From 2d9228c5f1c6739e7a76f5b4af98eb3611f5220a Mon Sep 17 00:00:00 2001
From: mitchell <mitchell@frost.(none)>
Date: Sat, 31 May 2008 18:40:23 -0400
Subject: [PATCH] Added XML Schema parser.

---
 ext/ohcount_native/ragel_parser.c             |   2 +
 ext/ohcount_native/ragel_parsers/xmlschema.rl | 136 ++++++++++++++++++
 2 files changed, 138 insertions(+)
 create mode 100644 ext/ohcount_native/ragel_parsers/xmlschema.rl

diff --git a/ext/ohcount_native/ragel_parser.c b/ext/ohcount_native/ragel_parser.c
index af78ac1..51c9be7 100644
--- a/ext/ohcount_native/ragel_parser.c
+++ b/ext/ohcount_native/ragel_parser.c
@@ -47,6 +47,7 @@
 //#include "vim_parser.h"
 #include "xml_parser.h"
 #include "xslt_parser.h"
+#include "xmlschema_parser.h"
 // END parser includes
 
 ParseResult *pr;
@@ -107,6 +108,7 @@ struct language languages[] = {
   //{ "vim", parse_vim },
   { "xml", parse_xml },
   { "xslt", parse_xslt },
+  { "xmlschema", parse_xmlschema },
 // END languages
   { "", NULL }
 };
diff --git a/ext/ohcount_native/ragel_parsers/xmlschema.rl b/ext/ohcount_native/ragel_parsers/xmlschema.rl
new file mode 100644
index 0000000..0041416
--- /dev/null
+++ b/ext/ohcount_native/ragel_parsers/xmlschema.rl
@@ -0,0 +1,136 @@
+// xmlschema.rl written by Mitchell Foral. mitchell<att>caladbolg<dott>net.
+
+/************************* Required for every parser *************************/
+#ifndef RAGEL_XMLSCHEMA_PARSER
+#define RAGEL_XMLSCHEMA_PARSER
+
+#include "ragel_parser_macros.h"
+
+// the name of the language
+const char *XMLSCHEMA_LANG = "xmlschema";
+
+// the languages entities
+const char *xmlschema_entities[] = {
+  "space", "comment", "doctype",
+  "tag", "entity", "any"
+};
+
+// constants associated with the entities
+enum {
+  XMLSCHEMA_SPACE = 0, XMLSCHEMA_COMMENT, XMLSCHEMA_DOCTYPE,
+  XMLSCHEMA_TAG, XMLSCHEMA_ENTITY, XMLSCHEMA_ANY
+};
+
+/*****************************************************************************/
+
+%%{
+  machine xmlschema;
+  write data;
+  include common "common.rl";
+
+  # Line counting machine
+
+  action xmlschema_ccallback {
+    switch(entity) {
+    case XMLSCHEMA_SPACE:
+      ls
+      break;
+    case XMLSCHEMA_ANY:
+      code
+      break;
+    case INTERNAL_NL:
+      std_internal_newline(XMLSCHEMA_LANG)
+      break;
+    case NEWLINE:
+      std_newline(XMLSCHEMA_LANG)
+      break;
+    case CHECK_BLANK_ENTRY:
+      check_blank_entry(XMLSCHEMA_LANG)
+    }
+  }
+
+  xmlschema_comment =
+    '<!--' @comment (
+      newline %{ entity = INTERNAL_NL; } %xmlschema_ccallback
+      |
+      ws
+      |
+      (nonnewline - ws) @comment
+    )* :>> '-->';
+
+  xmlschema_sq_str =
+    '\'' @code (
+      newline %{ entity = INTERNAL_NL; } %xmlschema_ccallback
+      |
+      ws
+      |
+      [^\r\n\f\t '\\] @code
+      |
+      '\\' nonnewline @code
+    )* '\'';
+  xmlschema_dq_str =
+    '"' @code (
+      newline %{ entity = INTERNAL_NL; } %xmlschema_ccallback
+      |
+      ws
+      |
+      [^\r\n\f\t "\\] @code
+      |
+      '\\' nonnewline @code
+    )* '"';
+  xmlschema_cdata_str =
+    '<![CDATA[' @code (
+      newline %{ entity = INTERNAL_NL; } %xmlschema_ccallback
+      |
+      ws
+      |
+      (nonnewline - ws) @code
+    )* :>> ']]>';
+  xmlschema_string = xmlschema_sq_str | xmlschema_dq_str | xmlschema_cdata_str;
+
+  xmlschema_line := |*
+    spaces             ${ entity = XMLSCHEMA_SPACE; } => xmlschema_ccallback;
+    xmlschema_comment;
+    xmlschema_string;
+    newline            ${ entity = NEWLINE;         } => xmlschema_ccallback;
+    ^space             ${ entity = XMLSCHEMA_ANY;   } => xmlschema_ccallback;
+  *|;
+
+  # Entity machine
+
+  action xmlschema_ecallback {
+    callback(XMLSCHEMA_LANG, entity, cint(ts), cint(te));
+  }
+
+  xmlschema_entity := 'TODO:';
+}%%
+
+/************************* Required for every parser *************************/
+
+/* Parses a string buffer with XML Schema markup.
+ *
+ * @param *buffer The string to parse.
+ * @param length The length of the string to parse.
+ * @param count Integer flag specifying whether or not to count lines. If yes,
+ *   uses the Ragel machine optimized for counting. Otherwise uses the Ragel
+ *   machine optimized for returning entity positions.
+ * @param *callback Callback function. If count is set, callback is called for
+ *   every line of code, comment, or blank with 'lcode', 'lcomment', and
+ *   'lblank' respectively. Otherwise callback is called for each entity found.
+ */
+void parse_xmlschema(char *buffer, int length, int count,
+  void (*callback) (const char *lang, const char *entity, int start, int end)
+  ) {
+  init
+
+  %% write init;
+  cs = (count) ? xmlschema_en_xmlschema_line : xmlschema_en_xmlschema_entity;
+  %% write exec;
+
+  // if no newline at EOF; callback contents of last line
+  if (count) { process_last_line(XMLSCHEMA_LANG) }
+}
+
+#endif
+
+/*****************************************************************************/
-- 
2.32.0.93.g670b81a890