Search in sources :

Example 41 with FieldSchema

use of org.apache.hadoop.hive.metastore.api.FieldSchema in project hive by apache.

the class TestHBaseStoreCached method listGetDropPartitionNames.

@Test
public void listGetDropPartitionNames() throws Exception {
    String dbName = "default";
    String tableName = "listParts";
    int startTime = (int) (System.currentTimeMillis() / 1000);
    List<FieldSchema> cols = new ArrayList<FieldSchema>();
    cols.add(new FieldSchema("col1", "int", "nocomment"));
    SerDeInfo serde = new SerDeInfo("serde", "seriallib", null);
    StorageDescriptor sd = new StorageDescriptor(cols, "file:/tmp", "input", "output", false, 0, serde, null, null, emptyParameters);
    List<FieldSchema> partCols = new ArrayList<FieldSchema>();
    partCols.add(new FieldSchema("pc", "string", ""));
    partCols.add(new FieldSchema("region", "string", ""));
    Table table = new Table(tableName, dbName, "me", startTime, startTime, 0, sd, partCols, emptyParameters, null, null, null);
    store.createTable(table);
    String[][] partVals = new String[][] { { "today", "north america" }, { "tomorrow", "europe" } };
    for (String[] pv : partVals) {
        List<String> vals = new ArrayList<String>();
        for (String v : pv) vals.add(v);
        StorageDescriptor psd = new StorageDescriptor(sd);
        psd.setLocation("file:/tmp/pc=" + pv[0] + "/region=" + pv[1]);
        Partition part = new Partition(vals, dbName, tableName, startTime, startTime, psd, emptyParameters);
        store.addPartition(part);
    }
    List<String> names = store.listPartitionNames(dbName, tableName, (short) -1);
    Assert.assertEquals(2, names.size());
    String[] resultNames = names.toArray(new String[names.size()]);
    Arrays.sort(resultNames);
    Assert.assertArrayEquals(resultNames, new String[] { "pc=today/region=north america", "pc=tomorrow/region=europe" });
    List<Partition> parts = store.getPartitionsByNames(dbName, tableName, names);
    Assert.assertArrayEquals(partVals[0], parts.get(0).getValues().toArray(new String[2]));
    Assert.assertArrayEquals(partVals[1], parts.get(1).getValues().toArray(new String[2]));
    store.dropPartitions(dbName, tableName, names);
    List<Partition> afterDropParts = store.getPartitions(dbName, tableName, -1);
    Assert.assertEquals(0, afterDropParts.size());
}
Also used : Partition(org.apache.hadoop.hive.metastore.api.Partition) Table(org.apache.hadoop.hive.metastore.api.Table) FieldSchema(org.apache.hadoop.hive.metastore.api.FieldSchema) SerDeInfo(org.apache.hadoop.hive.metastore.api.SerDeInfo) ArrayList(java.util.ArrayList) StorageDescriptor(org.apache.hadoop.hive.metastore.api.StorageDescriptor) Test(org.junit.Test)

Example 42 with FieldSchema

use of org.apache.hadoop.hive.metastore.api.FieldSchema in project hive by apache.

the class TestHiveMetaStoreChecker method createPartitionsDirectoriesOnFS.

/**
   * Creates partition sub-directories for a given table on the file system. Used to test the
   * use-cases when partitions for the table are not present in the metastore db
   *
   * @param table - Table which provides the base locations and partition specs for creating the
   *          sub-directories
   * @param numPartitions - Number of partitions to be created
   * @throws IOException
   */
private void createPartitionsDirectoriesOnFS(Table table, int numPartitions) throws IOException {
    String path = table.getDataLocation().toString();
    fs = table.getPath().getFileSystem(hive.getConf());
    int numPartKeys = table.getPartitionKeys().size();
    for (int i = 0; i < numPartitions; i++) {
        StringBuilder partPath = new StringBuilder(path);
        partPath.append(Path.SEPARATOR);
        for (int j = 0; j < numPartKeys; j++) {
            FieldSchema field = table.getPartitionKeys().get(j);
            partPath.append(field.getName());
            partPath.append('=');
            partPath.append("val_");
            partPath.append(i);
            if (j < (numPartKeys - 1)) {
                partPath.append(Path.SEPARATOR);
            }
        }
        createDirectory(partPath.toString());
    }
}
Also used : FieldSchema(org.apache.hadoop.hive.metastore.api.FieldSchema)

Example 43 with FieldSchema

use of org.apache.hadoop.hive.metastore.api.FieldSchema in project hive by apache.

the class TestHive method testIndex.

/**
   * Tests creating a simple index on a simple table.
   *
   * @throws Throwable
   */
public void testIndex() throws Throwable {
    try {
        // create a simple table
        String tableName = "table_for_testindex";
        String qTableName = MetaStoreUtils.DEFAULT_DATABASE_NAME + "." + tableName;
        try {
            hm.dropTable(MetaStoreUtils.DEFAULT_DATABASE_NAME, tableName);
        } catch (HiveException e) {
            e.printStackTrace();
            assertTrue("Unable to drop table", false);
        }
        Table tbl = new Table(MetaStoreUtils.DEFAULT_DATABASE_NAME, tableName);
        List<FieldSchema> fields = tbl.getCols();
        fields.add(new FieldSchema("col1", serdeConstants.INT_TYPE_NAME, "int -- first column"));
        fields.add(new FieldSchema("col2", serdeConstants.STRING_TYPE_NAME, "string -- second column"));
        fields.add(new FieldSchema("col3", serdeConstants.DOUBLE_TYPE_NAME, "double -- thrift column"));
        tbl.setFields(fields);
        tbl.setOutputFormatClass(HiveIgnoreKeyTextOutputFormat.class);
        tbl.setInputFormatClass(SequenceFileInputFormat.class);
        // create table
        try {
            hm.createTable(tbl);
        } catch (HiveException e) {
            e.printStackTrace();
            assertTrue("Unable to create table: " + tableName, false);
        }
        // Create a simple index
        String indexName = "index_on_table_for_testindex";
        String indexHandlerClass = HiveIndex.IndexType.COMPACT_SUMMARY_TABLE.getHandlerClsName();
        List<String> indexedCols = new ArrayList<String>();
        indexedCols.add("col1");
        String indexTableName = "index_on_table_for_testindex_table";
        String qIndexTableName = MetaStoreUtils.DEFAULT_DATABASE_NAME + "." + indexTableName;
        boolean deferredRebuild = true;
        String inputFormat = SequenceFileInputFormat.class.getName();
        String outputFormat = SequenceFileOutputFormat.class.getName();
        String serde = null;
        String storageHandler = null;
        String location = null;
        String collItemDelim = null;
        String fieldDelim = null;
        String fieldEscape = null;
        String lineDelim = null;
        String mapKeyDelim = null;
        String indexComment = null;
        Map<String, String> indexProps = null;
        Map<String, String> tableProps = null;
        Map<String, String> serdeProps = new HashMap<String, String>();
        hm.createIndex(qTableName, indexName, indexHandlerClass, indexedCols, qIndexTableName, deferredRebuild, inputFormat, outputFormat, serde, storageHandler, location, indexProps, tableProps, serdeProps, collItemDelim, fieldDelim, fieldEscape, lineDelim, mapKeyDelim, indexComment);
        // Retrieve and validate the index
        Index index = null;
        try {
            index = hm.getIndex(tableName, indexName);
            assertNotNull("Unable to fetch index", index);
            index.validate();
            assertEquals("Index names don't match for index: " + indexName, indexName, index.getIndexName());
            assertEquals("Table names don't match for index: " + indexName, tableName, index.getOrigTableName());
            assertEquals("Index table names didn't match for index: " + indexName, indexTableName, index.getIndexTableName());
            assertEquals("Index handler classes didn't match for index: " + indexName, indexHandlerClass, index.getIndexHandlerClass());
            assertEquals("Deferred rebuild didn't match for index: " + indexName, deferredRebuild, index.isDeferredRebuild());
        } catch (HiveException e) {
            System.err.println(StringUtils.stringifyException(e));
            assertTrue("Unable to fetch index correctly: " + indexName, false);
        }
        // Drop index
        try {
            hm.dropIndex(MetaStoreUtils.DEFAULT_DATABASE_NAME, tableName, indexName, false, true);
        } catch (HiveException e) {
            System.err.println(StringUtils.stringifyException(e));
            assertTrue("Unable to drop index: " + indexName, false);
        }
        boolean dropIndexException = false;
        try {
            hm.getIndex(tableName, indexName);
        } catch (HiveException e) {
            // Expected since it was just dropped
            dropIndexException = true;
        }
        assertTrue("Unable to drop index: " + indexName, dropIndexException);
        // Drop table
        try {
            hm.dropTable(tableName);
            Table droppedTable = hm.getTable(tableName, false);
            assertNull("Unable to drop table " + tableName, droppedTable);
        } catch (HiveException e) {
            System.err.println(StringUtils.stringifyException(e));
            assertTrue("Unable to drop table: " + tableName, false);
        }
    } catch (Throwable e) {
        System.err.println(StringUtils.stringifyException(e));
        System.err.println("testIndex failed");
        throw e;
    }
}
Also used : HashMap(java.util.HashMap) FieldSchema(org.apache.hadoop.hive.metastore.api.FieldSchema) ArrayList(java.util.ArrayList) HiveIndex(org.apache.hadoop.hive.ql.index.HiveIndex) Index(org.apache.hadoop.hive.metastore.api.Index)

Example 44 with FieldSchema

use of org.apache.hadoop.hive.metastore.api.FieldSchema in project hive by apache.

the class DelimitedInputWriter method getCols.

private ArrayList<String> getCols(Table table) {
    List<FieldSchema> cols = table.getSd().getCols();
    ArrayList<String> colNames = new ArrayList<String>(cols.size());
    for (FieldSchema col : cols) {
        colNames.add(col.getName().toLowerCase());
    }
    return colNames;
}
Also used : FieldSchema(org.apache.hadoop.hive.metastore.api.FieldSchema) ArrayList(java.util.ArrayList)

Example 45 with FieldSchema

use of org.apache.hadoop.hive.metastore.api.FieldSchema in project hive by apache.

the class TestMetastoreAuthorizationProvider method testSimplePrivileges.

public void testSimplePrivileges() throws Exception {
    if (!isTestEnabled()) {
        System.out.println("Skipping test " + this.getClass().getName());
        return;
    }
    String dbName = getTestDbName();
    String tblName = getTestTableName();
    String userName = setupUser();
    allowCreateDatabase(userName);
    CommandProcessorResponse ret = driver.run("create database " + dbName);
    assertEquals(0, ret.getResponseCode());
    Database db = msc.getDatabase(dbName);
    String dbLocn = db.getLocationUri();
    validateCreateDb(db, dbName);
    disallowCreateInDb(dbName, userName, dbLocn);
    disallowCreateDatabase(userName);
    driver.run("use " + dbName);
    ret = driver.run(String.format("create table %s (a string) partitioned by (b string)", tblName));
    assertEquals(1, ret.getResponseCode());
    // Even if table location is specified table creation should fail
    String tblNameLoc = tblName + "_loc";
    String tblLocation = new Path(dbLocn).getParent().toUri() + "/" + tblNameLoc;
    driver.run("use " + dbName);
    ret = driver.run(String.format("create table %s (a string) partitioned by (b string) location '" + tblLocation + "'", tblNameLoc));
    assertEquals(1, ret.getResponseCode());
    // failure from not having permissions to create table
    ArrayList<FieldSchema> fields = new ArrayList<FieldSchema>(2);
    fields.add(new FieldSchema("a", serdeConstants.STRING_TYPE_NAME, ""));
    Table ttbl = new Table();
    ttbl.setDbName(dbName);
    ttbl.setTableName(tblName);
    StorageDescriptor sd = new StorageDescriptor();
    ttbl.setSd(sd);
    sd.setCols(fields);
    sd.setParameters(new HashMap<String, String>());
    sd.getParameters().put("test_param_1", "Use this for comments etc");
    sd.setSerdeInfo(new SerDeInfo());
    sd.getSerdeInfo().setName(ttbl.getTableName());
    sd.getSerdeInfo().setParameters(new HashMap<String, String>());
    sd.getSerdeInfo().getParameters().put(org.apache.hadoop.hive.serde.serdeConstants.SERIALIZATION_FORMAT, "1");
    sd.getSerdeInfo().setSerializationLib(org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe.class.getName());
    sd.setInputFormat(HiveInputFormat.class.getName());
    sd.setOutputFormat(HiveOutputFormat.class.getName());
    ttbl.setPartitionKeys(new ArrayList<FieldSchema>());
    MetaException me = null;
    try {
        msc.createTable(ttbl);
    } catch (MetaException e) {
        me = e;
    }
    assertNoPrivileges(me);
    allowCreateInDb(dbName, userName, dbLocn);
    driver.run("use " + dbName);
    ret = driver.run(String.format("create table %s (a string) partitioned by (b string)", tblName));
    // now it succeeds.
    assertEquals(0, ret.getResponseCode());
    Table tbl = msc.getTable(dbName, tblName);
    validateCreateTable(tbl, tblName, dbName);
    // Table creation should succeed even if location is specified
    driver.run("use " + dbName);
    ret = driver.run(String.format("create table %s (a string) partitioned by (b string) location '" + tblLocation + "'", tblNameLoc));
    assertEquals(0, ret.getResponseCode());
    Table tblLoc = msc.getTable(dbName, tblNameLoc);
    validateCreateTable(tblLoc, tblNameLoc, dbName);
    String fakeUser = "mal";
    List<String> fakeGroupNames = new ArrayList<String>();
    fakeGroupNames.add("groupygroup");
    InjectableDummyAuthenticator.injectUserName(fakeUser);
    InjectableDummyAuthenticator.injectGroupNames(fakeGroupNames);
    InjectableDummyAuthenticator.injectMode(true);
    ret = driver.run(String.format("create table %s (a string) partitioned by (b string)", tblName + "mal"));
    assertEquals(1, ret.getResponseCode());
    ttbl.setTableName(tblName + "mal");
    me = null;
    try {
        msc.createTable(ttbl);
    } catch (MetaException e) {
        me = e;
    }
    assertNoPrivileges(me);
    disallowCreateInTbl(tbl.getTableName(), userName, tbl.getSd().getLocation());
    ret = driver.run("alter table " + tblName + " add partition (b='2011')");
    assertEquals(1, ret.getResponseCode());
    List<String> ptnVals = new ArrayList<String>();
    ptnVals.add("b=2011");
    Partition tpart = new Partition();
    tpart.setDbName(dbName);
    tpart.setTableName(tblName);
    tpart.setValues(ptnVals);
    tpart.setParameters(new HashMap<String, String>());
    tpart.setSd(tbl.getSd().deepCopy());
    tpart.getSd().setSerdeInfo(tbl.getSd().getSerdeInfo().deepCopy());
    tpart.getSd().setLocation(tbl.getSd().getLocation() + "/tpart");
    me = null;
    try {
        msc.add_partition(tpart);
    } catch (MetaException e) {
        me = e;
    }
    assertNoPrivileges(me);
    InjectableDummyAuthenticator.injectMode(false);
    allowCreateInTbl(tbl.getTableName(), userName, tbl.getSd().getLocation());
    ret = driver.run("alter table " + tblName + " add partition (b='2011')");
    assertEquals(0, ret.getResponseCode());
    allowDropOnTable(tblName, userName, tbl.getSd().getLocation());
    allowDropOnDb(dbName, userName, db.getLocationUri());
    ret = driver.run("drop database if exists " + getTestDbName() + " cascade");
    assertEquals(0, ret.getResponseCode());
    InjectableDummyAuthenticator.injectUserName(userName);
    InjectableDummyAuthenticator.injectGroupNames(Arrays.asList(ugi.getGroupNames()));
    InjectableDummyAuthenticator.injectMode(true);
    allowCreateDatabase(userName);
    driver.run("create database " + dbName);
    allowCreateInDb(dbName, userName, dbLocn);
    tbl.setTableType("EXTERNAL_TABLE");
    msc.createTable(tbl);
    disallowDropOnTable(tblName, userName, tbl.getSd().getLocation());
    ret = driver.run("drop table " + tbl.getTableName());
    assertEquals(1, ret.getResponseCode());
}
Also used : Path(org.apache.hadoop.fs.Path) Partition(org.apache.hadoop.hive.metastore.api.Partition) Table(org.apache.hadoop.hive.metastore.api.Table) CommandProcessorResponse(org.apache.hadoop.hive.ql.processors.CommandProcessorResponse) FieldSchema(org.apache.hadoop.hive.metastore.api.FieldSchema) SerDeInfo(org.apache.hadoop.hive.metastore.api.SerDeInfo) ArrayList(java.util.ArrayList) StorageDescriptor(org.apache.hadoop.hive.metastore.api.StorageDescriptor) HiveOutputFormat(org.apache.hadoop.hive.ql.io.HiveOutputFormat) HiveInputFormat(org.apache.hadoop.hive.ql.io.HiveInputFormat) Database(org.apache.hadoop.hive.metastore.api.Database) MetaException(org.apache.hadoop.hive.metastore.api.MetaException)

Aggregations

FieldSchema (org.apache.hadoop.hive.metastore.api.FieldSchema)407 ArrayList (java.util.ArrayList)254 Table (org.apache.hadoop.hive.metastore.api.Table)163 Test (org.junit.Test)160 StorageDescriptor (org.apache.hadoop.hive.metastore.api.StorageDescriptor)136 SerDeInfo (org.apache.hadoop.hive.metastore.api.SerDeInfo)118 Partition (org.apache.hadoop.hive.metastore.api.Partition)93 HashMap (java.util.HashMap)69 MetastoreCheckinTest (org.apache.hadoop.hive.metastore.annotation.MetastoreCheckinTest)44 List (java.util.List)42 MetaException (org.apache.hadoop.hive.metastore.api.MetaException)42 ColumnStatistics (org.apache.hadoop.hive.metastore.api.ColumnStatistics)40 Database (org.apache.hadoop.hive.metastore.api.Database)40 ColumnStatisticsObj (org.apache.hadoop.hive.metastore.api.ColumnStatisticsObj)39 IOException (java.io.IOException)36 ColumnStatisticsDesc (org.apache.hadoop.hive.metastore.api.ColumnStatisticsDesc)36 HiveException (org.apache.hadoop.hive.ql.metadata.HiveException)35 ColumnStatisticsData (org.apache.hadoop.hive.metastore.api.ColumnStatisticsData)34 Path (org.apache.hadoop.fs.Path)32 AggrStats (org.apache.hadoop.hive.metastore.api.AggrStats)32