/* //////////////////////////////////////////////////////////////////////////////////////
* implementation
*/
tb_xml_reader_ref_t tb_xml_reader_init()
{
// init reader
tb_xml_reader_impl_t* reader = tb_malloc0_type(tb_xml_reader_impl_t);
tb_assert_and_check_return_val(reader, tb_null);
// init string
tb_string_init(&reader->text);
tb_string_init(&reader->version);
tb_string_init(&reader->charset);
tb_string_init(&reader->element);
tb_string_init(&reader->element_name);
tb_string_init(&reader->attribute_name);
tb_string_init(&reader->attribute_data);
tb_string_cstrcpy(&reader->version, "2.0");
tb_string_cstrcpy(&reader->charset, "utf-8");
// init attributes
tb_size_t i = 0;
for (i = 0; i < TB_XML_READER_ATTRIBUTES_MAXN; i++)
{
tb_xml_node_ref_t node = (tb_xml_node_ref_t)(reader->attributes + i);
tb_string_init(&node->name);
tb_string_init(&node->data);
}
// ok
return (tb_xml_reader_ref_t)reader;
}
/* //////////////////////////////////////////////////////////////////////////////////////
* implementation
*/
tb_option_ref_t tb_option_init(tb_char_t const* name, tb_char_t const* help, tb_option_item_t const* opts)
{
// check
tb_assert_and_check_return_val(name && opts, tb_null);
// done
tb_bool_t ok = tb_false;
tb_option_impl_t* impl = tb_null;
do
{
// make option
impl = tb_malloc0_type(tb_option_impl_t);
tb_assert_and_check_break(impl);
// init option
impl->opts = opts;
impl->list = tb_oc_dictionary_init(TB_OC_DICTIONARY_SIZE_MICRO, tb_false);
tb_assert_and_check_break(impl->list);
// init name
tb_strlcpy(impl->name, name, sizeof(impl->name));
// init help
if (!tb_string_init(&impl->help)) break;
if (help) tb_string_cstrcpy(&impl->help, help);
// ok
ok = tb_true;
} while (0);
// failed?
if (!ok)
{
// exit it
if (impl) tb_option_exit((tb_option_ref_t)impl);
impl = tb_null;
}
// ok?
return (tb_option_ref_t)impl;
}
tb_size_t tb_xml_reader_next(tb_xml_reader_ref_t reader)
{
// check
tb_xml_reader_impl_t* impl = (tb_xml_reader_impl_t*)reader;
tb_assert_and_check_return_val(impl && impl->rstream, TB_XML_READER_EVENT_NONE);
// reset event
impl->event = TB_XML_READER_EVENT_NONE;
// next
while (!impl->event)
{
// peek character
tb_char_t* pc = tb_null;
if (!tb_stream_need(impl->rstream, (tb_byte_t**)&pc, 1) || !pc) break;
// is element?
if (*pc == '<')
{
// parse element: <...>
tb_char_t const* element = tb_xml_reader_element_parse(impl);
tb_assert_and_check_break(element);
// is document begin: <?xml version="..." charset=".." ?>
tb_size_t size = tb_string_size(&impl->element);
if (size > 4 && !tb_strnicmp(element, "?xml", 4))
{
// update event
impl->event = TB_XML_READER_EVENT_DOCUMENT;
// update version & charset
tb_xml_node_ref_t attr = (tb_xml_node_ref_t)tb_xml_reader_attributes(reader);
for (; attr; attr = attr->next)
{
if (!tb_string_cstricmp(&attr->name, "version")) tb_string_strcpy(&impl->version, &attr->data);
if (!tb_string_cstricmp(&attr->name, "encoding")) tb_string_strcpy(&impl->charset, &attr->data);
}
// transform stream => utf-8
if (tb_string_cstricmp(&impl->charset, "utf-8") && tb_string_cstricmp(&impl->charset, "utf8"))
{
// charset
tb_size_t charset = TB_CHARSET_TYPE_UTF8;
if (!tb_string_cstricmp(&impl->charset, "gb2312") || !tb_string_cstricmp(&impl->charset, "gbk"))
charset = TB_CHARSET_TYPE_GB2312;
else tb_trace_e("the charset: %s is not supported", tb_string_cstr(&impl->charset));
// init transform stream
if (charset != TB_CHARSET_TYPE_UTF8)
{
#ifdef TB_CONFIG_MODULE_HAVE_CHARSET
// init the filter stream
if (!impl->fstream) impl->fstream = tb_stream_init_filter_from_charset(impl->istream, charset, TB_CHARSET_TYPE_UTF8);
else
{
// ctrl stream
if (!tb_stream_ctrl(impl->fstream, TB_STREAM_CTRL_FLTR_SET_STREAM, impl->istream)) break;
// the filter
tb_stream_filter_ref_t filter = tb_null;
if (!tb_stream_ctrl(impl->fstream, TB_STREAM_CTRL_FLTR_GET_FILTER, &filter)) break;
tb_assert_and_check_break(filter);
// ctrl filter
if (!tb_stream_filter_ctrl(filter, TB_STREAM_FILTER_CTRL_CHARSET_SET_FTYPE, charset)) break;
}
// open the filter stream
if (impl->fstream && tb_stream_open(impl->fstream))
impl->rstream = impl->fstream;
tb_string_cstrcpy(&impl->charset, "utf-8");
#else
// trace
tb_trace_e("unicode type is not supported, please enable charset module config if you want to use it!");
#endif
}
}
}
// is document type: <!DOCTYPE ... >
else if (size > 8 && !tb_strnicmp(element, "!DOCTYPE", 8))
{
// update event
impl->event = TB_XML_READER_EVENT_DOCUMENT_TYPE;
}
// is element end: </name>
else if (size > 1 && element[0] == '/')
{
// check
tb_check_break(impl->level);
// update event
impl->event = TB_XML_READER_EVENT_ELEMENT_END;
// leave
impl->level--;
}
// is comment: <!-- text -->
else if (size >= 3 && !tb_strncmp(element, "!--", 3))
{
// no comment end?
if (element[size - 2] != '-' || element[size - 1] != '-')
{
// patch '>'
tb_string_chrcat(&impl->element, '>');
// seek to comment end
tb_char_t ch = '\0';
tb_int_t n = 0;
while ((ch = tb_stream_bread_s8(impl->rstream)))
{
// -->
if (n == 2 && ch == '>') break;
else
{
// append it
tb_string_chrcat(&impl->element, ch);
if (ch == '-') n++;
else n = 0;
}
}
// update event
if (ch != '\0') impl->event = TB_XML_READER_EVENT_COMMENT;
}
else impl->event = TB_XML_READER_EVENT_COMMENT;
}
// is cdata: <![CDATA[ text ]]>
else if (size >= 8 && !tb_strnicmp(element, "![CDATA[", 8))
{
if (element[size - 2] != ']' || element[size - 1] != ']')
{
// patch '>'
tb_string_chrcat(&impl->element, '>');
// seek to cdata end
tb_char_t ch = '\0';
tb_int_t n = 0;
while ((ch = tb_stream_bread_s8(impl->rstream)))
{
// ]]>
if (n == 2 && ch == '>') break;
else
{
// append it
tb_string_chrcat(&impl->element, ch);
if (ch == ']') n++;
else n = 0;
}
}
// update event
if (ch != '\0') impl->event = TB_XML_READER_EVENT_CDATA;
}
else impl->event = TB_XML_READER_EVENT_CDATA;
}
// is empty element: <name/>
else if (size > 1 && element[size - 1] == '/')
{
// update event
impl->event = TB_XML_READER_EVENT_ELEMENT_EMPTY;
}
// is element begin: <name>
else
{
// update event
impl->event = TB_XML_READER_EVENT_ELEMENT_BEG;
// enter
impl->level++;
}
// trace
// tb_trace_d("<%s>", element);
}
// is text: <> text </>
else if (*pc)
{
// parse text: <> ... <>
tb_char_t const* text = tb_xml_reader_text_parse(impl);
if (text && tb_string_cstrcmp(&impl->text, "\r\n") && tb_string_cstrcmp(&impl->text, "\n"))
impl->event = TB_XML_READER_EVENT_TEXT;
// trace
// tb_trace_d("%s", text);
}
else
{
// skip the invalid character
if (!tb_stream_skip(impl->rstream, 1)) break;
}
}
// ok?
return impl->event;
}
