blob: f7e55ec424e681a919fdd7723a06a340bf2c04a1 [file] [log] [blame]
String Data
-----------
Created by modifying Impala's HdfsAvroTableWriter(removed).
These files' schemas have a single nullable string column 's'.
negative_string_len.avro: contains two values, but the second value has a negative length.
invalid_union.avro: contains three values, all of which have an invalid union value (4,
which decodes to 2). Also has a single extra byte at the end (0x0a), which will be
interpreted as a malformed data block.
truncated_string.avro: contains one value, which is missing the last byte.
Float Data
----------
Created by modifying Impala's HdfsAvroTableWriter(removed).
These files' schemas have a single nullable float column 'c1'.
truncated_float.avro: contains two float values. The second is missing the last byte.
Bad Schema
----------
Created by editing the schema of a valid Avro file with vim.
invalid_decimal_schema.avro: two columns, name STRING and value DECIMAL(5,7).
The DECIMAL value is invalid.
Date Data
---------
out_of_range_date.avro:
Small avro table with a single DATE column. Contains 5 values, two of which are out of
range.
hive2_pre_gregorian_date.avro:
Small avro table with one DATE column, created by Hive 2.1.1.
Used to demonstrate avro interoperability issues between Hive and Impala for dates
before the introduction of Gregorian calendar in 1582-10-15.
hive3_pre_gregorian_date.avro:
Small avro table with one DATE column, created by Hive 3.1.0.
Used to demonstrate that pre-Gregorian avro dates written by Hive3 are read back correctly
by Impala.