use Carp;
- # Keywords from SQL-92, SQL-99 and SQL-2003.
+ # Keywords from SQL-92, SQL-99, SQL-2003, SQL-2008 and SQL-2011 specifics keywords.
use constant KEYWORDS => qw(
ABSOLUTE ACTION ADD AFTER ALL ALLOCATE ALTER AND ANY ARE ARRAY AS ASC
ASENSITIVE ASSERTION ASYMMETRIC AT ATOMIC AUTHORIZATION AVG BEFORE BEGIN
CHAR_LENGTH CHECK CLOB CLOSE COALESCE COLLATE COLLATION COLUMN COMMIT
CONDITION CONNECT CONNECTION CONSTRAINT CONSTRAINTS CONSTRUCTOR CONTAINS
CONTINUE CONVERT CORRESPONDING COUNT CREATE CROSS CUBE CURRENT CURRENT_DATE
- CURRENT_DEFAULT_TRANSFORM_GROUP CURRENT_PATH CURRENT_ROLE CURRENT_TIME
- CURRENT_TIMESTAMP CURRENT_TRANSFORM_GROUP_FOR_TYPE CURRENT_USER CURSOR
- CYCLE DATA DATE DAY DEALLOCATE DEC DECIMAL DECLARE DEFAULT DEFERRABLE
- DEFERRED DELETE DEPTH DEREF DESC DESCRIBE DESCRIPTOR DETERMINISTIC
+ CURRENT_DEFAULT_TRANSFORM_GROUP CURRENT_CATALOG CURRENT_PATH CURRENT_ROLE
+ CURRENT_SCHEMA CURRENT_TIME CURRENT_TIMESTAMP CURRENT_TRANSFORM_GROUP_FOR_TYPE
+ CURRENT_USER CURSOR CYCLE DATA DATE DAY DEALLOCATE DEC DECIMAL DECLARE DEFAULT
+ DEFERRABLE DEFERRED DELETE DEPTH DEREF DESC DESCRIBE DESCRIPTOR DETERMINISTIC
DIAGNOSTICS DISCONNECT DISTINCT DO DOMAIN DOUBLE DROP DYNAMIC EACH ELEMENT
ELSE ELSEIF END EPOCH EQUALS ESCAPE EXCEPT EXCEPTION EXEC EXECUTE EXISTS
EXIT EXTERNAL EXTRACT FALSE FETCH FILTER FIRST FLOAT FOR FOREIGN FOUND FREE
LIMIT LOCAL LOCALTIME LOCALTIMESTAMP LOCATOR LOOP LOWER MAP MATCH MAX
MEMBER MERGE METHOD MIN MINUTE MODIFIES MODULE MONTH MULTISET NAMES
NATIONAL NATURAL NCHAR NCLOB NEW NEXT NO NONE NOT NULL NULLIF NUMERIC
- OBJECT OCTET_LENGTH OF OLD ON ONLY OPEN OPTION OR ORDER ORDINALITY OUT
+ OBJECT OCTET_LENGTH OF OFFSET OLD ON ONLY OPEN OPTION OR ORDER ORDINALITY OUT
OUTER OUTPUT OVER OVERLAPS PAD PARAMETER PARTIAL PARTITION PATH POSITION
PRECISION PREPARE PRESERVE PRIMARY PRIOR PRIVILEGES PROCEDURE PUBLIC RANGE
READ READS REAL RECURSIVE REF REFERENCES REFERENCING RELATIVE RELEASE
UNTIL UPDATE UPPER USAGE USER USING VALUE VALUES VARCHAR VARYING VIEW WHEN
WHENEVER WHERE WHILE WINDOW WITH WITHIN WITHOUT WORK WRITE YEAR ZONE
);
-
+ use constant FUNCTIONS => qw();
sub tokenize_sql
{
my ($query, $remove_white_tokens) = @_;
my $re = qr{
- (
- (?:--)[\ \t\S]* # single line comments
- |
- (?:\-\|\-) # range operator "is adjacent to"
- |
- (?:\->>|\->|\#>>|\#>|\?\&|\?) # Json Operators
- |
- (?:\#<=|\#>=|\#<>|\#<|\#=) # compares tinterval and reltime
- |
- (?:>>=|<<=) # inet operators
- |
- (?:!!|\@\@\@) # deprecated factorial and full text search operators
- |
- (?:\|\|\/|\|\/) # square root and cube root
- |
- (?:\@\-\@|\@\@|\#\#|<\->|<<\||\|>>|\&<\||\&<|\|\&>|\&>|<\^|>\^|\?\#|\#|\?<\||\?\-\||\?\-|\?\|\||\?\||\@>|<\@|\~=)
- # Geometric Operators
- |
- (?:~<=~|~>=~|~>~|~<~) # string comparison for pattern matching operator families
- |
- (?:!~~|!~~\*|~~\*|~~) # LIKE operators
- |
- (?:!~\*|!~|~\*) # regular expression operators
- |
- (?:\*=|\*<>|\*<=|\*>=|\*<|\*>) # composite type comparison operators
- |
- (?:<>|<=>|>=|<=|==|!=|=|!|<<|>>|<|>|\|\||\||&&|&|-|\+|\*(?!/)|/(?!\*)|\%|~|\^|\?) # operators and tests
- |
- [\[\]\(\),;.] # punctuation (parenthesis, comma)
- |
- E\'\'(?!\') # empty escaped single quoted string
- |
- \'\'(?!\') # empty single quoted string
- |
- \"\"(?!\"") # empty double quoted string
- |
- "(?>(?:(?>[^"\\]+)|""|\\.)*)+" # anything inside double quotes, ungreedy
- |
- `(?>(?:(?>[^`\\]+)|``|\\.)*)+` # anything inside backticks quotes, ungreedy
- |
- E'(?>(?:(?>[^'\\]+)|''|\\.)*)+' # anything escaped inside single quotes, ungreedy.
- |
- '(?>(?:(?>[^'\\]+)|''|\\.)*)+' # anything inside single quotes, ungreedy.
- |
- /\*[\ \t\r\n\S]*?\*/ # C style comments
- |
- (?:[\w:@]+(?:\.(?:\w+|\*)?)*) # words, standard named placeholders, db.table.*, db.*
- |
- (?:\$\w+\$)
- |
- (?: \$_\$ | \$\d+ | \${1,2} | \$\w+\$ ) # dollar expressions - eg $_$ $3 $$ $BODY$
- |
- \n # newline
- |
- [\t\ ]+ # any kind of white spaces
- )
+ (
+ (?:--)[\ \t\S]* # single line comments
+ |
+ (?:\-\|\-) # range operator "is adjacent to"
+ |
+ (?:\->>|\->|\#>>|\#>|\?\&|\?) # Json Operators
+ |
+ (?:\#<=|\#>=|\#<>|\#<|\#=) # compares tinterval and reltime
+ |
+ (?:>>=|<<=) # inet operators
+ |
+ (?:!!|\@\@\@) # deprecated factorial and full text search operators
+ |
+ (?:\|\|\/|\|\/) # square root and cube root
+ |
+ (?:\@\-\@|\@\@|\#\#|<\->|<<\||\|>>|\&<\||\&<|\|\&>|\&>|<\^|>\^|\?\#|\#|\?<\||\?\-\||\?\-|\?\|\||\?\||\@>|<\@|\~=)
+ # Geometric Operators
+ |
+ (?:~<=~|~>=~|~>~|~<~) # string comparison for pattern matching operator families
+ |
+ (?:!~~|!~~\*|~~\*|~~) # LIKE operators
+ |
+ (?:!~\*|!~|~\*) # regular expression operators
+ |
+ (?:\*=|\*<>|\*<=|\*>=|\*<|\*>) # composite type comparison operators
+ |
+ (?:<>|<=>|>=|<=|==|!=|=|!|<<|>>|<|>|\|\||\||&&|&|-|\+|\*(?!/)|/(?!\*)|\%|~|\^|\?)
+ # operators and tests
+ |
+ [\[\]\(\),;.] # punctuation (parenthesis, comma)
+ |
+ E\'\'(?!\') # escape empty single quoted string
+ |
+ \'\'(?!\') # empty single quoted string
+ |
+ \"\"(?!\"") # empty double quoted string
+ |
+ "(?>(?:(?>[^"\\]+)|""|\\.)*)+"
+ # anything inside double quotes, ungreedy
+ |
+ `(?>(?:(?>[^`\\]+)|``|\\.)*)+`
+ # anything inside backticks quotes, ungreedy
+ |
+ E'(?>(?:(?>[^'\\]+)|''|\\.)*)+'
+ # anything escaped inside single quotes, ungreedy.
+ |
+ '(?>(?:(?>[^'\\]+)|''|\\.)*)+'
+ # anything inside single quotes, ungreedy.
+ |
+ /\*[\ \t\r\n\S]*?\*/ # C style comments
+ |
+ (?:[\w:@]+(?:\.(?:\w+|\*)?)*)
+ # words, standard named placeholders, db.table.*, db.*
+ |
+ (?:\$\w+\$)
+ |
+ (?: \$_\$ | \$\d+ | \${1,2} | \$\w+\$ )
+ # dollar expressions - eg $_$ $3 $$ $BODY$
+ |
+ \n # newline
+ |
+ [\t\ ]+ # any kind of white spaces
+ )
}smx;
my @query = ();
if ($remove_white_tokens) {
@query = grep(!/^[\s\n\r]*$/, @query);
}
-
return wantarray ? @query : \@query;
}
$self->{spaces} = 4 unless defined($self->{spaces});
$self->{space} = ' ' unless defined($self->{space});
$self->{break} = "\n" unless defined($self->{break});
+ $self->{break} = ' ' unless ($self->{spaces} != 0);
$self->{wrap} = {} unless defined($self->{wrap});
$self->{keywords} = [] unless defined($self->{keywords});
+ $self->{functions} = [] unless defined($self->{functions});
$self->{rules} = {} unless defined($self->{rules});
- $self->{uc_keywords} = 0 unless defined $self->{uc_keywords};
+ $self->{uc_keywords} = 0 unless defined($self->{uc_keywords});
+ $self->{uc_functions}= 0 unless defined($self->{uc_functions});
+ $self->{no_comments} = 0 unless defined($self->{no_comments});
- push(@{$self->{keywords}}, KEYWORDS);
+ push @{$self->{keywords}}, KEYWORDS;
+ push @{$self->{functions}}, FUNCTIONS;
# Initialize internal stuff.
$self->{_level} = 0;
+ @{$self->{have_from_clause}} = qw( extract overlay substring trim );
return $self;
}
$self->{_level_stack} = [];
$self->{_new_line} = 1;
- my $last = '';
+ my $last;
$self->{_tokens} = [tokenize_sql($self->query, 1)];
while (defined(my $token = $self->_token)) {
elsif ($token eq '(') {
$self->_add_token($token);
- $self->_new_line;
- push @{$self->{_level_stack}}, $self->{_level};
- $self->_over unless $last and uc($last) eq 'WHERE';
+ if ( ($self->_next_token ne ')') && ($self->_next_token ne '*') ) {
+ $self->{ '_has_from' } = 1 if ($last && grep(/^\Q$last\E$/i, @{$self->{have_from_clause}}));
+ push @{$self->{_level_stack}}, $self->{_level};
+ $self->_over unless $last and uc($last) eq 'WHERE';
+ }
}
elsif ($token eq ')') {
-# $self->_new_line;
- $self->{_level} = pop(@{$self->{_level_stack}}) || 0;
+ $self->{ '_has_from' } = 0;
+ if ( ($last ne '(') && ($last ne '*') ) {
+ $self->{_level} = pop(@{$self->{_level_stack}}) || 0;
+ }
$self->_add_token($token);
$self->_new_line if ($self->_next_token
and $self->_next_token !~ /^AS$/i
and $self->_next_token ne ','
);
}
-
elsif ($token eq ',') {
$self->_add_token($token);
- $self->_new_line;
+ $self->_new_line if (!$self->{ '_is_in_where' });
}
elsif ($token eq ';') {
+ $self->{ '_has_from' } = 0;
+ $self->{ '_is_in_where' } = 0;
$self->_add_token($token);
+ $self->{break} = "\n" unless ($self->{spaces} != 0);
$self->_new_line;
# End of statement; remove all indentation.
@{$self->{_level_stack}} = ();
$self->{_level} = 0;
+ $self->{break} = ' ' unless ($self->{spaces} != 0);
}
elsif ($token =~ /^(?:SELECT|FROM|WHERE|HAVING|BEGIN|SET)$/i) {
- $self->_back if ($last and $last ne '(' and $last ne 'FOR');
- $self->_new_line;
- $self->_add_token($token);
- $self->_new_line if ((($token ne 'SET') || $last) and $self->_next_token and $self->_next_token ne '(' and $self->_next_token ne ';');
- $self->_over;
+
+ if (($token =~ /^FROM$/i) && $self->{ '_has_from' } ) {
+ $self->{ '_has_from' } = 0;
+ $self->_new_line;
+ $self->_add_token( $token );
+ $self->_new_line;
+ }
+ else
+ {
+ # if we're not in a sub-select, make sure these always are
+ # at the far left (col 1)
+ $self->_back if ( $last and $last ne '(' and $last ne 'FOR' );
+
+ $self->_new_line;
+ $self->_add_token( $token );
+ $self->_new_line if ( ( ( $token ne 'SET' ) || $last ) and $self->_next_token and $self->_next_token ne '(' and $self->_next_token ne ';' );
+ $self->_over;
+ }
+ if ($token =~ /^WHERE$/i) {
+ $self->{ '_is_in_where' } = 1;
+ } else {
+ $self->{ '_is_in_where' } = 0;
+ }
+
}
elsif ($token =~ /^(?:GROUP|ORDER|LIMIT)$/i) {
$self->_back;
$self->_new_line;
$self->_add_token($token);
+ $self->{ '_is_in_where' } = 0;
}
elsif ($token =~ /^(?:BY)$/i) {
$self->_over;
}
- elsif ($token =~ /^(?:LEFT|RIGHT|INNER|OUTER|CROSS)$/i) {
- $self->_back;
- $self->_new_line;
+ elsif ($token =~ /^(?:LEFT|RIGHT|INNER|OUTER|CROSS|NATURAL)$/i) {
+ $self->_back unless $last and $last eq ')';
+ if ($token =~ /(?:LEFT|RIGHT|CROSS|NATURAL)$/i) {
+ $self->_new_line;
+ $self->_over if ($self->{_level} == 0);
+ }
+ if ( ($token =~ /(?:INNER|OUTER)$/i) && ($last !~ /(?:LEFT|RIGHT|CROSS|NATURAL)$/i) ) {
+ $self->_new_line;
+ $self->_over if ($self->{_level} == 0);
+ }
$self->_add_token($token);
- $self->_over;
}
elsif ($token =~ /^(?:JOIN)$/i) {
- if ($last and $last !~ /^(?:LEFT|RIGHT|INNER|OUTER|CROSS)$/) {
+ if (!$last or $last !~ /^(?:LEFT|RIGHT|INNER|OUTER|CROSS|NATURAL)$/i) {
$self->_new_line;
}
-
$self->_add_token($token);
+ if ( $last && $last =~ /^(?:INNER|OUTER)$/i ) {
+ $self->_over;
+ }
}
elsif ($token =~ /^(?:AND|OR)$/i) {
- $self->_new_line;
+ if (!$last or ($last !~ /^(?:CREATE)$/i) ) {
+ $self->_new_line;
+ }
$self->_add_token($token);
-# $self->_new_line;
}
elsif ($token =~ /^--/) {
if (!$self->{no_comments}) {
$self->_add_token($token);
+ $self->{break} = "\n" unless ($self->{spaces} != 0);
$self->_new_line;
+ $self->{break} = ' ' unless ($self->{spaces} != 0);
}
}
elsif ($token =~ /^\/\*.*\*\/$/s) {
if (!$self->{no_comments}) {
- $token =~ s/\n\s+\*/\n\*/gs;
+ $token =~ s/\n[\s\t]+\*/\n\*/gs;
$self->_new_line;
$self->_add_token($token);
+ $self->{break} = "\n" unless ($self->{spaces} != 0);
$self->_new_line;
+ $self->{break} = " " unless ($self->{spaces} != 0);
}
}
- elsif ($token =~ /^(?:FOR)$/i) {
- $self->_new_line;
- $self->_over;
- $self->_add_token($token);
- }
-
elsif ($token =~ /^(?:USING)$/i) {
$self->_new_line;
$self->_add_token($token);
}
if ($wrap) {
- $token = $wrap->[0] . $token . $wrap->[1];
+ $token = $wrap->[0] . $token . $wrap->[1];
}
}
my $last_is_dot = defined($last_token) && $last_token eq '.';
if (!$self->_is_punctuation($token) and !$last_is_dot) {
- $self->{_output} .= $self->_indent;
+
+ my $sp = $self->_indent;
+ if ( (!defined($last_token) || $last_token ne '(') && ($token ne ')') && ($token !~ /^::/) ) {
+ $self->{_output} .= $sp if (!defined($last_token) || $last_token ne '::');
+ }
+ $token =~ s/\n/\n$sp/gs;
}
# uppercase keywords
- $token = uc $token
- if $self->_is_keyword($token)
- and $self->{uc_keywords};
+ if ($self->{uc_keywords} && $self->_is_keyword($token)) {
+ $token = lc($token) if ($self->{uc_keywords} == 1);
+ $token = uc($token) if ($self->{uc_keywords} == 2);
+ $token = ucfirst(lc($token)) if ($self->{uc_keywords} == 3);
+ }
+ # uppercase functions
+ if ($self->{uc_functions} && (my $fct = $self->_is_function($token))) {
+ $token =~ s/$fct/\L$fct\E/i if ($self->{uc_functions} == 1);
+ $token =~ s/$fct/\U$fct\E/i if ($self->{uc_functions} == 2);
+ $fct = ucfirst(lc($fct));
+ $token =~ s/$fct/$fct/i if ($self->{uc_functions} == 3);
+ }
$self->{_output} .= $token;
+ $self->{_output} =~ s/\(\s+\(/\(\(/gs;
# This can't be the beginning of a new line anymore.
$self->{_new_line} = 0;
return ~~ grep {$_ eq uc($token)} @{$self->{keywords}};
}
+ # Check if a token is a known SQL function.
+ sub _is_function
+ {
+ my ($self, $token) = @_;
+
+ my @ret = grep($token =~ /\b[\.]*$_$/i, @{$self->{functions}});
+
+ return $ret[0];
+ }
+
# Add new keywords to highlight.
sub add_keywords
{
}
}
+ # Add new functions to highlight.
+ sub add_functions
+ {
+ my $self = shift;
+
+ for my $function (@_) {
+ push @{$self->{functions}}, ref($function) ? @{$function} : $function;
+ }
+ }
+
# Add new rules.
sub add_rule
{
return $rule if (grep {uc($token) eq uc($_)} @$list);
}
- return undef;
+ return;
}
sub _process_rule
projects / pgbadger / commitdiff