Keywords

This version of BASIC recognises 129 keywords, whose storage is worthy of explanation. First of all, there is a keyword directory containing 26 pointers to keyword groups, one for each letter of the alphabet.

Keyword Directory

Each entry in the directory points to the group of zero or more keywords beginning with that letter.

Addr.	Letter	Group Addr.	Addr.	Letter	Group Addr.
013B:	A	016F	0155:	N	02BA
013D:	B	0180	0157:	O	02CD
013F:	C	0181	0159:	P	02DD
0141:	D	01BD	015B:	Q	02F1
0143:	E	01FA	015D:	R	02F2
0145:	F	021F	015F:	S	0322
0147:	G	0235	0161:	T	0351
0149:	H	0247	0163:	U	0368
014B:	I	024C	0165:	V	0377
014D:	J	0262	0167:	W	0381
014F:	K	0263	0169:	X	038B
0151:	L	0268	016B:	Y	038F
0153:	M	029C	016D:	Z	0390

Thus you can see from the above that keywords beginning with 'A' are stored at 016F, keywords beginning with 'B' are stored at 0180, and so on.

Keyword Groups

There are 26 keyword groups, one for each letter of the alphabet. Because the first keyword letter is always known by the group the keyword belongs to, the first letter of each keyword is not stored. The keywords in a group are are concatenated and delimited as follows: the last byte of each keyword has it's topmost bit set to mark the end of the keyword. The byte following this is the keyword ID byte - an ID BASIC uses internally for processing keywords. The end of the group is marked with a null byte. Lets look at the raw bytes for the group of keywords beginning with 'A' which we know from the directory is stored at 016F:

016F: 4E C4 F7 42 D3 06 54 CE 0E 53 C3 15 55 54 CF AB 00

016F: 4E C4 F7 42 D3 06 54 CE 0E 53 C3 15 55 54 CF AB 00

So this is interpreted as:

Addr.	Bytes	Keyword	ID
016F:	4E C4 F7	(A)ND	F7
0172:	42 D3 06	(A)BS	06
0175:	54 CE 0E	(A)TN	0E
0178:	53 C3 15	(A)SC	15
017B:	55 54 CF AB	(A)UTO	AB
017F:	00 <-- end of group marker

This approach has two immediate benefits:
1. Parsing BASIC programs is faster: it's much quicker to search a mere handful of keywords than to iterate through all 129 of them.
2. Less space to required to store keywords. The keyword directory above occupies 26*3=78 bytes. By using such a directory, the first letter of each keyword is always known and therefore doesn't need to be stored. This saves one byte per keyword (ie 129 bytes), which is a profit of 51 bytes.

Those Keywords In Full

Here is the full list of keywords, in the order that BASIC has them stored. When this is finished, each keyword will be linked to the appropriate disassembly page.

Keyword	ID
AND	F7
ABS	06
ATN	0E
ASC	15
AUTO	AB
CONSOLE	A0
CLOSE	C3
CONT	9A
CLEAR	92
CLOAD	9C
CSAVE	9B
CINT	1C
CSNG	1D
CDBL	1E
CVI	2B
CVS	2C
CVD	2D
COS	0C
CHR$	16
DATA	84
DIM	86
DEFSTR	AD
DEFINT	AE
DEFSNG	AF
DEFDBL	B0
DSKO$	BC
DEF	98
DELETE	AA
DSKI$	2A
DSKF	2E
DSKINI	CC
END	81
ELSE	A2
ERASE	A6
EDIT	A7
ERROR	A8
ERL	D6
ERR	D7
EXP	0B
EOF	2F
EQV	FA
FOR	82
FIELD	C0
FILES	C6
FN	D3
FRE	0F
FIX	1F
GOTO	89
GO TO	89
GOSUB	8D
GET	C1
HEX$	1A
INPUT	85
IF	8B
INSTR	DA
INT	05
INP	10
IMP	FB
KILL	C8
LET	88
LINE	B1
LOAD	C4
LSET	C9
LPRINT	9E
LIST	93
LLIST	9F
LPOS	1B
LOG	0A
LOC	30
LEN	12
LEFT$	01
LOF	31
MOUNT	BE
MERGE	C5
MOD	FC
MKI$	32
MKS$	33
MKD$	34
MID$	03
NEXT	83
NULL	96
NAME	C7
NEW	94
NOT	D5
OUT	9D
ON	95
OPEN	BF
OR	F8
OCT$	19
PUT	C2
POKE	99
PRINT	91
POS	11
PEEK	17
READ	87
RUN	8A
RESTORE	8C
RETURN	8E
REM	8F
RESUME	A9
RSET	CA
RIGHT$	02
RND	08
RENUM	AC
STOP	90
SWAP	A5
SAVE	CB
SPC(	D4
STEP	D1
SGN	04
SQR	07
SIN	09
STR$	13
STRING$	D8
SPACE$	18
TRON	A3
TROFF	A4
TAB(	D0
TO	CE
THEN	CF
TAN	0D
UNLOAD	BD
USING	D9
USR	D2
VAL	14
VARPTR	DC
WIDTH	A1
WAIT	97
XOR	F9