Provide basic experimental formatter which supports invalid XML

Fixes eclipse-lemminx#1195

Signed-off-by: azerr <azerr@redhat.com>

org.eclipse.lemminx/src/main/java/org/eclipse/lemminx/XMLLanguageServer.java

@@ -275,10 +275,6 @@ public XMLLanguageService getXMLLanguageService() {
 		return xmlLanguageService;
 	}
 
-	public SharedSettings getSettings() {
-		return xmlTextDocumentService.getSharedSettings();
-	}
-
 	public ScheduledFuture<?> schedule(Runnable command, int delay, TimeUnit unit) {
 		return delayer.schedule(command, delay, unit);
 	}
@@ -292,7 +288,7 @@ public long getParentProcessId() {
 	public CompletableFuture<AutoCloseTagResponse> closeTag(TextDocumentPositionParams params) {
 		return xmlTextDocumentService.computeDOMAsync(params.getTextDocument(), (xmlDocument, cancelChecker) -> {
 			return getXMLLanguageService().doAutoClose(xmlDocument, params.getPosition(),
-					getSettings().getCompletionSettings(), cancelChecker);
+					getSharedSettings().getCompletionSettings(), cancelChecker);
 		});
 	}
 

org.eclipse.lemminx/src/main/java/org/eclipse/lemminx/XMLTextDocumentService.java

@@ -312,25 +312,17 @@ public CompletableFuture<List<Either<SymbolInformation, DocumentSymbol>>> docume
 
 	@Override
 	public CompletableFuture<List<? extends TextEdit>> formatting(DocumentFormattingParams params) {
-		return computeAsync((cancelChecker) -> {
-			TextDocument document = getDocument(params.getTextDocument());
-			if (document == null) {
-				return null;
-			}
+		return computeDOMAsync(params.getTextDocument(), (xmlDocument, cancelChecker) -> {
 			CompositeSettings settings = new CompositeSettings(getSharedSettings(), params.getOptions());
-			return getXMLLanguageService().format(document, null, settings);
+			return getXMLLanguageService().format(xmlDocument, null, settings);
 		});
 	}
 
 	@Override
 	public CompletableFuture<List<? extends TextEdit>> rangeFormatting(DocumentRangeFormattingParams params) {
-		return computeAsync((cancelChecker) -> {
-			TextDocument document = getDocument(params.getTextDocument());
-			if (document == null) {
-				return null;
-			}
+		return computeDOMAsync(params.getTextDocument(), (xmlDocument, cancelChecker) -> {
 			CompositeSettings settings = new CompositeSettings(getSharedSettings(), params.getOptions());
-			return getXMLLanguageService().format(document, params.getRange(), settings);
+			return getXMLLanguageService().format(xmlDocument, params.getRange(), settings);
 		});
 	}
 

org.eclipse.lemminx/src/main/java/org/eclipse/lemminx/commons/TextDocument.java

@@ -78,6 +78,12 @@ public String lineText(int lineNumber) throws BadLocationException {
 		return text.substring(line.offset, line.offset + line.length);
 	}
 
+	public int lineOffsetAt(int position) throws BadLocationException {
+		ILineTracker lineTracker = getLineTracker();
+		Line line = lineTracker.getLineInformationOfOffset(position);
+		return line.offset;
+	}
+
 	public String lineDelimiter(int lineNumber) throws BadLocationException {
 		ILineTracker lineTracker = getLineTracker();
 		String lineDelimiter = lineTracker.getLineDelimiter(lineNumber);

org.eclipse.lemminx/src/main/java/org/eclipse/lemminx/dom/DOMAttr.java

@@ -403,11 +403,14 @@ public int getStart() {
 	@Override
 	public int getEnd() {
 		if (nodeAttrValue != null) {
+			// <foo attr="value"| >
 			return nodeAttrValue.getEnd();
 		}
 		if (hasDelimiter()) {
+			// <foo attr=| >
 			return delimiter + 1;
 		}
+		// <foo attr| >
 		return nodeAttrName.getEnd();
 	}
 
@@ -452,4 +455,8 @@ public boolean equals(Object obj) {
 		return true;
 	}
 
+	public int getDelimiterOffset() {
+		return delimiter;
+	}
+
 }

org.eclipse.lemminx/src/main/java/org/eclipse/lemminx/dom/DOMElement.java

@@ -412,9 +412,11 @@ public boolean isOrphanEndTagOf(String tagName) {
 	}
 
 	/**
-	 * Returns the offset at which the given unclosed start tag should be closed with an angle bracket
+	 * Returns the offset at which the given unclosed start tag should be closed
+	 * with an angle bracket
 	 *
-	 * @returns the offset at which the given unclosed start tag should be closed with an angle bracket
+	 * @returns the offset at which the given unclosed start tag should be closed
+	 *          with an angle bracket
 	 */
 	public int getUnclosedStartTagCloseOffset() {
 		String documentText = getOwnerDocument().getText();
@@ -554,7 +556,7 @@ public boolean isEmpty() {
 		for (DOMNode child : getChildren()) {
 			if (child.isText()) {
 				DOMText text = (DOMText) child;
-				if (!text.isWhitespace()) {
+				if (!text.isElementContentWhitespace()) {
 					return false;
 				}
 			} else {
@@ -564,4 +566,11 @@ public boolean isEmpty() {
 		return true;
 	}
 
+	public int getOffsetAfterStartTag() {
+		if (hasTagName()) {
+			return getStartTagOpenOffset() + 1;
+		}
+		return getStartTagOpenOffset() + getTagName().length() + 1;
+	}
+
 }

org.eclipse.lemminx/src/main/java/org/eclipse/lemminx/dom/DOMText.java

@@ -12,6 +12,7 @@
  */
 package org.eclipse.lemminx.dom;
 
+import org.eclipse.lemminx.utils.StringUtils;
 import org.w3c.dom.DOMException;
 
 /**
@@ -61,7 +62,8 @@ public String getWholeText() {
 	 */
 	@Override
 	public boolean isElementContentWhitespace() {
-		throw new UnsupportedOperationException();
+		String text = getOwnerDocument().getOwnerDocument().getText();
+		return StringUtils.isWhitespace(text, getStart(), getEnd());
 	}
 
 	/*

org.eclipse.lemminx/src/main/java/org/eclipse/lemminx/extensions/contentmodel/ContentModelPlugin.java

@@ -25,6 +25,7 @@
 import org.eclipse.lemminx.extensions.contentmodel.participants.ContentModelCodeLensParticipant;
 import org.eclipse.lemminx.extensions.contentmodel.participants.ContentModelCompletionParticipant;
 import org.eclipse.lemminx.extensions.contentmodel.participants.ContentModelDocumentLinkParticipant;
+import org.eclipse.lemminx.extensions.contentmodel.participants.ContentModelFormatterParticipant;
 import org.eclipse.lemminx.extensions.contentmodel.participants.ContentModelHoverParticipant;
 import org.eclipse.lemminx.extensions.contentmodel.participants.ContentModelSymbolsProviderParticipant;
 import org.eclipse.lemminx.extensions.contentmodel.participants.ContentModelTypeDefinitionParticipant;
@@ -85,6 +86,8 @@ public class ContentModelPlugin implements IXMLExtension {
 
 	private DocumentTelemetryParticipant documentTelemetryParticipant;
 
+	private ContentModelFormatterParticipant formatterParticipant;
+
 	public ContentModelPlugin() {
 		completionParticipant = new ContentModelCompletionParticipant();
 		hoverParticipant = new ContentModelHoverParticipant();
@@ -207,7 +210,9 @@ public void start(InitializeParams params, XMLExtensionsRegistry registry) {
 		documentTelemetryParticipant = new DocumentTelemetryParticipant(registry.getTelemetryManager(),
 				contentModelManager);
 		registry.registerDocumentLifecycleParticipant(documentTelemetryParticipant);
-
+		formatterParticipant = new ContentModelFormatterParticipant(contentModelManager);
+		registry.registerFormatterParticipant(formatterParticipant);
+
 		// Register custom commands to re-validate XML files
 		IXMLCommandService commandService = registry.getCommandService();
 		if (commandService != null) {
@@ -236,7 +241,8 @@ public void stop(XMLExtensionsRegistry registry) {
 		registry.unregisterSymbolsProviderParticipant(symbolsProviderParticipant);
 		registry.unregisterCodeLensParticipant(codeLensParticipant);
 		registry.unregisterDocumentLifecycleParticipant(documentTelemetryParticipant);
-
+		registry.unregisterFormatterParticipant(formatterParticipant);
+
 		// Un-register custom commands to re-validate XML files
 		IXMLCommandService commandService = registry.getCommandService();
 		if (commandService != null) {

org.eclipse.lemminx/src/main/java/org/eclipse/lemminx/extensions/contentmodel/model/CMElementDeclaration.java

@@ -142,4 +142,22 @@ default String getName(String prefix) {
 	 */
 	String getDocumentURI();
 
+	/**
+	 * Returns true if the element is a string type (ex : xs:string) and false
+	 * otherwise.
+	 * 
+	 * @return true if the element is a string type (ex : xs:string) and false
+	 *         otherwise.
+	 */
+	boolean isStringType();
+
+	/**
+	 * Returns true if the element can contains text and element both and false
+	 * otherwise.
+	 * 
+	 * @return true if the element can contains text and element both and false
+	 *         otherwise.
+	 */
+	boolean isMixedContent();
+
 }

org.eclipse.lemminx/src/main/java/org/eclipse/lemminx/extensions/contentmodel/participants/ContentModelFormatterParticipant.java

@@ -0,0 +1,70 @@
+/*******************************************************************************
+* Copyright (c) 2022 Red Hat Inc. and others.
+* All rights reserved. This program and the accompanying materials
+* which accompanies this distribution, and is available at
+* http://www.eclipse.org/legal/epl-v20.html
+*
+* SPDX-License-Identifier: EPL-2.0
+*
+* Contributors:
+*     Red Hat Inc. - initial API and implementation
+*******************************************************************************/
+package org.eclipse.lemminx.extensions.contentmodel.participants;
+
+import java.util.Collection;
+
+import org.eclipse.lemminx.dom.DOMElement;
+import org.eclipse.lemminx.extensions.contentmodel.model.CMDocument;
+import org.eclipse.lemminx.extensions.contentmodel.model.CMElementDeclaration;
+import org.eclipse.lemminx.extensions.contentmodel.model.ContentModelManager;
+import org.eclipse.lemminx.services.extensions.format.IFormatterParticipant;
+import org.eclipse.lemminx.services.format.FormatElementCategory;
+import org.eclipse.lemminx.services.format.XMLFormattingConstraints;
+import org.eclipse.lemminx.settings.SharedSettings;
+
+/**
+ * Formatter participant which uses XSD/DTD grammar information to know the
+ * {@link FormatElementCategory} of a given element.
+ * 
+ * <p>
+ * 
+ * This participant is enabled when 'xml.format.grammarAwareFormatting' setting
+ * is set to true.
+ * 
+ * </p>
+ * 
+ * @author Angelo ZERR
+ *
+ */
+public class ContentModelFormatterParticipant implements IFormatterParticipant {
+
+	private final ContentModelManager contentModelManager;
+
+	public ContentModelFormatterParticipant(ContentModelManager contentModelManager) {
+		this.contentModelManager = contentModelManager;
+	}
+
+	@Override
+	public FormatElementCategory getFormatElementCategory(DOMElement element,
+			XMLFormattingConstraints parentConstraints, SharedSettings sharedSettings) {
+		boolean enabled = sharedSettings.getFormattingSettings().isGrammarAwareFormatting();
+		if (!enabled) {
+			return null;
+		}
+
+		Collection<CMDocument> cmDocuments = contentModelManager.findCMDocument(element);
+		for (CMDocument cmDocument : cmDocuments) {
+			CMElementDeclaration cmElement = cmDocument.findCMElement(element);
+			if (cmElement != null) {
+				if (cmElement.isStringType()) {
+					return FormatElementCategory.PreserveSpace;
+				}
+				if (cmElement.isMixedContent()) {
+					return FormatElementCategory.MixedContent;
+				}
+			}
+		}
+		return null;
+	}
+
+}

org.eclipse.lemminx/src/main/java/org/eclipse/lemminx/extensions/dtd/contentmodel/CMDTDElementDeclaration.java

@@ -146,4 +146,14 @@ public int getIndex() {
 	public String getDocumentURI() {
 		return document.getURI();
 	}
+
+	@Override
+	public boolean isStringType() {
+		return false;
+	}
+
+	@Override
+	public boolean isMixedContent() {
+		return super.type == XMLElementDecl.TYPE_MIXED;
+	}
 }

org.eclipse.lemminx/src/main/java/org/eclipse/lemminx/extensions/xsd/contentmodel/CMXSDElementDeclaration.java

@@ -202,11 +202,11 @@ public Collection<CMElementDeclaration> getPossibleElements(DOMElement parentEle
 	/**
 	 * Returns the possible elements declaration if the given declaration is an
 	 * xs:any and null otherwise.
-	 * 
+	 *
 	 * @param declaration the element, wildcard declaration.
 	 * @return the possible elements declaration if the given declaration is an
 	 *         xs:any and null otherwise.
-	 * 
+	 *
 	 */
 	private Collection<CMElementDeclaration> getXSAnyElements(Object declaration) {
 		short processContents = getXSAnyProcessContents(declaration);
@@ -230,7 +230,7 @@ private Collection<CMElementDeclaration> getXSAnyElements(Object declaration) {
 	/**
 	 * Returns the value of the xs:any/@processContents if the given element is a
 	 * xs:any and {@link #PC_UNKWOWN} otherwise.
-	 * 
+	 *
 	 * @param declaration the element, wildcard declaration.
 	 * @return the value of the xs:any/@processContents if the given element is a
 	 *         xs:any and {@link #PC_UNKWOWN} otherwise.
@@ -248,7 +248,7 @@ private static short getXSAnyProcessContents(Object declaration) {
 	/**
 	 * Returns list of element (QName) of child elements of the given parent element
 	 * upon the given offset
-	 * 
+	 *
 	 * @param parentElement the parent element
 	 * @param offset        the offset where child element must be belong to
 	 * @return list of element (QName) of child elements of the given parent element
@@ -355,7 +355,7 @@ public String getDocumentation(ISharedSettingsRequest request) {
 	/**
 	 * Returns list of xs:annotation from the element declaration or type
 	 * declaration.
-	 * 
+	 *
 	 * @return list of xs:annotation from the element declaration or type
 	 *         declaration.
 	 */
@@ -490,4 +490,31 @@ public String getDocumentURI() {
 		SchemaGrammar schemaGrammar = document.getOwnerSchemaGrammar(elementDeclaration);
 		return CMXSDDocument.getSchemaURI(schemaGrammar);
 	}
+
+	@Override
+	public boolean isStringType() {
+		XSTypeDefinition typeDefinition = elementDeclaration.getTypeDefinition();
+		if (typeDefinition != null) {
+			XSSimpleTypeDefinition simpleDefinition = null;
+			if (typeDefinition.getTypeCategory() == XSTypeDefinition.SIMPLE_TYPE) {
+				simpleDefinition = (XSSimpleTypeDefinition) typeDefinition;
+			} else if (typeDefinition.getTypeCategory() == XSTypeDefinition.COMPLEX_TYPE) {
+				simpleDefinition = ((XSComplexTypeDefinition) typeDefinition).getSimpleType();
+			}
+			if (simpleDefinition != null) {
+				return "string".equals(simpleDefinition.getName());
+			}
+		}
+		return false;
+	}
+
+	@Override
+	public boolean isMixedContent() {
+		XSTypeDefinition typeDefinition = elementDeclaration.getTypeDefinition();
+		if (typeDefinition != null && typeDefinition.getTypeCategory() == XSTypeDefinition.COMPLEX_TYPE) {
+			XSComplexTypeDefinition complexTypeDefinition = (XSComplexTypeDefinition) typeDefinition;
+			return complexTypeDefinition.getContentType() == XSComplexTypeDefinition.CONTENTTYPE_MIXED;
+		}
+		return false;
+	}
 }

org.eclipse.lemminx/src/main/java/org/eclipse/lemminx/extensions/xsi/settings/XSISchemaLocationSplit.java

@@ -24,14 +24,12 @@ public enum XSISchemaLocationSplit {
 
 	onElement, onPair, none;
 
-	private static final String XSI_SCHEMA_LOCATION_SPLIT = "xsiSchemaLocationSplit";
-
 	public static XSISchemaLocationSplit getSplit(XMLFormattingOptions formattingSettings) {
-		return getSplit(formattingSettings.getString(XSI_SCHEMA_LOCATION_SPLIT));
+		return getSplit(formattingSettings.getXsiSchemaLocationSplit());
 	}
 
 	public static void setSplit(XSISchemaLocationSplit split, XMLFormattingOptions formattingSettings) {
-		formattingSettings.putString(XSI_SCHEMA_LOCATION_SPLIT, split.name());
+		formattingSettings.setXsiSchemaLocationSplit(split.name());
 	}
 
 	public static XSISchemaLocationSplit getSplit(String value) {