From 59acd90b04c8b9701c345af56a8d2086275ca9d9 Mon Sep 17 00:00:00 2001
From: neauoire <aliceffekt@gmail.com>
Date: Sat, 18 Nov 2023 19:34:58 -0800
Subject: [PATCH] (lz) Progress

---
 cli/lz/ulzenc.c   | 54 +++++++++++++++++++++++++++++------------------
 cli/lz/ulzenc.tal | 32 +++++-----------------------
 2 files changed, 38 insertions(+), 48 deletions(-)

diff --git a/cli/lz/ulzenc.c b/cli/lz/ulzenc.c
index 38ba59b..f56f54f 100644
--- a/cli/lz/ulzenc.c
+++ b/cli/lz/ulzenc.c
@@ -22,23 +22,24 @@ uxn_lz_compress(const void *input, int input_size)
 	const unsigned char *dict, *dict_best = 0;
 	const unsigned char *in = input, *start = in, *end = in + input_size;
 	while(in != end) {
-
 		/* Get available dictionary size (history of original output) */
 		dict_len = (int)(in - start);
 		if(dict_len > 256)
 			dict_len = 256;
-
 		/* Size of the string to search for */
 		/* Limit string length to what we can fit in 14 bits, plus the minimum match length */
 		string_len = (int)(end - in);
 		if(string_len > 0x3FFF + MinMatchLength)
 			string_len = 0x3FFF + MinMatchLength;
-		/* DEBUG: printf("%04x %04x \n", dict_len, string_len); */
 
+		printf("DEBUG1: %04x %04x ", dict_len, string_len); 
+		
 		/* Iterate through the dictionary */
 		match_len = 0;
 		dict = in - dict_len;
-		for(; dict_len; dict++, dict_len -= 1) {
+
+		printf("[%c]", in[0]);
+		for(; dict_len; dict++, dict_len--) {
 			/* Find common prefix length with the string */
 			/* If we reach the end of the string, this is the best possible match. End. */
 			for(i = 0;; i++) {
@@ -48,8 +49,10 @@ uxn_lz_compress(const void *input, int input_size)
 					goto done_search;
 				}
 				/* Dictionary repeats if we hit the end */
-				if(in[i] != dict[i % dict_len])
+				/* printf("(#%d, %04x, %c, %c)", i, in-raw, in[i], dict[i % dict_len]); */
+				if(in[i] != dict[i % dict_len]){
 					break;
+				}
 			}
 			if(i > match_len) {
 				match_len = i;
@@ -57,8 +60,12 @@ uxn_lz_compress(const void *input, int input_size)
 			}
 		}
 
+
 	done_search:
 
+		printf("-> %04x \n", match_len);
+		
+
 		/* CPY */
 
 		if(match_len >= MinMatchLength) {
@@ -75,26 +82,31 @@ uxn_lz_compress(const void *input, int input_size)
 			*output_ptr++ = in - dict_best - 1;
 			in += match_len; /* Advance input by size of the match */
 			combine = 0;     /* Disable combining previous literal, if any */
-			continue;
+			
 		}
-
+		
 		/* LIT */
-
-		/* Combine with previous literal */
-		if(combine) {
-			if(++*combine == 127)
-				combine = 0;
+		else{
+			/* printf("LIT:%d\n", combine); */
+			/* Combine with previous literal */
+			if(combine) {
+				if(++*combine == 127)
+					combine = 0;
+			}
+			/* Start a new literal */
+			else {
+				/* Store this address, and later use it to increment the literal size. */
+				combine = output_ptr++;
+				/* The 0 here means literal of length 1. */
+				*combine = 0;
+			}
+			
+			/* Write 1 literal byte from the input to the output. */
+			*output_ptr++ = *in++;
+			/* printf(">> %d\n", in-raw); */
 		}
-		/* Start a new literal */
-		else {
-			/* Store this address, and later use it to increment the literal size. */
-			combine = output_ptr++;
-			/* The 0 here means literal of length 1. */
-			*combine = 0;
-		}
-		/* Write 1 literal byte from the input to the output. */
-		*output_ptr++ = *in++;
 	}
+	
 	return (int)(output_ptr - mem);
 }
 
diff --git a/cli/lz/ulzenc.tal b/cli/lz/ulzenc.tal
index 36616ac..7922bc0 100644
--- a/cli/lz/ulzenc.tal
+++ b/cli/lz/ulzenc.tal
@@ -52,7 +52,7 @@
 
 	ADD2k NIP2 SWP2
 	&w ( end* start* -- )
-
+		EQU2k ?&end
 		( | get available dictionary size )
 		DUP2 ;raw SUB2 #0100 LTH2k ?{ SWP2 }
 		POP2 .dict-len STZ2
@@ -71,6 +71,7 @@
 		LIT "[ #18 DEO
 		LDAk #18 DEO
 		LIT "] #18 DEO
+		#0a18 DEO
 
 		&for1 ( for ; dict_len; dict++, dict_len-- )
 			.dict-len LDZ2 #0000 EQU2 ?&end-for1
@@ -84,14 +85,6 @@
 					.dict LDZ2 .dict-best STZ2
 					POP2 !&done-search }
 
-				( Dictionary repeats if we hit the end )
-				LIT "( #18 DEO
-				DUP phex/b #2018 DEO
-				OVR2 ;raw SUB2 phex #2018 DEO
-				ADD2k LDA #18 DEO #2018 DEO
-				DUP2 .dict-len LDZ2 DIV2k MUL2 SUB2 .dict LDZ2 ADD2 LDA #18 DEO
-				LIT ") #18 DEO #0a18 DEO
-
 				( in[i] != dict[i % dict_len] break; )
 				( a ) ADD2k LDA STH
 				( b ) DUP2 .dict-len LDZ2 DIV2k MUL2 SUB2 .dict LDZ2 ADD2 LDA STHr
@@ -107,27 +100,12 @@
 
 			.dict LDZ2 INC2 .dict STZ2
 			.dict-len LDZ2 #0001 SUB2 .dict-len STZ2 !&for1
-
 		&end-for1
-
 	&done-search ( -- )
-
-		LIT "- #18 DEO
-		.match-len LDZ2 phex #0a18 DEO
-
-		( [ LIT &break $1 ] INCk ,&break STR #03 NEQ ?{ BREAK } )
-
-		
 		( CPY ) .match-len LDZ2 #0003 GTH2 ?op-cpy
-		( LIT ) op-lit
-
-		LIT "> #18 DEO
-		DUP2 ;raw SUB2 phex #0a18 DEO
-
-		#010f DEO BRK
-		NEQ2k ?&w
-	JMP2r
-
+		( LIT ) !op-lit
+	&end
+	POP2 POP2 JMP2r
 
 
 (