summaryrefslogtreecommitdiffstats
path: root/magic/Magdir/apache
diff options
context:
space:
mode:
Diffstat (limited to 'magic/Magdir/apache')
-rwxr-xr-xmagic/Magdir/apache28
1 files changed, 28 insertions, 0 deletions
diff --git a/magic/Magdir/apache b/magic/Magdir/apache
new file mode 100755
index 0000000..d896b50
--- /dev/null
+++ b/magic/Magdir/apache
@@ -0,0 +1,28 @@
+
+#------------------------------------------------------------------------------
+# $File: apache,v 1.1 2017/04/11 14:52:15 christos Exp $
+# apache: file(1) magic for Apache Big Data formats
+
+# Avro files
+0 string Obj Apache Avro
+>3 byte x version %d
+
+# ORC files
+# Important information is in file footer, which we can't index to :(
+0 string ORC Apache ORC
+
+# Parquet files
+0 string PAR1 Apache Parquet
+
+# Hive RC files
+0 string RCF Apache Hive RC file
+>3 byte x version %d
+
+# Sequence files (and the careless first version of RC file)
+
+0 string SEQ
+>3 byte <6 Apache Hadoop Sequence file version %d
+>3 byte >6 Apache Hadoop Sequence file version %d
+>3 byte =6
+>>5 string org.apache.hadoop.hive.ql.io.RCFile$KeyBuffer Apache Hive RC file version 0
+>>3 default x Apache Hadoop Sequence file version 6