aboutsummaryrefslogtreecommitdiffstats
path: root/README.md
diff options
context:
space:
mode:
authorrotem <rotem.hermon@gmail.com>2013-06-17 16:43:27 +0300
committerrotem <rotem.hermon@gmail.com>2013-06-17 16:43:27 +0300
commite5f8ad62804daa1a08d5fd22b79d73018fd7bea8 (patch)
treed096ab26949b43008d2dd8e30071ed8573a7800a /README.md
parent688265bc93ab91413a04c9d5772ec639c2c030a3 (diff)
downloadSpyGlass-e5f8ad62804daa1a08d5fd22b79d73018fd7bea8.tar.gz
SpyGlass-e5f8ad62804daa1a08d5fd22b79d73018fd7bea8.zip
formatting
Diffstat (limited to 'README.md')
-rw-r--r--README.md2
1 files changed, 2 insertions, 0 deletions
diff --git a/README.md b/README.md
index 6938e48..e6d9af8 100644
--- a/README.md
+++ b/README.md
@@ -234,6 +234,7 @@ HBaseRawSource is an alternative HBase source implementation that provides two m
* Passing the row object to the mapper to allow full customized processing (without the need to declare in advance the read columns).
**Passing a scan object**
+
HBaseRawSource object provides a helper function to encode a scan object as a base64 string, which you can pass to the source.
e.g.
@@ -245,6 +246,7 @@ e.g.
val hbaseSource = new HBaseRawSource("MY-TABLE", "hbase-local", Array("col-family"), base64Scan = scanner)
**Processing the rows**
+
The mapper function gets from HBaseRawSource a tuple containing two fields: (rowkey, row).
The first field is the row key, the second is the row Result object. You can then process the row as needed.
The sink will write the output fields as columns under the provided family and field name as the column name.