[CALCITE-6805] Support hex and unhex for Hive and Spark Libaray

apache · Jan 29, 2025 · 97f7e9b · 97f7e9b
1 parent d83122b
commit 97f7e9b
Show file tree

Hide file tree

Showing 6 changed files with 76 additions and 0 deletions.
diff --git a/core/src/main/java/org/apache/calcite/adapter/enumerable/RexImpTable.java b/core/src/main/java/org/apache/calcite/adapter/enumerable/RexImpTable.java
@@ -215,6 +215,7 @@
 import static org.apache.calcite.sql.fun.SqlLibraryOperators.FROM_BASE64;
 import static org.apache.calcite.sql.fun.SqlLibraryOperators.FROM_HEX;
 import static org.apache.calcite.sql.fun.SqlLibraryOperators.GETBIT;
+import static org.apache.calcite.sql.fun.SqlLibraryOperators.HEX;
 import static org.apache.calcite.sql.fun.SqlLibraryOperators.ILIKE;
 import static org.apache.calcite.sql.fun.SqlLibraryOperators.IS_INF;
 import static org.apache.calcite.sql.fun.SqlLibraryOperators.IS_NAN;
@@ -330,6 +331,7 @@
 import static org.apache.calcite.sql.fun.SqlLibraryOperators.TRANSLATE3;
 import static org.apache.calcite.sql.fun.SqlLibraryOperators.TRUNC_BIG_QUERY;
 import static org.apache.calcite.sql.fun.SqlLibraryOperators.TRY_CAST;
+import static org.apache.calcite.sql.fun.SqlLibraryOperators.UNHEX;
 import static org.apache.calcite.sql.fun.SqlLibraryOperators.UNIX_DATE;
 import static org.apache.calcite.sql.fun.SqlLibraryOperators.UNIX_MICROS;
 import static org.apache.calcite.sql.fun.SqlLibraryOperators.UNIX_MILLIS;
@@ -686,6 +688,8 @@ void populate1() {
       defineMethod(FROM_BASE64, BuiltInMethod.FROM_BASE64.method, NullPolicy.STRICT);
       defineMethod(TO_BASE32, BuiltInMethod.TO_BASE32.method, NullPolicy.STRICT);
       defineMethod(FROM_BASE32, BuiltInMethod.FROM_BASE32.method, NullPolicy.STRICT);
+      defineMethod(HEX, BuiltInMethod.HEX.method, NullPolicy.STRICT);
+      defineMethod(UNHEX, BuiltInMethod.UNHEX.method, NullPolicy.STRICT);
       defineMethod(TO_HEX, BuiltInMethod.TO_HEX.method, NullPolicy.STRICT);
       defineMethod(FROM_HEX, BuiltInMethod.FROM_HEX.method, NullPolicy.STRICT);
       defineMethod(MD5, BuiltInMethod.MD5.method, NullPolicy.STRICT);

diff --git a/core/src/main/java/org/apache/calcite/runtime/SqlFunctions.java b/core/src/main/java/org/apache/calcite/runtime/SqlFunctions.java
@@ -359,6 +359,22 @@ public static ByteString fromHex(String hex) {
     }
   }
 
+  /** SQL HEX(varchar) function. */
+  public static String hex(String value) {
+    return Hex.encodeHexString(value.getBytes(UTF_8));
+  }
+
+  /** SQL UNHEX(varchar) function. */
+  public static String unhex(String value) {
+    try {
+      //CHECKSTYLE: IGNORE 1
+      return new String(Hex.decodeHex(value), UTF_8);
+    } catch (DecoderException e) {
+      throw new IllegalArgumentException(
+          String.format(Locale.ROOT, "Failed to decode hex string: %s", value), e);
+    }
+  }
+
   /** SQL TO_HEX(binary) function. */
   public static String toHex(ByteString byteString) {
     return Hex.encodeHexString(byteString.getBytes());

diff --git a/core/src/main/java/org/apache/calcite/sql/fun/SqlLibraryOperators.java b/core/src/main/java/org/apache/calcite/sql/fun/SqlLibraryOperators.java
@@ -1890,6 +1890,26 @@ private static RelDataType deriveTypeMapFromEntries(SqlOperatorBinding opBinding
           OperandTypes.BINARY,
           SqlFunctionCategory.STRING);
 
+  /**
+   * The "HEX(string)" function; converts {@code string} into a hexadecimal varchar.
+   */
+  @LibraryOperator(libraries = {HIVE, SPARK})
+  public static final SqlFunction HEX =
+      SqlBasicFunction.create("HEX",
+          ReturnTypes.VARCHAR_NULLABLE,
+          OperandTypes.CHARACTER,
+          SqlFunctionCategory.STRING);
+
+  /**
+   * The "UNHEX(string)" function; converts {@code string} into a hexadecimal varchar.
+   */
+  @LibraryOperator(libraries = {HIVE, SPARK})
+  public static final SqlFunction UNHEX =
+      SqlBasicFunction.create("UNHEX",
+          ReturnTypes.VARCHAR_NULLABLE,
+          OperandTypes.CHARACTER,
+          SqlFunctionCategory.STRING);
+
   /** The "FORMAT_NUMBER(value, decimalOrFormat)" function. */
   @LibraryOperator(libraries = {HIVE, SPARK})
   public static final SqlFunction FORMAT_NUMBER =

diff --git a/core/src/main/java/org/apache/calcite/util/BuiltInMethod.java b/core/src/main/java/org/apache/calcite/util/BuiltInMethod.java
@@ -435,6 +435,8 @@ public enum BuiltInMethod {
   FROM_BASE64(SqlFunctions.class, "fromBase64", String.class),
   TO_BASE32(SqlFunctions.class, "toBase32", String.class),
   FROM_BASE32(SqlFunctions.class, "fromBase32", String.class),
+  HEX(SqlFunctions.class, "hex", String.class),
+  UNHEX(SqlFunctions.class, "unhex", String.class),
   TO_HEX(SqlFunctions.class, "toHex", ByteString.class),
   FROM_HEX(SqlFunctions.class, "fromHex", String.class),
   MD5(SqlFunctions.class, "md5", String.class),

diff --git a/site/_docs/reference.md b/site/_docs/reference.md
@@ -2922,6 +2922,8 @@ In the following:
 | b | FROM_BASE32(string)                            | Returns the decoded result of a base-32 *string* as a string
 | m | TO_BASE64(string)                              | Converts the *string* to base-64 encoded form and returns a encoded string
 | b m | FROM_BASE64(string)                          | Returns the decoded result of a base-64 *string* as a string. If the input argument is an invalid base-64 *string* the function returns `NULL`
+| h s | HEX(string)                                   | Converts *string* into a hexadecimal varchar
+| h s | UNHEX(string)                                 | Converts a hexadecimal-encoded *varchar* into varchar
 | b | TO_HEX(binary)                                 | Converts *binary* into a hexadecimal varchar
 | b | FROM_HEX(varchar)                              | Converts a hexadecimal-encoded *varchar* into bytes
 | b o p r s | LTRIM(string)                          | Returns *string* with all blanks removed from the start

diff --git a/testkit/src/main/java/org/apache/calcite/test/SqlOperatorTest.java b/testkit/src/main/java/org/apache/calcite/test/SqlOperatorTest.java
@@ -5823,6 +5823,38 @@ void testBitGetFunc(SqlOperatorFixture f, String functionName) {
     f0.forEachLibrary(libraries, consumer);
   }
 
+  @Test void testHex() {
+    SqlOperatorFixture sqlOperatorFixture = fixture();
+    final SqlOperatorFixture f0 = sqlOperatorFixture.setFor(SqlLibraryOperators.HEX);
+    f0.checkFails("^hex('')^",
+        "No match found for function signature HEX\\(<CHARACTER>\\)",
+        false);
+    final Consumer<SqlOperatorFixture> consumer = f -> {
+      f.checkString("hex('abc')",
+          "616263",
+          "VARCHAR NOT NULL");
+      f.checkString("hex('')", "", "VARCHAR NOT NULL");
+      f.checkNull("hex(cast(null as varbinary))");
+    };
+    f0.forEachLibrary(list(SqlLibrary.HIVE, SqlLibrary.SPARK), consumer);
+  }
+
+  @Test void testUnHex() {
+    SqlOperatorFixture sqlOperatorFixture = fixture();
+    final SqlOperatorFixture f0 = sqlOperatorFixture.setFor(SqlLibraryOperators.UNHEX);
+    f0.checkFails("^unhex('')^",
+        "No match found for function signature UNHEX\\(<CHARACTER>\\)",
+        false);
+    final Consumer<SqlOperatorFixture> consumer = f -> {
+      f.checkString("unhex('616263')",
+          "abc",
+          "VARCHAR NOT NULL");
+      f.checkString("unhex('')", "", "VARCHAR NOT NULL");
+      f.checkNull("unhex(cast(null as varbinary))");
+    };
+    f0.forEachLibrary(list(SqlLibrary.HIVE, SqlLibrary.SPARK), consumer);
+  }
+
   @Test void testToHex() {
     final SqlOperatorFixture f0 = fixture().setFor(SqlLibraryOperators.TO_HEX);
     f0.checkFails("^to_hex(x'')^",