Merge pull request #17 from typeable/cdata

Skip over CDATA
ndmitchell · Aug 25, 2024 · f4c36c1 · f4c36c1
2 parents 9aae52a + be5b46e
commit f4c36c1
Show file tree

Hide file tree

Showing 2 changed files with 34 additions and 15 deletions.
diff --git a/cbits/hexml.c b/cbits/hexml.c
@@ -466,26 +466,44 @@ static str parse_content(document* d)
             // have found a </
             break;
         }
-        else if (peek_at(d, 1) == '!' && peek_at(d, 2) == '-' && peek_at(d, 3) == '-')
+        else if (peek_at(d, 1) == '!')
         {
-            skip(d, 3);
-            // you can't reuse the two '-' characters for the closing as well
-            if (peek_at(d, 0) == '\0' || peek_at(d, 1) == '\0')
+            if (peek_at(d, 2) == '-' && peek_at(d, 3) == '-')
             {
-                set_error(d, "Didn't get a closing comment");
-                return start_end(0, 0);
-            }
-            skip(d, 2);
-            for (;;)
-            {
-                if (!find(d, '>'))
+                skip(d, 3);
+                // you can't reuse the two '-' characters for the closing as well
+                if (peek_at(d, 0) == '\0' || peek_at(d, 1) == '\0')
                 {
                     set_error(d, "Didn't get a closing comment");
                     return start_end(0, 0);
                 }
-                skip(d, 1);
-                if (peek_at(d, -3) == '-' && peek_at(d, -2) == '-')
-                    break;
+                skip(d, 2);
+                for (;;)
+                {
+                    if (!find(d, '>'))
+                    {
+                        set_error(d, "Didn't get a closing comment");
+                        return start_end(0, 0);
+                    }
+                    skip(d, 1);
+                    if (peek_at(d, -3) == '-' && peek_at(d, -2) == '-')
+                        break;
+                }
+            } else if (d->end - d->cursor >= 9 && memcmp(d->cursor + 2, "[CDATA[", 7) == 0) {
+                skip(d, 9);
+                for (;;)
+                {
+                    if (!find(d, '>'))
+                    {
+                        set_error(d, "Didn't close CDATA");
+                        return start_end(0, 0);
+                    }
+                    skip(d, 1);
+                    if (peek_at(d, -3) == ']' && peek_at(d, -2) == ']')
+                        break;
+                }
+            } else {
+                parse_tag(d);
             }
         }
         else

diff --git a/src/Main.hs b/src/Main.hs
@@ -20,6 +20,7 @@ examples =
     ,(False, "<test")
     ,(True, "<?xml version=\"1.1\"?>\n<greeting>Hello, world!</greeting>")
     ,(True, "<foo bar.baz=\"qux\"></foo>")
+    ,(True, "<test><![CDATA[foo<x>]]</y>]><z>]baz]]></test>")
     ]
 
 main :: IO ()
@@ -80,5 +81,5 @@ rerender = inside
                     | otherwise = error "Invalid name"
         validAttr x | BS.notElem '\"' x = x
                     | otherwise = error "Invalid attribute"
-        validStr x | BS.notElem '<' x || BS.isInfixOf "<!--" x = x
+        validStr x | BS.notElem '<' x || BS.isInfixOf "<!--" x || BS.isInfixOf "<![CDATA[" x = x
                    | otherwise = error $ show ("Invalid string", x)