hivemeta 0.0.1 → 0.0.2

data/CHANGELOG

@@ -1,2 +1,7 @@
+* 2011-05-04 - fsf
+- pull delimiter from SERDE_PARAMS and only default to \t
+- hivemeta_query displays delimiter info
+- removed unused Rakefile
+
 * 2011-05-01 - fsf
 - initial import

data/README

@@ -1,16 +1,22 @@
 hivemeta
 
 A ruby API for access to the Hive metastore.  Useful for querying columns
-in map/reduce applications.  Includes a demo application to spit out
+in Hadoop map/reduce applications.  Includes a demo application to spit out
 table information from the command-line via table name search or
 by the table's location in HDFS.
 
+Installation
+
+gem install hivemeta
+
+Usage
+
 streaming map/reduce code snippet:
 
 require 'hivemeta'
 
 h = HiveMeta::Connection.new(...) # see sample-mapper.rb for detail
-inventory = h.table 'sample_inventory'
+inv_table = h.table 'sample_inventory'
 
 STDIN.each_line do |line|
   begin
@@ -24,7 +30,6 @@ STDIN.each_line do |line|
   puts "#{item_id}\t#{count}" if count >= 1000
 end
 
-
 sample usage for the demo app:
 
 # query by table names

data/examples/hivemeta_query.rb

@@ -100,6 +100,15 @@ tables.each do |table|
   next if list_tables
   puts table.path
   next if list_paths
+
+  delim = table.delimiter
+  a = table.delimiter.ord
+  delim = '\\t' if a == 9
+  if (1..7).include? a
+    delim = "^#{('A'.ord + a - 1).chr}"
+  end
+  puts "delimiter: \"#{delim}\" (ASCII #{a})"
+
   tput_cols = `tput cols`.chomp.to_i rescue tput_cols = 0
 
   table.each_with_index do |col_name, i|

data/examples/sample-mapper.rb

@@ -13,7 +13,7 @@ db_name    = 'hivemeta'
 dbi_string = "DBI:Mysql:#{db_name}:#{db_host}"
 h = HiveMeta::Connection.new(dbi_string, db_user, db_pass)
 
-inventory = h.table 'sample_inventory'
+inv_table = h.table 'sample_inventory'
 
 STDIN.each_line do |line|
   begin

data/lib/hivemeta/connection.rb

@@ -64,9 +64,11 @@ module HiveMeta
 #p name
         table = Table.new(name)
 
-        sql = "select c.INTEGER_IDX, c.column_name, c.COMMENT, s.LOCATION
+        sql = "select c.INTEGER_IDX, c.column_name, c.COMMENT,
+          s.LOCATION, s.SD_ID
           from TBLS t, COLUMNS c, SDS s
-          where t.SD_ID = c.SD_ID and t.SD_ID = s.SD_ID and t.TBL_NAME = ?"
+          where t.SD_ID = c.SD_ID and t.SD_ID = s.SD_ID
+          and t.TBL_NAME = ?"
         query sql, name do |rec|
 #puts "REC:"
 #p rec
@@ -74,10 +76,24 @@ module HiveMeta
           col_name = rec[1]
           col_cmt  = rec[2]
           tbl_loc  = rec[3]
+          sd_id    = rec[4]
           table.columns[col_idx]  = col_name
           table.comments[col_idx] = col_cmt
-          table.path = tbl_loc
+          table.path      = tbl_loc
+
+        end
+
+        sql = "select sp.PARAM_VALUE
+          from SERDE_PARAMS sp, TBLS t
+          where t.SD_ID = sp.SERDE_ID
+          and PARAM_KEY = 'field.delim'
+        and t.TBL_NAME = ?"
+        results = query sql, name
+        if results and results[0] and results[0][0]
+          table.delimiter = results[0][0]
         end
+#puts "#{name}: found delim '#{table.delimiter}'" if results[0]
+#puts "#{name}: no delim" if not results[0]
 
         tables << table
       end

data/lib/hivemeta.rb

@@ -1,5 +1,5 @@
 require 'hivemeta/connection'
 
 module HiveMeta
-  VERSION = '0.0.1'
+  VERSION = '0.0.2'
 end

metadata

@@ -5,8 +5,8 @@ version: !ruby/object:Gem::Version
   segments: 
   - 0
   - 0
-  - 1
-  version: 0.0.1
+  - 2
+  version: 0.0.2
 platform: ruby
 authors: 
 - Frank Fejes
@@ -14,7 +14,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2011-05-01 00:00:00 -05:00
+date: 2011-05-04 00:00:00 -05:00
 default_executable: 
 dependencies: []