From a0c3927510c8c92925f602c999d75c3e86c165b0 Mon Sep 17 00:00:00 2001
From: drh <drh@noemail.net>
Date: Thu, 6 Jul 2017 16:33:14 +0000
Subject: [PATCH] Change the (machine-generated) keywordhash.h file to increase
 the scope of the tables used for keyword matching, so that the tables are
 accessible to functions other then keywordCode().

FossilOrigin-Name: c5ed5ebdf660501fde7cc8aefaaaeae2a68e5899a64ac93f26684842a235281d
---
 manifest             | 12 +++---
 manifest.uuid        |  2 +-
 tool/mkkeywordhash.c | 90 ++++++++++++++++++++++++--------------------
 3 files changed, 57 insertions(+), 47 deletions(-)

diff --git a/manifest b/manifest
index b35c0c6bbe..a6685636f4 100644
--- a/manifest
+++ b/manifest
@@ -1,5 +1,5 @@
-C More\scompact\simplementation\sof\sthe\stypeof()\sSQL\sfunction.
-D 2017-07-06T13:51:50.091
+C Change\sthe\s(machine-generated)\skeywordhash.h\sfile\sto\sincrease\sthe\sscope\sof\sthe\ntables\sused\sfor\skeyword\smatching,\sso\sthat\sthe\stables\sare\saccessible\sto\nfunctions\sother\sthen\skeywordCode().
+D 2017-07-06T16:33:14.765
 F Makefile.in 081e48dfe7f995d57ce1a88ddf4d2917b4349158648a6cd45b42beae30de3a12
 F Makefile.linux-gcc 7bc79876b875010e8c8f9502eb935ca92aa3c434
 F Makefile.msc 4ebb1d257cac7fb1bcb4ba59278416d410ff1c4bf59447a9c37a415f3516056a
@@ -1561,7 +1561,7 @@ F tool/logest.c 11346aa019e2e77a00902aa7d0cabd27bd2e8cca
 F tool/max-limits.c cbb635fbb37ae4d05f240bfb5b5270bb63c54439
 F tool/mkautoconfamal.sh e855df211ecbcc7131dee817110ff386cfb112f7
 F tool/mkctimec.tcl dd183b73ae1c28249669741c250525f0407e579a70482371668fd5f130d9feb3
-F tool/mkkeywordhash.c f7f3b342211ac6a14258b9726d5b97cf4f548f22
+F tool/mkkeywordhash.c 2e852ac0dfdc5af18886dc1ce7e9676d11714ae3df0a282dc7d90b3a0fe2033c
 F tool/mkmsvcmin.tcl cbd93f1cfa3a0a9ae56fc958510aa3fc3ac65e29cb111716199e3d0e66eefaa4
 F tool/mkopcodec.tcl d1b6362bd3aa80d5520d4d6f3765badf01f6c43c
 F tool/mkopcodeh.tcl a01d2c1d8a6205b03fc635adf3735b4c523befd3
@@ -1628,7 +1628,7 @@ F vsixtest/vsixtest.tcl 6a9a6ab600c25a91a7acc6293828957a386a8a93
 F vsixtest/vsixtest.vcxproj.data 2ed517e100c66dc455b492e1a33350c1b20fbcdc
 F vsixtest/vsixtest.vcxproj.filters 37e51ffedcdb064aad6ff33b6148725226cd608e
 F vsixtest/vsixtest_TemporaryKey.pfx e5b1b036facdb453873e7084e1cae9102ccc67a0
-P 06269257647db51fbc9f8cda88abac5db28b6f917a509768af7686dfa2b94511
-R 8d88458ca1178062a686c7622fab862c
+P efb4aab0caa4145732a5438cc2a193bc12c455b4007220564d240e75900ea8ad
+R 77ecc187246c1c337a58790c9818d2e8
 U drh
-Z 16d02781d00f9c4bbd9708369c77eae3
+Z a448086e00aa693e2477a7bd0d12a0df
diff --git a/manifest.uuid b/manifest.uuid
index 4e696488e9..eed343f04b 100644
--- a/manifest.uuid
+++ b/manifest.uuid
@@ -1 +1 @@
-efb4aab0caa4145732a5438cc2a193bc12c455b4007220564d240e75900ea8ad
\ No newline at end of file
+c5ed5ebdf660501fde7cc8aefaaaeae2a68e5899a64ac93f26684842a235281d
\ No newline at end of file
diff --git a/tool/mkkeywordhash.c b/tool/mkkeywordhash.c
index 7e5287ea54..42112718da 100644
--- a/tool/mkkeywordhash.c
+++ b/tool/mkkeywordhash.c
@@ -336,8 +336,8 @@ int main(int argc, char **argv){
   int count;
   int nChar;
   int totalLen = 0;
-  int aHash[1000];  /* 1000 is much bigger than nKeyword */
-  char zText[2000];
+  int aKWHash[1000];  /* 1000 is much bigger than nKeyword */
+  char zKWText[2000];
 
   /* Remove entries from the list of keywords that have mask==0 */
   for(i=j=0; i<nKeyword; i++){
@@ -441,13 +441,13 @@ int main(int argc, char **argv){
   bestSize = nKeyword;
   bestCount = nKeyword*nKeyword;
   for(i=nKeyword/2; i<=2*nKeyword; i++){
-    for(j=0; j<i; j++) aHash[j] = 0;
+    for(j=0; j<i; j++) aKWHash[j] = 0;
     for(j=0; j<nKeyword; j++){
       h = aKeywordTable[j].hash % i;
-      aHash[h] *= 2;
-      aHash[h]++;
+      aKWHash[h] *= 2;
+      aKWHash[h]++;
     }
-    for(j=count=0; j<i; j++) count += aHash[j];
+    for(j=count=0; j<i; j++) count += aKWHash[j];
     if( count<bestCount ){
       bestCount = count;
       bestSize = i;
@@ -455,30 +455,29 @@ int main(int argc, char **argv){
   }
 
   /* Compute the hash */
-  for(i=0; i<bestSize; i++) aHash[i] = 0;
+  for(i=0; i<bestSize; i++) aKWHash[i] = 0;
   for(i=0; i<nKeyword; i++){
     h = aKeywordTable[i].hash % bestSize;
-    aKeywordTable[i].iNext = aHash[h];
-    aHash[h] = i+1;
+    aKeywordTable[i].iNext = aKWHash[h];
+    aKWHash[h] = i+1;
   }
 
   /* Begin generating code */
   printf("%s", zHdr);
   printf("/* Hash score: %d */\n", bestCount);
-  printf("static int keywordCode(const char *z, int n, int *pType){\n");
-  printf("  /* zText[] encodes %d bytes of keywords in %d bytes */\n",
+  printf("/* zKWText[] encodes %d bytes of keyword text in %d bytes */\n",
           totalLen + nKeyword, nChar+1 );
   for(i=j=k=0; i<nKeyword; i++){
     Keyword *p = &aKeywordTable[i];
     if( p->substrId ) continue;
-    memcpy(&zText[k], p->zName, p->len);
+    memcpy(&zKWText[k], p->zName, p->len);
     k += p->len;
     if( j+p->len>70 ){
       printf("%*s */\n", 74-j, "");
       j = 0;
     }
     if( j==0 ){
-      printf("  /*   ");
+      printf("/*   ");
       j = 8;
     }
     printf("%s", p->zName);
@@ -487,16 +486,16 @@ int main(int argc, char **argv){
   if( j>0 ){
     printf("%*s */\n", 74-j, "");
   }
-  printf("  static const char zText[%d] = {\n", nChar);
-  zText[nChar] = 0;
+  printf("static const char zKWText[%d] = {\n", nChar);
+  zKWText[nChar] = 0;
   for(i=j=0; i<k; i++){
     if( j==0 ){
-      printf("    ");
+      printf("  ");
     }
-    if( zText[i]==0 ){
+    if( zKWText[i]==0 ){
       printf("0");
     }else{
-      printf("'%c',", zText[i]);
+      printf("'%c',", zKWText[i]);
     }
     j += 4;
     if( j>68 ){
@@ -505,23 +504,27 @@ int main(int argc, char **argv){
     }
   }
   if( j>0 ) printf("\n");
-  printf("  };\n");
+  printf("};\n");
 
-  printf("  static const unsigned char aHash[%d] = {\n", bestSize);
+  printf("/* aKWHash[i] is the hash value for the i-th keyword */\n");
+  printf("static const unsigned char aKWHash[%d] = {\n", bestSize);
   for(i=j=0; i<bestSize; i++){
-    if( j==0 ) printf("    ");
-    printf(" %3d,", aHash[i]);
+    if( j==0 ) printf("  ");
+    printf(" %3d,", aKWHash[i]);
     j++;
     if( j>12 ){
       printf("\n");
       j = 0;
     }
   }
-  printf("%s  };\n", j==0 ? "" : "\n");    
+  printf("%s};\n", j==0 ? "" : "\n");    
 
-  printf("  static const unsigned char aNext[%d] = {\n", nKeyword);
+  printf("/* aKWNext[] forms the hash collision chain.  If aKWHash[i]==0\n");
+  printf("** then the i-th keyword has no more hash collisions.  Otherwise,\n");
+  printf("** the next keyword with the same hash is aKWHash[i]-1. */\n");
+  printf("static const unsigned char aKWNext[%d] = {\n", nKeyword);
   for(i=j=0; i<nKeyword; i++){
-    if( j==0 ) printf("    ");
+    if( j==0 ) printf("  ");
     printf(" %3d,", aKeywordTable[i].iNext);
     j++;
     if( j>12 ){
@@ -529,11 +532,12 @@ int main(int argc, char **argv){
       j = 0;
     }
   }
-  printf("%s  };\n", j==0 ? "" : "\n");    
+  printf("%s};\n", j==0 ? "" : "\n");    
 
-  printf("  static const unsigned char aLen[%d] = {\n", nKeyword);
+  printf("/* aKWLen[i] is the length (in bytes) of the i-th keyword */\n");
+  printf("static const unsigned char aKWLen[%d] = {\n", nKeyword);
   for(i=j=0; i<nKeyword; i++){
-    if( j==0 ) printf("    ");
+    if( j==0 ) printf("  ");
     printf(" %3d,", aKeywordTable[i].len+aKeywordTable[i].prefix);
     j++;
     if( j>12 ){
@@ -541,11 +545,13 @@ int main(int argc, char **argv){
       j = 0;
     }
   }
-  printf("%s  };\n", j==0 ? "" : "\n");    
+  printf("%s};\n", j==0 ? "" : "\n");    
 
-  printf("  static const unsigned short int aOffset[%d] = {\n", nKeyword);
+  printf("/* aKWOffset[i] is the index into zKWText[] of the start of\n");
+  printf("** the text for the i-th keyword. */\n");
+  printf("static const unsigned short int aKWOffset[%d] = {\n", nKeyword);
   for(i=j=0; i<nKeyword; i++){
-    if( j==0 ) printf("    ");
+    if( j==0 ) printf("  ");
     printf(" %3d,", aKeywordTable[i].offset);
     j++;
     if( j>12 ){
@@ -553,12 +559,13 @@ int main(int argc, char **argv){
       j = 0;
     }
   }
-  printf("%s  };\n", j==0 ? "" : "\n");
+  printf("%s};\n", j==0 ? "" : "\n");
 
-  printf("  static const unsigned char aCode[%d] = {\n", nKeyword);
+  printf("/* aKWCode[i] is the parser symbol code for the i-th keyword */\n");
+  printf("static const unsigned char aKWCode[%d] = {\n", nKeyword);
   for(i=j=0; i<nKeyword; i++){
     char *zToken = aKeywordTable[i].zTokenType;
-    if( j==0 ) printf("    ");
+    if( j==0 ) printf("  ");
     printf("%s,%*s", zToken, (int)(14-strlen(zToken)), "");
     j++;
     if( j>=5 ){
@@ -566,17 +573,20 @@ int main(int argc, char **argv){
       j = 0;
     }
   }
-  printf("%s  };\n", j==0 ? "" : "\n");
-
+  printf("%s};\n", j==0 ? "" : "\n");
+  printf("/* Check to see if z[0..n-1] is a keyword. If it is, write the\n");
+  printf("** parser symbol code for that keyword into *pType.  Always\n");
+  printf("** return the integer n (the length of the token). */\n");
+  printf("static int keywordCode(const char *z, int n, int *pType){\n");
   printf("  int i, j;\n");
   printf("  const char *zKW;\n");
   printf("  if( n>=2 ){\n");
   printf("    i = ((charMap(z[0])*4) ^ (charMap(z[n-1])*3) ^ n) %% %d;\n",
           bestSize);
-  printf("    for(i=((int)aHash[i])-1; i>=0; i=((int)aNext[i])-1){\n");
-  printf("      if( aLen[i]!=n ) continue;\n");
+  printf("    for(i=((int)aKWHash[i])-1; i>=0; i=((int)aKWNext[i])-1){\n");
+  printf("      if( aKWLen[i]!=n ) continue;\n");
   printf("      j = 0;\n");
-  printf("      zKW = &zText[aOffset[i]];\n");
+  printf("      zKW = &zKWText[aKWOffset[i]];\n");
   printf("#ifdef SQLITE_ASCII\n");
   printf("      while( j<n && (z[j]&~0x20)==zKW[j] ){ j++; }\n");
   printf("#endif\n");
@@ -588,7 +598,7 @@ int main(int argc, char **argv){
     printf("      testcase( i==%d ); /* %s */\n",
            i, aKeywordTable[i].zOrigName);
   }
-  printf("      *pType = aCode[i];\n");
+  printf("      *pType = aKWCode[i];\n");
   printf("      break;\n");
   printf("    }\n");
   printf("  }\n");