| 1 |
#!/usr/bin/perl -w -s |
| 2 |
use XML::DT; |
| 3 |
use Data::Dumper; |
| 4 |
use File::Temp; |
| 5 |
use Term::ReadLine; |
| 6 |
#use locale; |
| 7 |
|
| 8 |
use strict; |
| 9 |
|
| 10 |
our ($latin1,$html,$show_att,$expand_att_id,$class); |
| 11 |
our ($lines,$t,$shell); |
| 12 |
|
| 13 |
my (@files)=@ARGV; |
| 14 |
@ARGV=(); |
| 15 |
|
| 16 |
|
| 17 |
|
| 18 |
mkxmltypes (@files); |
| 19 |
|
| 20 |
|
| 21 |
sub mkxmltypes { |
| 22 |
my %type=(); |
| 23 |
my @files = @_; |
| 24 |
my %root = (); |
| 25 |
my %att=(); |
| 26 |
my %dom=(); |
| 27 |
|
| 28 |
my %ele=(); |
| 29 |
my %elel=(); |
| 30 |
my %atl=(); |
| 31 |
my %handler=( |
| 32 |
# '-outputenc' => 'ISO-8859-1', |
| 33 |
'-default' => sub{ |
| 34 |
$c =~ s/,$//; |
| 35 |
if(not $class){ |
| 36 |
push(@{$type{$q}}, (eval("[$c]") || "?$c")); |
| 37 |
$elel{$q}++; |
| 38 |
if(ctxt(1)){ $ele{ctxt(1)}{$q} ++;} |
| 39 |
else { $root{$q}++} |
| 40 |
for(keys(%v)){ |
| 41 |
$atl{$_}++; |
| 42 |
$att{$q}{$_}{tipo($v{$_})||"_str"} ++ ; |
| 43 |
$dom{$q}{$_}{$v{$_}} ++ } |
| 44 |
"'$q',"; |
| 45 |
} |
| 46 |
else{ |
| 47 |
my $qcl=$q; |
| 48 |
if($v{class}){$qcl .="+$v{class}"} |
| 49 |
if($v{id} ){$qcl .="+$v{id}"} |
| 50 |
|
| 51 |
push(@{$type{$qcl}}, (eval("[$c]") || "?$c")); |
| 52 |
$elel{$qcl}++; |
| 53 |
if(ctxt(1)){ |
| 54 |
my $fcl=ctxt(1); |
| 55 |
if(father->{class}){$fcl .="+".father->{class}} |
| 56 |
if(father->{id} ){$fcl .="+".father->{id}} |
| 57 |
$ele{$fcl}{$qcl} ++;} |
| 58 |
else { $root{$qcl}++} |
| 59 |
|
| 60 |
for(keys(%v)){ |
| 61 |
next if ($_ eq "class"); |
| 62 |
next if ($_ eq "id"); |
| 63 |
$atl{$_}++; |
| 64 |
$att{$qcl}{$_}{tipo($v{$_})||"_str"} ++ ; |
| 65 |
$dom{$qcl}{$_}{$v{$_}} ++ } |
| 66 |
"'$qcl',"; |
| 67 |
} |
| 68 |
}, |
| 69 |
'-pcdata' => sub{ if ($c =~ /[^ \t\n]/){ $ele{ctxt(1)}{"#PCDATA"}=1; |
| 70 |
"'#PCDATA'," } else {""}}, |
| 71 |
); |
| 72 |
|
| 73 |
if ($html) { $handler{'-html'} = 1;} |
| 74 |
if($latin1) { $handler{'-inputenc'}='ISO-8859-1';} |
| 75 |
|
| 76 |
for my $fname (@files){ |
| 77 |
if($lines){ |
| 78 |
my $tmpfile = File::Temp->new( UNLINK => 0 )->filename; |
| 79 |
system("head -$lines $fname | xmllint --recover - > $tmpfile"); |
| 80 |
$fname = $tmpfile; |
| 81 |
} |
| 82 |
dt($fname,%handler); |
| 83 |
unlink($fname) if $lines; |
| 84 |
} |
| 85 |
|
| 86 |
### print "DEBUG",Dumper(\%att,\%ele,\%dom,\%atl); |
| 87 |
print "# ", join(" ",keys %root)," ...", scalar(localtime(time)) ,"\n"; |
| 88 |
|
| 89 |
my %resumofinal=(); |
| 90 |
for (keys %type){ |
| 91 |
my @tipo=(); |
| 92 |
for my $lista (@{$type{$_}}){ push (@tipo, processa($lista)) } |
| 93 |
$resumofinal{$_}=resumele(processa2([@tipo])).resumeatts($att{$_}); |
| 94 |
} |
| 95 |
|
| 96 |
if($shell){ shell($t,\%root,\%ele,\%att,\%dom,\%resumofinal,\%atl); } |
| 97 |
else{ pprint(\%resumofinal,ordem(\%ele,(($t) ||(keys %root) ))); } |
| 98 |
} |
| 99 |
|
| 100 |
sub shell{ |
| 101 |
my ($t,$root,$ele,$att,$dom,$resumofinal,$atl) = @_; |
| 102 |
my $last=(keys %$root)[0]; |
| 103 |
my $elepat = q{[\w:]+}; |
| 104 |
my $max = 10; |
| 105 |
my $term = new Term::ReadLine 'sample'; |
| 106 |
my $tas = $term->Attribs; |
| 107 |
$tas->{completion_entry_function}= $tas->{list_completion_function}; |
| 108 |
$tas->{completion_word} = [ keys(%$ele), keys(%$atl) ]; |
| 109 |
|
| 110 |
pprint($resumofinal,ordem($ele,(($t) ||(keys %$root) ))); |
| 111 |
while ( defined ($_ = $term->readline("\npfs> ")) ) { |
| 112 |
chomp(); |
| 113 |
$term->addhistory($_) if /\S/; |
| 114 |
s/^\s*(.*?)\s*$/$1/; |
| 115 |
if(/($elepat)\[\@?($elepat)\]/){ |
| 116 |
print resumeatt($att->{$1}{$2},$dom->{$1}{$2},$max); |
| 117 |
$last = $1} |
| 118 |
elsif(/\!max\s*=?\s*(\d+)/){$max=$1;} |
| 119 |
elsif(/\.($elepat)/){ |
| 120 |
print resumeatt($att->{$last}{$1},$dom->{$last}{$1},$max);} |
| 121 |
elsif(!$_ or defined $ele->{$_}) { |
| 122 |
$last=$_; |
| 123 |
pprint($resumofinal,ordem($ele,(($_) ||(keys %$root) ))); } |
| 124 |
else{ |
| 125 |
for my $e (keys %$att){ |
| 126 |
for my $a (keys %{$att->{$e}}){ |
| 127 |
print "$e($a):", |
| 128 |
resumeatt($att->{$e}{$a},$dom->{$e}{$a},$max) if($a eq $_) |
| 129 |
} |
| 130 |
} |
| 131 |
} |
| 132 |
} |
| 133 |
} |
| 134 |
|
| 135 |
sub ordem{ |
| 136 |
my ($rel,@st)=@_; |
| 137 |
my @r=(); |
| 138 |
my %visited = ('#PCDATA' => 1); |
| 139 |
|
| 140 |
while(@st){ |
| 141 |
my $next = shift(@st); |
| 142 |
next if $visited{$next}; |
| 143 |
push(@r,$next); |
| 144 |
$visited{$next} = 1; |
| 145 |
push(@st, (grep {! $visited{$_}} (keys %{$rel->{$next}}))); |
| 146 |
} |
| 147 |
\@r; |
| 148 |
} |
| 149 |
|
| 150 |
sub pprint{ |
| 151 |
my $r = shift; |
| 152 |
my $order = shift; |
| 153 |
for (@$order){ print "$_ \t=> $r->{$_}\n";} |
| 154 |
} |
| 155 |
|
| 156 |
sub resumeatts{ |
| 157 |
my $a=shift; |
| 158 |
my $r=""; |
| 159 |
for (keys(%{$a})) { |
| 160 |
if($expand_att_id){ $r .= "\n\t\t * $_:(".join(",",keys %{$a->{$_}}) . ")" } |
| 161 |
else { $r .= " * $_" } |
| 162 |
} |
| 163 |
$r |
| 164 |
} |
| 165 |
|
| 166 |
sub resumeatt{ |
| 167 |
my $a=shift; |
| 168 |
my $d=shift; |
| 169 |
my $max = shift(@_) || 10; |
| 170 |
my $r= join("|",keys %{$a}) ; |
| 171 |
my @domact = (grep {defined $_} ((keys %{$d}))[0..$max]); |
| 172 |
$domact[$max] = '...' if $domact[$max]; |
| 173 |
$r . " = {". join(",",@domact) . "}\n"; |
| 174 |
} |
| 175 |
|
| 176 |
sub processa{ |
| 177 |
my $a=shift; |
| 178 |
if( @$a == 0 ) { +{ _isa =>"empty"} } |
| 179 |
elsif( @$a == 1 && $a->[0] eq '#PCDATA') { |
| 180 |
+{ _isa =>"text" ,$a->[0] =>[1,1]} } |
| 181 |
elsif( @$a == 1 ) { +{ _isa =>"singleton",$a->[0] =>[1,1]} } |
| 182 |
else{ my %f = (); |
| 183 |
for (@$a){$f{$_}[0]++,$f{$_}[1]++} |
| 184 |
my $dif = scalar keys %f; |
| 185 |
if($dif == 1) { +{ _isa =>"seq", %f} ; } |
| 186 |
elsif($dif == @$a) { +{ _isa =>"tup", %f}; } |
| 187 |
elsif($f{'#PCDATA'}){ +{ _isa =>"mixed", %f }; } |
| 188 |
else { +{ _isa =>"mtup", %f } } |
| 189 |
} |
| 190 |
} |
| 191 |
|
| 192 |
sub processa2{ |
| 193 |
my $a=shift; |
| 194 |
if ( @$a == 0 ) { die("no sons????") } |
| 195 |
elsif( @$a == 1 ) { $a->[0] } |
| 196 |
else{ |
| 197 |
my %f = (); |
| 198 |
my %maybe = (); |
| 199 |
for (@$a){$f{sons2str($_)}++; |
| 200 |
$maybe{$_->{_isa}}++ } |
| 201 |
my $dif = scalar keys %f; |
| 202 |
if ($dif == 1) { $a->[0]; } |
| 203 |
elsif($maybe{mixed} || $maybe{text}){ +{%{join_sons($a)}, _isa=> "mixed"} } |
| 204 |
else { my %s= %{join_sons($a)}; |
| 205 |
if(keys %s == 1) { +{%s, _isa => "seq"}} |
| 206 |
else { +{%s, _isa => "mtup"} } |
| 207 |
} |
| 208 |
} |
| 209 |
} |
| 210 |
|
| 211 |
sub resumele{ |
| 212 |
my $a=shift; |
| 213 |
## print Dumper($a); |
| 214 |
my $i = $a->{_isa}; |
| 215 |
delete $a->{_isa}; |
| 216 |
if ($i eq "text") {"text"} |
| 217 |
elsif ($i eq "empty") {"empty"} |
| 218 |
elsif ($i eq "singleton") {join(", ", keys %{$a}) } |
| 219 |
elsif ($i eq "mixed") {delete $a->{'#PCDATA'}; |
| 220 |
if(keys %{$a}){ "mixed(".join(", ", keys %{$a}).")"} |
| 221 |
else {"text"} |
| 222 |
} |
| 223 |
elsif ($i eq "tup") {"tup(".join(", ", keys %{$a}).")"} |
| 224 |
elsif ($i eq "seq") {"seq(".join(", ", keys %{$a}).")"} |
| 225 |
else { my $r= "mtup("; |
| 226 |
for(sort keys %$a){ |
| 227 |
$r .= "$_, " if ( $a->{$_}[0] == 1 && $a->{$_}[1] == 1 ); |
| 228 |
$r .= "$_?, " if ( $a->{$_}[0] == 0 && $a->{$_}[1] == 1 ); |
| 229 |
$r .= "$_*, " if ( $a->{$_}[0] == 0 && $a->{$_}[1] > 1 ); |
| 230 |
$r .= "$_+, " if ( $a->{$_}[0] > 0 && $a->{$_}[1] > 1 ); |
| 231 |
$r .= Dumper($a) if($r =~ /\($/ ); |
| 232 |
} |
| 233 |
$r =~ s/, $//; |
| 234 |
$r.=")"; |
| 235 |
} |
| 236 |
} |
| 237 |
|
| 238 |
sub join_sons{ |
| 239 |
my $a = shift; |
| 240 |
my %final = ( map { ($_ => [$a->[0]{$_}[0], $a->[0]{$_}[1]])} |
| 241 |
grep {$_ ne "_isa"} keys %{$a->[0]}); |
| 242 |
my %todas=(); |
| 243 |
for (@$a){ |
| 244 |
my @novas=keys %{$_}; |
| 245 |
@todas{@novas}= @novas; |
| 246 |
for my $k (keys %todas){ |
| 247 |
next if $k =~ /_isa/; |
| 248 |
$final{$k}[0]=0 unless $final{$k}[0]; |
| 249 |
unless (exists $_->{$k}){ $final{$k}[0]=0; |
| 250 |
next} |
| 251 |
$final{$k}[1]=$_->{$k}[1] if $_->{$k}[1] > ($final{$k}[1] || 0); |
| 252 |
$final{$k}[0]=$_->{$k}[0] if $_->{$k}[0] < $final{$k}[0]; |
| 253 |
} |
| 254 |
} |
| 255 |
\%final |
| 256 |
} |
| 257 |
|
| 258 |
sub sons2str{ |
| 259 |
my $a = shift; |
| 260 |
join(' ',($a->{_isa},map { $_ . ($a->{$_}[0]==1 ? "" : "+") } |
| 261 |
grep {$_ ne "_isa"} sort keys %$a)); |
| 262 |
} |
| 263 |
|
| 264 |
sub tipo{ |
| 265 |
my $a=shift; |
| 266 |
for ($a){ |
| 267 |
if(/^\s*\d+\s*$/) {return "_int" } |
| 268 |
elsif(m{^\s*(https?|ftp|file)://\w[~&=?\w:/.-]+\s*$}i){return "_url" } |
| 269 |
elsif(/^\s*\d+\.\d+\s*$/) {return "_real" } |
| 270 |
elsif(/^\w+$/) {return "_id" } |
| 271 |
elsif(m{^\s*[\w.-]+\@\w[\w_:/.-]+\s*$}) {return "_email" } |
| 272 |
else {return undef } |
| 273 |
} |
| 274 |
} |
| 275 |
|
| 276 |
__END__ |
| 277 |
|
| 278 |
=head1 NAME |
| 279 |
|
| 280 |
mkxmltype - Make XML analysis using XML::DT |
| 281 |
|
| 282 |
=head1 SYNOPSIS |
| 283 |
|
| 284 |
mkxmltype <xmlfile> |
| 285 |
|
| 286 |
=head1 DESCRIPTION |
| 287 |
|
| 288 |
This command tries to infer DTD and Camlila-like types for a specific XML file; |
| 289 |
|
| 290 |
=head1 Options |
| 291 |
|
| 292 |
-latin1 input file encoding is forced to be latin1 |
| 293 |
-html uses html (libxml2) parser |
| 294 |
-show_att Show attribute values |
| 295 |
-expand_att_id |
| 296 |
-lines=20000 just reads the first 20000 lines of the XML file |
| 297 |
-t |
| 298 |
-class '<div class="a" id="b">'... is treated as 'div+a+b' |
| 299 |
-shell Enter interactive shell mode |
| 300 |
|
| 301 |
=head1 SEE ALSO |
| 302 |
|
| 303 |
XML::DT(1), mkdtskel(1), mkdtdskel and perl(1) |
| 304 |
|
| 305 |
=cut |