nokogiri 1.5.0.beta.1 → 1.5.0.beta.2

data/ext/java/nokogiri/XmlXpathContext.java

@@ -0,0 +1,130 @@
+package nokogiri;
+
+import java.util.Set;
+import java.util.regex.Matcher;
+import java.util.regex.Pattern;
+
+import javax.xml.xpath.XPath;
+import javax.xml.xpath.XPathConstants;
+import javax.xml.xpath.XPathExpression;
+import javax.xml.xpath.XPathExpressionException;
+import javax.xml.xpath.XPathFactory;
+
+import nokogiri.internals.NokogiriNamespaceContext;
+import nokogiri.internals.NokogiriXPathFunctionResolver;
+
+import org.jruby.Ruby;
+import org.jruby.RubyBoolean;
+import org.jruby.RubyClass;
+import org.jruby.RubyException;
+import org.jruby.RubyNumeric;
+import org.jruby.RubyObject;
+import org.jruby.RubyString;
+import org.jruby.anno.JRubyClass;
+import org.jruby.anno.JRubyMethod;
+import org.jruby.exceptions.RaiseException;
+import org.jruby.runtime.ThreadContext;
+import org.jruby.runtime.builtin.IRubyObject;
+import org.w3c.dom.NodeList;
+
+@JRubyClass(name="Nokogiri::XML::XPathContext")
+public class XmlXpathContext extends RubyObject {
+    private XmlNode context;
+    private XPath xpath;
+
+    public XmlXpathContext(Ruby ruby, RubyClass rubyClass, XmlNode context) {
+        super(ruby, rubyClass);
+        this.context = context;
+        this.xpath = XPathFactory.newInstance().newXPath();
+        this.xpath.setNamespaceContext(new NokogiriNamespaceContext());
+    }
+
+    @JRubyMethod(name = "new", meta = true)
+    public static IRubyObject rbNew(ThreadContext context, IRubyObject cls, IRubyObject node) {
+        XmlNode xmlNode = (XmlNode)node;
+        return new XmlXpathContext(context.getRuntime(), (RubyClass)cls, xmlNode);
+    }
+
+    @JRubyMethod
+    public IRubyObject evaluate(ThreadContext context, IRubyObject expr, IRubyObject handler) {
+        String src = expr.convertToString().asJavaString();
+        try {
+            if(!handler.isNil()) {
+            	if (!isContainsPrefix(src)) {
+                    Set<String> methodNames = handler.getMetaClass().getMethods().keySet();
+                    for (String name : methodNames) {
+                        src = src.replaceAll(name, NokogiriNamespaceContext.NOKOGIRI_PREFIX+":"+name);
+                    }
+                }
+                xpath.setXPathFunctionResolver(new NokogiriXPathFunctionResolver(handler));
+            }
+            XPathExpression xpathExpression = xpath.compile(src);
+            return node_set(context, xpathExpression);
+            //return new XmlXpath(context.getRuntime(), (RubyClass)context.getRuntime().getClassFromPath("Nokogiri::XML::XPath"), xpathExpression, this.context);
+        } catch (XPathExpressionException xpee) {
+            xpee = new XPathExpressionException(src);
+            RubyException e =
+                XmlSyntaxError.createXPathSyntaxError(getRuntime(), xpee);
+            throw new RaiseException(e);
+        }
+    }
+
+    protected IRubyObject node_set(ThreadContext rbctx, XPathExpression xpathExpression) {
+        XmlNodeSet result = null;
+        try {  
+            result = tryGetNodeSet(xpathExpression);
+//            result.relink_namespace(context);
+            result.setDocument(context.document(rbctx));
+            return result;
+        } catch (XPathExpressionException xpee) {
+            try {
+                return tryGetOpaqueValue(xpathExpression);
+            } catch (XPathExpressionException xpee_opaque) {
+                 RubyException e = XmlSyntaxError.createXPathSyntaxError(getRuntime(), xpee_opaque);
+                 throw new RaiseException(e);
+            }
+        }
+    }
+    
+    private XmlNodeSet tryGetNodeSet(XPathExpression xpathExpression) throws XPathExpressionException {
+        NodeList nodes = (NodeList)xpathExpression.evaluate(context.node, XPathConstants.NODESET);
+        return new XmlNodeSet(getRuntime(), nodes);       
+    }
+    
+    private static Pattern number_pattern = Pattern.compile("\\d.*");
+    private static Pattern boolean_pattern = Pattern.compile("true|false");
+    
+    private IRubyObject tryGetOpaqueValue(XPathExpression xpathExpression) throws XPathExpressionException {
+        String string = (String)xpathExpression.evaluate(context.node, XPathConstants.STRING);
+        if (doesMatch(number_pattern, string)) return RubyNumeric.dbl2num(getRuntime(), Double.parseDouble(string));
+        if (doesMatch(boolean_pattern, string)) return RubyBoolean.newBoolean(getRuntime(), Boolean.parseBoolean(string));
+        return RubyString.newString(getRuntime(), string);
+    }
+    
+    private boolean doesMatch(Pattern pattern, String string) {
+        Matcher m = pattern.matcher(string);
+        return m.matches();
+    }
+
+    private boolean isContainsPrefix(String str) {
+        Set<String> prefixes = ((NokogiriNamespaceContext)xpath.getNamespaceContext()).getAllPrefixes();
+        for (String prefix : prefixes) {
+            if (str.contains(prefix + ":")) {
+                return true;
+            }
+        }
+        return false;
+    }
+
+
+    @JRubyMethod
+    public IRubyObject evaluate(ThreadContext context, IRubyObject expr) {
+        return this.evaluate(context, expr, context.getRuntime().getNil());
+    }
+
+    @JRubyMethod
+    public IRubyObject register_ns(ThreadContext context, IRubyObject prefix, IRubyObject uri) {
+        ((NokogiriNamespaceContext) this.xpath.getNamespaceContext()).registerNamespace(prefix.convertToString().asJavaString(), uri.convertToString().asJavaString());
+        return this;
+    }
+}

data/ext/java/nokogiri/XsltStylesheet.java

@@ -0,0 +1,126 @@
+package nokogiri;
+
+import static nokogiri.internals.NokogiriHelpers.getNokogiriClass;
+
+import java.util.regex.Matcher;
+import java.util.regex.Pattern;
+
+import javax.xml.transform.Templates;
+import javax.xml.transform.Transformer;
+import javax.xml.transform.TransformerConfigurationException;
+import javax.xml.transform.TransformerException;
+import javax.xml.transform.TransformerFactory;
+import javax.xml.transform.dom.DOMResult;
+import javax.xml.transform.dom.DOMSource;
+
+import org.jruby.Ruby;
+import org.jruby.RubyArray;
+import org.jruby.RubyClass;
+import org.jruby.RubyObject;
+import org.jruby.anno.JRubyClass;
+import org.jruby.anno.JRubyMethod;
+import org.jruby.javasupport.util.RuntimeHelpers;
+import org.jruby.runtime.ThreadContext;
+import org.jruby.runtime.builtin.IRubyObject;
+import org.w3c.dom.Document;
+
+@JRubyClass(name="Nokogiri::XSLT::Stylesheet")
+public class XsltStylesheet extends RubyObject {
+
+    private Templates sheet;
+
+    public XsltStylesheet(Ruby ruby, RubyClass rubyClass) {
+        super(ruby, rubyClass);
+    }
+
+    private void addParametersToTransformer(ThreadContext context, Transformer transf, IRubyObject parameters) {
+        Ruby ruby = context.getRuntime();
+        RubyArray params = parameters.convertToArray();
+        int limit = params.getLength();
+        if(limit % 2 == 1) limit--;
+
+        for(int i = 0; i < limit; i+=2) {
+            String name = params.aref(ruby.newFixnum(i)).asJavaString();
+            String value = params.aref(ruby.newFixnum(i+1)).asJavaString();
+            transf.setParameter(name, unparseValue(value));
+        }
+    }
+    
+    private Pattern p = Pattern.compile("'.{1,}'");
+
+    private String unparseValue(String orig) {
+        Matcher m = p.matcher(orig);
+        if ((orig.startsWith("\"") && orig.endsWith("\"")) || m.matches()) {
+            orig = orig.substring(1, orig.length()-1);
+        }
+
+        return orig;
+    }
+
+    @JRubyMethod(meta = true)
+    public static IRubyObject parse_stylesheet_doc(ThreadContext context, IRubyObject cls, IRubyObject document) {
+        
+        Ruby ruby = context.getRuntime();
+
+        if(!(document instanceof XmlDocument)) {
+            throw ruby.newArgumentError("doc must be a Nokogiri::XML::Document instance");
+        }
+
+        XmlDocument xmlDoc = (XmlDocument) document;
+
+        RubyArray errors = (RubyArray) xmlDoc.getInstanceVariable("@errors");
+
+        if(!errors.isEmpty()) {
+            throw ruby.newRuntimeError(errors.first().asJavaString());
+        }
+        
+        Document doc = ((XmlDocument) xmlDoc.dup_implementation(context, true)).getDocument();
+
+        XsltStylesheet xslt = new XsltStylesheet(ruby, (RubyClass) cls);
+        try {
+            xslt.sheet = TransformerFactory.newInstance().newTemplates(new DOMSource(doc));
+        } catch (TransformerConfigurationException ex) {
+            ruby.newRuntimeError("could not parse xslt stylesheet");
+        }
+
+        return xslt;
+    }
+
+    @JRubyMethod
+    public IRubyObject serialize(ThreadContext context, IRubyObject doc) {
+        System.out.println("Serialize called in stylesheet");
+        return RuntimeHelpers.invoke(context,
+                RuntimeHelpers.invoke(context, doc, "root"),
+                "to_s");
+    }
+
+    @JRubyMethod(rest = true, required=1, optional=2)
+    public IRubyObject transform(ThreadContext context, IRubyObject[] args) {
+        Ruby ruby = context.getRuntime();
+
+        DOMSource docSource = new DOMSource(((XmlDocument) args[0]).getDocument());
+        DOMResult result = new DOMResult();
+
+        try{
+            Transformer transf = this.sheet.newTransformer();
+            if(args.length > 1) {
+                addParametersToTransformer(context, transf, args[1]);
+            }
+            transf.transform(docSource, result);
+        } catch(TransformerConfigurationException ex) {
+            throw ruby.newRuntimeError("Could not transform the document.");
+        } catch(TransformerException ex) {
+            throw ruby.newRuntimeError("Could not transform the document.");
+        }
+        
+        if ("html".equals(result.getNode().getFirstChild().getNodeName())) {
+            return new HtmlDocument(ruby,
+                    getNokogiriClass(ruby, "Nokogiri::HTML::Document"),
+                    (Document) result.getNode());
+        } else {
+            return new XmlDocument(ruby,
+                    getNokogiriClass(ruby, "Nokogiri::XML::Document"),
+                    (Document) result.getNode());
+        }
+    }
+}

data/ext/java/nokogiri/internals/HtmlDomParserContext.java

@@ -0,0 +1,181 @@
+package nokogiri.internals;
+
+import static nokogiri.internals.NokogiriHelpers.getNokogiriClass;
+import static nokogiri.internals.NokogiriHelpers.isNamespace;
+import nokogiri.HtmlDocument;
+import nokogiri.XmlDocument;
+
+import org.apache.xerces.parsers.DOMParser;
+import org.apache.xerces.xni.Augmentations;
+import org.apache.xerces.xni.QName;
+import org.apache.xerces.xni.XMLAttributes;
+import org.apache.xerces.xni.XNIException;
+import org.apache.xerces.xni.parser.XMLDocumentFilter;
+import org.apache.xerces.xni.parser.XMLParserConfiguration;
+import org.cyberneko.html.HTMLConfiguration;
+import org.cyberneko.html.filters.DefaultFilter;
+import org.jruby.Ruby;
+import org.jruby.RubyClass;
+import org.jruby.runtime.ThreadContext;
+import org.jruby.runtime.builtin.IRubyObject;
+import org.w3c.dom.Document;
+
+/**
+ *
+ * @author sergio
+ */
+public class HtmlDomParserContext extends XmlDomParserContext {
+    protected static final String PROPERTY_FILTERS =
+        "http://cyberneko.org/html/properties/filters";
+    protected static final String PROPERTY_ELEM_NAMES =
+        "http://cyberneko.org/html/properties/names/elems";
+    protected static final String PROPERTY_ATTRS_NAMES =
+        "http://cyberneko.org/html/properties/names/attrs";
+    protected static final String FEATURE_DOCUMENT_FRAGMENT =
+        "http://cyberneko.org/html/features/balance-tags/document-fragment";
+    protected static final String FEATURE_REPORT_ERRORS =
+        "http://cyberneko.org/html/features/report-errors";
+
+    public HtmlDomParserContext(Ruby runtime, IRubyObject options) {
+        super(runtime, options);
+    }
+    
+    public HtmlDomParserContext(Ruby runtime, IRubyObject encoding, IRubyObject options) {
+        super(runtime, encoding, options);
+    }
+
+    @Override
+    protected void initErrorHandler() {
+        if (continuesOnError()) {
+            errorHandler = new NokogiriNonStrictErrorHandler4NekoHtml();
+        } else if (options.noError) {
+            errorHandler = new NokogiriNonStrictErrorHandler4NekoHtml(options.noError);
+        } else {
+            errorHandler = new NokogiriStrictErrorHandler();
+        }
+    }
+
+    @Override
+    protected void initParser(Ruby runtime) {
+        XMLParserConfiguration config = new HTMLConfiguration();
+        XMLDocumentFilter removeNSAttrsFilter = new RemoveNSAttrsFilter();
+        XMLDocumentFilter elementValidityCheckFilter = new ElementValidityCheckFilter(errorHandler);
+        //XMLDocumentFilter[] filters = { removeNSAttrsFilter};
+        XMLDocumentFilter[] filters = { removeNSAttrsFilter,  elementValidityCheckFilter};
+
+        config.setErrorHandler(this.errorHandler);
+        parser = new DOMParser(config);
+
+        setProperty("http://cyberneko.org/html/properties/default-encoding", java_encoding);
+        setProperty(PROPERTY_ELEM_NAMES, "lower");
+        setProperty(PROPERTY_ATTRS_NAMES, "lower");
+        setFeature(FEATURE_REPORT_ERRORS, true);
+        setFeature("http://xml.org/sax/features/namespaces", false);
+        setProperty(PROPERTY_FILTERS, filters);
+    }
+
+    /**
+     * Enable NekoHTML feature for balancing tags in a document
+     * fragment.
+     */
+    public void enableDocumentFragment() {
+        setFeature(FEATURE_DOCUMENT_FRAGMENT, true);
+    }
+
+    @Override
+    protected XmlDocument getNewEmptyDocument(ThreadContext context) {
+        IRubyObject[] args = new IRubyObject[0];
+        return (XmlDocument) XmlDocument.rbNew(context,
+                    getNokogiriClass(context.getRuntime(), "Nokogiri::XML::Document"),
+                    args);
+    }
+
+    @Override
+    protected XmlDocument wrapDocument(ThreadContext context,
+                                       RubyClass klass,
+                                       Document doc) {
+        HtmlDocument htmlDocument = new HtmlDocument(context.getRuntime(), klass, doc);
+        htmlDocument.setEncoding(ruby_encoding);
+        return htmlDocument;
+    }
+
+    /**
+     * Filter to strip out attributes that pertain to XML namespaces.
+     *
+     * @author sergio
+     * @author Patrick Mahoney <pat@polycrystal.org>
+     */
+    public static class RemoveNSAttrsFilter extends DefaultFilter {
+        @Override
+        public void startElement(QName element, XMLAttributes attrs,
+                                 Augmentations augs) throws XNIException {
+            int i;
+            for (i = 0; i < attrs.getLength(); ++i) {
+                if (isNamespace(attrs.getQName(i))) {
+                    attrs.removeAttributeAt(i);
+                    --i;
+                }
+            }
+
+            element.uri = null;
+            super.startElement(element, attrs, augs);
+        }
+    }
+    
+    public static class ElementValidityCheckFilter extends DefaultFilter {
+        private NokogiriErrorHandler errorHandler;
+        
+        private ElementValidityCheckFilter(NokogiriErrorHandler errorHandler) {
+            this.errorHandler = errorHandler;
+        }
+        
+        // element names from xhtml1-strict.dtd
+        private static String[][] element_names = {
+                {"a", "abbr", "acronym", "address", "area"},
+                {"b", "base", "basefont", "bdo", "big", "blockquote", "body", "br", "button"},
+                {"caption", "cite", "code", "col", "colgroup"},
+                {"dd", "del", "dfn", "div", "dl", "dt"},
+                {"em"},
+                {"fieldset", "font", "form", "frame", "frameset"},
+                {}, // g
+                {"h1", "h2", "h3", "h4", "h5", "h6", "head", "hr", "html"},
+                {"i", "iframe", "img", "input", "ins"},
+                {}, // j
+                {"kbd"},
+                {"label", "legend", "li", "link"},
+                {"map", "meta"},
+                {"noframes", "noscript"},
+                {"object", "ol", "optgroup", "option"},
+                {"p", "param", "pre"},
+                {"q"},
+                {}, // r
+                {"s", "samp", "script", "select", "small", "span", "strike", "strong", "style", "sub", "sup"},
+                {"table", "tbody", "td", "textarea", "tfoot", "th", "thead", "title", "tr", "tt"},
+                {"u", "ul"},
+                {"var"},
+                {}, // w
+                {}, // x
+                {}, // y
+                {}  // z
+        };
+        
+        private boolean isValid(String testee) {
+            char[] c = testee.toCharArray();
+            int index = new Integer(c[0]) - 97;
+            for (int i=0; i<element_names[index].length; i++) {
+                if (testee.equals(element_names[index][i])) {
+                    return true;
+                }
+            }
+            return false;
+        }
+        
+        @Override
+        public void startElement(QName name, XMLAttributes attrs, Augmentations augs) throws XNIException {
+            if (!isValid(name.rawname)) {
+                errorHandler.addError(new Exception("Tag " + name.rawname + " invalid"));
+            }
+            super.startElement(name, attrs, augs);
+        }
+    }
+}

data/ext/java/nokogiri/internals/NokogiriDocumentCache.java

@@ -0,0 +1,39 @@
+package nokogiri.internals;
+
+import java.util.Hashtable;
+import nokogiri.XmlDocument;
+import org.w3c.dom.Document;
+
+/**
+ *
+ * @author sergio
+ */
+public class NokogiriDocumentCache {
+
+    private static NokogiriDocumentCache instance;
+    protected Hashtable<Document, XmlDocument> cache;
+
+    private NokogiriDocumentCache() {
+        this.cache = new Hashtable<Document, XmlDocument>();
+    }
+
+    public static NokogiriDocumentCache getInstance() {
+        if(instance == null) {
+            instance = new NokogiriDocumentCache();
+        }
+        return instance;
+    }
+
+    public XmlDocument getXmlDocument(Document doc) {
+        return this.cache.get(doc);
+    }
+
+    public void putDocument(Document doc, XmlDocument xmlDoc) {
+        this.cache.put(doc, xmlDoc);
+    }
+
+    public XmlDocument removeDocument(Document doc) {
+        return this.cache.remove(doc);
+    }
+
+}