]> granicus.if.org Git - curl/commitdiff
example code by Michael Wallner
authorDaniel Stenberg <daniel@haxx.se>
Tue, 12 Sep 2006 07:54:55 +0000 (07:54 +0000)
committerDaniel Stenberg <daniel@haxx.se>
Tue, 12 Sep 2006 07:54:55 +0000 (07:54 +0000)
docs/examples/10-at-a-time.c [new file with mode: 0644]

diff --git a/docs/examples/10-at-a-time.c b/docs/examples/10-at-a-time.c
new file mode 100644 (file)
index 0000000..be3b6fa
--- /dev/null
@@ -0,0 +1,163 @@
+/*****************************************************************************
+ *                                  _   _ ____  _
+ *  Project                     ___| | | |  _ \| |
+ *                             / __| | | | |_) | |
+ *                            | (__| |_| |  _ <| |___
+ *                             \___|\___/|_| \_\_____|
+ *
+ * $Id$
+ *
+ * Example application source code using the multi interface to download many
+ * files, but with a capped maximum amount of simultaneous transfers.
+ *
+ * Written by Michael Wallner
+ */
+
+#include <stdlib.h>
+#include <curl/multi.h>
+
+static const char *urls[] = {
+  "http://www.microsoft.com",
+  "http://www.opensource.org",
+  "http://www.google.com",
+  "http://www.yahoo.com",
+  "http://www.ibm.com",
+  "http://www.mysql.com",
+  "http://www.oracle.com",
+  "http://www.ripe.net",
+  "http://www.iana.org",
+  "http://www.amazon.com",
+  "http://www.netcraft.com",
+  "http://www.heise.de",
+  "http://www.chip.de",
+  "http://www.ca.com",
+  "http://www.cnet.com",
+  "http://www.news.com",
+  "http://www.cnn.com",
+  "http://www.wikipedia.org",
+  "http://www.dell.com",
+  "http://www.hp.com",
+  "http://www.cert.org",
+  "http://www.mit.edu",
+  "http://www.nist.gov",
+  "http://www.ebay.com",
+  "http://www.playstation.com",
+  "http://www.uefa.com",
+  "http://www.ieee.org",
+  "http://www.apple.com",
+  "http://www.sony.com",
+  "http://www.symantec.com",
+  "http://www.zdnet.com",
+  "http://www.fujitsu.com",
+  "http://www.supermicro.com",
+  "http://www.hotmail.com",
+  "http://www.ecma.com",
+  "http://www.bbc.co.uk",
+  "http://news.google.com",
+  "http://www.foxnews.com",
+  "http://www.msn.com",
+  "http://www.wired.com",
+  "http://www.sky.com",
+  "http://www.usatoday.com",
+  "http://www.cbs.com",
+  "http://www.nbc.com",
+  "http://slashdot.org",
+  "http://www.bloglines.com",
+  "http://www.techweb.com",
+  "http://www.newslink.org",
+  "http://www.un.org",
+};
+
+#define MAX 10 /* number of simultaneous transfers */
+#define CNT sizeof(urls)/sizeof(char*) /* total number of transfers to do */
+
+static int cb(char *d, size_t n, size_t l, void *p)
+{
+  /* take care of the data here, ignored in this example */
+  (void)d;
+  (void)p;
+  return n*l;
+}
+
+static void init(CURLM *cm, int i)
+{
+  CURL *eh = curl_easy_init();
+
+  curl_easy_setopt(eh, CURLOPT_WRITEFUNCTION, cb);
+  curl_easy_setopt(eh, CURLOPT_HEADER, 0);
+  curl_easy_setopt(eh, CURLOPT_URL, urls[i]);
+  curl_easy_setopt(eh, CURLOPT_PRIVATE, urls[i]);
+  curl_easy_setopt(eh, CURLOPT_VERBOSE, 0);
+
+  curl_multi_add_handle(cm, eh);
+}
+
+int main(void)
+{
+  CURLM *cm;
+  CURLMsg *msg;
+  long L;
+  unsigned int C=0;
+  int M, Q, U = -1;
+  fd_set R, W, E;
+  struct timeval T;
+
+  curl_global_init(CURL_GLOBAL_ALL);
+
+  cm = curl_multi_init();
+
+  for (C = 0; C < MAX; ++C) {
+    init(cm, C);
+  }
+
+  while (U) {
+    while (CURLM_CALL_MULTI_PERFORM == curl_multi_perform(cm, &U));
+
+    if (U) {
+      FD_ZERO(&R);
+      FD_ZERO(&W);
+      FD_ZERO(&E);
+
+      if (curl_multi_fdset(cm, &R, &W, &E, &M)) {
+        fprintf(stderr, "E: curl_multi_fdset\n");
+        return EXIT_FAILURE;
+      }
+
+      if (curl_multi_timeout(cm, &L)) {
+        fprintf(stderr, "E: curl_multi_timeout\n");
+        return EXIT_FAILURE;
+      }
+
+      T.tv_sec = L/1000;
+      T.tv_usec = (L%1000)*1000;
+
+      if (0 > select(M+1, &R, &W, &E, &T)) {
+        fprintf(stderr, "E: select\n");
+        return EXIT_FAILURE;
+      }
+    }
+
+    while ((msg = curl_multi_info_read(cm, &Q))) {
+      if (msg->msg == CURLMSG_DONE) {
+        char *url;
+        CURL *e = msg->easy_handle;
+        curl_easy_getinfo(msg->easy_handle, CURLINFO_PRIVATE, &url);
+        fprintf(stderr, "R: %d - %s <%s>\n",
+                msg->data.result, curl_easy_strerror(msg->data.result), url);
+        curl_multi_remove_handle(cm, e);
+        curl_easy_cleanup(e);
+      }
+      else {
+        fprintf(stderr, "E: CURLMsg (%d)\n", msg->msg);
+      }
+      if (C < CNT) {
+        init(cm, C++);
+      }
+    }
+  }
+
+  curl_multi_cleanup(cm);
+  curl_global_cleanup();
+
+  return EXIT_SUCCESS;
+}