Import hubbub -- an HTML parsing library.

Plenty of work still to do (like tree generation ;)

svn path=/trunk/hubbub/; revision=3359

1 files changed, 71 insertions, 0 deletions

diff --git a/src/tokeniser/tokeniser.h b/src/tokeniser/tokeniser.h
new file mode 100644
index 0000000..20bbe20
--- /dev/null
+++ b/src/tokeniser/tokeniser.h
@@ -0,0 +1,71 @@
+/*
+ * This file is part of Hubbub.
+ * Licensed under the MIT License,
+ *                http://www.opensource.org/licenses/mit-license.php
+ * Copyright 2007 John-Mark Bell <jmb@netsurf-browser.org>
+ */
+
+#ifndef hubbub_tokeniser_tokeniser_h_
+#define hubbub_tokeniser_tokeniser_h_
+
+#include <stdbool.h>
+#include <inttypes.h>
+
+#include <hubbub/errors.h>
+#include <hubbub/functypes.h>
+#include <hubbub/types.h>
+
+#include "input/inputstream.h"
+
+typedef struct hubbub_tokeniser hubbub_tokeniser;
+
+/**
+ * Hubbub tokeniser option types
+ */
+typedef enum hubbub_tokeniser_opttype {
+	HUBBUB_TOKENISER_TOKEN_HANDLER,
+	HUBBUB_TOKENISER_BUFFER_HANDLER,
+	HUBBUB_TOKENISER_ERROR_HANDLER,
+	HUBBUB_TOKENISER_CONTENT_MODEL,
+} hubbub_tokeniser_opttype;
+
+/**
+ * Hubbub tokeniser option parameters
+ */
+typedef union hubbub_tokeniser_optparams {
+	struct {
+		hubbub_token_handler handler;
+		void *pw;
+	} token_handler;
+
+	struct {
+		hubbub_buffer_handler handler;
+		void *pw;
+	} buffer_handler;
+
+	struct {
+		hubbub_error_handler handler;
+		void *pw;
+	} error_handler;
+
+	struct {
+		hubbub_content_model model;
+	} content_model;
+} hubbub_tokeniser_optparams;
+
+/* Create a hubbub tokeniser */
+hubbub_tokeniser *hubbub_tokeniser_create(hubbub_inputstream *input,
+		hubbub_alloc alloc, void *pw);
+/* Destroy a hubbub tokeniser */
+void hubbub_tokeniser_destroy(hubbub_tokeniser *tokeniser);
+
+/* Configure a hubbub tokeniser */
+hubbub_error hubbub_tokeniser_setopt(hubbub_tokeniser *tokeniser,
+		hubbub_tokeniser_opttype type,
+		hubbub_tokeniser_optparams *params);
+
+/* Process remaining data in the input stream */
+hubbub_error hubbub_tokeniser_run(hubbub_tokeniser *tokeniser);
+
+#endif
+