Paper on Lookahead TDFA: added benchmark results and graphs.

diff --git a/re2c/doc/tdfa/img/bench/data b/re2c/doc/tdfa/img/bench/data
new file mode 100644 (file)
index 0000000..d972222
--- /dev/null
+++ b/re2c/doc/tdfa/img/bench/data
@@ -0,0 +1,47 @@
+1  TDFA(0)  45    452  250  63   135  339   247  12.86  10.27  99.09   55.83
+2  TDFA(0)  18    70   32   15   31   41    31   7.66   5.47   71.60   33.90
+3  TDFA(0)  23    252  152  39   75   203   155  10.01  6.01   111.76  73.75
+4  TDFA(0)  16    26   17   11   19   23    19   8.34   3.55   102.72  59.84
+
+
+1  TDFA(1)  42    457  183  55   139  213   151  6.43   5.59   67.00   27.93
+2  TDFA(1)  16    73   33   15   35   41    31   5.30   3.83   63.30   26.74
+3  TDFA(1)  20    256  115  35   75   138   103  6.78   3.23   104.36  51.00
+4  TDFA(1)  13    28   19   11   19   25    23   6.04   3.12   100.28  47.85
+
+
+1  DFA      --    414  135  35   111  145   91   4.96   4.46   62.04   23.67
+2  DFA      --    69   25   15   31   31    23   4.90   3.34   62.00   23.59
+3  DFA      --    198  67   23   55   73    55   7.06   3.19   97.87   51.37
+4  DFA      --    22   10   11   15   14    15   5.89   2.66   97.95   47.01
+
+
+1  TDFA(0)  45    452  295  63   59   352   267  11.95  10.30  65.47   36.95
+2  TDFA(0)  18    70   31   15   19   31    31   7.12   7.30   31.81   17.44
+3  TDFA(0)  23    252  165  39   35   181   151  8.36   8.58   39.51   31.81
+4  TDFA(0)  16    26   20   11   11   22    23   7.14   6.67   23.19   18.73
+
+
+1  TDFA(1)  42    457  171  55   51   144   111  6.01   5.40   15.94   10.53
+2  TDFA(1)  16    73   29   15   19   29    27   5.24   4.43   13.50   8.84
+3  TDFA(1)  20    256  127  55   31   130   107  5.21   4.81   12.02   10.01
+4  TDFA(1)  13    28   17   11   11   19    19   4.02   3.08   8.56    6.90
+
+
+1  DFA      --    414  123  35   39   75    51   4.71   4.76   10.88   5.61
+2  DFA      --    69   19   11   15   15    15   4.64   3.94   11.00   5.77
+3  DFA      --    198  60   19   23   39    35   4.04   4.06   9.13    8.17
+4  DFA      --    22   7    11   11   8     11   3.90   2.52   8.00    4.40
+
+
+1  TDFA(0)  2054  625  816  275  267  1107  839  14.11  13.25  105.58  59.60
+2  TDFA(0)  72    106  57   23   55   73    55   8.61   6.77   72.96   34.63
+3  TDFA(0)  611   280  426  127  151  536   463  10.39  7.51   127.35  75.23
+4  TDFA(0)  79    29   33   19   23   43    39   7.43   4.05   105.06  61.74
+
+
+1  TDFA(1)  149   462  200  63   147  233   167  6.47   5.90   68.43   29.09
+2  TDFA(1)  44    82   39   19   43   49    39   6.00   5.39   63.79   27.37
+3  TDFA(1)  64    256  131  43   87   156   123  6.74   3.54   103.91  51.08
+4  TDFA(1)  40    31   28   15   23   36    31   6.27   3.32   101.79  48.15
+

diff --git a/re2c/doc/tdfa/img/bench/size_gcc_clang.png b/re2c/doc/tdfa/img/bench/size_gcc_clang.png
new file mode 100644 (file)
index 0000000..4a22142
Binary files /dev/null and b/re2c/doc/tdfa/img/bench/size_gcc_clang.png differ

diff --git a/re2c/doc/tdfa/img/bench/size_tcc_pcc.png b/re2c/doc/tdfa/img/bench/size_tcc_pcc.png
new file mode 100644 (file)
index 0000000..552b423
Binary files /dev/null and b/re2c/doc/tdfa/img/bench/size_tcc_pcc.png differ

diff --git a/re2c/doc/tdfa/img/bench/time_gcc_clang.png b/re2c/doc/tdfa/img/bench/time_gcc_clang.png
new file mode 100644 (file)
index 0000000..17f1c38
Binary files /dev/null and b/re2c/doc/tdfa/img/bench/time_gcc_clang.png differ

diff --git a/re2c/doc/tdfa/img/bench/time_tcc_pcc.png b/re2c/doc/tdfa/img/bench/time_tcc_pcc.png
new file mode 100644 (file)
index 0000000..24cfbac
Binary files /dev/null and b/re2c/doc/tdfa/img/bench/time_tcc_pcc.png differ

diff --git a/re2c/doc/tdfa/tdfa.tex b/re2c/doc/tdfa/tdfa.tex
index 2fb0c8c116cd9af67f0517905a566c6829e78381..f88f51cb87f2b9da9bfb89e063db6fa1a72a39c2 100644 (file)
--- a/re2c/doc/tdfa/tdfa.tex
+++ b/re2c/doc/tdfa/tdfa.tex
@@ -2155,7 +2155,7 @@ Tag $1$ is deterministic for both automata.
 \begin{multicols}{2}
 
 From these examples we can draw the following conclusions.
-First, TDFA(1) are generally better than TDFA(0): delaying register operations allows to get rid of many conflicts.
+First, TDFA(1) is generally better than TDFA(0): delaying register operations allows to get rid of many conflicts.
 Second, both kinds of automata are only suitable for RE with modest levels of ambiguity
 and low submatch detalisation: TDFA can be applied to full parsing, but other methods would probably outperform them.
 However, RE of such form are very common in practice and for them TDFA can be very efficient.
@@ -2271,7 +2271,7 @@ we can also use it for backup.
 
 \subsection*{Fixed tags}
 
-It may happen that two tags in TRE are bound: separated by a fixed number of characters, so that
+It may happen that two tags in TRE are separated by a fixed number of characters:
 each offset of one tag is equal to the corresponding offset of the other tag plus some static offset.
 %the value of one tag is always equal to the value of the other plus some static offset.
 In this case we can track only one of the tags; we say that the second tag is \emph{fixed} on the first one.
@@ -2489,28 +2489,20 @@ and visualized on subsequent plots.
     \hline \hline
     \multicolumn{12}{|c|}{re2c} \\
     \hline
-    TDFA(0) & 45 & 452 & 250 & 63 & 135 & 339 & 247 & 12.88 & 10.31 & 99.12 & 55.91 \\
-    TDFA(1) & 42 & 457 & 183 & 55 & 139 & 213 & 151 &  6.42 &  5.59 & 67.04 & 27.96 \\
-    DFA     & -- & 414 & 135 & 35 & 111 & 145 &  91 &  4.96 &  4.46 & 62.15 & 23.74 \\
-%    TDFA(0) & 45 & 452 & 255712 & 63544 & 137320 & 346408 & 252024 & 12.88 & 10.31 & 99.12 & 55.91 \\
-%    TDFA(1) & 42 & 457 & 186600 & 55352 & 141416 & 217160 & 153720 &  6.42 &  5.59 & 67.04 & 27.96 \\
-%    DFA     & -- & 414 & 137816 & 34864 & 112728 & 148048 &  92256 &  4.96 &  4.46 & 62.15 & 23.74 \\
+    TDFA(0) & 45 & 452 & 250 & 63 & 135 & 339 & 247 & 12.86 & 10.27 & 99.09 & 55.83 \\
+    TDFA(1) & 42 & 457 & 183 & 55 & 139 & 213 & 151 &  6.43 &  5.59 & 67.00 & 27.93 \\
+    DFA     & -- & 414 & 135 & 35 & 111 & 145 &  91 &  4.96 &  4.46 & 62.04 & 23.67 \\
     \hline \hline
     \multicolumn{12}{|c|}{re2c -b} \\
     \hline
-    TDFA(0) & 45 & 452 & 295 & 63 & 59 & 352 & 267 & 11.96 & 10.31 & 65.53 & 36.98 \\
-    TDFA(1) & 42 & 457 & 171 & 55 & 51 & 144 & 111 &  6.01 &  5.40 & 15.96 & 10.59 \\
-    DFA     & -- & 414 & 123 & 35 & 39 &  75 &  51 &  4.73 &  4.78 & 10.93 &  5.63 \\
-%    TDFA(0) & 45 & 452 & 301968 & 63544 & 59496 & 360136 & 272504 & 11.96 & 10.31 & 65.53 & 36.98 \\
-%    TDFA(1) & 42 & 457 & 174903 & 55352 & 51304 & 147016 & 112760 &  6.01 &  5.40 & 15.96 & 10.59 \\
-%    DFA     & -- & 414 & 125389 & 34864 & 39000 &  76272 &  51296 &  4.73 &  4.78 & 10.93 &  5.63 \\
+    TDFA(0) & 45 & 452 & 295 & 63 & 59 & 352 & 267 & 11.95 & 10.30 & 65.47 & 36.95 \\
+    TDFA(1) & 42 & 457 & 171 & 55 & 51 & 144 & 111 &  6.01 &  5.40 & 15.94 & 10.53 \\
+    DFA     & -- & 414 & 123 & 35 & 39 &  75 &  51 &  4.71 &  4.76 & 10.88 &  5.61 \\
     \hline \hline
     \multicolumn{12}{|c|}{re2c --no-optimize-tags} \\
     \hline
-    TDFA(0) & 2054 & 625 & 816 & 275 & 267 & 1107 & 839 & 14.14 & 13.24 & 105.87 & 59.71 \\
-    TDFA(1) &  149 & 462 & 200 &  63 & 147 &  233 & 167 &  6.64 &  5.90 &  68.50 & 29.39 \\
-%    TDFA(0) & 2054 & 625 & 835285 & 280632 & 272488 & 1132616 & 858232 & 14.14 & 13.24 & 105.87 & 59.71 \\
-%    TDFA(1) &  149 & 462 & 204119 &  63544 & 149608 &  238568 & 170104 &  6.64 &  5.90 &  68.50 & 29.39 \\
+    TDFA(0) & 2054 & 625 & 816 & 275 & 267 & 1107 & 839 & 14.11 & 13.25 & 105.58 & 59.60 \\
+    TDFA(1) &  149 & 462 & 200 &  63 & 147 &  233 & 167 &  6.47 &  5.90 &  68.43 & 29.09 \\
     \hline
     \end{tabular}\\*
     \medskip
@@ -2536,28 +2528,20 @@ and visualized on subsequent plots.
     \hline \hline
     \multicolumn{12}{|c|}{re2c} \\
     \hline
-    TDFA(0) & 18 & 70 & 32 & 15 & 31 & 41 & 31 & 7.65 & 5.50 & 71.60 & 33.96 \\
-    TDFA(1) & 16 & 73 & 33 & 15 & 35 & 41 & 31 & 5.31 & 3.83 & 63.36 & 26.78 \\
-    DFA     & -- & 69 & 25 & 15 & 31 & 31 & 23 & 4.90 & 3.34 & 62.12 & 23.64 \\
-%    TDFA(0) &  &  &           & 14392 & 30816 & 41160 & 30840 & 7.65 & 5.50 & 71.60 & 33.96 \\
-%    TDFA(1) &  &  &           & 14392 & 34912 & 41704 & 30840 & 5.31 & 3.83 & 63.36 & 26.78 \\
-%    DFA     & -- & 69 & 24937 & 14384 & 30808 & 31280 & 22624 & 4.90 & 3.34 & 62.12 & 23.64 \\
+    TDFA(0) & 18 & 70 & 32 & 15 & 31 & 41 & 31 & 7.66 & 5.47 & 71.60 & 33.90 \\
+    TDFA(1) & 16 & 73 & 33 & 15 & 35 & 41 & 31 & 5.30 & 3.83 & 63.30 & 26.74 \\
+    DFA     & -- & 69 & 25 & 15 & 31 & 31 & 23 & 4.90 & 3.34 & 62.00 & 23.59 \\
     \hline \hline
     \multicolumn{12}{|c|}{re2c -b} \\
     \hline
-    TDFA(0) & 18 & 70 & 31 & 15 & 19 & 31 & 31 & 7.12 & 7.31 & 31.85 & 17.47 \\
-    TDFA(1) & 16 & 73 & 29 & 15 & 19 & 29 & 27 & 5.25 & 4.42 & 13.52 &  8.86 \\
-    DFA     & -- & 69 & 19 & 11 & 15 & 15 & 15 & 4.66 & 3.96 & 11.00 &  5.79 \\
-%    TDFA(0) &  &  &           & 14392 & 18528 & 31336 & 30840 & 7.12 & 7.31 & 31.85 & 17.47 \\
-%    TDFA(1) &  &  &           & 14392 & 18528 & 29288 & 26744 & 5.25 & 4.42 & 13.52 &  8.86 \\
-%    DFA     & -- & 69 & 18472 & 10288 & 14424 & 14832 & 14432 & 4.66 & 3.96 & 11.00 &  5.79 \\
+    TDFA(0) & 18 & 70 & 31 & 15 & 19 & 31 & 31 & 7.12 & 7.30 & 31.81 & 17.44 \\
+    TDFA(1) & 16 & 73 & 29 & 15 & 19 & 29 & 27 & 5.24 & 4.43 & 13.50 &  8.84 \\
+    DFA     & -- & 69 & 19 & 11 & 15 & 15 & 15 & 4.64 & 3.94 & 11.00 &  5.77 \\
     \hline \hline
     \multicolumn{12}{|c|}{re2c --no-optimize-tags} \\
     \hline
-    TDFA(0) & 72 & 106 & 57 & 23 & 55 & 73 & 55 & 8.61 & 6.77 & 73.05 & 34.68 \\
-    TDFA(1) & 44 &  82 & 39 & 19 & 43 & 49 & 39 & 6.01 & 5.38 & 63.87 & 27.44 \\
-%    TDFA(0) & 72 & 106 & 57956 & 22584 & 55400 & 73928 & 55416 & 8.61 & 6.77 & 73.05 & 34.68 \\
-%    TDFA(1) & 44 &  82 & 39674 & 18488 & 43112 & 49480 & 39032 & 6.01 & 5.38 & 63.87 & 27.44 \\
+    TDFA(0) & 72 & 106 & 57 & 23 & 55 & 73 & 55 & 8.61 & 6.77 & 72.96 & 34.63 \\
+    TDFA(1) & 44 &  82 & 39 & 19 & 43 & 49 & 39 & 6.00 & 5.39 & 63.79 & 27.37 \\
     \hline
     \end{tabular}\\*
     \medskip
@@ -2584,28 +2568,20 @@ and visualized on subsequent plots.
     \hline \hline
     \multicolumn{12}{|c|}{re2c} \\
     \hline
-    TDFA(0) & 23 & 252 & 152 & 39 & 75 & 203 & 155 & 10.03 & 6.10 & 111.90 & 73.81 \\
-    TDFA(1) & 20 & 256 & 115 & 35 & 75 & 138 & 103 &  6.75 & 3.24 & 104.56 & 50.90 \\
-    DFA     & -- & 198 &  67 & 23 & 55 &  73 &  55 &  7.05 & 3.21 &  97.89 & 51.43 \\
-%    TDFA(0) & 23 & 252 & 154776 & 38960 & 75864 & 207600 & 157792 & 10.03 & 6.10 & 111.90 & 73.81 \\
-%    TDFA(1) & 20 & 256 & 117498 & 34864 & 75864 & 140560 & 104544 &  6.75 & 3.24 & 104.56 & 50.90 \\
-%    DFA     & -- & 198 &  67617 & 22576 & 55384 &  74384 &  55392 &  7.05 & 3.21 &  97.89 & 51.43 \\
+    TDFA(0) & 23 & 252 & 152 & 39 & 75 & 203 & 155 & 10.01 & 6.01 & 111.76 & 73.75 \\
+    TDFA(1) & 20 & 256 & 115 & 35 & 75 & 138 & 103 &  6.78 & 3.23 & 104.36 & 51.00 \\
+    DFA     & -- & 198 &  67 & 23 & 55 &  73 &  55 &  7.06 & 3.19 &  97.87 & 51.37 \\
     \hline \hline
     \multicolumn{12}{|c|}{re2c -b} \\
     \hline
-    TDFA(0) & 23 & 252 & 165 & 39 & 35 & 181 & 151 & 8.40 & 8.56 & 39.56 & 31.84 \\
-    TDFA(1) & 20 & 256 & 127 & 55 & 31 & 130 & 107 & 5.23 & 4.83 & 12.04 & 10.02 \\
-    DFA     & -- & 198 &  60 & 19 & 23 &  39 &  35 & 4.05 & 4.08 &  9.23 &  8.19 \\
-%    TDFA(0) & 23 & 252 & 168684 & 38960 & 34904 & 186704 & 153696 & 8.40 & 8.56 & 39.56 & 31.84 \\
-%    TDFA(1) & 20 & 256 & 129322 & 55344 & 30808 & 132912 & 108640 & 5.23 & 4.83 & 12.04 & 10.02 \\
-%    DFA     & -- & 198 &  60759 & 18480 & 22616 &  39376 &  34912 & 4.05 & 4.08 &  9.23 &  8.19 \\
+    TDFA(0) & 23 & 252 & 165 & 39 & 35 & 181 & 151 & 8.36 & 8.58 & 39.51 & 31.81 \\
+    TDFA(1) & 20 & 256 & 127 & 55 & 31 & 130 & 107 & 5.21 & 4.81 & 12.02 & 10.01 \\
+    DFA     & -- & 198 &  60 & 19 & 23 &  39 &  35 & 4.04 & 4.06 &  9.13 &  8.17 \\
     \hline \hline
     \multicolumn{12}{|c|}{re2c --no-optimize-tags} \\
     \hline
-    TDFA(0) & 611 & 280 & 426 & 127 & 151 & 536 & 463 & 10.41 & 7.56 & 127.48 & 75.46 \\
-    TDFA(1) &  64 & 256 & 131 &  43 &  87 & 156 & 123 &  6.74 & 3.55 & 103.98 & 51.12 \\
-%    TDFA(0) & 611 & 280 & 435350 & 129072 & 153696 & 548272 & 473184 & 10.41 & 7.56 & 127.48 & 75.46 \\
-%    TDFA(1) &  64 & 256 & 133518 &  43056 &  88160 & 159248 & 125024 &  6.74 & 3.55 & 103.98 & 51.12 \\
+    TDFA(0) & 611 & 280 & 426 & 127 & 151 & 536 & 463 & 10.39 & 7.51 & 127.35 & 75.23 \\
+    TDFA(1) &  64 & 256 & 131 &  43 &  87 & 156 & 123 &  6.74 & 3.54 & 103.91 & 51.08 \\
     \hline
     \end{tabular}\\*
     \medskip
@@ -2632,28 +2608,20 @@ and visualized on subsequent plots.
     \hline \hline
     \multicolumn{12}{|c|}{re2c} \\
     \hline
-    TDFA(0) & 16 & 26 & 17 & 11 & 19 & 23 & 19 & 8.34 & 3.57 & 102.84 & 59.88 \\
-    TDFA(1) & 13 & 28 & 19 & 11 & 19 & 25 & 23 & 6.06 & 3.14 & 100.33 & 48.02 \\
-    DFA     & -- & 22 & 10 & 11 & 15 & 14 & 15 & 5.91 & 2.68 &  98.10 & 47.25 \\
-%    TDFA(0) &    &  &  & 10288 & 18520 & 22960 & 18528 & 8.34 & 3.57 & 102.84 & 59.88 \\
-%    TDFA(1) &    &  &  & 10288 & 18520 & 25424 & 22624 & 6.06 & 3.14 & 100.33 & 48.02 \\
-%    DFA     & -- &  &  & 10288 & 14424 & 14256 & 14432 & 5.91 & 2.68 &  98.10 & 47.25 \\
+    TDFA(0) & 16 & 26 & 17 & 11 & 19 & 23 & 19 & 8.34 & 3.55 & 102.72 & 59.84 \\
+    TDFA(1) & 13 & 28 & 19 & 11 & 19 & 25 & 23 & 6.04 & 3.12 & 100.28 & 47.85 \\
+    DFA     & -- & 22 & 10 & 11 & 15 & 14 & 15 & 5.89 & 2.66 &  97.95 & 47.01 \\
     \hline \hline
     \multicolumn{12}{|c|}{re2c -b} \\
     \hline
-    TDFA(0) & 16 & 26 & 20 & 11 & 11 & 22 & 23 & 7.17 & 6.66 & 23.21 & 18.77 \\
-    TDFA(1) & 13 & 28 & 17 & 11 & 11 & 19 & 19 & 4.05 & 3.09 &  8.59 &  6.94 \\
-    DFA     & -- & 22 &  7 & 11 & 11 &  8 & 11 & 3.92 & 2.56 &  8.06 &  4.42 \\
-%    TDFA(0) &    &  &       & 10288 & 10328 & 22352 & 22624 & 7.17 & 6.66 & 23.21 & 18.77 \\
-%    TDFA(1) &    &  &       & 10288 & 10328 & 18960 & 18528 & 4.05 & 3.09 &  8.59 &  6.94 \\
-%    DFA     & -- &  &  6483 & 10288 & 10328 &  7888 & 10336 & 3.92 & 2.56 &  8.06 &  4.42 \\
+    TDFA(0) & 16 & 26 & 20 & 11 & 11 & 22 & 23 & 7.14 & 6.67 & 23.19 & 18.73 \\
+    TDFA(1) & 13 & 28 & 17 & 11 & 11 & 19 & 19 & 4.02 & 3.08 &  8.56 &  6.90 \\
+    DFA     & -- & 22 &  7 & 11 & 11 &  8 & 11 & 3.90 & 2.52 &  8.00 &  4.40 \\
     \hline \hline
     \multicolumn{12}{|c|}{re2c --no-optimize-tags} \\
     \hline
-    TDFA(0) & 79 & 29 & 33 & 19 & 23 & 43 & 39 & 7.46 & 3.94 & 105.22 & 61.72 \\
-    TDFA(1) & 40 & 31 & 28 & 15 & 23 & 36 & 31 & 6.29 & 3.33 & 102.00 & 48.22 \\
-%    TDFA(0) & 79 & 29 & 33745 & 18480 & 22624 & 43504 & 39008 & 7.46 & 3.94 & 105.22 & 61.72 \\
-%    TDFA(1) & 40 & 31 & 28013 & 14384 & 22624 & 36080 & 30816 & 6.29 & 3.33 & 102.00 & 48.22 \\
+    TDFA(0) & 79 & 29 & 33 & 19 & 23 & 43 & 39 & 7.43 & 4.05 & 105.06 & 61.74 \\
+    TDFA(1) & 40 & 31 & 28 & 15 & 23 & 36 & 31 & 6.27 & 3.32 & 101.79 & 48.15 \\
     \hline
     \end{tabular}\\*
     \medskip
@@ -2745,6 +2713,7 @@ Premnogoe spasibo drugu na bukvu S ! ! ! :)
 
     \item \! [Cox10] Russ Cox, \textit{"Regular Expression Matching in the Wild"}, March 2010, \\
         https://swtch.com/\textasciitilde rsc/regexp/regexp3.html
+    \item https://github.com/google/re2/issues/146
 
 \end{enumerate}