Add spell checking for the source code.

10 files changed, 397 insertions, 13 deletions

diff --git a/dist/api_data.py b/dist/api_data.py
index 7cd35df2e3d..333afe8ca44 100644
--- a/dist/api_data.py
+++ b/dist/api_data.py
@@ -168,7 +168,7 @@ methods = {
 
 'session.checkpoint' : Method([
 	Config('archive', 'false', r'''
-		remove log files no longer required for transactional durabilty''',
+		remove log files no longer required for transactional durability''',
 		type='boolean'),
 	Config('force', 'false', r'''
 		write a new checkpoint even if nothing has changed since the last
diff --git a/dist/s_all b/dist/s_all
index 6f4694d5f91..3c374a61b7b 100644
--- a/dist/s_all
+++ b/dist/s_all
@@ -56,6 +56,7 @@ run "sh ./s_define" "Checking for unused #defines"
 run "sh ./s_funcs" "Checking for unused functions"
 run "sh ./s_printf" "Checking for non-portable printf/scanf"
 run "sh ./s_longlines" "Checking for long lines"
+run "sh ./s_string" "Checking string spelling"
 run "sh ./s_style" "Checking style"
 run "sh ./s_typedef -c" "Checking for unused typedefs"
 run "sh ./s_types" "Checking for old-style types"
diff --git a/dist/s_string b/dist/s_string
new file mode 100644
index 00000000000..a42f968a555
--- /dev/null
+++ b/dist/s_string
@@ -0,0 +1,23 @@
+#!/bin/sh -
+#
+# Check spelling in comments and quoted strings from the source files.
+
+t=__wt.$$
+trap 'rm -f $t; exit 0' 0 1 2 3 13 15
+
+check() {
+	aspell --mode=ccpp --lang=en list < ../$1 |
+	sort -u |
+	comm -23 /dev/stdin s_string.ok > $t
+	test -s $t && {
+		echo "==== $1"
+		cat $t
+	}
+}
+
+l="`cd .. && echo src/include/*.[hi] src/include/*.in`"
+for i in $l `sed -e '/^[a-z]/! d' filelist`; do
+	check $i
+done
+
+exit 0
diff --git a/dist/s_string.ok b/dist/s_string.ok
new file mode 100644
index 00000000000..a6175edca52
--- /dev/null
+++ b/dist/s_string.ok
@@ -0,0 +1,360 @@
+AAAAA
+AAAAAA
+AAAAAAAAAA
+AAAAAAAAAAAAA
+AAAAAAAAAAAAAAAA
+AAAAAAAAAAAAAAAAAA
+ADDR
+API
+APIs
+Alloc
+Athlon
+BBBBB
+BBBBBB
+BBBBBBBBBB
+BBBBBBBBBBBBB
+BBBBBBBBBBBBBBBB
+BBBBBBBBBBBBBBBBBB
+BITSTRING
+BSR
+BTREE
+BUF
+BUFs
+Bsearch
+Btree
+CELL's
+CELLs
+COL's
+CONCAT
+CURSORs
+Checksum
+Config
+DATAITEMs
+DESC
+EINTR
+ENOMEM
+ETIME
+ETIMEDOUT
+Enqueue
+Env
+Eron
+FALLTHROUGH
+FH
+FLD
+FOREACH
+FREELIST
+Fasttrack
+FreeBSD
+FreeBSD's
+Freelist
+IMPL
+IMPL's
+INDX
+INIT
+INITIALIZER
+INSERT's
+JPEG
+KV
+Kanowski's
+Kounavis
+LF
+LRU
+LSN
+LSN's
+LSNs
+Lookup
+MBUF
+MEM
+MVCC
+Mewhort
+Mutex
+Mutexes
+NONINFRINGEMENT
+NOTFOUND
+NOTREACHED
+NUL
+NULLs
+NoAddr
+OVFL
+PADDR
+PAGE's
+POSIX
+PSIZE
+PTR
+Pagesize
+PlatformSDK
+Pre
+Qsort
+READONLY
+REF's
+REQ
+RLE
+RPC
+Recurse
+Redistributions
+Resize
+SCHED
+SLIST
+STAILQ
+SYS
+Solaris
+Split's
+Subtree
+Subtrees
+TAILQ
+TODO
+TOOSMALL
+UINT
+UTF
+UnixLib
+Vc
+Vixie
+VxWorks
+WIREDTIGER
+WinNT
+WiredTiger
+WiredTiger's
+XP
+abcdef
+addfrag
+addl
+addr
+addrs
+alfred
+alloc
+allocsize
+ap
+api
+bigram
+bitpos
+bitstring
+bitwise
+bnd
+boolean
+breakpoint
+bsearch
+bt
+btcur
+btree
+buf
+builtin
+calloc
+checkfrag
+checksum
+checksums
+cksum
+clr
+cmp
+colgroup
+cond
+conf
+config
+conn
+connectionp
+crc
+cref
+ctype
+curbtree
+curbulk
+curconfig
+curdump
+cursorp
+curstat
+datalen
+datasets
+del
+delfmt
+dequeue
+dequeued
+der
+dereference
+desc
+dest
+dlclose
+dlh
+dlopen
+dlsym
+dsk
+dumpable
+dup
+encodings
+endian
+english
+enqueue
+enum's
+eof
+eop
+errhandler
+errv
+errx
+esc
+exactp
+extern
+fcntl
+fh
+filename
+fileops
+filesize
+fixup
+fmt
+fp
+free'd
+freelist
+fstat
+fsync
+fsyncs
+ftruncate
+func
+getone
+getones
+getraw
+gobare
+goesc
+gostring
+gostruct
+goutf
+hrow
+html
+huffman
+iSh
+ikey
+imref
+incr
+indices
+indx
+init
+initsize
+inmem
+intl
+intpack
+ints
+inuse
+ispo
+keyname
+len
+lex
+llll
+logput
+lookup
+lookups
+lru
+lsn
+lu
+majorp
+malloc
+mb
+mem
+membar
+memcpy
+metadata
+mfence
+minorp
+minprefix
+msg
+mtx
+multiprocess
+mutex
+mutexes
+namespaces
+nbits
+negint
+newname
+nl
+nlpo
+nul
+numSymbols
+numbare
+offpage
+oldname
+ovfl
+packv
+patchp
+pfx
+posint
+printf
+pthread
+py
+qdown
+qsort
+qup
+realloc
+recno
+req
+ret
+retp
+rle
+rpc
+rref
+runlength
+savepoints
+sb
+schemas
+scr
+searchable
+serializable
+sessionp
+sizeof
+sizev
+slvg
+snprintf
+sp
+src
+srch
+srvr
+sset
+stbar
+stdarg
+stderr
+stdout
+strdup
+strerror
+stringin
+struct
+structs
+subtree
+subtrees
+superset
+sys
+t's
+tV
+tablename
+th
+thazard
+tinsert
+toffpage
+trecno
+trepeat
+trk
+trk's
+tt
+tupdate
+tworkq
+txn
+typedef
+uB
+uint
+unbare
+uncompressing
+undef
+unesc
+uninstantiated
+unjams
+unpackv
+upd
+uri
+utf
+va
+versa
+vpack
+vsize
+vunpack
+vupdate
+wiredtiger
+workQ
+workq
+wq
+wrapup
+xff
+xxxx
+xxxxx
+xxxxxx
diff --git a/src/btree/bt_alloc.c b/src/btree/bt_alloc.c
index 6ef41b54ea0..877284ab699 100644
--- a/src/btree/bt_alloc.c
+++ b/src/btree/bt_alloc.c
@@ -124,7 +124,7 @@ __wt_block_free(WT_SESSION_IMPL *session, uint32_t addr, uint32_t size)
 
 combine:/*
 	 * Insert the entry at the appropriate place in the address list after
-	 * checking to see if it ajoins adjacent entries.
+	 * checking to see if it adjoins adjacent entries.
 	 */
 	TAILQ_FOREACH(fe, &btree->freeqa, qa) {
 		/*
@@ -173,7 +173,7 @@ combine:/*
 
 		/*
 		 * The freed entry must appear before the list entry, but does
-		 * not ajoin it.  Insert the freed entry before the list entry.
+		 * not adjoin it. Insert the freed entry before the list entry.
 		 */
 		TAILQ_INSERT_BEFORE(fe, new, qa);
 		break;
diff --git a/src/btree/bt_reconcile.c b/src/btree/bt_reconcile.c
index 4f9293c07df..366a11d049b 100644
--- a/src/btree/bt_reconcile.c
+++ b/src/btree/bt_reconcile.c
@@ -159,8 +159,8 @@ typedef struct {
 	 *
 	 * We obviously could maintain this list on a per-split-chunk basis:
 	 * copy it into the WT_BOUNDARY structure if the boundary is reached.
-	 * We're binarily searching this array during fixup and I don't expect
-	 * it to be large.
+	 * We're doing a binary search of this array during fixup and I don't
+	 * expect it to be large.
 	 */
 	WT_REF **imref;			/* In-memory subtree reference list */
 	uint32_t imref_next;		/* Next list slot */
@@ -328,14 +328,14 @@ __rec_init(WT_SESSION_IMPL *session, uint32_t flags)
 	r->locked = LF_ISSET(WT_REC_LOCKED);
 
 	/*
-	 * During internal page reconcilition we track referenced objects that
+	 * During internal page reconciliation we track referenced objects that
 	 * are discarded when the page is discarded.  That tracking is done per
 	 * reconciliation run, initialize it before anything else.
 	 */
 	__rec_discard_init(r);
 
 	/*
-	 * During internal page reconcilition we track in-memory subtrees we
+	 * During internal page reconciliation we track in-memory subtrees we
 	 * find in the page.  That tracking is done per reconciliation run,
 	 * initialize it before anything else.
 	 */
@@ -2937,7 +2937,7 @@ __rec_key_state_update(WT_RECONCILE *r, int ovfl_key)
 	WT_BUF *a;
 
 	/*
-	 * If writing an overflow key onto the page, dan't update the "last key"
+	 * If writing an overflow key onto the page, don't update the "last key"
 	 * value, and leave the state of prefix compression alone.  (If we are
 	 * currently doing prefix compression, we have a key state which will
 	 * continue to work, we're just skipping the key just created because
diff --git a/src/btree/row_key.c b/src/btree/row_key.c
index 51a422b51f9..0d26a854fae 100644
--- a/src/btree/row_key.c
+++ b/src/btree/row_key.c
@@ -68,7 +68,7 @@ __wt_row_key(
 		 * read/write them during reconciliation simply because their
 		 * prefix might change.  That means we can't use instantiated
 		 * overflow keys to figure out the prefix for other keys,
-		 * specifially, in this code when we're looking for a key we
+		 * specifically, in this code when we're looking for a key we
 		 * can roll-forward to figure out the target key's prefix,
 		 * instantiated overflow keys aren't useful.
 		 *
diff --git a/src/include/btree.h b/src/include/btree.h
index 21bdd92662b..9d55852f5ae 100644
--- a/src/include/btree.h
+++ b/src/include/btree.h
@@ -340,7 +340,7 @@ struct __wt_page {
 
 #define	WT_PAGE_BULK_LOAD	0x01	/* Page bulk loaded */
 #define	WT_PAGE_CACHE_COUNTED	0x02	/* Page counted in cache stats */
-#define	WT_PAGE_DELETED		0x04	/* Page was empty at reconcilation */
+#define	WT_PAGE_DELETED		0x04	/* Page was empty at reconciliation */
 #define	WT_PAGE_INITIAL_EMPTY	0x08	/* Empty page created during open */
 #define	WT_PAGE_MODIFIED	0x10	/* Page is modified */
 #define	WT_PAGE_PINNED		0x20	/* Page is pinned */
diff --git a/src/include/wiredtiger.in b/src/include/wiredtiger.in
index f48f7e8a4b3..44db5d9803b 100644
--- a/src/include/wiredtiger.in
+++ b/src/include/wiredtiger.in
@@ -720,7 +720,7 @@ struct wt_session {
 	 * @param session the session handle
 	 * @configstart{session.checkpoint, see dist/api_data.py}
 	 * @config{archive, remove log files no longer required for
-	 * transactional durabilty}
+	 * transactional durability}
 	 * @config{force, write a new checkpoint even if nothing has changed
 	 * since the last one}
 	 * @config{flush_cache, flush the cache}
diff --git a/src/support/pow.c b/src/support/pow.c
index d458c4fbbff..c004531ade0 100644
--- a/src/support/pow.c
+++ b/src/support/pow.c
@@ -16,7 +16,7 @@
  * Note that in the edge case where v is 0, it returns 0, which isn't a power of
  * 2; you might append the expression v += (v == 0) to remedy this if it
  * matters.  It would be faster by 2 operations to use the formula and the
- * log base 2 methed that uses a lookup table, but in some situations, lookup
+ * log base 2 method that uses a lookup table, but in some situations, lookup
  * tables are not suitable, so the above code may be best. (On a Athlon XP 2100+
  * I've found the above shift-left and then OR code is as fast as using a single
  * BSR assembly language instruction, which scans in reverse to find the highest
@@ -24,7 +24,7 @@
  * and then adding one, which results in carries that set all of the lower bits
  * to 0 and one bit beyond the highest set bit to 1. If the original number was
  * a power of 2, then the decrement will reduce it to one less, so that we round
- * up to the same original value.  Devised by Sean Anderson, Sepember 14, 2001.
+ * up to the same original value.  Devised by Sean Anderson, September 14, 2001.
  * Pete Hart pointed me to a couple newsgroup posts by him and William Lewis in
  * February of 1997, where they arrive at the same algorithm.
  *	http://graphics.stanford.edu/~seander/bithacks.html