxURI::ParsePort(const wxChar* uri)
 558 {
 559     wxASSERT(uri != NULL);
 560
 561     // port          = *DIGIT
 562     if(*uri == wxT(':'))
 563     {
 564         ++uri;
 565         while(IsDigit(*uri))
 566         {
 567             m_port += *uri++;
 568         }
 569
 570         //mark the port as valid
 571         m_fields |= wxURI_PORT;
 572     }
 573
 574     return uri;
 575 }
 576
 577 const wxChar* wxURI::ParsePath(const wxChar* uri, bool bReference, bool bNormalize)
 578 {
 579     wxASSERT(uri != NULL);
 580
 581     //copy of the uri - used for figuring out
 582     //length of each component
 583     const wxChar* uricopy = uri;
 584
 585     /// hier-part     = "//" authority path-abempty
 586     ///               / path-absolute
 587     ///               / path-rootless
 588     ///               / path-empty
 589     ///
 590     /// relative-part = "//" authority path-abempty
 591     ///               / path-absolute
 592     ///               / path-noscheme
 593     ///               / path-empty
 594     ///
 595     /// path-abempty  = *( "/" segment )
 596     /// path-absolute = "/" [ segment-nz *( "/" segment ) ]
 597     /// path-noscheme = segment-nz-nc *( "/" segment )
 598     /// path-rootless = segment-nz *( "/" segment )
 599     /// path-empty    = 0<pchar>
 600     ///
 601     /// segment       = *pchar
 602     /// segment-nz    = 1*pchar
 603     /// segment-nz-nc = 1*( unreserved / pct-encoded / sub-delims / "@" )
 604     ///               ; non-zero-length segment without any colon ":"
 605     ///
 606     /// pchar         = unreserved / pct-encoded / sub-delims / ":" / "@"
 607     if (*uri == wxT('/'))
 608     {
 609         m_path += *uri++;
 610
 611         while(*uri && *uri != wxT('#') && *uri != wxT('?'))
 612         {
 613             if( IsUnreserved(*uri) || IsSubDelim(*uri) || IsEscape(uri) ||
 614                 *uri == wxT(':') || *uri == wxT('@') || *uri == wxT('/'))
 615                 m_path += *uri++;
 616             else
 617                 Escape(m_path, *uri++);
 618         }
 619
 620         if (bNormalize)
 621         {
 622             wxStringBufferLength theBuffer(m_path, m_path.length() + 1);
 623 #if wxUSE_STL
 624             wxTmemcpy(theBuffer, m_path.c_str(), m_path.length()+1);
 625 #endif
 626             Normalize(theBuffer, true);
 627             theBuffer.SetLength(wxStrlen(theBuffer));
 628         }
 629         //mark the path as valid
 630         m_fields |= wxURI_PATH;
 631     }
 632     else if(*uri) //Relative path
 633     {
 634         if (bReference)
 635         {
 636             //no colon allowed
 637             while(*uri && *uri != wxT('#') && *uri != wxT('?'))
 638             {
 639                 if(IsUnreserved(*uri) || IsSubDelim(*uri) || IsEscape(uri) ||
 640                   *uri == wxT('@') || *uri == wxT('/'))
 641                     m_path += *uri++;
 642                 else
 643                     Escape(m_path, *uri++);
 644             }
 645         }
 646         else
 647         {
 648             while(*uri && *uri != wxT('#') && *uri != wxT('?'))
 649             {
 650                 if(IsUnreserved(*uri) || IsSubDelim(*uri) || IsEscape(uri) ||
 651                    *uri == wxT(':') || *uri == wxT('@') || *uri == wxT('/'))
 652                     m_path += *uri++;
 653                 else
 654                     Escape(m_path, *uri++);
 655             }
 656         }
 657
 658         if (uri != uricopy)
 659         {
 660             if (bNormalize)
 661             {
 662                 wxStringBufferLength theBuffer(m_path, m_path.length() + 1);
 663 #if wxUSE_STL
 664                 wxTmemcpy(theBuffer, m_path.c_str(), m_path.length()+1);
 665 #endif
 666                 Normalize(theBuffer);
 667                 theBuffer.SetLength(wxStrlen(theBuffer));
 668             }
 669
 670             //mark the path as valid
 671             m_fields |= wxURI_PATH;
 672         }
 673     }
 674
 675     return uri;
 676 }
 677
 678
 679 const wxChar* wxURI::ParseQuery(const wxChar* uri)
 680 {
 681     wxASSERT(uri != NULL);
 682
 683     // query         = *( pchar / "/" / "?" )
 684     if (*uri == wxT('?'))
 685     {
 686         ++uri;
 687         while(*uri && *uri != wxT('#'))
 688         {
 689             if (IsUnreserved(*uri) || IsSubDelim(*uri) || IsEscape(uri) ||
 690                 *uri == wxT(':') || *uri == wxT('@') || *uri == wxT('/') || *uri == wxT('?'))
 691                   m_query += *uri++;
 692             else
 693                   Escape(m_query, *uri++);
 694         }
 695
 696         //mark the server as valid
 697         m_fields |= wxURI_QUERY;
 698     }
 699
 700     return uri;
 701 }
 702
 703
 704 const wxChar* wxURI::ParseFragment(const wxChar* uri)
 705 {
 706     wxASSERT(uri != NULL);
 707
 708     // fragment      = *( pchar / "/" / "?" )
 709     if (*uri == wxT('#'))
 710     {
 711         ++uri;
 712         while(*uri)
 713         {
 714             if (IsUnreserved(*uri) || IsSubDelim(*uri) || IsEscape(uri) ||
 715                 *uri == wxT(':') || *uri == wxT('@') || *uri == wxT('/') || *uri == wxT('?'))
 716                   m_fragment += *uri++;
 717             else
 718                   Escape(m_fragment, *uri++);
 719         }
 720
 721         //mark the server as valid
 722         m_fields |= wxURI_FRAGMENT;
 723     }
 724
 725     return uri;
 726 }
 727
 728 // ---------------------------------------------------------------------------
 729 // Resolve
 730 //
 731 // Builds missing components of this uri from a base uri
 732 //
 733 // A version of the algorithm outlined in the RFC is used here
 734 // (it is shown in comments)
 735 //
 736 // Note that an empty URI inherits all components
 737 // ---------------------------------------------------------------------------
 738
 739 void wxURI::Resolve(const wxURI& base, int flags)
 740 {
 741     wxASSERT_MSG(!base.IsReference(),
 742                 wxT("wxURI to inherit from must not be a reference!"));
 743
 744     // If we arn't being strict, enable the older (pre-RFC2396)
 745     // loophole that allows this uri to inherit other
 746     // properties from the base uri - even if the scheme
 747     // is defined
 748     if ( !(flags & wxURI_STRICT) &&
 749             HasScheme() && base.HasScheme() &&
 750                 m_scheme == base.m_scheme )
 751     {
 752         m_fields -= wxURI_SCHEME;
 753     }
 754
 755
 756     // Do nothing if this is an absolute wxURI
 757     //    if defined(R.scheme) then
 758     //       T.scheme    = R.scheme;
 759     //       T.authority = R.authority;
 760     //       T.path      = remove_dot_segments(R.path);
 761     //       T.query     = R.query;
 762     if (HasScheme())
 763     {
 764         return;
 765     }
 766
 767     //No scheme - inherit
 768     m_scheme = base.m_scheme;
 769     m_fields |= wxURI_SCHEME;
 770
 771     // All we need to do for relative URIs with an
 772     // authority component is just inherit the scheme
 773     //       if defined(R.authority) then
 774     //          T.authority = R.authority;
 775     //          T.path      = remove_dot_segments(R.path);
 776     //          T.query     = R.query;
 777     if (HasServer())
 778     {
 779         return;
 780     }
 781
 782     //No authority - inherit
 783     if (base.HasUserInfo())
 784     {
 785         m_userinfo = base.m_userinfo;
 786         m_fields |= wxURI_USERINFO;
 787     }
 788
 789     m_server = base.m_server;
 790     m_hostType = base.m_hostType;
 791     m_fields |= wxURI_SERVER;
 792
 793     if (base.HasPort())
 794     {
 795         m_port = base.m_port;
 796         m_fields |= wxURI_PORT;
 797     }
 798
 799
 800     // Simple path inheritance from base
 801     if (!HasPath())
 802     {
 803         //             T.path = Base.path;
 804         m_path = base.m_path;
 805         m_fields |= wxURI_PATH;
 806
 807
 808         //             if defined(R.query) then
 809         //                T.query = R.query;
 810         //             else
 811         //                T.query = Base.query;
 812         //             endif;
 813         if (!HasQuery())
 814         {
 815             m_query = base.m_query;
 816             m_fields |= wxURI_QUERY;
 817         }
 818     }
 819     else
 820     {
 821         //             if (R.path starts-with "/") then
 822         //                T.path = remove_dot_segments(R.path);
 823         //             else
 824         //                T.path = merge(Base.path, R.path);
 825         //                T.path = remove_dot_segments(T.path);
 826         //             endif;
 827         //             T.query = R.query;
 828         if (m_path[0u] != wxT('/'))
 829         {
 830             //Merge paths
 831             const wxChar* op = m_path.c_str();
 832             const wxChar* bp = base.m_path.c_str() + base.m_path.Length();
 833
 834             //not a ending directory?  move up
 835             if (base.m_path[0] && *(bp-1) != wxT('/'))
 836                 UpTree(base.m_path, bp);
 837
 838             //normalize directories
 839             while(*op == wxT('.') && *(op+1) == wxT('.') &&
 840                        (*(op+2) == '\0' || *(op+2) == wxT('/')) )
 841             {
 842                 UpTree(base.m_path, bp);
 843
 844                 if (*(op+2) == '\0')
 845                     op += 2;
 846                 else
 847                     op += 3;
 848             }
 849
 850             m_path = base.m_path.substr(0, bp - base.m_path.c_str()) +
 851                     m_path.substr((op - m_path.c_str()), m_path.Length());
 852         }
 853     }
 854
 855     //T.fragment = R.fragment;
 856 }
 857
 858 // ---------------------------------------------------------------------------
 859 // UpTree
 860 //
 861 // Moves a URI path up a directory
 862 // ---------------------------------------------------------------------------
 863
 864 //static
 865 void wxURI::UpTree(const wxChar* uristart, const wxChar*& uri)
 866 {
 867     if (uri != uristart && *(uri-1) == wxT('/'))
 868     {
 869         uri -= 2;
 870     }
 871
 872     for(;uri != uristart; --uri)
 873     {
 874         if (*uri == wxT('/'))
 875         {
 876             ++uri;
 877             break;
 878         }
 879     }
 880
 881     //!!!TODO:HACK!!!//
 882     if (uri == uristart && *uri == wxT('/'))
 883         ++uri;
 884     //!!!//
 885 }
 886
 887 // ---------------------------------------------------------------------------
 888 // Normalize
 889 //
 890 // Normalizes directories in-place
 891 //
 892 // I.E. ./ and . are ignored
 893 //
 894 // ../ and .. are removed if a directory is before it, along
 895 // with that directory (leading .. and ../ are kept)
 896 // ---------------------------------------------------------------------------
 897
 898 //static
 899 void wxURI::Normalize(wxChar* s, bool bIgnoreLeads)
 900 {
 901     wxChar* cp = s;
 902     wxChar* bp = s;
 903
 904     if(s[0] == wxT('/'))
 905         ++bp;
 906
 907     while(*cp)
 908     {
 909         if (*cp == wxT('.') && (*(cp+1) == wxT('/') || *(cp+1) == '\0')
 910             && (bp == cp || *(cp-1) == wxT('/')))
 911         {
 912             //. _or_ ./  - ignore
 913             if (*(cp+1) == '\0')
 914                 cp += 1;
 915             else
 916                 cp += 2;
 917         }
 918         else if (*cp == wxT('.') && *(cp+1) == wxT('.') &&
 919                 (*(cp+2) == wxT('/') || *(cp+2) == '\0')
 920                 && (bp == cp || *(cp-1) == wxT('/')))
 921         {
 922             //.. _or_ ../ - go up the tree
 923             if (s != bp)
 924             {
 925                 UpTree((const wxChar*)bp, (const wxChar*&)s);
 926
 927                 if (*(cp+2) == '\0')
 928                     cp += 2;
 929                 else
 930                     cp += 3;
 931             }
 932             else if (!bIgnoreLeads)
 933
 934             {
 935                 *bp++ = *cp++;
 936                 *bp++ = *cp++;
 937                 if (*cp)
 938                     *bp++ = *cp++;
 939
 940                 s = bp;
 941             }
 942             else
 943             {
 944                 if (*(cp+2) == '\0')
 945                     cp += 2;
 946                 else
 947                     cp += 3;
 948             }
 949         }
 950         else
 951             *s++ = *cp++;
 952     }
 953
 954     *s = '\0';
 955 }
 956
 957 // ---------------------------------------------------------------------------
 958 // ParseH16
 959 //
 960 // Parses 1 to 4 hex values.  Returns true if the first character of the input
 961 // string is a valid hex character.  It is the caller's responsability to move
 962 // the input string back to its original position on failure.
 963 // ---------------------------------------------------------------------------
 964
 965 bool wxURI::ParseH16(const wxChar*& uri)
 966 {
 967     // h16           = 1*4HEXDIG
 968     if(!IsHex(*++uri))
 969         return false;
 970
 971     if(IsHex(*++uri) && IsHex(*++uri) && IsHex(*++uri))
 972         ++uri;
 973
 974     return true;
 975 }
 976
 977 // ---------------------------------------------------------------------------
 978 // ParseIPXXX
 979 //
 980 // Parses a certain version of an IP address and moves the input string past
 981 // it.  Returns true if the input  string contains the proper version of an ip
 982 // address.  It is the caller's responsability to move the input string back
 983 // to its original position on failure.
 984 // ---------------------------------------------------------------------------
 985
 986 bool wxURI::ParseIPv4address(const wxChar*& uri)
 987 {
 988     //IPv4address   = dec-octet "." dec-octet "." dec-octet "." dec-octet
 989     //
 990     //dec-octet     =      DIGIT                    ; 0-9
 991     //                / %x31-39 DIGIT               ; 10-99
 992     //                / "1" 2DIGIT                  ; 100-199
 993     //                / "2" %x30-34 DIGIT           ; 200-249
 994     //                / "25" %x30-35                ; 250-255
 995     size_t iIPv4 = 0;
 996     if (IsDigit(*uri))
 997     {
 998         ++iIPv4;
 999
1000
1001         //each ip part must be between 0-255 (dupe of version in for loop)
1002         if( IsDigit(*++uri) && IsDigit(*++uri) &&
1003            //100 or less  (note !)
1004            !( (*(uri-2) < wxT('2')) ||
1005            //240 or less
1006              (*(uri-2) == wxT('2') &&
1007                (*(uri-1) < wxT('5') || (*(uri-1) == wxT('5') && *uri <= wxT('5')))
1008              )
1009             )
1010           )
1011         {
1012             return false;
1013         }
1014
1015         if(IsDigit(*uri))++uri;
1016
1017         //compilers should unroll this loop
1018         for(; iIPv4 < 4; ++iIPv4)
1019         {
1020             if (*uri != wxT('.') || !IsDigit(*++uri))
1021                 break;
1022
1023             //each ip part must be between 0-255
1024             if( IsDigit(*++uri) && IsDigit(*++uri) &&
1025                //100 or less  (note !)
1026                !( (*(uri-2) < wxT('2')) ||
1027                //240 or less
1028                  (*(uri-2) == wxT('2') &&
1029                    (*(uri-1) < wxT('5') || (*(uri-1) == wxT('5') && *uri <= wxT('5')))
1030                  )
1031                 )
1032               )
1033             {
1034                 return false;
1035             }
1036             if(IsDigit(*uri))++uri;
1037         }
1038     }
1039     return iIPv4 == 4;
1040 }
1041
1042 bool wxURI::ParseIPv6address(const wxChar*& uri)
1043 {
1044     // IPv6address   =                            6( h16 ":" ) ls32
1045     //               /                       "::" 5( h16 ":" ) ls32
1046     //               / [               h16 ] "::" 4( h16 ":" ) ls32
1047     //               / [ *1( h16 ":" ) h16 ] "::" 3( h16 ":" ) ls32
1048     //               / [ *2( h16 ":" ) h16 ] "::" 2( h16 ":" ) ls32
1049     //               / [ *3( h16 ":" ) h16 ] "::"    h16 ":"   ls32
1050     //               / [ *4( h16 ":" ) h16 ] "::"              ls32
1051     //               / [ *5( h16 ":" ) h16 ] "::"              h16
1052     //               / [ *6( h16 ":" ) h16 ] "::"
1053
1054     size_t numPrefix = 0,
1055               maxPostfix;
1056
1057     bool bEndHex = false;
1058
1059     for( ; numPrefix < 6; ++numPrefix)
1060     {
1061         if(!ParseH16(uri))
1062         {
1063             --uri;
1064             bEndHex = true;
1065             break;
1066         }
1067
1068         if(*uri != wxT(':'))
1069         {
1070             break;
1071         }
1072     }
1073
1074     if(!bEndHex && !ParseH16(uri))
1075     {
1076         --uri;
1077
1078         if (numPrefix)
1079             return false;
1080
1081         if (*uri == wxT(':'))
1082         {
1083             if (*++uri != wxT(':'))
1084                 return false;
1085
1086             maxPostfix = 5;
1087         }
1088         else
1089             maxPostfix = 6;
1090     }
1091     else
1092     {
1093         if (*uri != wxT(':') || *(uri+1) != wxT(':'))
1094         {
1095             if (numPrefix != 6)
1096                 return false;
1097
1098             while (*--uri != wxT(':')) {}
1099             ++uri;
1100
1101             const wxChar* uristart = uri;
1102             //parse ls32
1103             // ls32          = ( h16 ":" h16 ) / IPv4address
1104             if (ParseH16(uri) && *uri == wxT(':') && ParseH16(uri))
1105                 return true;
1106
1107             uri = uristart;
1108
1109             if (ParseIPv4address(uri))
1110                 return true;
1111             else
1112                 return false;
1113         }
1114         else
1115         {
1116             uri += 2;
1117
1118             if (numPrefix > 3)
1119                 maxPostfix = 0;
1120             else
1121                 maxPostfix = 4 - numPrefix;
1122         }
1123     }
1124
1125     bool bAllowAltEnding = maxPostfix == 0;
1126
1127     for(; maxPostfix != 0; --maxPostfix)
1128     {
1129         if(!ParseH16(uri) || *uri != wxT(':'))
1130             return false;
1131     }
1132
1133     if(numPrefix <= 4)
1134     {
1135         const wxChar* uristart = uri;
1136         //parse ls32
1137         // ls32          = ( h16 ":" h16 ) / IPv4address
1138         if (ParseH16(uri) && *uri == wxT(':') && ParseH16(uri))
1139             return true;
1140
1141         uri = uristart;
1142
1143         if (ParseIPv4address(uri))
1144             return true;
1145
1146         uri = uristart;
1147
1148         if (!bAllowAltEnding)
1149             return false;
1150     }
1151
1152     if(numPrefix <= 5 && ParseH16(uri))
1153         return true;
1154
1155     return true;
1156 }
1157
1158 bool wxURI::ParseIPvFuture(const wxChar*& uri)
1159 {
1160     // IPvFuture     = "v" 1*HEXDIG "." 1*( unreserved / sub-delims / ":" )
1161     if (*++uri != wxT('v') || !IsHex(*++uri))
1162         return false;
1163
1164     while (IsHex(*++uri)) {}
1165
1166     if (*uri != wxT('.') || !(IsUnreserved(*++uri) || IsSubDelim(*uri) || *uri == wxT(':')))
1167         return false;
1168
1169     while(IsUnreserved(*++uri) || IsSubDelim(*uri) || *uri == wxT(':')) {}
1170
1171     return true;
1172 }
1173
1174
1175 // ---------------------------------------------------------------------------
1176 // CharToHex
1177 //
1178 // Converts a character into a numeric hexidecimal value, or 0 if the
1179 // passed in character is not a valid hex character
1180 // ---------------------------------------------------------------------------
1181
1182 //static
1183 wxChar wxURI::CharToHex(const wxChar& c)
1184 {
1185     if ((c >= wxT('A')) && (c <= wxT('Z'))) return wxChar(c - wxT('A') + 0x0A);
1186     if ((c >= wxT('a')) && (c <= wxT('z'))) return wxChar(c - wxT('a') + 0x0a);
1187     if ((c >= wxT('0')) && (c <= wxT('9'))) return wxChar(c - wxT('0') + 0x00);
1188
1189     return 0;
1190 }
1191
1192 // ---------------------------------------------------------------------------
1193 // IsXXX
1194 //
1195 // Returns true if the passed in character meets the criteria of the method
1196 // ---------------------------------------------------------------------------
1197
1198 //! unreserved    = ALPHA / DIGIT / "-" / "." / "_" / "~"
1199 bool wxURI::IsUnreserved (const wxChar& c)
1200 {   return IsAlpha(c) || IsDigit(c) ||
1201            c == wxT('-') ||
1202            c == wxT('.') ||
1203            c == wxT('_') ||
1204            c == wxT('~') //tilde
1205            ;
1206 }
1207
1208 bool wxURI::IsReserved (const wxChar& c)
1209 {
1210     return IsGenDelim(c) || IsSubDelim(c);
1211 }
1212
1213 //! gen-delims    = ":" / "/" / "?" / "#" / "[" / "]" / "@"
1214 bool wxURI::IsGenDelim (const wxChar& c)
1215 {
1216     return c == wxT(':') ||
1217            c == wxT('/') ||
1218            c == wxT('?') ||
1219            c == wxT('#') ||
1220            c == wxT('[') ||
1221            c == wxT(']') ||
1222            c == wxT('@');
1223 }
1224
1225 //! sub-delims    = "!" / "$" / "&" / "'" / "(" / ")"
1226 //!               / "*" / "+" / "," / ";" / "="
1227 bool wxURI::IsSubDelim (const wxChar& c)
1228 {
1229     return c == wxT('!') ||
1230            c == wxT('$') ||
1231            c == wxT('&') ||
1232            c == wxT('\'') ||
1233            c == wxT('(') ||
1234            c == wxT(')') ||
1235            c == wxT('*') ||
1236            c == wxT('+') ||
1237            c == wxT(',') ||
1238            c == wxT(';') ||
1239            c == wxT('=')
1240            ;
1241 }
1242
1243 bool wxURI::IsHex(const wxChar& c)
1244 {   return IsDigit(c) || (c >= wxT('a') && c <= wxT('f')) || (c >= wxT('A') && c <= wxT('F')); }
1245
1246 bool wxURI::IsAlpha(const wxChar& c)
1247 {   return (c >= wxT('a') && c <= wxT('z')) || (c >= wxT('A') && c <= wxT('Z'));  }
1248
1249 bool wxURI::IsDigit(const wxChar& c)
1250 {   return c >= wxT('0') && c <= wxT('9');        }
1251
1252
1253 // ---------------------------------------------------------------------------
1254 //
1255 //                        wxURL Compatibility
1256 //
1257 // ---------------------------------------------------------------------------
1258
1259 #if wxUSE_URL
1260
1261 #if WXWIN_COMPATIBILITY_2_4
1262
1263 #include "wx/url.h"
1264
1265 wxString wxURL::GetProtocolName() const
1266 {
1267     return m_scheme;
1268 }
1269
1270 wxString wxURL::GetHostName() const
1271 {
1272     return m_server;
1273 }
1274
1275 wxString wxURL::GetPath() const
1276 {
1277     return m_path;
1278 }
1279
1280 //Note that this old code really doesn't convert to a URI that well and looks
1281 //more like a dirty hack than anything else...
1282
1283 wxString wxURL::ConvertToValidURI(const wxString& uri, const wxChar* delims)
1284 {
1285   wxString out_str;
1286   wxString hexa_code;
1287   size_t i;
1288
1289   for (i = 0; i < uri.Len(); i++)
1290   {
1291     wxChar c = uri.GetChar(i);
1292
1293     if (c == wxT(' '))
1294     {
1295       // GRG, Apr/2000: changed to "%20" instead of '+'
1296
1297       out_str += wxT("%20");
1298     }
1299     else
1300     {
1301       // GRG, Apr/2000: modified according to the URI definition (RFC 2396)
1302       //
1303       // - Alphanumeric characters are never escaped
1304       // - Unreserved marks are never escaped
1305       // - Delimiters must be escaped if they appear within a component
1306       //     but not if they are used to separate components. Here we have
1307       //     no clear way to distinguish between these two cases, so they
1308       //     are escaped unless they are passed in the 'delims' parameter
1309       //     (allowed delimiters).
1310
1311       static const wxChar marks[] = wxT("-_.!~*()'");
1312
1313       if ( !wxIsalnum(c) && !wxStrchr(marks, c) && !wxStrchr(delims, c) )
1314       {
1315         hexa_code.Printf(wxT("%%%02X"), c);
1316         out_str += hexa_code;
1317       }
1318       else
1319       {
1320         out_str += c;
1321       }
1322     }
1323   }
1324
1325   return out_str;
1326 }
1327
1328 wxString wxURL::ConvertFromURI(const wxString& uri)
1329 {
1330     return wxURI::Unescape(uri);
1331 }
1332
1333 #endif //WXWIN_COMPATIBILITY_2_4
1334
1335 #endif //wxUSE_URL
1336
1337 //end of uri.cpp
1338
1339
1340