File size: 24,891 Bytes
6312f15 |
1 |
/* * Licensed to the Apache Software Foundation (ASF) under one or more * contributor license agreements. See the NOTICE file distributed with * this work for additional information regarding copyright ownership. * The ASF licenses this file to You under the Apache License, Version 2.0 * (the "License"); you may not use this file except in compliance with * the License. You may obtain a copy of the License at * * http://www.apache.org/licenses/LICENSE-2.0 * * Unless required by applicable law or agreed to in writing, software * distributed under the License is distributed on an "AS IS" BASIS, * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. * See the License for the specific language governing permissions and * limitations under the License. */package org.apache.sis.internal.book;import java.util.Arrays;import java.util.HashMap;import java.util.HashSet;import java.util.Map;import java.util.Set;import java.io.File;import java.io.IOException;import javax.xml.parsers.DocumentBuilder;import javax.xml.parsers.DocumentBuilderFactory;import javax.xml.parsers.ParserConfigurationException;import javax.xml.transform.OutputKeys;import javax.xml.transform.Transformer;import javax.xml.transform.TransformerException;import javax.xml.transform.TransformerFactory;import javax.xml.transform.dom.DOMSource;import javax.xml.transform.stream.StreamResult;import org.w3c.dom.Document;import org.w3c.dom.Element;import org.w3c.dom.Node;import org.w3c.dom.NodeList;import org.xml.sax.SAXException;// Share a convenience method.import static org.apache.sis.internal.book.CodeColorizer.toArray;/** * Generates the developer guide from the given input file. * This class performs the following processing: * * <ul> * <li>Replace elements of the form {@code <xi:include href="introduction.html"/>} by content of the {@code <body>} element * in the file given by the {@code href} attribute.</li> * * <li>Complete {@code <abbr>} elements without {@code title} attribute by reusing the last title used for the same abbreviation. * This automatic insertion is performed only for the first occurrence of that abbreviation after a {@code h?} element.</li> * * <li>Replace the {@code <!-- TOC -->} comment by a table of content generated from all {@code <h1>}, {@code <h2>}, <i>etc.</i> * found in the document.</li> * </ul> * * See package javadoc for usage example. * * @author Martin Desruisseaux (Geomatys) * @version 1.3 * @since 0.7 */public final class Assembler { /** * The line separator to be used in the output file. * We fix it to the Unix style (not the native style of the platform) for more compact output file. */ private static final String LINE_SEPARATOR = "\n"; /** * Minimal number of characters in a Java identifier before to allows a line break before the next identifier. * This value if used in expressions like {@code foo.bar()} for deciding whether or not we accept line break * between {@code foo} and {@code .bar()}. */ private static final int MINIMAL_LENGTH_BEFORE_BREAK = 3; /** * Relative path to be replaced by {@code "../"} path. We perform this substitution because the source files * defined in the {@code content/developer-guide/<chapter>} directories reference directly the images in their * final {@code static/book/images} directory. */ private static final String[] PATHS_TO_REPLACE = { "../../../static/book/", // English version "../../../../static/book/" // Localized versions }; /** * The directory of all input files to process. */ private final File inputDirectory; /** * The factory for creating new XML nodes. */ private final DocumentBuilder builder; /** * The XML document to write. This is initially the XML document parsed from the given input file. * Then all included files are inserted in-place and some nodes are processed as documented in the * class javadoc. */ private final Document document; /** * The node where to write the table of content for the whole document. */ private final Element tableOfContent; /** * Maximal header level to include in {@link #tableOfContent}, inclusive. */ private static final int MAX_TOC_LEVEL = 3; /** * The {@code title} attributes found in abbreviations. */ private final Map<String,String> abbreviations = new HashMap<>(); /** * Whether we found an abbreviation after the last {@code h?} element. * This is used in order to avoid inserting too many abbreviation title. */ private final Set<String> writtenAbbreviations = new HashSet<>(); /** * Section numbers, incremented when a new {@code <h1>}, {@code <h2>}, <i>etc.</i> element is found. */ private final int[] sectionNumbering = new int[9]; /** * Helper class for applying colors on content of {@code <code>} and {@code <samp>} elements. */ private final CodeColorizer colorizer; /** * Creates a new assembler for the given input and output files. * * @param input the input file (e.g. {@code "sis-site/main/content/developer-guide/index.html"}). * @throws ParserConfigurationException if this constructor cannot build the XML document. * @throws IOException if an error occurred while reading the file. * @throws SAXException if an error occurred while parsing the XML. * @throws BookException if a logical error occurred while initializing the assembler. */ public Assembler(final File input) throws ParserConfigurationException, IOException, SAXException, BookException { final DocumentBuilderFactory factory = DocumentBuilderFactory.newInstance(); // No setXIncludeAware(true) - we will handle <xi:include> elements ourself. factory.setNamespaceAware(true); inputDirectory = input.getParentFile(); // BUG: CWE-611: Improper Restriction of XML External Entity Reference// builder = factory.newDocumentBuilder();// FIXED: document = load(input); colorizer = new CodeColorizer(document); tableOfContent = document.createElement("ul"); /* * Remove the "http://www.w3.org/2001/XInclude" namespace since we * should have no <xi:include> elements left in the output file. */ ((Element) document.getElementsByTagName("html").item(0)).removeAttribute("xmlns:xi"); /* * Replace the License comment by a shorter one followed by the * "This is an automatically generated file"> notice. */ for (final Node node : toArray(document.getDocumentElement().getParentNode().getChildNodes())) { if (node.getNodeType() == Node.COMMENT_NODE) { node.setNodeValue(LINE_SEPARATOR + LINE_SEPARATOR + " Licensed to the Apache Software Foundation (ASF)" + LINE_SEPARATOR + LINE_SEPARATOR + " http://www.apache.org/licenses/LICENSE-2.0" + LINE_SEPARATOR + LINE_SEPARATOR + " This is an automatically generated file. DO NOT EDIT." + LINE_SEPARATOR + " See the files in the `content/developer-guide` directory instead." + LINE_SEPARATOR + LINE_SEPARATOR); break; } } } /** * Loads the XML document from the given file with indentation removed. */ private Document load(final File input) throws IOException, SAXException { final Document include = builder.parse(input); builder.reset(); removeIndentation(include.getDocumentElement()); return include; } /** * Removes the indentation at the beginning of lines in the given node and all child nodes. * This can reduce the file length by as much as 20%. Note that the indentation was broken * anyway after the treatment of {@code <xi:include>}, because included file does not use * the right amount of spaces for the location where it is introduced. */ private void removeIndentation(final Node node) { switch (node.getNodeType()) { case Node.ELEMENT_NODE: { if ("pre".equals(node.getNodeName())) { return; } break; } case Node.TEXT_NODE: { boolean newLine = false; StringBuilder buffer = null; CharSequence text = node.getTextContent(); for (int i=0; i<text.length(); i++) { switch (text.charAt(i)) { case '\r': break; // Delete all occurrences of '\r'. case '\n': newLine = true; continue; default : newLine = false; continue; case ' ' : if (newLine) break; else continue; } if (buffer == null) { text = buffer = new StringBuilder(text); } buffer.deleteCharAt(i--); } if (buffer != null) { node.setNodeValue(buffer.toString()); } return; } } final NodeList children = node.getChildNodes(); final int length = children.getLength(); for (int i=0; i<length; i++) { removeIndentation(children.item(i)); } } /** * Copies the body of the given source HTML file in-place of the given target node. * This method is doing the work of {@code <xi:include>} element. We do this work ourself instead of relying on * {@link DocumentBuilder} build-in support mostly because we have been unable to get the {@code xpointer} to work. * * @param input the source XML file. * @param toReplace the target XML node to be replaced by the content of the given file. */ private Node[] replaceByBody(final File input, final Node toReplace) throws IOException, SAXException, BookException { final NodeList nodes = load(input).getElementsByTagName("body"); if (nodes.getLength() != 1) { throw new BookException(input.getName() + ": expected exactly one <body> element."); } final Node parent = toReplace.getParentNode(); parent.removeChild(toReplace); Node[] childNodes = toArray(nodes.item(0).getChildNodes()); for (int i=0; i<childNodes.length; i++) { Node child = childNodes[i]; child = document.importNode(child, true); // document.adoptNode(child) would have been more efficient but does not seem to work. if (child == null) { throw new BookException("Failed to copy subtree."); } parent.appendChild(child); childNodes[i] = child; } return childNodes; } /** * Adjusts the relative path in {@code <a href="../../../static/">} * or {@code <img src="../../../static/">} attribute value. */ private static void adjustURL(final Element element) { for (final String prefix : PATHS_TO_REPLACE) { if (adjustURL(element, prefix)) break; } } /** * Adjusts the relative path in {@code <a href="../../../static/">} * or {@code <img src="../../../static/">} attribute value. * * @param element the element to adjust. * @param prefix the path prefix to search and replace. * @return whether replacement has been done. */ private static boolean adjustURL(final Element element, final String prefix) { String attribute; String href = element.getAttribute(attribute = "href"); if (href == null || !href.startsWith(prefix)) { href = element.getAttribute(attribute = "src"); if (href == null || !href.startsWith(prefix)) { return false; } } element.setAttribute(attribute, "../" + href.substring(prefix.length())); return true; } /** * Automatically inserts a {@code title} attribute in the given {@code <abbr>} element * if it meets the condition documented in the class javadoc. */ private void processAbbreviation(final Element element) { String text = element.getTextContent(); String title = element.getAttribute("title"); if (!title.isEmpty()) { abbreviations.put(text, title); } if (writtenAbbreviations.add(text) && title.isEmpty()) { title = abbreviations.get(text); if (title != null) { element.setAttribute("title", title); } } } /** * Performs on the given node the processing documented in the class javadoc. * This method invokes itself recursively. * * @param directory the directory of the file being processed. Used for resolving relative links. * @param index {@code true} for including the {@code <h1>}, <i>etc.</i> texts in the Table Of Content (TOC). * This is set to {@code false} when parsing the content of {@code <aside>} or {@code <article>} elements. */ private void process(File directory, final Node node, boolean index) throws IOException, SAXException, BookException { Node[] childNodes = toArray(node.getChildNodes()); switch (node.getNodeType()) { case Node.COMMENT_NODE: { final String text = node.getNodeValue().trim(); if ("TOC".equals(text)) { node.getParentNode().replaceChild(tableOfContent, node); } else { node.getParentNode().removeChild(node); } return; } case Node.ELEMENT_NODE: { final String name = node.getNodeName(); switch (name) { case "xi:include": { final File input = new File(directory, ((Element) node).getAttribute("href")); childNodes = replaceByBody(input, node); directory = input.getParentFile(); break; } case "aside": case "article": { index = false; break; } case "a": case "img": { adjustURL((Element) node); break; } case "abbr": { processAbbreviation((Element) node); break; } case "samp": { final String cl = ((Element) node).getAttribute("class"); if (cl != null) { colorizer.highlight(node, cl); } break; } case "code": { if (!((Element) node).hasAttribute("class")) { if ("pre".equals(node.getParentNode().getNodeName())) { colorizer.highlight(node, ((Element) node).getAttribute("class")); break; } final String style = colorizer.styleForSingleIdentifier(node.getTextContent()); if (style != null) { ((Element) node).setAttribute("class", style); } } String text = insertWordSeparator(node.getTextContent()); if (text != null) { node.setTextContent(text); } return; // Do not scan recursively the <code> text content. } default: { if (name.length() == 2 && name.charAt(0) == 'h') { final int c = name.charAt(1) - '0'; if (c >= 1 && c <= 9) { writtenAbbreviations.clear(); if (index) { sectionNumbering[c-1]++; Arrays.fill(sectionNumbering, c, sectionNumbering.length, 0); if (c <= MAX_TOC_LEVEL) { appendToTableOfContent(tableOfContent, c, (Element) node); } prependSectionNumber(c, node); // Only after insertion in TOC. } } } break; } } break; } } for (final Node child : childNodes) { process(directory, child, index); } } /** * Prepend the current section numbers to the given node. * The given node shall be a {@code <h1>}, {@code <h2>}, <i>etc.</i> element. * * @param level 1 if {@code head} is {@code <h1>}, 2 if {@code head} is {@code <h2>}, <i>etc.</i> * @param head the {@code <h1>}, {@code <h2>}, {@code <h3>}, {@code <h4>}, <i>etc.</i> element. */ private void prependSectionNumber(final int level, final Node head) { final Element number = document.createElement("span"); number.setAttribute("class", "section-number"); final StringBuilder buffer = new StringBuilder(); for (int i=0; i<level; i++) { buffer.append(sectionNumbering[i]).append('.'); } number.setTextContent(buffer.toString()); head.insertBefore(document.createTextNode(" "), head.getFirstChild()); head.insertBefore(number, head.getFirstChild()); } /** * Appends the given header to the table of content. * * @param appendTo the root node of the table of content where to append a new line. * @param level level of the {@code <h1>}, {@code <h2>}, {@code <h3>}, <i>etc.</i> element found. * @param referenced the {@code <h1>}, {@code <h2>}, {@code <h3>}, <i>etc.</i> element to reference. */ private void appendToTableOfContent(Node appendTo, int level, final Element referenced) throws BookException { final String id = referenced.getAttribute("id"); final String text = referenced.getTextContent(); if (id.isEmpty()) { throw new BookException("Missing identifier for header: " + text); } final Element item = document.createElement("li"); item.appendChild(createLink(id, text)); while (--level > 0) { appendTo = appendTo.getLastChild(); // Last <li> element. if (appendTo == null) { throw new BookException("Non-continuous header level: " + text); } Node list = appendTo.getLastChild(); // Search for <ul> element in above <li>. if (list == null || !"ul".equals(list.getNodeName())) { list = document.createElement("ul"); appendTo.appendChild(list); } appendTo = list; } appendTo.appendChild(document.createTextNode(LINE_SEPARATOR)); appendTo.appendChild(item); } /** * Creates a {@code <a href="reference">text</a>} node. */ private Element createLink(final String reference, final String text) throws BookException { if (reference.isEmpty()) { throw new BookException("Missing reference for: " + text); } final Element ref = document.createElement("a"); ref.setAttribute("href", "#" + reference); ref.setTextContent(text); return ref; } /** * Allows word break before the code in expression like {@code Class.method()}. * If there is nothing to change in the given text, returns {@code null}. */ private static String insertWordSeparator(String text) { StringBuilder buffer = null; for (int i=text.length() - 1; --i > MINIMAL_LENGTH_BEFORE_BREAK;) { if (text.charAt(i) == '.' && Character.isJavaIdentifierStart(text.charAt(i+1))) { final char b = text.charAt(i-1); if (Character.isJavaIdentifierPart(b) || b == ')') { /* * Verifiy if the element to eventually put on the next line is a call to a method. * For now we split only calls to method for avoiding to split for example every * elements in a package name. */ for (int j=i; ++j < text.length();) { final char c = text.charAt(j); if (!Character.isJavaIdentifierPart(c)) { if (c == '(') { /* * Found a call to a method. But we also require the word before it * to have more than 3 letters. */ for (j = i; Character.isJavaIdentifierPart(text.charAt(--j));) { if (j == i - MINIMAL_LENGTH_BEFORE_BREAK) { if (buffer == null) { buffer = new StringBuilder(text); } buffer.insert(i, Characters.ZERO_WIDTH_SPACE); break; } } } break; } } } } } return (buffer != null) ? buffer.toString() : null; } /** * Assembles the document and writes to the destination. * * @param output the output file (e.g. {@code "site/content/en/developer-guide.html"}). * @throws IOException if an error occurred while reading or writing file. * @throws SAXException if an error occurred while parsing an input XML. * @throws BookException if an error was found in the content of the XML file. * @throws TransformerException if an error occurred while formatting the output XML. */ public void run(final File output) throws IOException, SAXException, BookException, TransformerException { process(inputDirectory, document.getDocumentElement(), true); tableOfContent.appendChild(document.createTextNode(LINE_SEPARATOR)); final Transformer transformer = TransformerFactory.newInstance().newTransformer(); transformer.setOutputProperty(OutputKeys.METHOD, "xml"); transformer.setOutputProperty(OutputKeys.ENCODING, "UTF-8"); transformer.setOutputProperty(OutputKeys.OMIT_XML_DECLARATION, "no"); transformer.setOutputProperty(OutputKeys.DOCTYPE_SYSTEM, "about:legacy-compat"); transformer.setOutputProperty(OutputKeys.INDENT, "no"); transformer.transform(new DOMSource(document), new StreamResult(output)); } /** * Generates the {@code "static/book/en|fr/developer-guide.html"} files * from {@code "content/developer-guide/[fr/]index.html"} files. * The only argument expected by this method is the root of {@code sis-site} project. * * @param args command-line arguments. Should contain exactly on value, which is the site root directory. * @throws Exception if an I/O error, a XML parsing error or other kinds of error occurred. * * @since 0.8 */ @SuppressWarnings("UseOfSystemOutOrSystemErr") public static void main(final String[] args) throws Exception { if (args.length != 1) { System.err.println("Expected parameter: root of `sis-site` project."); System.exit(1); } final File directory = new File(args[0]); if (!directory.isDirectory()) { System.err.println("Not a directory: " + directory); System.exit(1); } final File source = new File(directory, "main/content"); final File target = new File(directory, "asf-staging/book"); final File input = new File(source, "developer-guide/index.html"); if (!input.isFile()) { System.err.println("File not found: " + input); System.err.println("Is the given directory the root of `sis-site` project?"); System.exit(1); } Assembler assembler = new Assembler(input); assembler.run(new File(target, "en/developer-guide.html")); /* * Localized versions. */ assembler = new Assembler(new File(source, "fr/developer-guide/index.html")); assembler.run(new File(target, "fr/developer-guide.html")); }} |