[Htmlparser-cvs] htmlparser/src/org/htmlparser/scanners BodyScanner.java,1.20,1.21 BulletScanner.jav
Brought to you by:
derrickoswald
Menu
▾
▴
[Htmlparser-cvs] htmlparser/src/org/htmlparser/scanners BodyScanner.java,1.20,1.21 BulletScanner.java,1.25,1.26 CompositeTagScanner.java,1.74,1.75 FormScanner.java,1.49,1.50 HeadScanner.java,1.17,1.18 LabelScanner.java,1.35,1.36 LinkScanner.java,1.57,1.58 OptionTagScanner.java,1.36,1.37 SelectTagScanner.java,1.34,1.35 TableColumnScanner.java,1.37,1.38 TableRowScanner.java,1.40,1.41 TableScanner.java,1.39,1.40 TextareaTagScanner.java,1.31,1.32 TitleScanner.java,1.33,1.34
Update of /cvsroot/htmlparser/htmlparser/src/org/htmlparser/scanners
In directory sc8-pr-cvs1:/tmp/cvs-serv15003/scanners
Modified Files:
BodyScanner.java BulletScanner.java CompositeTagScanner.java
FormScanner.java HeadScanner.java LabelScanner.java
LinkScanner.java OptionTagScanner.java SelectTagScanner.java
TableColumnScanner.java TableRowScanner.java TableScanner.java
TextareaTagScanner.java TitleScanner.java
Log Message:
Replaced isAllowSelfChildren() using tagEnders set.
Index: BodyScanner.java
===================================================================
RCS file: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/scanners/BodyScanner.java,v
retrieving revision 1.20
retrieving revision 1.21
diff -C2 -d -r1.20 -r1.21
*** BodyScanner.java 26 Oct 2003 19:46:19 -0000 1.20
--- BodyScanner.java 28 Oct 2003 10:31:02 -0000 1.21
***************
*** 43,47 ****
{
private static final String MATCH_NAME [] = {"BODY"};
- private static final String ENDERS [] = {};
private static final String END_TAG_ENDERS [] = {"HTML"};
--- 43,46 ----
***************
*** 53,57 ****
public BodyScanner(String filter)
{
! super(filter,MATCH_NAME,ENDERS,END_TAG_ENDERS,false);
}
--- 52,56 ----
public BodyScanner(String filter)
{
! super(filter,MATCH_NAME,MATCH_NAME,END_TAG_ENDERS);
}
Index: BulletScanner.java
===================================================================
RCS file: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/scanners/BulletScanner.java,v
retrieving revision 1.25
retrieving revision 1.26
diff -C2 -d -r1.25 -r1.26
*** BulletScanner.java 28 Oct 2003 03:04:18 -0000 1.25
--- BulletScanner.java 28 Oct 2003 10:31:02 -0000 1.26
***************
*** 54,58 ****
public BulletScanner(String filter)
{
! super(filter, MATCH_STRING, ENDERS, END_TAG_ENDERS, false);
}
--- 54,58 ----
public BulletScanner(String filter)
{
! super(filter, MATCH_STRING, ENDERS, END_TAG_ENDERS);
}
Index: CompositeTagScanner.java
===================================================================
RCS file: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/scanners/CompositeTagScanner.java,v
retrieving revision 1.74
retrieving revision 1.75
diff -C2 -d -r1.74 -r1.75
*** CompositeTagScanner.java 28 Oct 2003 03:04:18 -0000 1.74
--- CompositeTagScanner.java 28 Oct 2003 10:31:02 -0000 1.75
***************
*** 48,52 ****
* <li>Tags which will trigger a match</li>
* <li>Tags which when encountered before a legal end tag, should force a correction</li>
- * <li>Preventing more tags of its own type to appear as children
* </ul>
* Here are examples of each:<BR>
--- 48,51 ----
***************
*** 79,90 ****
* This is useful when you know that a certain tag can never hold children of its own type.
* e.g. <FORM> can never have more form tags within it. If it does, it is an error and should
! * be corrected. The default behavior is to allow nesting.
* <pre>
* MyScanner extends CompositeTagScanner {
* private static final String [] MATCH_IDS = { "FORM" };
- * private static final String [] ENDERS = {};
* private static final String [] END_TAG_ENDERS = { "BODY", "HTML" };
* MyScanner() {
! * super(MATCH_IDS, ENDERS,END_TAG_ENDERS, false);
* }
* ...
--- 78,88 ----
* This is useful when you know that a certain tag can never hold children of its own type.
* e.g. <FORM> can never have more form tags within it. If it does, it is an error and should
! * be corrected. Specify the tagEnders set to contain (at least) the match ids.
* <pre>
* MyScanner extends CompositeTagScanner {
* private static final String [] MATCH_IDS = { "FORM" };
* private static final String [] END_TAG_ENDERS = { "BODY", "HTML" };
* MyScanner() {
! * super(MATCH_IDS, MATCH_IDS, END_TAG_ENDERS, false);
* }
* ...
***************
*** 96,100 ****
{
protected String [] nameOfTagToMatch;
- private boolean allowSelfChildren;
protected Set tagEnderSet;
private Set endTagEnderSet;
--- 94,97 ----
***************
*** 111,127 ****
}
- public CompositeTagScanner(String [] nameOfTagToMatch, String [] tagEnders, boolean allowSelfChildren)
- {
- this("",nameOfTagToMatch,tagEnders,allowSelfChildren);
- }
-
public CompositeTagScanner(String filter, String [] nameOfTagToMatch)
{
! this(filter,nameOfTagToMatch,new String [] {},true);
! }
!
! public CompositeTagScanner(String filter, String [] nameOfTagToMatch, String [] tagEnders)
! {
! this(filter,nameOfTagToMatch,tagEnders,true);
}
--- 108,114 ----
}
public CompositeTagScanner(String filter, String [] nameOfTagToMatch)
{
! this(filter,nameOfTagToMatch,new String [] {});
}
***************
*** 129,136 ****
String filter,
String [] nameOfTagToMatch,
! String [] tagEnders,
! boolean allowSelfChildren)
{
! this(filter,nameOfTagToMatch,tagEnders,new String[] {}, allowSelfChildren);
}
--- 116,122 ----
String filter,
String [] nameOfTagToMatch,
! String [] tagEnders)
{
! this(filter,nameOfTagToMatch,tagEnders,new String[] {});
}
***************
*** 139,146 ****
String [] nameOfTagToMatch,
String [] tagEnders,
! String [] endTagEnders,
! boolean allowSelfChildren)
{
! this(filter,nameOfTagToMatch,tagEnders,endTagEnders, allowSelfChildren, false);
}
--- 125,131 ----
String [] nameOfTagToMatch,
String [] tagEnders,
! String [] endTagEnders)
{
! this(filter,nameOfTagToMatch,tagEnders,endTagEnders, false);
}
***************
*** 171,180 ****
String [] tagEnders,
String [] endTagEnders,
- boolean allowSelfChildren,
boolean balance_quotes)
{
super(filter);
this.nameOfTagToMatch = nameOfTagToMatch;
- this.allowSelfChildren = allowSelfChildren;
this.balance_quotes = balance_quotes;
this.tagEnderSet = new HashSet();
--- 156,163 ----
***************
*** 193,200 ****
* If it's not an empty XML tag, the lexer is repeatedly asked for
* subsequent nodes until an end tag is found or a node is encountered
! * that matches the tag ender set or end tag ender set, or a node of
! * the same type is found and {@link #isAllowSelfChildren} returns
! * <code>false</code>. In all but the first case, a virtual end tag
! * is created. Each node found that is not the end tag is added to
* the list of children.<p>
* The scanner's {@link #createTag} method is called with details about
--- 176,182 ----
* If it's not an empty XML tag, the lexer is repeatedly asked for
* subsequent nodes until an end tag is found or a node is encountered
! * that matches the tag ender set or end tag ender set.
! * In the latter case, a virtual end tag is created.
! * Each node found that is not the end tag is added to
* the list of children.<p>
* The scanner's {@link #createTag} method is called with details about
***************
*** 213,216 ****
--- 195,199 ----
NodeList nodeList;
Tag endTag;
+ String match;
String name;
TagScanner scanner;
***************
*** 220,223 ****
--- 203,207 ----
nodeList = new NodeList ();
endTag = null;
+ match = tag.getTagName ();
if (tag.isEmptyXmlTag ())
***************
*** 234,248 ****
name = next.getTagName ();
// check for normal end tag
! if (next.isEndTag () && name.equals (tag.getTagName ()))
{
endTag = next;
node = null;
}
! else if (isTagToBeEndedFor (next) || // check DTD
! ( // check for child of same name not allowed
! !(next.isEndTag ()) &&
! !isAllowSelfChildren () &&
! name.equals (tag.getTagName ())
! ))
{
// insert a virtual end tag and backup one node
--- 218,227 ----
name = next.getTagName ();
// check for normal end tag
! if (next.isEndTag () && name.equals (match))
{
endTag = next;
node = null;
}
! else if (isTagToBeEndedFor (next)) // check DTD
{
// insert a virtual end tag and backup one node
***************
*** 338,341 ****
--- 317,321 ----
ret = false;
+
name = tag.getTagName ();
if (tag.isEndTag ())
***************
*** 345,353 ****
return (ret);
- }
-
- public final boolean isAllowSelfChildren()
- {
- return allowSelfChildren;
}
}
--- 325,328 ----
Index: FormScanner.java
===================================================================
RCS file: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/scanners/FormScanner.java,v
retrieving revision 1.49
retrieving revision 1.50
diff -C2 -d -r1.49 -r1.50
*** FormScanner.java 28 Oct 2003 03:04:18 -0000 1.49
--- FormScanner.java 28 Oct 2003 10:31:02 -0000 1.50
***************
*** 47,51 ****
public static final String PREVIOUS_DIRTY_LINK_MESSAGE="Encountered a form tag after an open link tag.\nThere should have been an end tag for the link before the form tag began.\nCorrecting this..";
private boolean linkScannerAlreadyOpen=false;
! private static final String [] formTagEnders = {"HTML","BODY"};
/**
--- 47,51 ----
public static final String PREVIOUS_DIRTY_LINK_MESSAGE="Encountered a form tag after an open link tag.\nThere should have been an end tag for the link before the form tag began.\nCorrecting this..";
private boolean linkScannerAlreadyOpen=false;
! private static final String [] formTagEnders = {"FORM","HTML","BODY"};
/**
***************
*** 64,68 ****
public FormScanner(String filter, Parser parser)
{
! super(filter,MATCH_ID,formTagEnders,false);
parser.addScanner(new InputTagScanner("-i"));
parser.addScanner(new TextareaTagScanner("-t"));
--- 64,68 ----
public FormScanner(String filter, Parser parser)
{
! super(filter,MATCH_ID,formTagEnders);
parser.addScanner(new InputTagScanner("-i"));
parser.addScanner(new TextareaTagScanner("-t"));
Index: HeadScanner.java
===================================================================
RCS file: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/scanners/HeadScanner.java,v
retrieving revision 1.17
retrieving revision 1.18
diff -C2 -d -r1.17 -r1.18
*** HeadScanner.java 26 Oct 2003 19:46:19 -0000 1.17
--- HeadScanner.java 28 Oct 2003 10:31:02 -0000 1.18
***************
*** 45,49 ****
{
private static final String MATCH_NAME [] = {"HEAD"};
! private static final String ENDERS [] = {"BODY"};
private static final String END_TAG_ENDERS [] = {"HTML"};
--- 45,49 ----
{
private static final String MATCH_NAME [] = {"HEAD"};
! private static final String ENDERS [] = {"HEAD","BODY"};
private static final String END_TAG_ENDERS [] = {"HTML"};
***************
*** 55,59 ****
public HeadScanner(String filter)
{
! super(filter,MATCH_NAME,ENDERS,END_TAG_ENDERS,false);
}
--- 55,59 ----
public HeadScanner(String filter)
{
! super(filter,MATCH_NAME,ENDERS,END_TAG_ENDERS);
}
Index: LabelScanner.java
===================================================================
RCS file: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/scanners/LabelScanner.java,v
retrieving revision 1.35
retrieving revision 1.36
diff -C2 -d -r1.35 -r1.36
*** LabelScanner.java 26 Oct 2003 19:46:20 -0000 1.35
--- LabelScanner.java 28 Oct 2003 10:31:02 -0000 1.36
***************
*** 43,51 ****
public LabelScanner() {
! super(MATCH_NAME,new String [] {},false);
}
public LabelScanner(String filter) {
! super(filter,MATCH_NAME,new String [] {},false);
}
--- 43,51 ----
public LabelScanner() {
! super(MATCH_NAME,MATCH_NAME);
}
public LabelScanner(String filter) {
! super(filter,MATCH_NAME,MATCH_NAME);
}
Index: LinkScanner.java
===================================================================
RCS file: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/scanners/LinkScanner.java,v
retrieving revision 1.57
retrieving revision 1.58
diff -C2 -d -r1.57 -r1.58
*** LinkScanner.java 27 Oct 2003 02:18:04 -0000 1.57
--- LinkScanner.java 28 Oct 2003 10:31:02 -0000 1.58
***************
*** 54,58 ****
public static final String LINK_SCANNER_ID = "A";
public LinkProcessor processor;
! private final static String ENDERS [] = { "TD","TR","FORM","LI","BODY", "HTML" };
private final static String ENDTAG_ENDERS [] = { "TD","TR","FORM","LI","BODY", "HTML" };
--- 54,58 ----
public static final String LINK_SCANNER_ID = "A";
public LinkProcessor processor;
! private final static String ENDERS [] = { "A","TD","TR","FORM","LI","BODY", "HTML" };
private final static String ENDTAG_ENDERS [] = { "TD","TR","FORM","LI","BODY", "HTML" };
***************
*** 68,72 ****
*/
public LinkScanner(String filter) {
! super(filter,MATCH_NAME,ENDERS,ENDTAG_ENDERS, false);
processor = new LinkProcessor();
}
--- 68,72 ----
*/
public LinkScanner(String filter) {
! super(filter,MATCH_NAME,ENDERS,ENDTAG_ENDERS);
processor = new LinkProcessor();
}
Index: OptionTagScanner.java
===================================================================
RCS file: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/scanners/OptionTagScanner.java,v
retrieving revision 1.36
retrieving revision 1.37
diff -C2 -d -r1.36 -r1.37
*** OptionTagScanner.java 28 Oct 2003 03:04:18 -0000 1.36
--- OptionTagScanner.java 28 Oct 2003 10:31:02 -0000 1.37
***************
*** 44,48 ****
public OptionTagScanner(String filter) {
! super(filter, MATCH_NAME, ENDERS, END_TAG_ENDERS, false);
}
--- 44,48 ----
public OptionTagScanner(String filter) {
! super(filter, MATCH_NAME, ENDERS, END_TAG_ENDERS);
}
Index: SelectTagScanner.java
===================================================================
RCS file: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/scanners/SelectTagScanner.java,v
retrieving revision 1.34
retrieving revision 1.35
diff -C2 -d -r1.34 -r1.35
*** SelectTagScanner.java 28 Oct 2003 03:04:18 -0000 1.34
--- SelectTagScanner.java 28 Oct 2003 10:31:02 -0000 1.35
***************
*** 48,52 ****
public SelectTagScanner(String filter)
{
! super(filter, MATCH_NAME, ENDERS, END_TAG_ENDERS, false);
}
--- 48,52 ----
public SelectTagScanner(String filter)
{
! super(filter, MATCH_NAME, ENDERS, END_TAG_ENDERS);
}
Index: TableColumnScanner.java
===================================================================
RCS file: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/scanners/TableColumnScanner.java,v
retrieving revision 1.37
retrieving revision 1.38
diff -C2 -d -r1.37 -r1.38
*** TableColumnScanner.java 26 Oct 2003 19:46:21 -0000 1.37
--- TableColumnScanner.java 28 Oct 2003 10:31:02 -0000 1.38
***************
*** 44,48 ****
public TableColumnScanner(String filter) {
! this(filter, MATCH_STRING, new String[] {}, new String[] {}, false);
}
--- 44,48 ----
public TableColumnScanner(String filter) {
! this(filter, MATCH_STRING, MATCH_STRING, new String[] {});
}
***************
*** 51,63 ****
String[] nameOfTagToMatch,
String [] tagEnders,
! String [] endTagEnders,
! boolean allowSelfChildren) {
super(
filter,
nameOfTagToMatch,
tagEnders,
! endTagEnders,
! allowSelfChildren
! );
}
public Tag createTag(Page page, int start, int end, Vector attributes, Tag startTag, Tag endTag, NodeList children) throws ParserException
--- 51,60 ----
String[] nameOfTagToMatch,
String [] tagEnders,
! String [] endTagEnders) {
super(
filter,
nameOfTagToMatch,
tagEnders,
! endTagEnders);
}
public Tag createTag(Page page, int start, int end, Vector attributes, Tag startTag, Tag endTag, NodeList children) throws ParserException
Index: TableRowScanner.java
===================================================================
RCS file: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/scanners/TableRowScanner.java,v
retrieving revision 1.40
retrieving revision 1.41
diff -C2 -d -r1.40 -r1.41
*** TableRowScanner.java 26 Oct 2003 19:46:21 -0000 1.40
--- TableRowScanner.java 28 Oct 2003 10:31:02 -0000 1.41
***************
*** 46,50 ****
public TableRowScanner(String filter,Parser parser) {
! this(filter, parser, MATCH_STRING, new String[] {}, new String[] {}, false);
}
--- 46,50 ----
public TableRowScanner(String filter,Parser parser) {
! this(filter, parser, MATCH_STRING, MATCH_STRING, new String[] {});
}
***************
*** 54,66 ****
String[] nameOfTagToMatch,
String [] tagEnders,
! String [] endTagEnders,
! boolean allowSelfChildren) {
super(
filter,
nameOfTagToMatch,
tagEnders,
! endTagEnders,
! allowSelfChildren
! );
parser.addScanner(new TableColumnScanner());
}
--- 54,63 ----
String[] nameOfTagToMatch,
String [] tagEnders,
! String [] endTagEnders) {
super(
filter,
nameOfTagToMatch,
tagEnders,
! endTagEnders);
parser.addScanner(new TableColumnScanner());
}
Index: TableScanner.java
===================================================================
RCS file: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/scanners/TableScanner.java,v
retrieving revision 1.39
retrieving revision 1.40
diff -C2 -d -r1.39 -r1.40
*** TableScanner.java 26 Oct 2003 19:46:21 -0000 1.39
--- TableScanner.java 28 Oct 2003 10:31:02 -0000 1.40
***************
*** 48,54 ****
public TableScanner(Parser parser,String filter) {
! super(filter, MATCH_STRING, ENDERS, ENDTAG_ENDERS, true);
parser.addScanner(new TableRowScanner(parser));
-
}
--- 48,53 ----
public TableScanner(Parser parser,String filter) {
! super(filter, MATCH_STRING, ENDERS, ENDTAG_ENDERS);
parser.addScanner(new TableRowScanner(parser));
}
Index: TextareaTagScanner.java
===================================================================
RCS file: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/scanners/TextareaTagScanner.java,v
retrieving revision 1.31
retrieving revision 1.32
diff -C2 -d -r1.31 -r1.32
*** TextareaTagScanner.java 28 Oct 2003 03:04:18 -0000 1.31
--- TextareaTagScanner.java 28 Oct 2003 10:31:02 -0000 1.32
***************
*** 45,49 ****
public TextareaTagScanner(String filter)
{
! super(filter, MATCH_NAME, ENDERS, END_TAG_ENDERS, false);
}
--- 45,49 ----
public TextareaTagScanner(String filter)
{
! super(filter, MATCH_NAME, ENDERS, END_TAG_ENDERS);
}
Index: TitleScanner.java
===================================================================
RCS file: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/scanners/TitleScanner.java,v
retrieving revision 1.33
retrieving revision 1.34
diff -C2 -d -r1.33 -r1.34
*** TitleScanner.java 27 Oct 2003 02:18:04 -0000 1.33
--- TitleScanner.java 28 Oct 2003 10:31:02 -0000 1.34
***************
*** 42,50 ****
public class TitleScanner extends CompositeTagScanner {
private static final String MATCH_NAME [] = {"TITLE"};
! private static final String ENDERS [] = {"BODY"};
private static final String END_TAG_ENDERS [] = {"HEAD", "HTML"};
public TitleScanner(String filter) {
! super(filter,MATCH_NAME,ENDERS,END_TAG_ENDERS,false);
}
--- 42,50 ----
public class TitleScanner extends CompositeTagScanner {
private static final String MATCH_NAME [] = {"TITLE"};
! private static final String ENDERS [] = {"TITLE","BODY"};
private static final String END_TAG_ENDERS [] = {"HEAD", "HTML"};
public TitleScanner(String filter) {
! super(filter,MATCH_NAME,ENDERS,END_TAG_ENDERS);
}
|
