CMakeLua/Utilities/cmcurl-7.19.0/docs/examples/htmltitle.cc

   1 /*****************************************************************************
   2  *                                  _   _ ____  _
   3  *  Project                     ___| | | |  _ \| |
   4  *                             / __| | | | |_) | |
   5  *                            | (__| |_| |  _ <| |___
   6  *                             \___|\___/|_| \_\_____|
   7  *
   8  * $Id: htmltitle.cc,v 1.1.1.1 2008-09-23 16:32:05 hoffman Exp $
   9  */
  10
  11 // Get a web page, parse it with libxml.
  12 //
  13 // Written by Lars Nilsson
  14 //
  15 // GNU C++ compile command line suggestion (edit paths accordingly):
  16 //
  17 // g++ -Wall -I/opt/curl/include -I/opt/libxml/include/libxml2 htmltitle.cc \
  18 // -o htmltitle -L/opt/curl/lib -L/opt/libxml/lib -lcurl -lxml2
  19
  20 #include <stdio.h>
  21 #include <string.h>
  22 #include <stdlib.h>
  23 #include <string>
  24 #include <curl/curl.h>
  25 #include <libxml/HTMLparser.h>
  26
  27 //
  28 //  Case-insensitive string comparison
  29 //
  30
  31 #ifdef _MSC_VER
  32 #define COMPARE(a, b) (!stricmp((a), (b)))
  33 #else
  34 #define COMPARE(a, b) (!strcasecmp((a), (b)))
  35 #endif
  36
  37 //
  38 //  libxml callback context structure
  39 //
  40
  41 struct Context
  42 {
  43   Context(): addTitle(false) { }
  44
  45   bool addTitle;
  46   std::string title;
  47 };
  48
  49 //
  50 //  libcurl variables for error strings and returned data
  51
  52 static char errorBuffer[CURL_ERROR_SIZE];
  53 static std::string buffer;
  54
  55 //
  56 //  libcurl write callback function
  57 //
  58
  59 static int writer(char *data, size_t size, size_t nmemb,
  60                   std::string *writerData)
  61 {
  62   if (writerData == NULL)
  63     return 0;
  64
  65   writerData->append(data, size*nmemb);
  66
  67   return size * nmemb;
  68 }
  69
  70 //
  71 //  libcurl connection initialization
  72 //
  73
  74 static bool init(CURL *&conn, char *url)
  75 {
  76   CURLcode code;
  77
  78   conn = curl_easy_init();
  79
  80   if (conn == NULL)
  81   {
  82     fprintf(stderr, "Failed to create CURL connection\n");
  83
  84     exit(EXIT_FAILURE);
  85   }
  86
  87   code = curl_easy_setopt(conn, CURLOPT_ERRORBUFFER, errorBuffer);
  88   if (code != CURLE_OK)
  89   {
  90     fprintf(stderr, "Failed to set error buffer [%d]\n", code);
  91
  92     return false;
  93   }
  94
  95   code = curl_easy_setopt(conn, CURLOPT_URL, url);
  96   if (code != CURLE_OK)
  97   {
  98     fprintf(stderr, "Failed to set URL [%s]\n", errorBuffer);
  99
 100     return false;
 101   }
 102
 103   code = curl_easy_setopt(conn, CURLOPT_FOLLOWLOCATION, 1L);
 104   if (code != CURLE_OK)
 105   {
 106     fprintf(stderr, "Failed to set redirect option [%s]\n", errorBuffer);
 107
 108     return false;
 109   }
 110
 111   code = curl_easy_setopt(conn, CURLOPT_WRITEFUNCTION, writer);
 112   if (code != CURLE_OK)
 113   {
 114     fprintf(stderr, "Failed to set writer [%s]\n", errorBuffer);
 115
 116     return false;
 117   }
 118
 119   code = curl_easy_setopt(conn, CURLOPT_WRITEDATA, &buffer);
 120   if (code != CURLE_OK)
 121   {
 122     fprintf(stderr, "Failed to set write data [%s]\n", errorBuffer);
 123
 124     return false;
 125   }
 126
 127   return true;
 128 }
 129
 130 //
 131 //  libxml start element callback function
 132 //
 133
 134 static void StartElement(void *voidContext,
 135                          const xmlChar *name,
 136                          const xmlChar **attributes)
 137 {
 138   Context *context = (Context *)voidContext;
 139
 140   if (COMPARE((char *)name, "TITLE"))
 141   {
 142     context->title = "";
 143     context->addTitle = true;
 144   }
 145 }
 146
 147 //
 148 //  libxml end element callback function
 149 //
 150
 151 static void EndElement(void *voidContext,
 152                        const xmlChar *name)
 153 {
 154   Context *context = (Context *)voidContext;
 155
 156   if (COMPARE((char *)name, "TITLE"))
 157     context->addTitle = false;
 158 }
 159
 160 //
 161 //  Text handling helper function
 162 //
 163
 164 static void handleCharacters(Context *context,
 165                              const xmlChar *chars,
 166                              int length)
 167 {
 168   if (context->addTitle)
 169     context->title.append((char *)chars, length);
 170 }
 171
 172 //
 173 //  libxml PCDATA callback function
 174 //
 175
 176 static void Characters(void *voidContext,
 177                        const xmlChar *chars,
 178                        int length)
 179 {
 180   Context *context = (Context *)voidContext;
 181
 182   handleCharacters(context, chars, length);
 183 }
 184
 185 //
 186 //  libxml CDATA callback function
 187 //
 188
 189 static void cdata(void *voidContext,
 190                   const xmlChar *chars,
 191                   int length)
 192 {
 193   Context *context = (Context *)voidContext;
 194
 195   handleCharacters(context, chars, length);
 196 }
 197
 198 //
 199 //  libxml SAX callback structure
 200 //
 201
 202 static htmlSAXHandler saxHandler =
 203 {
 204   NULL,
 205   NULL,
 206   NULL,
 207   NULL,
 208   NULL,
 209   NULL,
 210   NULL,
 211   NULL,
 212   NULL,
 213   NULL,
 214   NULL,
 215   NULL,
 216   NULL,
 217   NULL,
 218   StartElement,
 219   EndElement,
 220   NULL,
 221   Characters,
 222   NULL,
 223   NULL,
 224   NULL,
 225   NULL,
 226   NULL,
 227   NULL,
 228   NULL,
 229   cdata,
 230   NULL
 231 };
 232
 233 //
 234 //  Parse given (assumed to be) HTML text and return the title
 235 //
 236
 237 static void parseHtml(const std::string &html,
 238                       std::string &title)
 239 {
 240   htmlParserCtxtPtr ctxt;
 241   Context context;
 242
 243   ctxt = htmlCreatePushParserCtxt(&saxHandler, &context, "", 0, "",
 244                                   XML_CHAR_ENCODING_NONE);
 245
 246   htmlParseChunk(ctxt, html.c_str(), html.size(), 0);
 247   htmlParseChunk(ctxt, "", 0, 1);
 248
 249   htmlFreeParserCtxt(ctxt);
 250
 251   title = context.title;
 252 }
 253
 254 int main(int argc, char *argv[])
 255 {
 256   CURL *conn = NULL;
 257   CURLcode code;
 258   std::string title;
 259
 260   // Ensure one argument is given
 261
 262   if (argc != 2)
 263   {
 264     fprintf(stderr, "Usage: %s <url>\n", argv[0]);
 265
 266     exit(EXIT_FAILURE);
 267   }
 268
 269   curl_global_init(CURL_GLOBAL_DEFAULT);
 270
 271   // Initialize CURL connection
 272
 273   if (!init(conn, argv[1]))
 274   {
 275     fprintf(stderr, "Connection initializion failed\n");
 276
 277     exit(EXIT_FAILURE);
 278   }
 279
 280   // Retrieve content for the URL
 281
 282   code = curl_easy_perform(conn);
 283   curl_easy_cleanup(conn);
 284
 285   if (code != CURLE_OK)
 286   {
 287     fprintf(stderr, "Failed to get '%s' [%s]\n", argv[1], errorBuffer);
 288
 289     exit(EXIT_FAILURE);
 290   }
 291
 292   // Parse the (assumed) HTML code
 293
 294   parseHtml(buffer, title);
 295
 296   // Display the extracted title
 297
 298   printf("Title: %s\n", title.c_str());
 299
 300   return EXIT_SUCCESS;
 301 }