getbodytext.c

/*
 * Print entire HTML text after processed JavaScript
 *
 * build:
 *   FLAGS=`pkg-config --cflags --libs gtk+-x11-2.0 glib-2.0 webkit-1.0`
 *   gcc -Wall $FLAGS getbodytext.c -o getbodytext
 *
 * usage:
 *   /usr/bin/xvfb-run ./getbodytext test.html
 *
 */

#include <glib.h>
#include <glib/gprintf.h>
#include <gtk/gtk.h>
#include <gdk/gdk.h>
#include <webkit/webkit.h>
#include <JavaScriptCore/JavaScript.h>

static const gchar code[] =
  "(function () {"
  "  var scripts = document.getElementsByTagName(\"script\");"
  "  for (var i = 0; i < scripts.length; i += 1) {"
  "     scripts[i].innerHTML = \"\";"
  "  }"
  "  var range = document.createRange();"
  "  range.selectNodeContents(document.body);"
  "  return range.toString();"
  "})()";
/* const gchar code[] = "document.body.innerHTML"; */


static GString *
toGString(JSStringRef jsstr)
{
  size_t size;
  gchar * buf;

  size = JSStringGetMaximumUTF8CStringSize(jsstr);
  buf = g_alloca(size);
  JSStringGetUTF8CString(jsstr, buf, size);
  return g_string_new(buf);
}
/*
 * see: http://webkitgtk.org/reference/webkitgtk-WebKitWebView.html signals
 */
static void
loaded(WebKitWebView * view, WebKitWebFrame * frame, gpointer data)
{
  /*
  const gchar * title;
  */
  JSGlobalContextRef gc;
  JSStringRef script;
  JSValueRef result;

  /*
  title = webkit_web_frame_get_title(frame);
  g_printf("%s\n", title);
  */
  gc = webkit_web_frame_get_global_context(frame);

  script = JSStringCreateWithUTF8CString(code);
  result = JSEvaluateScript(gc, script, NULL, NULL, 0, NULL);
  JSStringRelease(script);

  if (JSValueIsString(gc, result)) {
    JSStringRef jsstr;
    GString * gstr;

    jsstr = JSValueToStringCopy(gc, result, NULL);
    gstr = toGString(jsstr);
    JSStringRelease(jsstr);
    g_printf("%s\n", gstr->str);
    g_string_free(gstr, TRUE);
  }

  gtk_main_quit();
}

int
main(int argc, char * argv[])
{
  GtkWidget * window;
  GtkWidget * webview;
  GString * url;

  g_thread_init(NULL);
  gdk_threads_init();
  gdk_threads_enter();
  gtk_init(&argc, &argv);

  url = g_string_new(argv[1]);
  if (!g_str_has_prefix(url->str, "http")) {
    url = g_string_prepend(url, "file://");
  }

  window = gtk_window_new(GTK_WINDOW_TOPLEVEL);
  webview = webkit_web_view_new();
  gtk_container_add(GTK_CONTAINER(window), webview);

  gtk_signal_connect(GTK_OBJECT(webview), "load-finished",
                     GTK_SIGNAL_FUNC(loaded), NULL);

  webkit_web_view_open(WEBKIT_WEB_VIEW(webview), url->str);
  g_string_free(url, TRUE);

  gtk_widget_show_all(window);
  gtk_main();
  gdk_threads_leave();
  return 0;
}