Update of /cvsroot/htmlparser/htmlparser/src/org/htmlparser
In directory sc8-pr-cvs1:/tmp/cvs-serv2842/src/org/htmlparser
Modified Files:
Parser.java StringNode.java
Log Message:
added non breaking space converting decorator
Index: Parser.java
===================================================================
RCS file: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/Parser.java,v
retrieving revision 1.47
retrieving revision 1.48
diff -C2 -d -r1.47 -r1.48
*** Parser.java 25 Jun 2003 03:46:37 -0000 1.47
--- Parser.java 25 Jun 2003 05:03:08 -0000 1.48
***************
*** 194,197 ****
--- 194,203 ----
*/
private boolean shouldRemoveEscapeCharacters = false;
+
+ /**
+ * Flag to tell the parser to convert non breaking space
+ * (i.e. \u00a0) to a space (" "). If true, this will happen inside StringNode's toPlainTextString.
+ */
+ private boolean shouldConvertNonBreakingSpace = false;
/**
***************
*** 1239,1241 ****
--- 1245,1256 ----
return shouldRemoveEscapeCharacters;
}
+
+ public void setNonBreakSpaceConversion(boolean shouldConvertNonBreakSpace) {
+ this.shouldConvertNonBreakingSpace = shouldConvertNonBreakSpace;
+ }
+
+ public boolean shouldConvertNonBreakingSpace() {
+ return shouldConvertNonBreakingSpace;
+ }
+
}
Index: StringNode.java
===================================================================
RCS file: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/StringNode.java,v
retrieving revision 1.25
retrieving revision 1.26
diff -C2 -d -r1.25 -r1.26
*** StringNode.java 25 Jun 2003 03:56:20 -0000 1.25
--- StringNode.java 25 Jun 2003 05:03:08 -0000 1.26
***************
*** 60,64 ****
public static Node createStringNode(
StringBuffer textBuffer, int textBegin, int textEnd,
! boolean shouldDecode, boolean shouldRemoveEscapeCharacters) {
Node newNode = new StringNode(textBuffer, textBegin, textEnd);
if (shouldDecode)
--- 60,65 ----
public static Node createStringNode(
StringBuffer textBuffer, int textBegin, int textEnd,
! boolean shouldDecode, boolean shouldRemoveEscapeCharacters,
! boolean shouldConvertNonBlankSpace) {
Node newNode = new StringNode(textBuffer, textBegin, textEnd);
if (shouldDecode)
***************
*** 66,69 ****
--- 67,72 ----
if (shouldRemoveEscapeCharacters)
newNode = new EscapeCharacterRemovingNode(newNode);
+ if (shouldConvertNonBlankSpace)
+ newNode = new NonBreakingSpaceConvertingNode(newNode);
return newNode;
}
|
