xmltooling/util/CurlURLInputStream.cpp

   1 /*
   2  * Licensed to the Apache Software Foundation (ASF) under one or more
   3  * contributor license agreements.  See the NOTICE file distributed with
   4  * this work for additional information regarding copyright ownership.
   5  * The ASF licenses this file to You under the Apache License, Version 2.0
   6  * (the "License"); you may not use this file except in compliance with
   7  * the License.  You may obtain a copy of the License at
   8  *
   9  *      http://www.apache.org/licenses/LICENSE-2.0
  10  *
  11  * Unless required by applicable law or agreed to in writing, software
  12  * distributed under the License is distributed on an "AS IS" BASIS,
  13  * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  14  * See the License for the specific language governing permissions and
  15  * limitations under the License.
  16  */
  17
  18 /*
  19  * $Id$
  20  */
  21
  22 #include "internal.h"
  23
  24 #include <curl/curl.h>
  25
  26 #include <xercesc/util/XercesDefs.hpp>
  27 #include <xercesc/util/XMLNetAccessor.hpp>
  28 #include <xercesc/util/XMLString.hpp>
  29 #include <xercesc/util/XMLExceptMsgs.hpp>
  30 #include <xercesc/util/Janitor.hpp>
  31 #include <xercesc/util/XMLUniDefs.hpp>
  32 #include <xercesc/util/TransService.hpp>
  33 #include <xercesc/util/TranscodingException.hpp>
  34 #include <xercesc/util/PlatformUtils.hpp>
  35
  36 #include <xmltooling/util/CurlURLInputStream.hpp>
  37
  38 using namespace xmltooling;
  39 using namespace xercesc;
  40
  41 CurlURLInputStream::CurlURLInputStream(const XMLURL& urlSource, const XMLNetHTTPInfo* httpInfo/*=0*/)
  42       : fMemoryManager(urlSource.getMemoryManager())
  43       , fURLSource(urlSource)
  44       , fURL(0)
  45       , fInputStream(NULL)
  46       , m_log(logging::Category::getInstance(XMLTOOLING_LOGCAT".libcurl.NetAccessor"))
  47 {
  48         // Get the text of the URL we're going to use
  49         fURL.reset(XMLString::transcode(fURLSource.getURLText(), fMemoryManager), fMemoryManager);
  50 }
  51
  52
  53 CurlURLInputStream::~CurlURLInputStream()
  54 {
  55     delete fInputStream;
  56 }
  57
  58
  59 size_t CurlURLInputStream::staticWriteCallback(void* ptr, size_t size, size_t nmemb, void* stream)
  60 {
  61     size_t len = size*nmemb;
  62     reinterpret_cast<std::stringstream*>(stream)->write(reinterpret_cast<const char*>(ptr),len);
  63     return len;
  64 }
  65
  66
  67 xsecsize_t CurlURLInputStream::readBytes(XMLByte* const toFill, const xsecsize_t maxToRead)
  68 {
  69     if (!fInputStream) {
  70         // Allocate the curl easy handle.
  71         CURL* fEasy = curl_easy_init();
  72         if (!fEasy)
  73             ThrowXMLwithMemMgr1(NetAccessorException, XMLExcepts::NetAcc_InternalError, "unable to allocate libcurl handle", fMemoryManager);
  74
  75         m_log.debug("libcurl trying to fetch %s", fURL.get());
  76
  77         // Set URL option
  78         curl_easy_setopt(fEasy, CURLOPT_URL, fURL.get());
  79         curl_easy_setopt(fEasy, CURLOPT_WRITEDATA, &fUnderlyingStream);
  80         curl_easy_setopt(fEasy, CURLOPT_WRITEFUNCTION, staticWriteCallback);
  81         curl_easy_setopt(fEasy, CURLOPT_CONNECTTIMEOUT, 30);
  82         curl_easy_setopt(fEasy, CURLOPT_TIMEOUT, 60);
  83         curl_easy_setopt(fEasy, CURLOPT_SSL_VERIFYHOST, 0);
  84         curl_easy_setopt(fEasy, CURLOPT_SSL_VERIFYPEER, 0);
  85         curl_easy_setopt(fEasy, CURLOPT_NOPROGRESS, 1);
  86         curl_easy_setopt(fEasy, CURLOPT_NOSIGNAL, 1);
  87         curl_easy_setopt(fEasy, CURLOPT_FAILONERROR, 1);
  88
  89         char curl_errorbuf[CURL_ERROR_SIZE];
  90         curl_errorbuf[0]=0;
  91         curl_easy_setopt(fEasy,CURLOPT_ERRORBUFFER,curl_errorbuf);
  92
  93         // Fetch the data.
  94         if (curl_easy_perform(fEasy) != CURLE_OK) {
  95             curl_easy_cleanup(fEasy);
  96             ThrowXMLwithMemMgr1(NetAccessorException, XMLExcepts::NetAcc_InternalError, curl_errorbuf, fMemoryManager);
  97         }
  98
  99         curl_easy_cleanup(fEasy);
 100
 101         /*
 102         switch (msg->data.result)
 103         {
 104         case CURLE_OK:
 105             // We completed successfully. runningHandles should have dropped to zero, so we'll bail out below...
 106             break;
 107
 108         case CURLE_UNSUPPORTED_PROTOCOL:
 109             ThrowXMLwithMemMgr(MalformedURLException, XMLExcepts::URL_UnsupportedProto, fMemoryManager);
 110             break;
 111
 112         case CURLE_COULDNT_RESOLVE_HOST:
 113         case CURLE_COULDNT_RESOLVE_PROXY:
 114             ThrowXMLwithMemMgr1(NetAccessorException,  XMLExcepts::NetAcc_TargetResolution, fURLSource.getHost(), fMemoryManager);
 115             break;
 116
 117         case CURLE_COULDNT_CONNECT:
 118             ThrowXMLwithMemMgr1(NetAccessorException, XMLExcepts::NetAcc_ConnSocket, fURLSource.getURLText(), fMemoryManager);
 119
 120         case CURLE_RECV_ERROR:
 121             ThrowXMLwithMemMgr1(NetAccessorException, XMLExcepts::NetAcc_ReadSocket, fURLSource.getURLText(), fMemoryManager);
 122             break;
 123
 124         default:
 125             m_log.error("curl NetAccessor encountered error from libcurl (%d)", msg->data.result);
 126             ThrowXMLwithMemMgr1(NetAccessorException, XMLExcepts::NetAcc_InternalError, fURLSource.getURLText(), fMemoryManager);
 127             break;
 128         }
 129         */
 130
 131         fInputStream = new (fMemoryManager) StreamInputSource::StreamBinInputStream(fUnderlyingStream);
 132     }
 133
 134     // Defer to the stream wrapper.
 135     return fInputStream->readBytes(toFill, maxToRead);
 136 }