stevedlawrence commented on code in PR #821:
URL: https://github.com/apache/daffodil/pull/821#discussion_r944810630
##########
daffodil-runtime1/src/main/scala/org/apache/daffodil/infoset/JDOMInfosetOutputter.scala:
##########
@@ -48,14 +48,57 @@ class JDOMInfosetOutputter extends InfosetOutputter
true
}
- def startSimple(diSimple: DISimple): Boolean = {
+ def startSimple(diSimple: DISimple, xmlOutputStyle: String): Boolean = {
val elem = createElement(diSimple)
+ val correctFormat = new StringBuilder("");
+ val cdataIntro = "<![CDATA["
+ val cdataOutro = "]]>"
+ var charEntMode: Boolean = false
if (diSimple.hasValue) {
val text =
if (diSimple.erd.optPrimType.get.isInstanceOf[NodeInfo.String.Kind]) {
- remapped(diSimple.dataValueAsString)
+ val s = remapped(diSimple.dataValueAsString)
+
+ if(xmlOutputStyle == "prettyPrintSafe"){
+ val newData = s.replaceAll(">",">").replace("\\r\\n",
"")
+ val readyForCDATA = newData.replaceAll("0x","&#x")
+ //Figure out what mode you have to be in
+ if(readyForCDATA(0) == '&') {
+ charEntMode = true
+ } else {
+ charEntMode = false
+ correctFormat.append(cdataIntro)
+ }
+
+ //Traverse over the string surrounding correct areas with CDATA
info
+ for(c <- readyForCDATA) {
+ if(charEntMode) {
+ correctFormat.append(c)
+ if(c == ';'){
+ correctFormat.append(cdataIntro)
+ charEntMode = false
+ }
+ } else {
+ if(c == '&'){
+ correctFormat.append(cdataOutro)
+ charEntMode = true
+ }
+ correctFormat.append(c)
+ }
+ }
+
+ //You are done with the string. If you are still a non
+ //char ent then close and finish up.
+ if(!charEntMode){
+ correctFormat.append(cdataOutro)
+ }
+
Review Comment:
In this example, is `&` actually in the data, or is the data actually
`something & else`. In either case though, aren't these both valid:
```xml
<![CDATA[something & else]]>
<![CDATA[something & else]]>
```
Aren't things like entities the xml tags (e.g. `<foo>`) not interpreted in
CDATA blocks? So this is valid too:
```xml
<![CDATA[blah <foo> bar <bazr>]]>
```
Doesn't CDATA tell the parser to ignore special characters and just treat
them as normal chars until it finds the closing `]]>`?
Seems like we want to take exactly what came out of the Daffodil parse, not
do any remapping of entities, and just wrap it in the CDATA brackets?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
