format.java

一个用于搜索本地文件内容的小型搜索引擎

    }

    /**
     * This will set whether JAXP TrAX processing instructions for
     * disabling/enabling output escaping are ignored.  Disabling
     * output escaping allows using XML text as element content and
     * outputing it verbatim, i.e. as element children would be.
     * <p>
     * When processed, these processing instructions are removed from
     * the generated XML text and control whether the element text
     * content is output verbatim or with escaping of the pre-defined
     * entities in XML 1.0.  The text to be output verbatim shall be
     * surrounded by the
     * <code>&lt;?javax.xml.transform.disable-output-escaping ?&gt;</code>
     * and <code>&lt;?javax.xml.transform.enable-output-escaping ?&gt;</code>
     * PIs.</p>
     * <p>
     * When ignored, the processing instructions are present in the
     * generated XML text and the pre-defined entities in XML 1.0 are
     * escaped.
     * <p>
     * Default: <code>false</code>.</p>
     *
     * @param ignoreTrAXEscapingPIs <code>boolean</code> indicating
     *        whether or not TrAX ouput escaping PIs are ignored.
     *
     * @see javax.xml.transform.Result#PI_ENABLE_OUTPUT_ESCAPING
     * @see javax.xml.transform.Result#PI_DISABLE_OUTPUT_ESCAPING
     */
    public void setIgnoreTrAXEscapingPIs(boolean ignoreTrAXEscapingPIs) {
        this.ignoreTrAXEscapingPIs = ignoreTrAXEscapingPIs;
    }

    /**
     * Returns whether JAXP TrAX processing instructions for
     * disabling/enabling output escaping are ignored.
     *
     * @return whether or not TrAX ouput escaping PIs are ignored.
     */
    public boolean getIgnoreTrAXEscapingPIs() {
        return ignoreTrAXEscapingPIs;
    }

    /**
     * This sets the text output style.  Options are available as static
     * {@link TextMode} instances.  The default is {@link TextMode#PRESERVE}.
     *
     * @return a pointer to this Format for chaining
     */
    public Format setTextMode(Format.TextMode mode) {
        this.mode = mode;
        return this;
    }

    /**
     * Returns the current text output style.
     *
     * @return the current text output style
     */
    public Format.TextMode getTextMode() {
        return mode;
    }

    /**
     * This will set the indent <code>String</code> to use; this
     * is usually a <code>String</code> of empty spaces. If you pass
     * the empty string (""), then no indentation will happen but newlines
     * will still be generated.  Passing null will result in no indentation
     * and no newlines generated.  Default: none (null)
     *
     * @param indent <code>String</code> to use for indentation.
     * @return a pointer to this Format for chaining
     */
    public Format setIndent(String indent) {
        this.indent = indent;
        return this;
    }

    /**
     * Returns the indent string in use.
     *
     * @return the indent string in use
     */
    public String getIndent() {
        return indent;
    }

    /**
     * Sets the output encoding.  The name should be an accepted XML
     * encoding.
     *
     * @param encoding the encoding format.  Use XML-style names like
     *                 "UTF-8" or "ISO-8859-1" or "US-ASCII"
     * @return a pointer to this Format for chaining
     */
    public Format setEncoding(String encoding) {
        this.encoding = encoding;
        escapeStrategy = new DefaultEscapeStrategy(encoding);
        return this;
    }

    /**
     * Returns the configured output encoding.
     *
     * @return the output encoding
     */
    public String getEncoding() {
        return encoding;
    }

    protected Object clone() {
        Format format = null;

        try {
            format = (Format) super.clone();
        }
        catch (CloneNotSupportedException ce) {
        }

        return format;
    }


    /**
     * Handle common charsets quickly and easily.  Use reflection
     * to query the JDK 1.4 CharsetEncoder class for unknown charsets.
     * If JDK 1.4 isn't around, default to no special encoding.
     */
    class DefaultEscapeStrategy implements EscapeStrategy {
        private int bits;
        Object encoder;
        Method canEncode;

        public DefaultEscapeStrategy(String encoding) {
            if ("UTF-8".equalsIgnoreCase(encoding) ||
                    "UTF-16".equalsIgnoreCase(encoding)) {
                bits = 16;
            }
            else if ("ISO-8859-1".equalsIgnoreCase(encoding) ||
                    "Latin1".equalsIgnoreCase(encoding)) {
                bits = 8;
            }
            else if ("US-ASCII".equalsIgnoreCase(encoding) ||
                    "ASCII".equalsIgnoreCase(encoding)) {
                bits = 7;
            }
            else {
                bits = 0;
                //encoder = Charset.forName(encoding).newEncoder();
                try {
                    Class charsetClass = Class.forName("java.nio.charset.Charset");
                    Class encoderClass = Class.forName("java.nio.charset.CharsetEncoder");
                    Method forName = charsetClass.getMethod("forName", new Class[]{String.class});
                    Object charsetObj = forName.invoke(null, new Object[]{encoding});
                    Method newEncoder = charsetClass.getMethod("newEncoder", null);
                    encoder = newEncoder.invoke(charsetObj, null);
                    canEncode = encoderClass.getMethod("canEncode", new Class[]{char.class});
                }
                catch (Exception ignored) {
                }
            }
        }

        public boolean shouldEscape(char ch) {
            if (bits == 16) {
                return false;
            }
            if (bits == 8) {
                if ((int) ch > 255)
                    return true;
                else
                    return false;
            }
            if (bits == 7) {
                if ((int) ch > 127)
                    return true;
                else
                    return false;
            }
            else {
                if (canEncode != null && encoder != null) {
                    try {
                        Boolean val = (Boolean) canEncode.invoke(encoder, new Object[]{new Character(ch)});
                        return !val.booleanValue();
                    }
                    catch (Exception ignored) {
                    }
                }
                // Return false if we don't know.  This risks not escaping
                // things which should be escaped, but also means people won't
                // start getting loads of unnecessary escapes.
                return false;
            }
        }
    }


    /**
     * Class to signify how text should be handled on output.  The following
     * table provides details.
     *
     * <table>
     *   <tr>
     *     <th align="left">
     *       Text Mode
     *     </th>
     *     <th>
     *       Resulting behavior.
     *     </th>
     *   </tr>
     *
     *   <tr valign="top">
     *     <td>
     *       <i>PRESERVE (Default)</i>
     *     </td>
     *     <td>
     *       All content is printed in the format it was created, no whitespace
     *       or line separators are are added or removed.
     *     </td>
     *   </tr>
     *
     *   <tr valign="top">
     *     <td>
     *       TRIM_FULL_WHITE
     *     </td>
     *     <td>
     *       Content between tags consisting of all whitespace is not printed.
     *       If the content contains even one non-whitespace character, it is
     *       printed verbatim, whitespace and all.
     *     </td>
     *   </tr>
     *
     *   <tr valign="top">
     *     <td>
     *       TRIM
     *     </td>
     *     <td>
     *       Same as TrimAllWhite, plus leading/trailing whitespace are
     *       trimmed.
     *     </td>
     *   </tr>
     *
     *   <tr valign="top">
     *     <td>
     *       NORMALIZE
     *     </td>
     *     <td>
     *       Same as TextTrim, plus addition interior whitespace is compressed
     *       to a single space.
     *     </td>
     *   </tr>
     * </table>
     *
     * In most cases textual content is aligned with the surrounding tags
     * (after the appropriate text mode is applied). In the case where the only
     * content between the start and end tags is textual, the start tag, text,
     * and end tag are all printed on the same line. If the document being
     * output already has whitespace, it's wise to turn on TRIM mode so the
     * pre-existing whitespace can be trimmed before adding new whitespace.
     * <p>
     * When a element has a xml:space attribute with the value of "preserve",
     * all formating is turned off and reverts back to the default until the
     * element and its contents have been printed. If a nested element contains
     * another xml:space with the value "default" formatting is turned back on
     * for the child element and then off for the remainder of the parent
     * element.
     */
    public static class TextMode {
        /**
         * Mode for literal text preservation.
         */
        public static final TextMode PRESERVE = new TextMode("PRESERVE");

        /**
         * Mode for text trimming (left and right trim).
         */
        public static final TextMode TRIM = new TextMode("TRIM");

        /**
         * Mode for text normalization (left and right trim plus internal
         * whitespace is normalized to a single space.
         * @see org.jdom.Element#getTextNormalize
         */
        public static final TextMode NORMALIZE = new TextMode("NORMALIZE");

        /**
         * Mode for text trimming of content consisting of nothing but
         * whitespace but otherwise not changing output.
         */
        public static final TextMode TRIM_FULL_WHITE =
                new TextMode("TRIM_FULL_WHITE");

        private final String name;

        private TextMode(String name) {
            this.name = name;
        }

        public String toString() {
            return name;
        }
    }
}