Examples with ParserCSSCharsetDetectorTokenManager com.helger.css.parser.ParserCSSCharsetDetectorTokenManager used on opensource projects

Example 1 with ParserCSSCharsetDetectorTokenManager

use of com.helger.css.parser.ParserCSSCharsetDetectorTokenManager in project ph-css by phax.

the class CSSReader method getCharsetDeclaredInCSS.

/**
 * Determine the charset to read the CSS file. The logic is as follows:
 * <ol>
 * <li>Determine the charset used to read the @charset from the stream. If a
 * BOM is present and a matching Charset is present, this charset is used. As
 * a fallback the CSS file is initially read with ISO-8859-1.</li>
 * <li>If the CSS content contains a valid @charset rule, the defined charset
 * is returned even if a different BOM is present.</li>
 * <li>If the CSS content does not contain a valid @charset rule than the
 * charset of the BOM is returned (if any).</li>
 * <li>Otherwise <code>null</code> is returned.</li>
 * </ol>
 *
 * @param aISP
 *        The input stream provider to read from. May not be <code>null</code>
 *        .
 * @return <code>null</code> if the input stream could not be opened or if
 *         neither a BOM nor a charset is specified. Otherwise a non-
 *         <code>null</code> Charset is returned.
 * @throws IllegalStateException
 *         if an invalid charset is supplied
 */
@Nullable
public static Charset getCharsetDeclaredInCSS(@Nonnull final IHasInputStream aISP) {
    ValueEnforcer.notNull(aISP, "InputStreamProvider");
    // Try to open input stream
    final InputStream aIS = aISP.getInputStream();
    if (aIS == null)
        return null;
    final InputStreamAndCharset aISAndBOM = CharsetHelper.getInputStreamAndCharsetFromBOM(aIS);
    final Charset aBOMCharset = aISAndBOM.getCharset();
    Charset aStreamCharset = aBOMCharset;
    if (aStreamCharset == null) {
        // Always read as ISO-8859-1 as everything contained in the CSS charset
        // declaration can be handled by this charset
        // A known problem is when the file is UTF-16, UTF-16BE, UTF-16LE etc.
        // encoded. In this case a BOM must be present to read the file correctly!
        aStreamCharset = StandardCharsets.ISO_8859_1;
    }
    final Reader aReader = StreamHelper.createReader(aISAndBOM.getInputStream(), aStreamCharset);
    try {
        // Read with the Stream charset
        final CSSCharStream aCharStream = new CSSCharStream(aReader);
        final ParserCSSCharsetDetectorTokenManager aTokenHdl = new ParserCSSCharsetDetectorTokenManager(aCharStream);
        final ParserCSSCharsetDetector aParser = new ParserCSSCharsetDetector(aTokenHdl);
        final String sCharsetName = aParser.styleSheetCharset().getText();
        if (sCharsetName == null) {
            // No charset specified - use the one from the BOM (may be null)
            return aBOMCharset;
        }
        // Remove leading and trailing quotes from value
        final String sPlainCharsetName = CSSParseHelper.extractStringValue(sCharsetName);
        final Charset aReadCharset = CharsetHelper.getCharsetFromName(sPlainCharsetName);
        if (aBOMCharset != null && !aBOMCharset.equals(aReadCharset)) {
            // BOM charset different from read charset
            s_aLogger.warn("The charset found in the CSS data (" + aReadCharset.name() + ") differs from the charset determined by the BOM (" + aBOMCharset.name() + ") -> Using the read charset");
        }
        return aReadCharset;
    } catch (final ParseException ex) {
        // grammar!
        throw new IllegalStateException("Failed to parse CSS charset definition", ex);
    } catch (final Throwable ex) {
        // As e.g. indicated by https://github.com/phax/ph-css/issues/9
        throw new IllegalStateException("Failed to parse CSS charset definition", ex);
    } finally {
        StreamHelper.close(aReader);
    }
}