formdata.c

harvest是一个下载html网页得机器人

            /* CMC: Added support for buffer uploads */
            char * buffer, long bufferlength,

            char *contenttype,
            long flags,
            struct curl_slist* contentHeader,
            char *showfilename,
            struct curl_httppost *parent_post,
            struct curl_httppost **httppost,
            struct curl_httppost **last_post)
{
  struct curl_httppost *post;
  post = (struct curl_httppost *)malloc(sizeof(struct curl_httppost));
  if(post) {
    memset(post, 0, sizeof(struct curl_httppost));
    post->name = name;
    post->namelength = name?(namelength?namelength:(long)strlen(name)):0;
    post->contents = value;
    post->contentslength = contentslength;

    /* CMC: Added support for buffer uploads */
    post->buffer = buffer;
    post->bufferlength = bufferlength;

    post->contenttype = contenttype;
    post->contentheader = contentHeader;
    post->showfilename = showfilename;
    post->flags = flags;
  }
  else
    return NULL;
  
  if (parent_post) {
    /* now, point our 'more' to the original 'more' */
    post->more = parent_post->more;
    
    /* then move the original 'more' to point to ourselves */
    parent_post->more = post;	         
  }
  else {
    /* make the previous point to this */
    if(*last_post)
      (*last_post)->next = post;
    else
      (*httppost) = post;
    
    (*last_post) = post;  
  }
  return post;
}

/***************************************************************************
 *
 * AddFormInfo()
 *	
 * Adds a FormInfo structure to the list presented by parent_form_info.
 *
 * Returns newly allocated FormInfo on success and NULL if malloc failed/
 * parent_form_info is NULL.
 *
 ***************************************************************************/
static FormInfo * AddFormInfo(char *value,
                              char *contenttype,
                              FormInfo *parent_form_info)
{
  FormInfo *form_info;
  form_info = (FormInfo *)malloc(sizeof(FormInfo));
  if(form_info) {
    memset(form_info, 0, sizeof(FormInfo));
    if (value)
      form_info->value = value;
    if (contenttype)
      form_info->contenttype = contenttype;
    form_info->flags = HTTPPOST_FILENAME;
  }
  else
    return NULL;
  
  if (parent_form_info) {
    /* now, point our 'more' to the original 'more' */
    form_info->more = parent_form_info->more;
    
    /* then move the original 'more' to point to ourselves */
    parent_form_info->more = form_info;
  }
  else
    return NULL;

  return form_info;
}

/***************************************************************************
 *
 * ContentTypeForFilename()
 *	
 * Provides content type for filename if one of the known types (else
 * (either the prevtype or the default is returned).
 *
 * Returns some valid contenttype for filename.
 *
 ***************************************************************************/
static const char * ContentTypeForFilename (const char *filename,
					    const char *prevtype)
{
  const char *contenttype = NULL;
  unsigned int i;
  /*
   * No type was specified, we scan through a few well-known
   * extensions and pick the first we match!
   */
  struct ContentType {
    const char *extension;
    const char *type;
  };
  static struct ContentType ctts[]={
    {".gif",  "image/gif"},
    {".jpg",  "image/jpeg"},
    {".jpeg", "image/jpeg"},
    {".txt",  "text/plain"},
    {".html", "text/html"}
  };
  
  if(prevtype)
    /* default to the previously set/used! */
    contenttype = prevtype;
  else
    /* It seems RFC1867 defines no Content-Type to default to
       text/plain so we don't actually need to set this: */
    contenttype = HTTPPOST_CONTENTTYPE_DEFAULT;
  
  for(i=0; i<sizeof(ctts)/sizeof(ctts[0]); i++) {
    if(strlen(filename) >= strlen(ctts[i].extension)) {
      if(strequal(filename +
		  strlen(filename) - strlen(ctts[i].extension),
		  ctts[i].extension)) {
	contenttype = ctts[i].type;
	break;
      }	      
    }
  }
  /* we have a contenttype by now */
  return contenttype;
}

/***************************************************************************
 *
 * AllocAndCopy()
 *	
 * Copies the data currently available under *buffer using newly allocated
 * buffer (that becomes *buffer). Uses buffer_length if not null, else
 * uses strlen to determine the length of the buffer to be copied
 *
 * Returns 0 on success and 1 if the malloc failed.
 *
 ***************************************************************************/
static int AllocAndCopy (char **buffer, int buffer_length)
{
  const char *src = *buffer;
  int length, add = 0;
  if (buffer_length)
    length = buffer_length;
  else {
    length = strlen(*buffer);
    add = 1;
  }
  *buffer = (char*)malloc(length+add);
  if (!*buffer)
    return 1;
  memcpy(*buffer, src, length);
  /* if length unknown do null termination */
  if (add)
    (*buffer)[length] = '\0';
  return 0;
}

/***************************************************************************
 *
 * FormAdd()
 *	
 * Stores a 'name=value' formpost parameter and builds the appropriate
 * linked list.
 *
 * Has two principal functionalities: using files and byte arrays as
 * post parts. Byte arrays are either copied or just the pointer is stored
 * (as the user requests) while for files only the filename and not the
 * content is stored.
 *
 * While you may have only one byte array for each name, multiple filenames
 * are allowed (and because of this feature CURLFORM_END is needed after
 * using CURLFORM_FILE).
 *
 * Examples:
 *
 * Simple name/value pair with copied contents:
 * curl_formadd (&post, &last, CURLFORM_COPYNAME, "name",
 * CURLFORM_COPYCONTENTS, "value", CURLFORM_END);
 *
 * name/value pair where only the content pointer is remembered:
 * curl_formadd (&post, &last, CURLFORM_COPYNAME, "name",
 * CURLFORM_PTRCONTENTS, ptr, CURLFORM_CONTENTSLENGTH, 10, CURLFORM_END);
 * (if CURLFORM_CONTENTSLENGTH is missing strlen () is used)
 *
 * storing a filename (CONTENTTYPE is optional!):
 * curl_formadd (&post, &last, CURLFORM_COPYNAME, "name",
 * CURLFORM_FILE, "filename1", CURLFORM_CONTENTTYPE, "plain/text",
 * CURLFORM_END);
 *
 * storing multiple filenames:
 * curl_formadd (&post, &last, CURLFORM_COPYNAME, "name",
 * CURLFORM_FILE, "filename1", CURLFORM_FILE, "filename2", CURLFORM_END);
 *
 * Returns:
 * CURL_FORMADD_OK             on success
 * CURL_FORMADD_MEMORY         if the FormInfo allocation fails
 * CURL_FORMADD_OPTION_TWICE   if one option is given twice for one Form
 * CURL_FORMADD_NULL           if a null pointer was given for a char
 * CURL_FORMADD_MEMORY         if the allocation of a FormInfo struct failed
 * CURL_FORMADD_UNKNOWN_OPTION if an unknown option was used
 * CURL_FORMADD_INCOMPLETE     if the some FormInfo is not complete (or an error)
 * CURL_FORMADD_MEMORY         if a HttpPost struct cannot be allocated
 * CURL_FORMADD_MEMORY         if some allocation for string copying failed.
 * CURL_FORMADD_ILLEGAL_ARRAY  if an illegal option is used in an array
 *
 ***************************************************************************/

static
CURLFORMcode FormAdd(struct curl_httppost **httppost,
                     struct curl_httppost **last_post,
                     va_list params)
{
  FormInfo *first_form, *current_form, *form;
  CURLFORMcode return_value = CURL_FORMADD_OK;
  const char *prevtype = NULL;
  struct curl_httppost *post = NULL;
  CURLformoption option;
  struct curl_forms *forms = NULL;
  char *array_value=NULL; /* value read from an array */

  /* This is a state variable, that if TRUE means that we're parsing an
     array that we got passed to us. If FALSE we're parsing the input
     va_list arguments. */
  bool array_state = FALSE;

  /*
   * We need to allocate the first struct to fill in.
   */
  first_form = (FormInfo *)malloc(sizeof(struct FormInfo));
  if(first_form) {
    memset(first_form, 0, sizeof(FormInfo));
    current_form = first_form;
  }
  else
    return CURL_FORMADD_MEMORY;

  /*
   * Loop through all the options set.
   */
  while (1) {

    /* break if we have an error to report */
    if (return_value != CURL_FORMADD_OK)
      break;

    /* first see if we have more parts of the array param */
    if ( array_state ) {
      /* get the upcoming option from the given array */
      option = forms->option;
      array_value = (char *)forms->value;

      forms++; /* advance this to next entry */
      if (CURLFORM_END == option) {
        /* end of array state */
        array_state = FALSE;
        continue;
      }
    }
    else {
      /* This is not array-state, get next option */
      option = va_arg(params, CURLformoption);
      if (CURLFORM_END == option)
        break;
    }

    switch (option) {
    case CURLFORM_ARRAY:
      if(array_state)
        /* we don't support an array from within an array */
        return_value = CURL_FORMADD_ILLEGAL_ARRAY;
      else {
        forms = va_arg(params, struct curl_forms *);
        if (forms)
          array_state = TRUE;
        else
          return_value = CURL_FORMADD_NULL;
      }
      break;

      /*
       * Set the Name property.
       */
    case CURLFORM_PTRNAME:
      current_form->flags |= HTTPPOST_PTRNAME; /* fall through */
    case CURLFORM_COPYNAME:
      if (current_form->name)
        return_value = CURL_FORMADD_OPTION_TWICE;
      else {
        char *name = array_state?
          array_value:va_arg(params, char *);
        if (name)
          current_form->name = name; /* store for the moment */
        else
          return_value = CURL_FORMADD_NULL;
      }
      break;
    case CURLFORM_NAMELENGTH:
      if (current_form->namelength)
        return_value = CURL_FORMADD_OPTION_TWICE;
      else
        current_form->namelength =
          array_state?(long)array_value:va_arg(params, long);
      break;

      /*
       * Set the contents property.
       */
    case CURLFORM_PTRCONTENTS:
      current_form->flags |= HTTPPOST_PTRCONTENTS; /* fall through */
    case CURLFORM_COPYCONTENTS:
      if (current_form->value)
        return_value = CURL_FORMADD_OPTION_TWICE;
      else {
        char *value =
          array_state?array_value:va_arg(params, char *);
        if (value)
          current_form->value = value; /* store for the moment */
        else
          return_value = CURL_FORMADD_NULL;
      }
      break;
    case CURLFORM_CONTENTSLENGTH:
      if (current_form->contentslength)
        return_value = CURL_FORMADD_OPTION_TWICE;
      else
        current_form->contentslength =
          array_state?(long)array_value:va_arg(params, long);
      break;

      /* Get contents from a given file name */
    case CURLFORM_FILECONTENT:
      if (current_form->flags != 0)
        return_value = CURL_FORMADD_OPTION_TWICE;
      else {
        char *filename = array_state?
          array_value:va_arg(params, char *);
        if (filename) {
          current_form->value = strdup(filename);
          current_form->flags |= HTTPPOST_READFILE;
        }
        else
          return_value = CURL_FORMADD_NULL;
      }
      break;

      /* We upload a file */
    case CURLFORM_FILE:
      {
        char *filename = array_state?array_value:
          va_arg(params, char *);

        if (current_form->value) {
          if (current_form->flags & HTTPPOST_FILENAME) {
            if (filename) {
              if (!(current_form = AddFormInfo(strdup(filename),
                                               NULL, current_form)))
                return_value = CURL_FORMADD_MEMORY;
            }
            else
              return_value = CURL_FORMADD_NULL;
          }
          else
            return_value = CURL_FORMADD_OPTION_TWICE;
        }
        else {
          if (filename)
            current_form->value = strdup(filename);
          else
            return_value = CURL_FORMADD_NULL;
          current_form->flags |= HTTPPOST_FILENAME;
        }
        break;
      }

    /* CMC: Added support for buffer uploads */
    case CURLFORM_BUFFER:
      {
        char *filename = array_state?array_value:
          va_arg(params, char *);

        if (current_form->value) {
          if (current_form->flags & HTTPPOST_BUFFER) {