Java Code Examples for org.apache.tika.metadata.Metadata#isMultiValued()

The following examples show how to use org.apache.tika.metadata.Metadata#isMultiValued() . You can vote up the ones you like or vote down the ones you don't like, and go to the original project or source file by following the links above each example. You may check out the related API usage on the sidebar.
Example 1
Source File: TikaPoweredMetadataExtracter.java    From alfresco-repository with GNU Lesser General Public License v3.0 6 votes vote down vote up 
private String getMetadataValue(Metadata metadata, String key)
{
    if (metadata.isMultiValued(key))
    {
        String[] parts = metadata.getValues(key);
        
        // use Set to prevent duplicates
        Set<String> value = new LinkedHashSet<String>(parts.length);
        
        for (int i = 0; i < parts.length; i++)
        {
            value.add(parts[i]);
        }
        
        String valueStr = value.toString();
        
        // remove leading/trailing braces []
        return valueStr.substring(1, valueStr.length() - 1);
    }
    else
    {
        return metadata.get(key);
    }
}
Example 2
Source File: ExtractMediaMetadata.java    From localization_nifi with Apache License 2.0 4 votes vote down vote up 
private Map<String, String> tika_parse(InputStream sourceStream, String prefix, Integer maxAttribs,
                                       Integer maxAttribLen) throws IOException, TikaException, SAXException {
    final Metadata metadata = new Metadata();
    final TikaInputStream tikaInputStream = TikaInputStream.get(sourceStream);
    autoDetectParser.parse(tikaInputStream, new DefaultHandler(), metadata);

    final Map<String, String> results = new HashMap<>();
    final Pattern metadataKeyFilter = metadataKeyFilterRef.get();
    final StringBuilder dataBuilder = new StringBuilder();
    for (final String key : metadata.names()) {
        if (metadataKeyFilter != null && !metadataKeyFilter.matcher(key).matches()) {
            continue;
        }
        dataBuilder.setLength(0);
        if (metadata.isMultiValued(key)) {
            for (String val : metadata.getValues(key)) {
                if (dataBuilder.length() > 1) {
                    dataBuilder.append(", ");
                }
                if (dataBuilder.length() + val.length() < maxAttribLen) {
                    dataBuilder.append(val);
                } else {
                    dataBuilder.append("...");
                    break;
                }
            }
        } else {
            dataBuilder.append(metadata.get(key));
        }
        if (prefix == null) {
            results.put(key, dataBuilder.toString().trim());
        } else {
            results.put(prefix + key, dataBuilder.toString().trim());
        }

        // cutoff at max if provided
        if (maxAttribs != null && results.size() >= maxAttribs) {
            break;
        }
    }
    return results;
}
Example 3
Source File: ExtractMediaMetadata.java    From nifi with Apache License 2.0 4 votes vote down vote up 
private Map<String, String> tika_parse(InputStream sourceStream, String prefix, Integer maxAttribs,
                                       Integer maxAttribLen) throws IOException, TikaException, SAXException {
    final Metadata metadata = new Metadata();
    final TikaInputStream tikaInputStream = TikaInputStream.get(sourceStream);
    try {
        autoDetectParser.parse(tikaInputStream, new DefaultHandler(), metadata);
    } finally {
        tikaInputStream.close();
    }

    final Map<String, String> results = new HashMap<>();
    final Pattern metadataKeyFilter = metadataKeyFilterRef.get();
    final StringBuilder dataBuilder = new StringBuilder();
    for (final String key : metadata.names()) {
        if (metadataKeyFilter != null && !metadataKeyFilter.matcher(key).matches()) {
            continue;
        }
        dataBuilder.setLength(0);
        if (metadata.isMultiValued(key)) {
            for (String val : metadata.getValues(key)) {
                if (dataBuilder.length() > 1) {
                    dataBuilder.append(", ");
                }
                if (dataBuilder.length() + val.length() < maxAttribLen) {
                    dataBuilder.append(val);
                } else {
                    dataBuilder.append("...");
                    break;
                }
            }
        } else {
            dataBuilder.append(metadata.get(key));
        }
        if (prefix == null) {
            results.put(key, dataBuilder.toString().trim());
        } else {
            results.put(prefix + key, dataBuilder.toString().trim());
        }

        // cutoff at max if provided
        if (maxAttribs != null && results.size() >= maxAttribs) {
            break;
        }
    }
    return results;
}