expand an reuse UDFLike#likePatternToRegExp

apache · Jun 29, 2024 · 14e9f76 · 14e9f76
1 parent 2c22ae8
commit 14e9f76
Show file tree

Hide file tree

Showing 11 changed files with 21 additions and 47 deletions.
diff --git a/ql/src/java/org/apache/hadoop/hive/ql/ddl/database/show/ShowDatabasesOperation.java b/ql/src/java/org/apache/hadoop/hive/ql/ddl/database/show/ShowDatabasesOperation.java
@@ -45,7 +45,7 @@ public int execute() throws HiveException {
     List<String> databases = context.getDb().getAllDatabases();
     if (desc.getPattern() != null) {
       LOG.debug("pattern: {}", desc.getPattern());
-      Pattern pattern = Pattern.compile(UDFLike.likePatternToRegExp(desc.getPattern()), Pattern.CASE_INSENSITIVE);
+      Pattern pattern = Pattern.compile(UDFLike.likePatternToRegExp(desc.getPattern(), true), Pattern.CASE_INSENSITIVE);
       databases = databases.stream().filter(name -> pattern.matcher(name).matches()).collect(Collectors.toList());
     }
 

diff --git a/...rc/java/org/apache/hadoop/hive/ql/ddl/dataconnector/show/ShowDataConnectorsOperation.java b/...rc/java/org/apache/hadoop/hive/ql/ddl/dataconnector/show/ShowDataConnectorsOperation.java
@@ -45,7 +45,7 @@ public int execute() throws HiveException {
     List<String> connectors = context.getDb().getAllDataConnectorNames();
     if (desc.getPattern() != null) {
       LOG.debug("pattern: {}", desc.getPattern());
-      Pattern pattern = Pattern.compile(UDFLike.likePatternToRegExp(desc.getPattern()), Pattern.CASE_INSENSITIVE);
+      Pattern pattern = Pattern.compile(UDFLike.likePatternToRegExp(desc.getPattern(), true), Pattern.CASE_INSENSITIVE);
       connectors = connectors.stream().filter(name -> pattern.matcher(name).matches()).collect(Collectors.toList());
     }
 

diff --git a/...c/java/org/apache/hadoop/hive/ql/ddl/table/info/show/status/ShowTableStatusOperation.java b/...c/java/org/apache/hadoop/hive/ql/ddl/table/info/show/status/ShowTableStatusOperation.java
@@ -63,7 +63,7 @@ public int execute() throws HiveException {
       LOG.debug("pattern: {}", desc.getPattern());
       List<String> tableNames = context.getDb().getTablesForDb(desc.getDbName(), null);
       if (desc.getPattern() != null) {
-        Pattern pattern = Pattern.compile(UDFLike.likePatternToRegExp(desc.getPattern()), Pattern.CASE_INSENSITIVE);
+        Pattern pattern = Pattern.compile(UDFLike.likePatternToRegExp(desc.getPattern(), true), Pattern.CASE_INSENSITIVE);
         tableNames = tableNames.stream()
             .filter(name -> pattern.matcher(name).matches())
             .collect(Collectors.toList());

diff --git a/ql/src/java/org/apache/hadoop/hive/ql/ddl/table/info/show/tables/ShowTablesOperation.java b/ql/src/java/org/apache/hadoop/hive/ql/ddl/table/info/show/tables/ShowTablesOperation.java
@@ -24,16 +24,15 @@
 import java.io.DataOutputStream;
 import java.util.ArrayList;
 import java.util.Collections;
-import java.util.Comparator;
 import java.util.List;
 import java.util.TreeMap;
 
 import org.apache.hadoop.fs.Path;
 import org.apache.hadoop.hive.metastore.TableType;
-import org.apache.hadoop.hive.metastore.utils.MetaStoreUtils;
 import org.apache.hadoop.hive.ql.ErrorMsg;
 import org.apache.hadoop.hive.ql.ddl.DDLOperation;
 import org.apache.hadoop.hive.ql.metadata.HiveException;
+import org.apache.hadoop.hive.ql.udf.UDFLike;
 
 /**
  * Operation process showing the tables.
@@ -59,7 +58,7 @@ public int execute() throws HiveException {
   }
 
   private void showTables() throws HiveException {
-    String pattern = MetaStoreUtils.convertSqlPatternToRegExp(desc.getPattern());
+    String pattern = UDFLike.likePatternToRegExp(desc.getPattern(), false);
     List<String> tableNames = new ArrayList<>(
         context.getDb().getTablesByType(desc.getDbName(), pattern, desc.getTypeFilter()));
     Collections.sort(tableNames);
@@ -75,13 +74,12 @@ private void showTables() throws HiveException {
 
   private void showTablesExtended() throws HiveException {
     TreeMap<String, String> tableNameToType = new TreeMap<>();
-    String pattern = MetaStoreUtils.convertSqlPatternToRegExp(desc.getPattern());
+    String pattern = UDFLike.likePatternToRegExp(desc.getPattern(), false);
     TableType typeFilter = desc.getTypeFilter();
     TableType[] tableTypes = typeFilter == null ? TableType.values() : new TableType[]{typeFilter};
     for (TableType tableType : tableTypes) {
-      String typeString = tableType.toString();
       List<String> tables = context.getDb().getTablesByType(desc.getDbName(), pattern, tableType);
-      tables.forEach(name -> tableNameToType.put(name, typeString));
+      tables.forEach(name -> tableNameToType.put(name, tableType.toString()));
     }
     LOG.debug("Found {} table(s) matching the SHOW EXTENDED TABLES statement.", tableNameToType.size());
 

diff --git a/.../org/apache/hadoop/hive/ql/ddl/view/materialized/show/ShowMaterializedViewsOperation.java b/.../org/apache/hadoop/hive/ql/ddl/view/materialized/show/ShowMaterializedViewsOperation.java
@@ -53,7 +53,7 @@ public int execute() throws HiveException {
     List<Table> viewObjects = new ArrayList<>(
         context.getDb().getMaterializedViewObjectsByPattern(desc.getDbName(), null));
     if (desc.getPattern() != null) {
-      Pattern pattern = Pattern.compile(UDFLike.likePatternToRegExp(desc.getPattern()), Pattern.CASE_INSENSITIVE);
+      Pattern pattern = Pattern.compile(UDFLike.likePatternToRegExp(desc.getPattern(), true), Pattern.CASE_INSENSITIVE);
       viewObjects = viewObjects.stream()
           .filter(object -> pattern.matcher(object.getTableName()).matches())
           .collect(Collectors.toList());

diff --git a/ql/src/java/org/apache/hadoop/hive/ql/ddl/view/show/ShowViewsOperation.java b/ql/src/java/org/apache/hadoop/hive/ql/ddl/view/show/ShowViewsOperation.java
@@ -24,13 +24,13 @@
 
 import org.apache.hadoop.fs.Path;
 import org.apache.hadoop.hive.metastore.TableType;
-import org.apache.hadoop.hive.metastore.utils.MetaStoreUtils;
 import org.apache.hadoop.hive.ql.ErrorMsg;
 import org.apache.hadoop.hive.ql.ddl.DDLOperation;
 import org.apache.hadoop.hive.ql.ddl.DDLOperationContext;
 import org.apache.hadoop.hive.ql.ddl.ShowUtils;
 import org.apache.hadoop.hive.ql.ddl.table.info.show.tables.ShowTablesFormatter;
 import org.apache.hadoop.hive.ql.metadata.HiveException;
+import org.apache.hadoop.hive.ql.udf.UDFLike;
 
 /**
  * Operation process showing the views.
@@ -46,7 +46,7 @@ public int execute() throws HiveException {
       throw new HiveException(ErrorMsg.DATABASE_NOT_EXISTS, desc.getDbName());
     }
 
-    String pattern = MetaStoreUtils.convertSqlPatternToRegExp(desc.getPattern());
+    String pattern = UDFLike.likePatternToRegExp(desc.getPattern(), false);
     List<String> viewNames = context.getDb().getTablesByType(desc.getDbName(), pattern, TableType.VIRTUAL_VIEW);
     Collections.sort(viewNames);
     LOG.debug("Found {} view(s) matching the SHOW VIEWS statement.", viewNames.size());

diff --git a/ql/src/java/org/apache/hadoop/hive/ql/exec/FunctionRegistry.java b/ql/src/java/org/apache/hadoop/hive/ql/exec/FunctionRegistry.java
@@ -858,7 +858,7 @@ public static Set<String> getFunctionNamesByLikePattern(String funcPatternStr) {
     Set<String> allFuncs = getFunctionNames();
     String[] subpatterns = funcPatternStr.trim().split("\\|");
     for (String subpattern : subpatterns) {
-      subpattern = "(?i)" + UDFLike.likePatternToRegExp(subpattern);
+      subpattern = "(?i)" + UDFLike.likePatternToRegExp(subpattern, true);
       try {
         Pattern patternObj = Pattern.compile(subpattern);
         for (String funcName : allFuncs) {

diff --git a/...va/org/apache/hadoop/hive/ql/exec/vector/expressions/FilterStringColLikeStringScalar.java b/...va/org/apache/hadoop/hive/ql/exec/vector/expressions/FilterStringColLikeStringScalar.java
@@ -82,7 +82,7 @@ String format(String pattern) {
     COMPLEX(ComplexChecker.class) {
       @Override
       String format(String pattern) {
-        return "^" + UDFLike.likePatternToRegExp(pattern) + "$";
+        return "^" + UDFLike.likePatternToRegExp(pattern, true) + "$";
       }
     },
     // Accepts chained LIKE patterns without escaping like "abc%def%ghi%" and

diff --git a/ql/src/java/org/apache/hadoop/hive/ql/udf/UDFLike.java b/ql/src/java/org/apache/hadoop/hive/ql/udf/UDFLike.java
@@ -61,7 +61,10 @@ private enum PatternType {
   public UDFLike() {
   }
 
-  public static String likePatternToRegExp(String likePattern) {
+  public static String likePatternToRegExp(String likePattern, boolean quote) {
+    if (likePattern == null) {
+      return null;
+    }
     StringBuilder sb = new StringBuilder();
     for (int i = 0; i < likePattern.length(); i++) {
       // Make a special case for "\\_" and "\\%"
@@ -77,9 +80,9 @@ public static String likePatternToRegExp(String likePattern) {
       if (n == '_') {
         sb.append(".");
       } else if (n == '%') {
-        sb.append(".*?");
+        sb.append(".*");
       } else {
-        sb.append(Pattern.quote(Character.toString(n)));
+        sb.append(quote ? Pattern.quote(Character.toString(n)) : n);
       }
     }
     return sb.toString();
@@ -184,7 +187,7 @@ public BooleanWritable evaluate(Text s, Text likePattern) {
 
       parseSimplePattern(strLikePattern);
       if (type == PatternType.COMPLEX) {
-        p = Pattern.compile(likePatternToRegExp(strLikePattern), Pattern.DOTALL);
+        p = Pattern.compile(likePatternToRegExp(strLikePattern, true), Pattern.DOTALL);
       }
     }
 

diff --git a/.../metastore-common/src/main/java/org/apache/hadoop/hive/metastore/HiveMetaStoreClient.java b/.../metastore-common/src/main/java/org/apache/hadoop/hive/metastore/HiveMetaStoreClient.java
@@ -2902,7 +2902,7 @@ public List<String> getTables(String catName, String dbName, String tablePattern
     Database db = null;
     try {
       db = getDatabase(catName, dbName);
-    } catch (Exception e) { /* appears exception is not thrown currently if db doesnt exist */ }
+    } catch (NoSuchObjectException e) { /* appears exception is not thrown currently if db doesnt exist */ }
 
     if (MetaStoreUtils.isDatabaseRemote(db)) {
       // TODO: remote database does not support list table names by pattern yet.
@@ -2930,7 +2930,7 @@ public List<String> getTables(String catName, String dbName, String tablePattern
       String[] patterns = tablePattern.split("\\|");
       for (String pattern : patterns) {
         pattern = "(?i)" + pattern.replaceAll("\\.\\*", "\\*").replaceAll("\\*", ".*");
-        String filter = hive_metastoreConstants.HIVE_FILTER_FIELD_TABLE_NAME + " like \"" + pattern + "\"";
+        String filter = String.format("%s like \"%s\"", hive_metastoreConstants.HIVE_FILTER_FIELD_TABLE_NAME, pattern);
         tables.addAll(listTableNamesByFilter(catName, dbName, filter, (short) -1));
       }
     }

diff --git a/...metastore-common/src/main/java/org/apache/hadoop/hive/metastore/utils/MetaStoreUtils.java b/...metastore-common/src/main/java/org/apache/hadoop/hive/metastore/utils/MetaStoreUtils.java
@@ -1337,33 +1337,6 @@ public static String getHttpPath(String httpPath) {
     return httpPath;
   }
 
-  public static String convertSqlPatternToRegExp(String likePattern) {
-    if (likePattern == null) {
-      return null;
-    }
-    StringBuilder sb = new StringBuilder();
-    for (int i = 0; i < likePattern.length(); i++) {
-      // Make a special case for "\\_" and "\\%"
-      char n = likePattern.charAt(i);
-      if (n == '\\'
-          && i + 1 < likePattern.length()
-          && (likePattern.charAt(i + 1) == '_' || likePattern.charAt(i + 1) == '%')) {
-        sb.append(likePattern.charAt(i + 1));
-        i++;
-        continue;
-      }
-
-      if (n == '_') {
-        sb.append(".");
-      } else if (n == '%') {
-        sb.append(".*");
-      } else {
-        sb.append(n);
-      }
-    }
-    return sb.toString();
-  }
-
   public static boolean isDatabaseRemote(Database db) {
     return db != null && db.getType() == DatabaseType.REMOTE;
   }