• Main Page
  • Related Pages
  • Namespaces
  • Classes
  • Files
  • File List
  • File Members

escaping.cc

Go to the documentation of this file.
00001 
00002 // -*- mode: c++; c-basic-offset:4 -*-
00003 
00004 // This file is part of libdap, A C++ implementation of the OPeNDAP Data
00005 // Access Protocol.
00006 
00007 // Copyright (c) 2002,2003 OPeNDAP, Inc.
00008 // Author: James Gallagher <jgallagher@opendap.org>
00009 //
00010 // This library is free software; you can redistribute it and/or
00011 // modify it under the terms of the GNU Lesser General Public
00012 // License as published by the Free Software Foundation; either
00013 // version 2.1 of the License, or (at your option) any later version.
00014 //
00015 // This library is distributed in the hope that it will be useful,
00016 // but WITHOUT ANY WARRANTY; without even the implied warranty of
00017 // MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
00018 // Lesser General Public License for more details.
00019 //
00020 // You should have received a copy of the GNU Lesser General Public
00021 // License along with this library; if not, write to the Free Software
00022 // Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
00023 //
00024 // You can contact OPeNDAP, Inc. at PO Box 112, Saunderstown, RI. 02874-0112.
00025 
00026 // Copyright (c) 1996, California Institute of Technology.
00027 // ALL RIGHTS RESERVED.   U.S. Government Sponsorship acknowledged.
00028 //
00029 // Please read the full copyright notice in the file COPYRIGHT_URI
00030 // in this directory.
00031 //
00032 // Author: Todd Karakashian, NASA/Jet Propulsion Laboratory
00033 //         Todd.K.Karakashian@jpl.nasa.gov
00034 //
00035 // $RCSfile: escaping.cc,v $ - Miscellaneous routines for OPeNDAP HDF server
00036 //
00037 // These two routines are for escaping/unescaping strings that are identifiers
00038 // in DAP2
00039 // id2www() -- escape (using WWW hex codes) non-allowable characters in a
00040 // DAP2 identifier
00041 // www2id() -- given an WWW hexcode escaped identifier, restore it
00042 //
00043 // These two routines are for escaping/unescaping strings storing attribute
00044 // values.  They use traditional octal escapes (\nnn) because they are
00045 // intended to be viewed by a user
00046 // escattr() -- escape (using traditional octal backslash) non-allowable
00047 // characters in the value of a DAP2 attribute
00048 // unescattr() -- given an octally escaped string, restore it
00049 //
00050 // These are routines used by the above, not intended to be called directly:
00051 //
00052 // hexstring()
00053 // unhexstring()
00054 // octstring()
00055 // unoctstring()
00056 //
00057 // -Todd
00058 
00059 #include <ctype.h>
00060 
00061 #include <iomanip>
00062 #include <string>
00063 #include <sstream>
00064 
00065 #include "GNURegex.h"
00066 #include "Error.h"
00067 #include "InternalErr.h"
00068 // #define DODS_DEBUG
00069 #include "debug.h"
00070 
00071 using namespace std;
00072 
00073 namespace libdap {
00074 
00075 // The next four functions were originally defined static, but I removed that
00076 // to make testing them (see generalUtilTest.cc) easier to write. 5/7/2001
00077 // jhrg
00078 
00079 string
00080 hexstring(unsigned char val)
00081 {
00082     ostringstream buf;
00083     buf << hex << setw(2) << setfill('0')
00084     << static_cast<unsigned int>(val);
00085 
00086     return buf.str();
00087 }
00088 
00089 string
00090 unhexstring(string s)
00091 {
00092     int val;
00093     istringstream ss(s);
00094     ss >> hex >> val;
00095     char tmp_str[2];
00096     tmp_str[0] = static_cast<char>(val);
00097     tmp_str[1] = '\0';
00098     return string(tmp_str);
00099 }
00100 
00101 string
00102 octstring(unsigned char val)
00103 {
00104     ostringstream buf;
00105     buf << oct << setw(3) << setfill('0')
00106     << static_cast<unsigned int>(val);
00107 
00108     return buf.str();
00109 }
00110 
00111 string
00112 unoctstring(string s)
00113 {
00114     int val;
00115 
00116     istringstream ss(s);
00117     ss >> oct >> val;
00118 
00119     DBG(cerr << "unoctstring: " << val << endl);
00120 
00121     char tmp_str[2];
00122     tmp_str[0] = static_cast<char>(val);
00123     tmp_str[1] = '\0';
00124     return string(tmp_str);
00125 }
00126 
00151 string
00152 id2www(string in, const string &allowable)
00153 {
00154     string::size_type i = 0;
00155 
00156     while ((i = in.find_first_not_of(allowable, i)) != string::npos) {
00157         in.replace(i, 1, "%" + hexstring(in[i]));
00158         i++;
00159     }
00160 
00161     return in;
00162 }
00163 
00174 string
00175 id2www_ce(string in, const string &allowable)
00176 {
00177     return id2www(in, allowable);
00178 }
00179 
00212 string
00213 www2id(const string &in, const string &escape, const string &except)
00214 {
00215     string::size_type i = 0;
00216     string res = in;
00217     while ((i = res.find_first_of(escape, i)) != string::npos) {
00218         if (except.find(res.substr(i, 3)) != string::npos) {
00219             i += 3;
00220             continue;
00221         }
00222         res.replace(i, 3, unhexstring(res.substr(i + 1, 2)));
00223     }
00224 
00225     return res;
00226 }
00227 
00228 static string
00229 entity(char c)
00230 {
00231     switch (c) {
00232     case '>': return "&gt;";
00233     case '<': return "&lt;";
00234     case '&': return "&amp;";
00235     case '\'': return "&apos;";
00236     case '\"': return "&quot;";
00237     default:
00238         throw InternalErr(__FILE__, __LINE__, "Unrecognized character.");
00239     }
00240 }
00241 
00242 // Assumption: There are always exactly two octal digits in the input
00243 // and two hex digits in the result.
00244 string
00245 octal_to_hex(const string &octal_digits)
00246 {
00247     int val;
00248 
00249     istringstream ss(octal_digits);
00250     ss >> oct >> val;
00251 
00252     ostringstream ds;
00253     ds << hex << setw(2) << setfill('0') << val;
00254     return ds.str();
00255 }
00256 
00263 string
00264 id2xml(string in, const string &not_allowed)
00265 {
00266     string::size_type i = 0;
00267 
00268     while ((i = in.find_first_of(not_allowed, i)) != string::npos) {
00269         in.replace(i, 1, entity(in[i]));
00270         ++i;
00271     }
00272 #if 0
00273     // Removed the encoding of octal escapes. This function is used by
00274     // AttrTable to encode the stuff that is the value of the <value>
00275     // element in the DDX. The problem is that some of the values are not
00276     // valid UTF-8 and that makes a XML parser gag.; ticket 1512.
00277     // jhrg 3/19/10
00278 
00279     // OK, now scan for octal escape sequences like \\012 (where the '\'
00280     // is itself escaped). This type of attribute value comes from the netCDF
00281     // handler and maybe others. Assumption: The '\' will always appear as
00282     // in its escaped form: '\\'. NB: Both backslashes must be escaped in the
00283     // C++ string.
00284     string octal_escape = "\\\\";
00285     i = 0;
00286     string::size_type length = in.length();
00287     while ((i = in.find(octal_escape, i)) != string::npos) {
00288         // Get the three octal digits following the '\\0'
00289         string::size_type j = i + 2;
00290         if (j + 1 >= length)  // Check that we're not past the end
00291             break;
00292         string octal_digits = in.substr(j, 3);
00293         // convert to a &#xdd; XML escape
00294         string hex_escape = string("&#x");
00295         hex_escape.append(octal_to_hex(octal_digits));
00296         hex_escape.append(string(";"));
00297 
00298         // replace the octal escape with an XML/hex escape
00299         in.replace(i, 5, hex_escape);
00300 
00301         // increment i
00302         i += 6;
00303     }
00304 #endif
00305     return in;
00306 }
00307 
00313 string
00314 xml2id(string in)
00315 {
00316     string::size_type i = 0;
00317 
00318     while ((i = in.find("&gt;", i)) != string::npos)
00319         in.replace(i, 4, ">");
00320 
00321     i = 0;
00322     while ((i = in.find("&lt;", i)) != string::npos)
00323         in.replace(i, 4, "<");
00324 
00325     i = 0;
00326     while ((i = in.find("&amp;", i)) != string::npos)
00327         in.replace(i, 5, "&");
00328 
00329     i = 0;
00330     while ((i = in.find("&apos;", i)) != string::npos)
00331         in.replace(i, 6, "'");
00332 
00333     i = 0;
00334     while ((i = in.find("&quot;", i)) != string::npos)
00335         in.replace(i, 6, "\"");
00336 
00337     return in;
00338 }
00339 
00345 string
00346 esc2underscore(string s)
00347 {
00348     string::size_type pos;
00349     while ((pos = s.find('%')) != string::npos)
00350         s.replace(pos, 3, "_");
00351 
00352     return s;
00353 }
00354 
00355 
00359 string
00360 escattr(string s)
00361 {
00362     const string printable = " ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789~`!@#$%^&*()_-+={[}]|\\:;<,>.?/'\"";
00363     const string ESC = "\\";
00364     const string DOUBLE_ESC = ESC + ESC;
00365     const string QUOTE = "\"";
00366     const string ESCQUOTE = ESC + QUOTE;
00367 
00368     // escape non-printing characters with octal escape
00369     string::size_type ind = 0;
00370     while ((ind = s.find_first_not_of(printable, ind)) != s.npos)
00371         s.replace(ind, 1, ESC + octstring(s[ind]));
00372 
00373     // escape \ with a second backslash
00374     ind = 0;
00375     while ((ind = s.find(ESC, ind)) != s.npos) {
00376         s.replace(ind, 1, DOUBLE_ESC);
00377         ind += DOUBLE_ESC.length();
00378     }
00379 
00380     // escape " with backslash
00381     ind = 0;
00382     while ((ind = s.find(QUOTE, ind)) != s.npos) {
00383         s.replace(ind, 1, ESCQUOTE);
00384         ind += ESCQUOTE.length();
00385     }
00386 
00387     return s;
00388 }
00389 
00398 string
00399 unescattr(string s)
00400 {
00401     Regex octal("\\\\[0-3][0-7][0-7]");  // matches 4 characters
00402     Regex esc_quote("\\\\\"");  // matches 3 characters
00403     Regex esc_esc("\\\\\\\\");      // matches 2 characters
00404     const string ESC = "\\";
00405     const string QUOTE = "\"";
00406     int matchlen;
00407     unsigned int index;
00408 
00409     DBG(cerr << "0XX" << s << "XXX" << endl);
00410     // unescape any escaped backslashes
00411     index = esc_esc.search(s.c_str(), s.length(), matchlen, 0);
00412     while (index < s.length()) {
00413         DBG(cerr << "1aXX" << s << "XXX index: " << index << endl);
00414         s.replace(index, 2, ESC);
00415         DBG(cerr << "1bXX" << s << "XXX index: " << index << endl);
00416         index = esc_esc.search(s.c_str(), s.length(), matchlen, 0);
00417     }
00418 
00419     // unescape any escaped double quote characters
00420     index = esc_quote.search(s.c_str(), s.length(), matchlen, 0);
00421     while (index < s.length()) {
00422         s.replace(index, 2, QUOTE);
00423         DBG(cerr << "2XX" << s << "XXX index: " << index << endl);
00424         index = esc_quote.search(s.c_str(), s.length(), matchlen, 0);
00425     }
00426 
00427     // unescape octal characters
00428     index = octal.search(s.c_str(), s.length(), matchlen, 0);
00429     while (index < s.length()) {
00430         s.replace(index, 4, unoctstring(s.substr(index + 1, 3)));
00431         DBG(cerr << "3XX" << s << "XXX index: " << index << endl);
00432         index = octal.search(s.c_str(), s.length(), matchlen, 0);
00433     }
00434 
00435     DBG(cerr << "4XX" << s << "XXX" << endl);
00436     return s;
00437 }
00438 
00439 string
00440 munge_error_message(string msg)
00441 {
00442     // First, add enclosing quotes if needed.
00443     if (*msg.begin() != '"')
00444         msg.insert(msg.begin(), '"');
00445     if (*(msg.end() - 1) != '"')
00446         msg += "\"";
00447 
00448     // Now escape any internal double quotes that aren't escaped.
00449     string::iterator miter;
00450     for (miter = msg.begin() + 1; miter != msg.end() - 1; miter++)
00451         if (*miter == '"' && *(miter - 1) != '\\')
00452             miter = msg.insert(miter, '\\');
00453 
00454     return msg;
00455 }
00456 
00461 string
00462 escape_double_quotes(string source)
00463 {
00464     string::size_type idx = 0;
00465     while((idx = source.find('\"', idx)) != string::npos) {
00466         source.replace(idx, 1, "\\\""); // a backslash and a double quote
00467         idx += 2;
00468     }
00469 
00470     return source;
00471 }
00472 
00478 string
00479 unescape_double_quotes(string source)
00480 {
00481     string::size_type idx = 0;
00482     while((idx = source.find("\\\"", idx)) != string::npos) {
00483         source.replace(idx, 2, "\""); // a backslash and a double quote
00484         ++idx;
00485     }
00486 
00487     return source;
00488 }
00489 
00490 } // namespace libdap
00491 

Generated on Mon Sep 6 2010 06:34:01 for libdap++ by  doxygen 1.7.1