/**
 * Helper class for document processing and manipulation in the Fess search system.
 * This class provides utilities for processing document content, titles, and digests,
 * handling text normalization, content extraction, and similar document hash encoding/decoding.
 * It also manages document processing requests and integrates with the crawler system.
 *
 */
public class DocumentHelper {

- **SitemapsHelper**: Sitemap XML parsing, index handling
- **MimeTypeHelper**: MIME detection via Tika
- **EncodingHelper**: Charset detection with BOM
- **UrlConvertHelper**: URL normalization
- **ContentLengthHelper**: Content length limits per MIME type

---

## Development Workflow

### Build Commands

```bash
mvn clean install              # Build all

Aliases:
├── Search Alias: {baseIndex}          (read operations)
└── Update Alias: {baseIndex}.update   (write operations)
```

### Text Processing Pipeline

```
Input Text
    ↓
[Normalization] - NormalizerChain
    ↓
[Reading Conversion] - ReadingConverterChain
    ↓
[Analysis] - SuggestAnalyzer
```

Note: Suggester has TWO ReadingConverter instances:
- `readingConverter` - For query/metadata fields

/**
 * Builds OpenSearch queries for suggestion requests.
 *
 * <p>This class encapsulates the query building logic extracted from SuggestRequest,
 * including reading conversion, normalization, and function score query construction.</p>
 */
public class SuggestQueryBuilder {

    private final ReadingConverter readingConverter;
    private final Normalizer normalizer;

          "internal-instrumentation-api",
          "java-api-extractor",
          "java-compiler-plugin",
          "javadoc",
          "kotlin-dsl-integ-tests",
          "kotlin-dsl-plugins",
          "normalization-java"
        ],
        "parallelizationMethod": {
          "name": "TestDistribution"
        }
      },
      {
        "subprojects": [
          "build-cache-example-client",

    "name": "native",
    "path": "platforms/core-runtime/native",
    "unitTests": true,
    "functionalTests": false,
    "crossVersionTests": false
  },
  {
    "name": "normalization-java",
    "path": "platforms/jvm/normalization-java",
    "unitTests": true,
    "functionalTests": false,
    "crossVersionTests": false
  },
  {
    "name": "performance",
    "path": "testing/performance",

            assertNotNull(languages);
            if (languages.length > 0) {
                // Note: The actual order might be different due to normalization
                boolean hasEn = false, hasJa = false;
                for (String lang : languages) {
                    if ("en".equals(lang))
                        hasEn = true;

  /** Returns the domain name, normalized to all lower case. */
  @Override
  public String toString() {
    return name;
  }

  /**
   * Equality testing is based on the text supplied by the caller, after normalization as described
   * in the class documentation. For example, a non-ASCII Unicode domain name and the Punycode
   * version of the same domain name would not be considered equal.
   */
  @Override

            <option value="$PROJECT_DIR$/platforms/jvm/language-java" />
            <option value="$PROJECT_DIR$/platforms/jvm/language-jvm" />
            <option value="$PROJECT_DIR$/platforms/jvm/normalization-java" />
            <option value="$PROJECT_DIR$/platforms/jvm/platform-jvm" />
            <option value="$PROJECT_DIR$/platforms/jvm/plugins-application" />

            dimension if necessary. Use this mode on rank 1 inputs for the old
            `"binary"` behavior of one-hot encoding a batch of scalars.
        *   `Normalization` will no longer automatically uprank rank 1 inputs,
            allowing normalization of unbatched multi-dimensional samples.

*   `tf.lite`:

    *   The recommended Android NDK version for building TensorFlow Lite has