Add a "size of checkpoint" statistic.

Merge statistics from file and LSM sources into a "data source" statistic structure.  Rename and regroup some shared stastistics.  Make statistics constant names upper case.  Add a helper to the Python API to lookup in a cursor in a simple expression.
closes #232

1 files changed, 25 insertions, 49 deletions

diff --git a/dist/stat.py b/dist/stat.py
index bbaaa08b06c..a70c09e8a63 100644
--- a/dist/stat.py
+++ b/dist/stat.py
@@ -1,24 +1,20 @@
 # Read the source files and output the statistics #defines and allocation code.
 
 import re, string, sys, textwrap
-from operator import attrgetter
 from dist import compare_srcfile
 from dist import source_paths_list
 
 # Read the source files.
-from stat_data import btree_stats, connection_stats, lsm_stats
+from stat_data import dsrc_stats, connection_stats
 
-# print_struct --
-#	Print the structures for the stat.h file.
-def print_struct(title, name, list):
+def print_struct(title, name, stats):
+	'''Print the structures for the stat.h file.'''
 	f.write('/*\n')
-	f.write(' * Statistics entries for ' + title + ' handle.\n')
+	f.write(' * Statistics entries for ' + title + '.\n')
 	f.write(' */\n')
 	f.write('struct __wt_' + name + '_stats {\n')
 
-	# Sort the structure fields by their description, so the eventual
-	# disply is sorted by string.
-	for l in sorted(list, key=attrgetter('desc')):
+	for l in stats:
 		f.write('\tWT_STATS ' + l.name + ';\n')
 	f.write('};\n\n')
 
@@ -35,59 +31,41 @@ for line in open('../src/include/stat.h', 'r'):
 	elif line.count('Statistics section: BEGIN'):
 		f.write('\n')
 		skip = 1
-		print_struct('BTREE', 'btree', btree_stats)
-		print_struct('CONNECTION', 'connection', connection_stats)
-		print_struct('LSM', 'lsm', lsm_stats)
+		print_struct('data sources', 'dsrc', dsrc_stats)
+		print_struct('connections', 'connection', connection_stats)
 f.close()
 compare_srcfile(tmp_file, '../src/include/stat.h')
 
-# print_define --
-#	Print the #defines for the wiredtiger.in file.
-def print_define():
-	# Sort the structure fields by their description so they match
-	# the structure lists.
+def print_defines():
+	'''Print the #defines for the wiredtiger.in file.'''
 	f.write('''
 /*!
- * @name Statistics for connection handles
+ * @name Connection statistics
  * @anchor statistics_keys
  * @anchor statistics_conn
- * Statistics in WiredTiger are accessed through cursors with \c "statistics:"
- * URIs.  Individual statistics can be queried through the cursor using the
- * following keys.
+ * Statistics are accessed through cursors with \c "statistics:" URIs.
+ * Individual statistics can be queried through the cursor using the following
+ * keys.
  * @{
  */
 ''')
-	for v, l in enumerate(sorted(connection_stats, key=attrgetter('desc'))):
+	for v, l in enumerate(connection_stats):
 		f.write('/*! %s */\n' % '\n * '.join(textwrap.wrap(l.desc, 70)))
-		f.write('#define\tWT_STAT_' + l.name + "\t" *
-		    max(1, 6 - int((len('WT_STAT_') + len(l.name)) / 8)) +
+		f.write('#define\tWT_STAT_CONN_' + l.name.upper() + "\t" *
+		    max(1, 6 - int((len('WT_STAT_CONN_' + l.name)) / 8)) +
 		    str(v) + '\n')
 	f.write('''
 /*!
  * @}
- * @name Statistics for file objects
- * @anchor statistics_file
+ * @name Statistics for data sources
+ * @anchor statistics_dsrc
  * @{
  */
 ''')
-	for v, l in enumerate(sorted(btree_stats, key=attrgetter('desc'))):
+	for v, l in enumerate(dsrc_stats):
 		f.write('/*! %s */\n' % '\n * '.join(textwrap.wrap(l.desc, 70)))
-		f.write('#define\tWT_STAT_' + l.name + "\t" *
-		    max(1, 6 - int((len('WT_STAT_') + len(l.name)) / 8)) +
-		    str(v) + '\n')
-	f.write('/*! @} */\n')
-	f.write('''
-/*!
- * @}
- * @name Statistics for lsm objects
- * @anchor statistics_lsm
- * @{
- */
-''')
-	for v, l in enumerate(sorted(lsm_stats, key=attrgetter('desc'))):
-		f.write('/*! %s */\n' % '\n * '.join(textwrap.wrap(l.desc, 70)))
-		f.write('#define\tWT_STAT_' + l.name + "\t" *
-		    max(1, 6 - int((len('WT_STAT_') + len(l.name)) / 8)) +
+		f.write('#define\tWT_STAT_DSRC_' + l.name.upper() + "\t" *
+		    max(1, 6 - int((len('WT_STAT_DSRC_' + l.name)) / 8)) +
 		    str(v) + '\n')
 	f.write('/*! @} */\n')
 
@@ -104,14 +82,13 @@ for line in open('../src/include/wiredtiger.in', 'r'):
 	elif line.count('Statistics section: BEGIN'):
 		f.write(' */\n')
 		skip = 1
-		print_define()
+		print_defines()
 		f.write('/*\n')
 f.close()
 compare_srcfile(tmp_file, '../src/include/wiredtiger.in')
 
-# print_func --
-#	Print the functions for the stat.c file.
 def print_func(name, list):
+	'''Print the functions for the stat.c file.'''
 	f.write('''
 int
 __wt_stat_alloc_''' + name + '''_stats(WT_SESSION_IMPL *session, WT_''' +
@@ -145,7 +122,7 @@ __wt_stat_clear_''' + name + '''_stats(WT_STATS *stats_arg)
 	for l in sorted(list):
 		# Items marked permanent aren't cleared by the stat clear
 		# methods.
-		if not l.config.count('perm'):
+		if not l.flags.get('perm', 0):
 			f.write('\tstats->' + l.name + '.v = 0;\n');
 	f.write('}\n')
 
@@ -154,8 +131,7 @@ f = open(tmp_file, 'w')
 f.write('/* DO NOT EDIT: automatically built by dist/stat.py. */\n\n')
 f.write('#include "wt_internal.h"\n')
 
-print_func('btree', btree_stats)
+print_func('dsrc', dsrc_stats)
 print_func('connection', connection_stats)
-print_func('lsm', lsm_stats)
 f.close()
 compare_srcfile(tmp_file, '../src/support/stat.c')