Examples with TikaException org.apache.tika.exception.TikaException used on opensource projects

Example 36 with TikaException

use of org.apache.tika.exception.TikaException in project tika by apache.

the class MicrosoftTranslator method translate.

/**
     * Use the Microsoft service to translate the given text from the given source language to the given target.
     * You must set the client keys in translator.microsoft.properties.
     *
     * @param text The text to translate.
     * @param sourceLanguage The input text language (for example, "en").
     * @param targetLanguage The desired language to translate to (for example, "fr").
     * @return The translated text. If translation is unavailable, returns the unchanged text.
     * @throws Exception
     * @see org.apache.tika.language.translate.Translator
     * @since Tika 1.6
     */
public String translate(String text, String sourceLanguage, String targetLanguage) throws TikaException, IOException {
    if (!available)
        return text;
    Language source = Language.fromString(sourceLanguage);
    Language target = Language.fromString(targetLanguage);
    Translate.setClientId(clientId);
    Translate.setClientSecret(clientSecret);
    try {
        return Translate.execute(text, source, target);
    } catch (Exception e) {
        throw new TikaException("Error with Microsoft Translation: " + e.getMessage());
    }
}

Example 37 with TikaException

use of org.apache.tika.exception.TikaException in project tika by apache.

the class YandexTranslator method translate.

@Override
public String translate(String text, String sourceLanguage, String targetLanguage) throws TikaException, IOException {
    if (!this.isAvailable()) {
        return text;
    }
    WebClient client = WebClient.create(YANDEX_TRANSLATE_URL_BASE);
    String langCode;
    if (sourceLanguage == null) {
        //Translate Service will identify source language
        langCode = targetLanguage;
    } else {
        //Source language is well known
        langCode = sourceLanguage + '-' + targetLanguage;
    }
    //TODO Add support for text over 10k characters
    Response response = client.accept(MediaType.APPLICATION_JSON).query("key", this.apiKey).query("lang", langCode).query("text", text).get();
    BufferedReader reader = new BufferedReader(new InputStreamReader((InputStream) response.getEntity(), UTF_8));
    String line = null;
    StringBuffer responseText = new StringBuffer();
    while ((line = reader.readLine()) != null) {
        responseText.append(line);
    }
    try {
        ObjectMapper mapper = new ObjectMapper();
        JsonNode jsonResp = mapper.readTree(responseText.toString());
        if (!jsonResp.findValuesAsText("code").isEmpty()) {
            String code = jsonResp.findValuesAsText("code").get(0);
            if (code.equals("200")) {
                return jsonResp.findValue("text").get(0).asText();
            } else {
                throw new TikaException(jsonResp.findValue("message").get(0).asText());
            }
        } else {
            throw new TikaException("Return message not recognized: " + responseText.toString().substring(0, Math.min(responseText.length(), 100)));
        }
    } catch (JsonParseException e) {
        throw new TikaException("Error requesting translation from '" + sourceLanguage + "' to '" + targetLanguage + "', JSON response from Lingo24 is not well formatted: " + responseText.toString());
    }
}

Example 38 with TikaException

use of org.apache.tika.exception.TikaException in project tika by apache.

the class JoshuaNetworkTranslator method translate.

/**
   * <p>Initially then check if the source language has been provided.
   * If no source language (or a null value) has been provided then
   * we make an attempt to guess the source using Tika's
   * {@link org.apache.tika.langdetect.OptimaizeLangDetector}. If we
   * are still unable to guess the language then we return the source
   * text.</p>
   * 
   * <p>We then process the input text into a new string consisting of 
   * sentences, one per line e.g. insert \n between the presence of '.'</p>
   * 
   * @see org.apache.tika.language.translate.Translator#translate
   * (java.lang.String, java.lang.String, java.lang.String)
   */
@Override
public String translate(String text, String sourceLanguage, String targetLanguage) throws TikaException, IOException {
    //create networkURI
    if (!networkServer.endsWith("/")) {
        networkURI = networkServer + "/" + targetLanguage;
    } else {
        networkURI = networkServer + targetLanguage;
    }
    if (!this.isAvailable())
        return text;
    //make an attempt to guess language if one is not provided.
    if (sourceLanguage == null)
        sourceLanguage = detectLanguage(text).getLanguage();
    //process input text into sentences, one per line 
    // e.g. insert \n between the presence of '.'
    StringBuilder sb = new StringBuilder(text);
    int i = 0;
    while ((i = sb.indexOf(".", i + 1)) != -1) {
        sb.replace(i, i + 1, "\n");
    }
    String inputText = sb.toString();
    WebClient client;
    final List<Object> providers = new ArrayList<>();
    JacksonJsonProvider jacksonJsonProvider = new JacksonJsonProvider();
    providers.add(jacksonJsonProvider);
    client = WebClient.create(networkURI, providers);
    ObjectMapper requestMapper = new ObjectMapper();
    ObjectNode jsonNode = requestMapper.createObjectNode();
    jsonNode.put("inputLanguage", sourceLanguage);
    jsonNode.put("inputText", inputText);
    //make the reuest
    Response response = client.accept(MediaType.APPLICATION_JSON).type(MediaType.APPLICATION_JSON).post(jsonNode);
    BufferedReader reader = new BufferedReader(new InputStreamReader((InputStream) response.getEntity(), UTF_8));
    String line;
    StringBuilder responseText = new StringBuilder();
    while ((line = reader.readLine()) != null) {
        responseText.append(line);
    }
    try {
        ObjectMapper responseMapper = new ObjectMapper();
        JsonNode jsonResp = responseMapper.readTree(responseText.toString());
        if (jsonResp.findValuesAsText("outputText") != null) {
            return jsonResp.findValuesAsText("outputText").get(0);
        } else {
            throw new TikaException(jsonResp.findValue("message").get(0).asText());
        }
    } catch (JsonParseException e) {
        throw new TikaException("Error requesting translation from '" + sourceLanguage + "' to '" + targetLanguage + "', JSON response " + "from Joshua REST Server is not well formatted: " + responseText.toString());
    }
}

Example 39 with TikaException

use of org.apache.tika.exception.TikaException in project tika by apache.

the class TikaToXMP method getConverter.

/**
     * Retrieve a specific converter according to the mimetype
     *
     * @param mimetype
     *            the Mimetype
     * @return the converter or null, if none exists
     * @throws TikaException
     */
public static ITikaToXMPConverter getConverter(String mimetype) throws TikaException {
    if (mimetype == null) {
        throw new IllegalArgumentException("mimetype must not be null");
    }
    ITikaToXMPConverter converter = null;
    MediaType type = MediaType.parse(mimetype);
    if (type != null) {
        Class<? extends ITikaToXMPConverter> clazz = getConverterMap().get(type);
        if (clazz != null) {
            try {
                converter = clazz.newInstance();
            } catch (Exception e) {
                throw new TikaException("TikaToXMP converter class cannot be instantiated for mimetype: " + type.toString(), e);
            }
        }
    }
    return converter;
}

Example 40 with TikaException

use of org.apache.tika.exception.TikaException in project tika by apache.

the class NetCDFParser method parse.

/*
     * (non-Javadoc)
     * 
     * @see org.apache.tika.parser.Parser#parse(java.io.InputStream,
     * org.xml.sax.ContentHandler, org.apache.tika.metadata.Metadata,
     * org.apache.tika.parser.ParseContext)
     */
public void parse(InputStream stream, ContentHandler handler, Metadata metadata, ParseContext context) throws IOException, SAXException, TikaException {
    TemporaryResources tmp = TikaInputStream.isTikaInputStream(stream) ? null : new TemporaryResources();
    TikaInputStream tis = TikaInputStream.get(stream, tmp);
    NetcdfFile ncFile = null;
    try {
        ncFile = NetcdfFile.open(tis.getFile().getAbsolutePath());
        metadata.set("File-Type-Description", ncFile.getFileTypeDescription());
        // first parse out the set of global attributes
        for (Attribute attr : ncFile.getGlobalAttributes()) {
            Property property = resolveMetadataKey(attr.getFullName());
            if (attr.getDataType().isString()) {
                metadata.add(property, attr.getStringValue());
            } else if (attr.getDataType().isNumeric()) {
                int value = attr.getNumericValue().intValue();
                metadata.add(property, String.valueOf(value));
            }
        }
        XHTMLContentHandler xhtml = new XHTMLContentHandler(handler, metadata);
        xhtml.startDocument();
        xhtml.newline();
        xhtml.element("h1", "dimensions");
        xhtml.startElement("ul");
        xhtml.newline();
        for (Dimension dim : ncFile.getDimensions()) {
            xhtml.element("li", dim.getFullName() + " = " + dim.getLength());
        }
        xhtml.endElement("ul");
        xhtml.element("h1", "variables");
        xhtml.startElement("ul");
        xhtml.newline();
        for (Variable var : ncFile.getVariables()) {
            xhtml.startElement("li");
            xhtml.characters(var.getDataType() + " " + var.getNameAndDimensions());
            xhtml.newline();
            List<Attribute> attributes = var.getAttributes();
            if (!attributes.isEmpty()) {
                xhtml.startElement("ul");
                for (Attribute element : attributes) {
                    xhtml.element("li", element.toString());
                }
                xhtml.endElement("ul");
            }
            xhtml.endElement("li");
        }
        xhtml.endElement("ul");
        xhtml.endDocument();
    } catch (IOException e) {
        throw new TikaException("NetCDF parse error", e);
    } finally {
        if (ncFile != null) {
            ncFile.close();
        }
        if (tmp != null) {
            tmp.dispose();
        }
    }
}