summaryrefslogtreecommitdiff
path: root/t/op/numconvert.t
blob: fedef70d40d84014b42e858dad772fa8cd6e1615 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
#!./perl

#
# test the conversion operators
#
# Notations:
#
# "N p i N vs N N":  Apply op-N, then op-p, then op-i, then reporter-N
# Compare with application of op-N, then reporter-N
# Right below are descriptions of different ops and reporters.

# We do not use these subroutines any more, sub overhead makes a "switch"
# solution better:

# obviously, 0, 1 and 2, 3 are destructive.  (XXXX 64-bit? 4 destructive too)

# *0 = sub {--$_[0]};		# -
# *1 = sub {++$_[0]};		# +

# # Converters
# *2 = sub { $_[0] = $max_uv & $_[0]}; # U
# *3 = sub { use integer; $_[0] += $zero}; # I
# *4 = sub { $_[0] += $zero};	# N
# *5 = sub { $_[0] = "$_[0]" };	# P

# # Side effects
# *6 = sub { $max_uv & $_[0]};	# u
# *7 = sub { use integer; $_[0] + $zero};	# i
# *8 = sub { $_[0] + $zero};	# n
# *9 = sub { $_[0] . "" };	# p

# # Reporters
# sub a2 { sprintf "%u", $_[0] }	# U
# sub a3 { sprintf "%d", $_[0] }	# I
# sub a4 { sprintf "%g", $_[0] }	# N
# sub a5 { "$_[0]" }		# P

BEGIN {
    chdir 't' if -d 't';
    @INC = '../lib';
}

use strict 'vars';

my $max_chain = $ENV{PERL_TEST_NUMCONVERTS} || 2;

# Bulk out if unsigned type is hopelessly wrong:
my $max_uv1 = ~0;
my $max_uv2 = sprintf "%u", $max_uv1 ** 6; # 6 is an arbitrary number here
my $big_iv = do {use integer; $max_uv1 * 16}; # 16 is an arbitrary number here
my $max_uv_less3 = $max_uv1 - 3;

print "# max_uv1 = $max_uv1, max_uv2 = $max_uv2, big_iv = $big_iv\n";
print "# max_uv_less3 = $max_uv_less3\n";
if ($max_uv1 ne $max_uv2 or $big_iv > $max_uv1 or $max_uv1 == $max_uv_less3) {
  print "1..0 # skipped: unsigned perl arithmetic is not sane";
  eval { require Config; import Config };
  use vars qw(%Config);
  if ($Config{d_quad} eq 'define') {
      print " (common in 64-bit platforms)";
  }
  print "\n";
  exit 0;
}
if ($max_uv_less3 =~ tr/0-9//c) {
  print "1..0 # skipped: this perl stringifies large unsigned integers using E notation\n";
  exit 0;
}

my $st_t = 4*4;			# We try 4 initializers and 4 reporters

my $num = 0;
$num += 10**$_ - 4**$_ for 1.. $max_chain;
$num *= $st_t;
print "1..$num\n";		# In fact 15 times more subsubtests...

my $max_uv = ~0;
my $max_iv = int($max_uv/2);
my $zero = 0;

my $l_uv = length $max_uv;
my $l_iv = length $max_iv;

# Hope: the first digits are good
my $larger_than_uv = substr 97 x 100, 0, $l_uv;
my $smaller_than_iv = substr 12 x 100, 0, $l_iv;
my $yet_smaller_than_iv = substr 97 x 100, 0, ($l_iv - 1);

my @list = (1, $yet_smaller_than_iv, $smaller_than_iv, $max_iv, $max_iv + 1,
	    $max_uv, $max_uv + 1);
unshift @list, (reverse map -$_, @list), 0; # 15 elts
@list = map "$_", @list; # Normalize

print "# @list\n";

# need to special case ++ for max_uv, as ++ "magic" on a string gives
# another string, whereas ++ magic on a string used as a number gives
# a number. Not a problem when NV preserves UV, but if it doesn't then
# stringification of the latter gives something in e notation.

my $max_uv_pp = "$max_uv"; $max_uv_pp++;
my $max_uv_p1 = "$max_uv"; $max_uv_p1+=0; $max_uv_p1++;

# Also need to cope with %g notation for max_uv_p1 that actually gives an
# integer less than max_uv because of correct rounding for the limited
# precisision. This bites for 12 byte long doubles and 8 byte UVs

my $temp = $max_uv_p1;
my $max_uv_p1_as_iv;
{use integer; $max_uv_p1_as_iv = 0 + sprintf "%s", $temp}
my $max_uv_p1_as_uv = 0 | sprintf "%s", $temp;

my @opnames = split //, "-+UINPuinp";

# @list = map { 2->($_), 3->($_), 4->($_), 5->($_),  } @list; # Prepare input

#print "@list\n";
#print "'@ops'\n";

my $test = 1;
my $nok;
for my $num_chain (1..$max_chain) {
  my @ops = map [split //], grep /[4-9]/,
    map { sprintf "%0${num_chain}d", $_ }  0 .. 10**$num_chain - 1;

  #@ops = ([]) unless $num_chain;
  #@ops = ([6, 4]);

  # print "'@ops'\n";
  for my $op (@ops) {
    for my $first (2..5) {
      for my $last (2..5) {
	$nok = 0;
	my @otherops = grep $_ <= 3, @$op;
	my @curops = ($op,\@otherops);

	for my $num (@list) {
	  my $inpt;
	  my @ans;

	  for my $short (0, 1) {
	    # undef $inpt;	# Forget all we had - some bugs were masked

	    $inpt = $num;	# Try to not contaminate $num...
	    $inpt = "$inpt";
	    if ($first == 2) {
	      $inpt = $max_uv & $inpt; # U 2
	    } elsif ($first == 3) {
	      use integer; $inpt += $zero; # I 3
	    } elsif ($first == 4) {
	      $inpt += $zero;	# N 4
	    } else {
	      $inpt = "$inpt";	# P 5
	    }

	    # Saves 20% of time - not with this logic:
	    #my $tmp = $inpt;
	    #my $tmp1 = $num;
	    #next if $num_chain > 1
	    #  and "$tmp" ne "$tmp1"; # Already the coercion gives problems...

	    for my $curop (@{$curops[$short]}) {
	      if ($curop < 5) {
		if ($curop < 3) {
		  if ($curop == 0) {
		    --$inpt;	# - 0
		  } elsif ($curop == 1) {
		    ++$inpt;	# + 1
		  } else {
		    $inpt = $max_uv & $inpt; # U 2
		  }
		} elsif ($curop == 3) {
		  use integer; $inpt += $zero;
		} else {
		  $inpt += $zero; # N 4
		}
	      } elsif ($curop < 8) {
		if ($curop == 5) {
		  $inpt = "$inpt"; # P 5
		} elsif ($curop == 6) {
		  $max_uv & $inpt; # u 6
		} else {
		  use integer; $inpt + $zero;
		}
	      } elsif ($curop == 8) {
		$inpt + $zero;	# n 8
	      } else {
		$inpt . "";	# p 9
	      }
	    }

	    if ($last == 2) {
	      $inpt = sprintf "%u", $inpt; # U 2
	    } elsif ($last == 3) {
	      $inpt = sprintf "%d", $inpt; # I 3
	    } elsif ($last == 4) {
	      $inpt = sprintf "%g", $inpt; # N 4
	    } else {
	      $inpt = "$inpt";	# P 5
	    }
	    push @ans, $inpt;
	  }
	  if ($ans[0] ne $ans[1]) {
	    print "# '$ans[0]' ne '$ans[1]',\t$num\t=> @opnames[$first,@{$curops[0]},$last] vs @opnames[$first,@{$curops[1]},$last]\n";
	    # XXX ought to check that "+" was in the list of opnames
	    if ((($ans[0] eq $max_uv_pp) and ($ans[1] eq $max_uv_p1))
		or (($ans[1] eq $max_uv_pp) and ($ans[0] eq $max_uv_p1))) {
	      # string ++ versus numeric ++. Tolerate this little
	      # bit of insanity
	      print "# ok, as string ++ of max_uv is \"$max_uv_pp\", numeric is $max_uv_p1\n"
	    } elsif ($opnames[$last] eq 'I' and $ans[1] eq "-1"
		     and $ans[0] eq $max_uv_p1_as_iv) {
              # Max UV plus 1 is NV. This NV may stringify in E notation.
              # And the number of decimal digits shown in E notation will depend
              # on the binary digits in the mantissa. And it may be that
              # (say)  18446744073709551616 in E notation is truncated to
              # (say) 1.8446744073709551e+19 (say) which gets converted back
              # as    1.8446744073709551000e+19
              # ie    18446744073709551000
              # which isn't the integer we first had.
              # But each step of conversion is correct. So it's not an error.
              # (Only shows up for 64 bit UVs and NVs with 64 bit mantissas,
              #  and on Crays (64 bit integers, 48 bit mantissas) IIRC)
	      print "# ok, \"$max_uv_p1\" correctly converts to IV \"$max_uv_p1_as_iv\"\n";
	    } elsif ($opnames[$last] eq 'U' and $ans[1] eq ~0
		     and $ans[0] eq $max_uv_p1_as_uv) {
              # as aboce
	      print "# ok, \"$max_uv_p1\" correctly converts to UV \"$max_uv_p1_as_uv\"\n";
	    } elsif (grep {/^N$/} @opnames[@{$curops[0]}]
		     and $ans[0] == $ans[1] and $ans[0] <= ~0
                     # First must be in E notation (ie not just digits) and
                     # second must still be an integer.
		     # eg 1.84467440737095516e+19
		     # 1.84467440737095516e+19 for 64 bit mantissa is in the
		     # integer range, so 1.84467440737095516e+19 + 0 is treated
		     # as integer addition. [should it be?]
		     # and 18446744073709551600 + 0 is 18446744073709551600
		     # Which isn't the string you first thought of.
                     # I can't remember why there isn't symmetry in this
                     # exception, ie why only the first ops are tested for 'N'
                     and $ans[0] != /^-?\d+$/ and $ans[1] !~ /^-?\d+$/) {
	      print "# ok, numerically equal - notation changed due to adding zero\n";
	    } else {
	      $nok++,
	    }
	  }
	}
        if ($nok) {
          print "not ok $test\n";
        } else {
          print "ok $test\n";
        }
	#print $txt if $nok;
	$test++;
      }
    }
  }
}