Examples with CorpusConfig annis.service.objects.CorpusConfig used on opensource projects

Example 1 with CorpusConfig

use of annis.service.objects.CorpusConfig in project ANNIS by korpling.

the class QueryDaoImpl method getCorpusConfigurations.

@Override
public CorpusConfigMap getCorpusConfigurations() {
    List<AnnisCorpus> annisCorpora = listCorpora();
    CorpusConfigMap cConfigs = new CorpusConfigMap();
    if (annisCorpora != null) {
        for (AnnisCorpus c : annisCorpora) {
            try {
                Properties p = getCorpusConfiguration(c.getName());
                if (p != null) {
                    CorpusConfig corpusConfig = new CorpusConfig();
                    corpusConfig.setConfig(p);
                    cConfigs.put(c.getName(), corpusConfig);
                }
            } catch (FileNotFoundException ex) {
                log.error("no corpus.properties found for {}", c.getName());
            }
        }
    }
    return cConfigs;
}

Example 2 with CorpusConfig

use of annis.service.objects.CorpusConfig in project ANNIS by korpling.

the class SaltBasedExporter method convertSaltProject.

/**
 * Iterates over all matches (modelled as corpus graphs) and calls {@link #convertText(de.hu_berlin.german.korpling.saltnpepper.salt.saltCommon.sDocumentStructure.SDocumentGraph, java.util.List, java.util.Map, int, java.io.Writer) } for
 * the single document graph.
 * @param p
 * @param annoKeys
 * @param args
 * @param alignmc
 * @param offset
 * @param out
 */
// invokes the createAdjacencyMatrix method, if nodeCount != null or outputText otherwise
private void convertSaltProject(SaltProject p, List<String> annoKeys, Map<String, String> args, boolean alignmc, int offset, Map<String, CorpusConfig> corpusConfigs, Writer out, Integer nodeCount) throws IOException, IllegalArgumentException {
    int recordNumber = offset;
    if (p != null && p.getCorpusGraphs() != null) {
        Map<String, String> spanAnno2order = null;
        boolean virtualTokenizationFromNamespace = false;
        Set<String> corpusNames = CommonHelper.getToplevelCorpusNames(p);
        if (!corpusNames.isEmpty()) {
            CorpusConfig config = corpusConfigs.get(corpusNames.iterator().next());
            if (config != null) {
                if ("true".equalsIgnoreCase(config.getConfig("virtual_tokenization_from_namespace"))) {
                    virtualTokenizationFromNamespace = true;
                } else {
                    String mappingRaw = config.getConfig("virtual_tokenization_mapping");
                    if (mappingRaw != null) {
                        spanAnno2order = new HashMap<>();
                        for (String singleMapping : Splitter.on(',').split(mappingRaw)) {
                            List<String> mappingParts = Splitter.on('=').splitToList(singleMapping);
                            if (mappingParts.size() >= 2) {
                                spanAnno2order.put(mappingParts.get(0), mappingParts.get(1));
                            }
                        }
                    }
                }
            }
        }
        for (SCorpusGraph corpusGraph : p.getCorpusGraphs()) {
            if (corpusGraph.getDocuments() != null) {
                for (SDocument doc : corpusGraph.getDocuments()) {
                    if (virtualTokenizationFromNamespace) {
                        TimelineReconstructor.removeVirtualTokenizationUsingNamespace(doc.getDocumentGraph());
                    } else if (spanAnno2order != null) {
                        // there is a definition how to map the virtual tokenization to a real one
                        TimelineReconstructor.removeVirtualTokenization(doc.getDocumentGraph(), spanAnno2order);
                    }
                    if (nodeCount != null) {
                        createAdjacencyMatrix(doc.getDocumentGraph(), args, recordNumber++, nodeCount);
                    } else {
                        outputText(doc.getDocumentGraph(), alignmc, recordNumber++, out);
                    }
                }
            }
        }
    }
}

Example 3 with CorpusConfig

use of annis.service.objects.CorpusConfig in project ANNIS by korpling.

the class SearchOptionsPanel method mergeConfigs.

/**
 * Builds a single config for selection of one or muliple corpora.
 *
 * @param corpora Specifies the combination of corpora, for which the config
 * is calculated.
 * @param corpusConfigurations  A map containg the known corpus configurations.
 * @return A new config which takes into account the segementation of all
 * selected corpora.
 */
private CorpusConfig mergeConfigs(Set<String> corpora, CorpusConfigMap corpusConfigurations) {
    CorpusConfig corpusConfig = new CorpusConfig();
    // calculate the left and right context.
    String leftCtx = mergeConfigValue(KEY_MAX_CONTEXT_LEFT, corpora, corpusConfigurations);
    String rightCtx = mergeConfigValue(KEY_MAX_CONTEXT_RIGHT, corpora, corpusConfigurations);
    corpusConfig.setConfig(KEY_MAX_CONTEXT_LEFT, leftCtx);
    corpusConfig.setConfig(KEY_MAX_CONTEXT_RIGHT, rightCtx);
    // calculate the default-context
    corpusConfig.setConfig(KEY_CONTEXT_STEPS, mergeConfigValue(KEY_CONTEXT_STEPS, corpora, corpusConfigurations));
    corpusConfig.setConfig(KEY_DEFAULT_CONTEXT, mergeConfigValue(KEY_DEFAULT_CONTEXT, corpora, corpusConfigurations));
    // get the results per page
    corpusConfig.setConfig(KEY_RESULT_PER_PAGE, mergeConfigValue(KEY_RESULT_PER_PAGE, corpora, corpusConfigurations));
    corpusConfig.setConfig(KEY_DEFAULT_CONTEXT_SEGMENTATION, checkSegments(KEY_DEFAULT_CONTEXT_SEGMENTATION, corpora, corpusConfigurations));
    corpusConfig.setConfig(KEY_DEFAULT_BASE_TEXT_SEGMENTATION, checkSegments(KEY_DEFAULT_BASE_TEXT_SEGMENTATION, corpora, corpusConfigurations));
    return corpusConfig;
}

Example 4 with CorpusConfig

use of annis.service.objects.CorpusConfig in project ANNIS by korpling.

the class SearchOptionsPanel method checkSegments.

/**
 * Checks, if all selected corpora have the same default segmentation layer.
 * If not the tok layer is taken, because every corpus has this one.
 *
 * @param key the key for the segementation config, must be
 * {@link #KEY_DEFAULT_BASE_TEXT_SEGMENTATION} or
 * {@link #KEY_DEFAULT_CONTEXT_SEGMENTATION}.
 * @param corpora the corpora which has to be checked.
 * @return "tok" or a segment which is defined in all corpora.
 */
private String checkSegments(String key, Set<String> corpora, CorpusConfigMap corpusConfigurations) {
    String segmentation = null;
    for (String corpus : corpora) {
        CorpusConfig c = null;
        if (corpusConfigurations.containsConfig(corpus)) {
            c = corpusConfigurations.get(corpus);
        } else {
            c = corpusConfigurations.get(DEFAULT_CONFIG);
        }
        // do nothing if not even default config is set
        if (c == null) {
            continue;
        }
        String tmpSegment = c.getConfig(key);
        /**
         * If no segment is set in the corpus config use always the tok segment.
         */
        if (tmpSegment == null) {
            return corpusConfigurations.get(DEFAULT_CONFIG).getConfig(key);
        }
        if (segmentation == null) {
            segmentation = tmpSegment;
            continue;
        }
        if (// return the default config
        !segmentation.equals(tmpSegment)) {
            return corpusConfigurations.get(DEFAULT_CONFIG).getConfig(key);
        }
    }
    if (segmentation == null) {
        return corpusConfigurations.get(DEFAULT_CONFIG).getConfig(key);
    } else {
        return segmentation;
    }
}

Example 5 with CorpusConfig

use of annis.service.objects.CorpusConfig in project ANNIS by korpling.

the class DocBrowserController method docsAvailable.

public boolean docsAvailable(String id) {
    if (ui != null) {
        CorpusConfig corpusConfig = ui.getCorpusConfigWithCache(id);
        if (corpusConfig != null) {
            if (corpusConfig.containsKey("browse-documents")) {
                return Boolean.parseBoolean(corpusConfig.getConfig("browse-documents"));
            } else // get the default config
            {
                corpusConfig = ui.getCorpusConfigWithCache(Helper.DEFAULT_CONFIG);
                boolean browseDocuments = Boolean.parseBoolean(corpusConfig.getConfig("browse-documents", "true"));
                return browseDocuments;
            }
        }
    }
    return true;
}