X-Git-Url: https://thelambdalab.xyz/gitweb/index.cgi?p=forth.jl.git;a=blobdiff_plain;f=src%2Flib.4th;h=2312aeb1727ca682657c07e0d484dfb5eb7c7386;hp=7a3833d1e48ebee2f94c8fb4540a755f8d59e1c9;hb=b593ff258f6b62e7bd7c00b566d8f4f812b0f95a;hpb=3de60226999be55f55a944418c6794faf623509d diff --git a/src/lib.4th b/src/lib.4th index 7a3833d..2312aeb 100644 --- a/src/lib.4th +++ b/src/lib.4th @@ -1,3 +1,9 @@ +: \ IMMEDIATE + #TIB @ >IN ! +; \ We can now comment! + +\ BASIC DEFINITIONS ---------------------------------------------------------------------- + : / /MOD SWAP DROP ; : MOD /MOD DROP ; : */ -ROT * SWAP / ; @@ -8,32 +14,37 @@ : FALSE 0 ; : NOT 0= ; -: CELLS ; \ Allow for slightly more portable code +\ Translate a number of cells into memory units +\ (in our case 1 cell = 1 memory unit) +: CELLS ; + +\ Since the smallest unit of memory in our system is 64 bits and since strings +\ are stored as arrays of 64 bit integers, the character store/fetch words are +\ just aliases of the standard store/fetch words. +: C! ! ; +: C@ @ ; +: C, , ; -: DEPTH PSP@ PSP0 @ - ; +: DEPTH PSP@ PSP0 - ; : '\n' 10 ; : BL 32 ; -: LITERAL IMMEDIATE ' LIT , , ; +: LITERAL IMMEDIATE ['] LIT , , ; -: ':' [ CHAR : ] LITERAL ; -: ';' [ CHAR ; ] LITERAL ; -: '(' [ CHAR ( ] LITERAL ; -: ')' [ CHAR ) ] LITERAL ; -: '<' [ CHAR < ] LITERAL ; -: '>' [ CHAR > ] LITERAL ; -: '"' [ CHAR " ] LITERAL ; -: 'A' [ CHAR A ] LITERAL ; -: '0' [ CHAR 0 ] LITERAL ; -: '-' [ CHAR - ] LITERAL ; -: '.' [ CHAR . ] LITERAL ; +: ' BL WORD FIND >CFA ; + +: CHAR BL WORD 1+ @ ; +: [CHAR] IMMEDIATE + CHAR + ['] LIT , , +; : CR '\n' emit ; : SPACE BL emit ; : [COMPILE] IMMEDIATE - WORD \ get the next word + BL WORD \ get the next word FIND \ find it in the dictionary >CFA \ get its codeword , \ and compile that @@ -45,73 +56,78 @@ , \ compile it ; -: DEBUGON TRUE DEBUG ! ; -: DEBUGOFF FALSE DEBUG ! ; - \ CONTROL STRUCTURES ---------------------------------------------------------------------- : IF IMMEDIATE - ' 0BRANCH , \ compile 0BRANCH - HERE @ \ save location of the offset on the stack + ['] 0BRANCH , \ compile 0BRANCH + HERE \ save location of the offset on the stack 0 , \ compile a dummy offset ; : THEN IMMEDIATE DUP - HERE @ SWAP - \ calculate the offset from the address saved on the stack + HERE SWAP - \ calculate the offset from the address saved on the stack SWAP ! \ store the offset in the back-filled location ; : ELSE IMMEDIATE - ' BRANCH , \ definite branch to just over the false-part - HERE @ \ save location of the offset on the stack + ['] BRANCH , \ definite branch to just over the false-part + HERE \ save location of the offset on the stack 0 , \ compile a dummy offset SWAP \ now back-fill the original (IF) offset DUP \ same as for THEN word above - HERE @ SWAP - + HERE SWAP - SWAP ! ; : BEGIN IMMEDIATE - HERE @ \ save location on the stack + HERE \ save location on the stack ; : UNTIL IMMEDIATE - ' 0BRANCH , \ compile 0BRANCH - HERE @ - \ calculate the offset from the address saved on the stack + ['] 0BRANCH , \ compile 0BRANCH + HERE - \ calculate the offset from the address saved on the stack , \ compile the offset here ; : AGAIN IMMEDIATE - ' BRANCH , \ compile BRANCH - HERE @ - \ calculate the offset back + ['] BRANCH , \ compile BRANCH + HERE - \ calculate the offset back , \ compile the offset here ; : WHILE IMMEDIATE - ' 0BRANCH , \ compile 0BRANCH - HERE @ \ save location of the offset2 on the stack + ['] 0BRANCH , \ compile 0BRANCH + HERE \ save location of the offset2 on the stack 0 , \ compile a dummy offset2 ; : REPEAT IMMEDIATE - ' BRANCH , \ compile BRANCH + ['] BRANCH , \ compile BRANCH SWAP \ get the original offset (from BEGIN) - HERE @ - , \ and compile it after BRANCH + HERE - , \ and compile it after BRANCH DUP - HERE @ SWAP - \ calculate the offset2 + HERE SWAP - \ calculate the offset2 SWAP ! \ and back-fill it in the original location ; : UNLESS IMMEDIATE - ' NOT , \ compile NOT (to reverse the test) + ['] NOT , \ compile NOT (to reverse the test) [COMPILE] IF \ continue by calling the normal IF ; : DO IMMEDIATE - ' >R , ' >R , - ' LIT , HERE @ 0 , ' >R , - HERE @ + ['] LIT , -1 , [COMPILE] IF + ['] >R , ['] >R , + ['] LIT , HERE 0 , ['] >R , + HERE +; + +: ?DO IMMEDIATE + ['] 2DUP , ['] - , [COMPILE] IF + ['] >R , ['] >R , + ['] LIT , HERE 0 , ['] >R , + HERE ; : I RSP@ 3 - @ ; @@ -119,43 +135,96 @@ : J RSP@ 6 - @ ; : ?LEAVE IMMEDIATE - ' 0BRANCH , 13 , - ' R> , ' RDROP , ' RDROP , - ' LIT , HERE @ 7 + , ' DUP , ' -ROT , ' - , ' SWAP , ' ! , - ' BRANCH , + ['] 0BRANCH , 13 , + ['] R> , ['] RDROP , ['] RDROP , + ['] LIT , HERE 7 + , ['] DUP , ['] -ROT , ['] - , ['] SWAP , ['] ! , + ['] BRANCH , 0 , ; : LEAVE IMMEDIATE - ' LIT , -1 , + ['] LIT , -1 , [COMPILE] ?LEAVE ; : +LOOP IMMEDIATE - ' R> , ' SWAP , ' R> , ' SWAP , ' R> , ' SWAP , ' + , ' 2DUP , ' - , - ' SWAP , ' >R , ' SWAP , ' >R , ' SWAP , ' >R , - ' 0<= , ' 0BRANCH , - HERE @ - , - ' RDROP , ' RDROP , ' RDROP , - HERE @ SWAP ! + + ['] DUP , \ Store copy of increment + + ['] R> , ['] SWAP , ['] R> , ['] SWAP , ['] R> , ['] SWAP , ['] + , ['] 2DUP , ['] - , + ['] SWAP , ['] >R , ['] SWAP , ['] >R , ['] SWAP , ['] >R , + + \ Condition differently depending on sign of increment + ['] SWAP , ['] 0>= , [COMPILE] IF + ['] 0<= , + [COMPILE] ELSE + ['] 0> , + [COMPILE] THEN + + \ Branch back to begining of loop kernel + ['] 0BRANCH , HERE - , + + \ Clean up + ['] RDROP , ['] RDROP , ['] RDROP , + + \ Record address of loop end for any LEAVEs to use + HERE SWAP ! + + [COMPILE] ELSE + ['] 2DROP , \ Clean up if loop was entirely skipped (?DO) + [COMPILE] THEN ; : LOOP IMMEDIATE - ' LIT , 1 , + ['] LIT , 1 , [COMPILE] +LOOP ; + +\ CASE ------------------------------------------------------------------------ + +: CASE IMMEDIATE + 0 \ push 0 to mark the bottom of the stack +; + +: OF IMMEDIATE + ['] OVER , \ compile OVER + ['] = , \ compile = + [COMPILE] IF \ compile IF + ['] DROP , \ compile DROP +; + +: ENDOF IMMEDIATE + [COMPILE] ELSE \ ENDOF is the same as ELSE +; + +: ENDCASE IMMEDIATE + ['] DROP , \ compile DROP + + \ keep compiling THEN until we get to our zero marker + BEGIN + ?DUP + WHILE + [COMPILE] THEN + REPEAT +; + + \ COMMENTS ---------------------------------------------------------------------- : ( IMMEDIATE 1 \ allowed nested parens by keeping track of depth BEGIN - KEY \ read next character - DUP '(' = IF \ open paren? + >IN @ #TIB @ >= IF \ End of TIB? + QUERY \ Get next line + THEN + + TIB >IN @ + @ 1 >IN +! + DUP [CHAR] ( = IF \ open paren? DROP \ drop the open paren 1+ \ depth increases ELSE - ')' = IF \ close paren? + [CHAR] ) = IF \ close paren? 1- \ depth decreases THEN THEN @@ -171,12 +240,22 @@ PSP@ SWAP - ( add to the stack pointer ) @ ( and fetch ) ; +: ROLL ( x_u x_u-1... x_0 u -- x_u-1 ... x_0 x_u ) + 1+ DUP PICK SWAP ( x_u x_u-1 ... x_0 x_u u+1 ) + PSP@ 1- SWAP - PSP@ 2- SWAP + DO + i 1+ @ i ! + LOOP + SWAP DROP +; ( With the looping constructs, we can now write SPACES, which writes n spaces to stdout. ) : SPACES ( n -- ) - 0 DO - SPACE - LOOP + DUP 0> IF + 0 DO SPACE LOOP + ELSE + DROP + THEN ; ( Standard words for manipulating BASE. ) @@ -217,10 +296,10 @@ ( print the remainder ) DUP 10 < IF - '0' ( decimal digits 0..9 ) + [CHAR] 0 ( decimal digits 0..9 ) ELSE 10 - ( hex and beyond digits A..Z ) - 'A' + [CHAR] A THEN + EMIT @@ -273,7 +352,7 @@ SWAP ( u flag ) IF ( was it negative? print the - character ) - '-' EMIT + [CHAR] - EMIT THEN U. @@ -282,8 +361,8 @@ : . 0 .R SPACE ; : .S ( -- ) - '<' EMIT DEPTH U. '>' EMIT SPACE - PSP0 @ 1+ + [CHAR] < EMIT DEPTH U. [CHAR] > EMIT SPACE + PSP0 1+ BEGIN DUP PSP@ 2 - <= WHILE @@ -314,13 +393,367 @@ THEN ; -: ROLL ( x_u x_u-1... x_0 u -- x_u-1 ... x_0 x_u ) - 1+ DUP PICK SWAP ( x_u x_u-1 ... x_0 x_u u+1 ) - PSP@ 1- SWAP - PSP@ 2- SWAP + +( STRINGS ---------------------------------------------------------------------- ) + + +( Block copy, however, is important and novel: ) +: CMOVE ( src dest length -- ) + + DUP 0<= IF + EXIT + THEN + + -ROT OVER - ( length src (dest-src) ) + -ROT DUP ROT + SWAP ( (dest-src) (src+length) src ) + DO - i 1+ @ i ! + I @ ( (dest-src) i@ ) + OVER I + ( (dest-src) i@ (dest-src+i) ) + ! ( (dest-src) ) + LOOP + + DROP +; + +( C, appends a byte to the current compiled word. ) +: C, + HERE C! + 1 H +! +; + +: S" IMMEDIATE ( -- addr len ) + STATE @ IF ( compiling? ) + ['] LITSTRING , ( compile LITSTRING ) + HERE ( save the address of the length word on the stack ) + 0 , ( dummy length - we don't know what it is yet ) + + BEGIN + >IN @ #TIB @ >= IF \ End of TIB? + QUERY \ Get next line + THEN + + TIB >IN @ + @ 1 >IN +! \ Get char from TIB + + DUP [CHAR] " <> + WHILE + C, ( copy character ) + REPEAT + DROP ( drop the double quote character at the end ) + DUP ( get the saved address of the length word ) + HERE SWAP - ( calculate the length ) + 1- ( subtract 1 (because we measured from the start of the length word) ) + SWAP ! ( and back-fill the length location ) + ELSE ( immediate mode ) + HERE ( get the start address of the temporary space ) + + BEGIN + >IN @ #TIB @ >= IF \ End of TIB? + QUERY \ Get next line + THEN + + TIB >IN @ + @ 1 >IN +! \ Get char from TIB + + DUP [CHAR] " <> + WHILE + OVER C! ( save next character ) + 1+ ( increment address ) + REPEAT + DROP ( drop the final " character ) + HERE - ( calculate the length ) + HERE ( push the start address ) + SWAP ( addr len ) + THEN +; + +: ." IMMEDIATE ( -- ) + [COMPILE] S" ( read the string, and compile LITSTRING, etc. ) + ['] TYPE , ( compile the final TYPE ) +; + +: .( + BEGIN + >IN @ #TIB @ >= IF \ End of TIB? + QUERY \ Get next line + THEN + + TIB >IN @ + @ 1 >IN +! \ Get char from TIB + + DUP [CHAR] ) = IF + DROP ( drop the double quote character ) + EXIT ( return from this function ) + THEN + EMIT + AGAIN +; + +( Converts address of counted string into address of + start of string and length of string. ) +: COUNT ( addr1 -- addr2 n ) + DUP 1+ SWAP @ ; + + +( CONSTANTS AND VARIABLES ------------------------------------------------------ ) + +: CONSTANT + CREATE , +DOES> @ +; + +: ALLOT ( n -- ) + H +! ( adds n to H, after this the old value of H is still on the stack ) +; + +: VARIABLE + CREATE + 1 CELLS ALLOT ( allocate 1 cell of memory, push the pointer to this memory ) +; + +: VALUE ( n -- ) + CREATE , +DOES> @ +; + +: TO IMMEDIATE ( n -- ) + BL WORD ( get the name of the value ) + FIND ( look it up in the dictionary ) + >PFA ( get a pointer to the first data field (the 'LIT') ) + STATE @ IF ( compiling? ) + ['] LIT , ( compile LIT ) + , ( compile the address of the value ) + ['] ! , ( compile ! ) + ELSE ( immediate mode ) + ! ( update it straightaway ) + THEN +; + +( x +TO VAL adds x to VAL ) +: +TO IMMEDIATE + BL WORD ( get the name of the value ) + FIND ( look it up in the dictionary ) + >PFA ( get a pointer to the first data field (the 'LIT') ) + STATE @ IF ( compiling? ) + ['] LIT , ( compile LIT ) + , ( compile the address of the value ) + ['] +! , ( compile +! ) + ELSE ( immediate mode ) + +! ( update it straightaway ) + THEN +; + +( Fill u ints, starting at a, with the value b ) +: FILL ( a u b -- ) + -ROT OVER + SWAP ?DO + DUP I ! LOOP + DROP +; + +: ERASE ( a u -- ) + 0 FILL +; + +( PRINTING THE DICTIONARY ------------------------------------------------------ ) + +: ID. + 1+ ( skip over the link pointer ) + DUP @ ( get the flags/length byte ) + F_LENMASK AND ( mask out the flags - just want the length ) + + BEGIN + DUP 0> ( length > 0? ) + WHILE + SWAP 1+ ( addr len -- len addr+1 ) + DUP @ ( len addr -- len addr char | get the next character) + DUP 32 >= OVER 127 <= AND IF + EMIT ( len addr char -- len addr | and print it) + ELSE + BASE @ SWAP HEX + ." \x" 0 .R + BASE ! + THEN + SWAP 1- ( len addr -- addr len-1 | subtract one from length ) + REPEAT + 2DROP ( len addr -- ) +; + +: ?HIDDEN + 1+ ( skip over the link pointer ) + @ ( get the flags/length byte ) + F_HIDDEN AND ( mask the F_HIDDEN flag and return it (as a truth value) ) +; +: ?IMMEDIATE + 1+ ( skip over the link pointer ) + @ ( get the flags/length byte ) + F_IMMED AND ( mask the F_IMMED flag and return it (as a truth value) ) +; + +: WORDS + LATEST @ ( start at LATEST dictionary entry ) + BEGIN + ?DUP ( while link pointer is not null ) + WHILE + DUP ?HIDDEN NOT IF ( ignore hidden words ) + DUP ID. ( but if not hidden, print the word ) + SPACE + THEN + @ ( dereference the link pointer - go to previous word ) + REPEAT + CR +; + + +( FORGET ---------------------------------------------------------------------- ) + +: FORGET + BL WORD FIND ( find the word, gets the dictionary entry address ) + DUP @ LATEST ! ( set LATEST to point to the previous word ) + H ! ( and store H with the dictionary address ) +; + +( DUMP ------------------------------------------------------------------------ ) + +\ TODO! + + +( DECOMPILER ------------------------------------------------------------------ ) + +: CFA> + LATEST @ ( start at LATEST dictionary entry ) + BEGIN + ?DUP ( while link pointer is not null ) + WHILE + 2DUP SWAP ( cfa curr curr cfa ) + < IF ( current dictionary entry < cfa? ) + NIP ( leave curr dictionary entry on the stack ) + EXIT + THEN + @ ( follow link pointer back ) + REPEAT + DROP ( restore stack ) + 0 ( sorry, nothing found ) +; + +: SEE + BL WORD DUP FIND ( find the dictionary entry to decompile ) + + ?DUP 0= IF + ." Word '" COUNT TYPE ." ' not found in dictionary." + EXIT + THEN + SWAP DROP + + ( Now we search again, looking for the next word in the dictionary. This gives us + the length of the word that we will be decompiling. (Well, mostly it does). ) + HERE ( address of the end of the last compiled word ) + LATEST @ ( word last curr ) + BEGIN + 2 PICK ( word last curr word ) + OVER ( word last curr word curr ) + <> ( word last curr word<>curr? ) + WHILE ( word last curr ) + NIP ( word curr ) + DUP @ ( word curr prev (which becomes: word last curr) ) + REPEAT + + DROP ( at this point, the stack is: start-of-word end-of-word ) + SWAP ( end-of-word start-of-word ) + + DUP >CFA @ CASE + DOCOL OF + \ Colon definition + [CHAR] : EMIT SPACE DUP ID. SPACE + DUP ?IMMEDIATE IF ." IMMEDIATE " THEN CR + ENDOF + DOVAR OF + \ Variable definition + ." Variable " DUP ID. CR + 2DROP EXIT + ENDOF + DOCON OF + \ Constant definition + ." Constant " DUP ID. CR + 2DROP EXIT + ENDOF + + \ Unknown codeword + ." Primitive or word with unrecognized codeword." CR + DROP 2DROP EXIT + ENDCASE + + ( begin the definition with : NAME [IMMEDIATE] ) + ( [CHAR] : EMIT SPACE DUP ID. SPACE + DUP ?IMMEDIATE IF ." IMMEDIATE " THEN CR 4 ) + + 4 SPACES + + >PFA ( get the data address, ie. points after DOCOL | end-of-word start-of-data ) + + ( now we start decompiling until we hit the end of the word ) + BEGIN ( end start ) + 2DUP > + WHILE + DUP @ ( end start codeword ) + + CASE + ['] LIT OF ( is it LIT ? ) + 1+ DUP @ ( get next word which is the integer constant ) + . ( and print it ) + ENDOF + ['] LITSTRING OF ( is it LITSTRING ? ) + [CHAR] S EMIT [CHAR] " EMIT SPACE ( print S" ) + 1+ DUP @ ( get the length word ) + SWAP 1+ SWAP ( end start+1 length ) + 2DUP TYPE ( print the string ) + [CHAR] " EMIT SPACE ( finish the string with a final quote ) + + ( end start+1+len, aligned ) + 1- ( because we're about to add 4 below ) + ENDOF + ['] 0BRANCH OF ( is it 0BRANCH ? ) + ." 0BRANCH ( " + 1+ DUP @ ( print the offset ) + . + ." ) " + ENDOF + ['] BRANCH OF ( is it BRANCH ? ) + ." BRANCH ( " + 1+ DUP @ ( print the offset ) + . + ." ) " + ENDOF + ['] ['] OF ( is it ['] ? ) + ." ['] " + 1+ DUP @ ( get the next codeword ) + CFA> ( and force it to be printed as a dictionary entry ) + ID. SPACE + ENDOF + ['] EXIT OF ( is it EXIT? ) + ( We expect the last word to be EXIT, and if it is then we don't print it + because EXIT is normally implied by ;. EXIT can also appear in the middle + of words, and then it needs to be printed. ) + 2DUP ( end start end start ) + 1+ ( end start end start+1 ) + <> IF ( end start | we're not at the end ) + ." EXIT " + THEN + ENDOF + ( default case: ) + DUP ( in the default case we always need to DUP before using ) + CFA> ( look up the codeword to get the dictionary entry ) + ID. SPACE ( and print it ) + ENDCASE + + 1+ ( end start+1 ) + REPEAT + + [CHAR] ; EMIT CR + + 2DROP ( restore stack ) ; +( MEMORY ------------------------------------------------------------------ ) + +: UNUSED ( -- cells ) + MEMSIZE HERE - ;