Initial revision

[helm.git] / helm / DEVEL / pxp / netstring / nethtml.mli

diff --git a/helm/DEVEL/pxp/netstring/nethtml.mli b/helm/DEVEL/pxp/netstring/nethtml.mli
new file mode 100644 (file)
index 0000000..d7af381
--- /dev/null
+++ b/helm/DEVEL/pxp/netstring/nethtml.mli
@@ -0,0 +1,72 @@
+(* $Id$
+ * ----------------------------------------------------------------------
+ *
+ *)
+
+
+(* The type 'document' represents parsed HTML documents. 
+ * Element (name, args, subnodes): is an element node for an element of
+ *   type 'name' (i.e. written <name ...>...</name>) with arguments 'args'
+ *   and subnodes 'subnodes' (the material within the element). The arguments
+ *   are simply name/value pairs. Entity references (something like %xy;)
+ *   occuring in the values are NOT resolved.
+ *   Arguments without values (e.g. <select name="x" multiple>: here,
+ *   "multiple" is such an argument) are represented as (name,name), i.e. the
+ *   name is returned as value.
+ *   As argument names are case-insensitive, the names are all lowercase.
+ * Data s: is a character data node. Again, entity references are contained
+ *   as such and not as what they mean.
+ *)
+
+type document =
+    Element of (string  *  (string*string) list  *  document list)
+  | Data of string
+;;
+
+
+val no_end_tag : string list ref;;
+  (* List of tags which are always empty. This variable is pre-configured,
+   * but you may want to change it.
+   * It is important to know which elements are always empty, because HTML
+   * allows it to omit the end tag for them. For example, 
+   * <a><b>x</a> is parsed as 
+   *   Element("a",[],[ Element("b",[],[]); Data "x" ])
+   * if we know that "a" is an empty element, but it is wrongly parsed as
+   *   Element("a",[],[ Element("b",[], [ Data "x"]) ])
+   * if "a" is actually empty but we do not know it.
+   * An example of such a tag is "br".
+   *)
+
+val special_tag : string list ref;;
+  (* List of tags with a special rule for recognizing the end.
+   * This variable is pre-configured, but you may want to change it.
+   * The special rule is that the metacharacters '<', '>' and so on lose
+   * their meaning within the element, and that only the corresponding 
+   * end tag stops this kind of scanning. An example is the element
+   * "javascript". Inner elements are not recognized, and the element
+   * can only be ended by </javascript>. (Other elements are also ended
+   * if an embracing element ends, e.g. "j" in <k><j></k>!)
+   *
+   * Note that comments are not recognized within special elements;
+   * comments are returned as character material.
+   *)
+
+val parse_string : string -> document list
+  (* Parses the HTML document from a string and returns it *)
+
+val parse_file : in_channel -> document list
+  (* Parses the HTML document from a file and returns it *)
+
+
+(* ======================================================================
+ * History:
+ * 
+ * $Log$
+ * Revision 1.1  2000/11/17 09:57:28  lpadovan
+ * Initial revision
+ *
+ * Revision 1.1  2000/03/03 01:07:25  gerd
+ *     Initial revision.
+ *
+ * 
+ *)