Format scripts in awk, csh, ksh, perl, sh


I am used to relay on tools for formatting and indenting SQL or PL/SQL code.
Looking for the same for linux scripts (especially bash) I found the following code
fmt.script

#!/usr/bin/env perl
# fmt.script - format scripts in awk, csh, ksh, perl, sh
#
# we do:
# standardize indentation (an indent is one tab by default)
# strip trailing whitespace
# change ${var} to $var where possible
# change ">x" to "> x" for shell scripts
# change "[ ... ]" to "test ..." for Bourne shell scripts
#
# we may do someday, but these are harder:
# convert $VAR to $var unless a setenv or default environment variable
# possibly prepending stuff from template.sh
# "if ... \nthen", for ... in\ndo", "while/until ... \ndo", "fn()\n{"
#
# to have fmt.script reformat itself (a fair test, yes?) try:
# fmt.script fmt.script fmt.script.new # use tabs for indents
# fmt.script -s4 fmt.script fmt.script.new # indent is fou

# variable initialization
$RCSversion = q$Revision: 1.17 $;
($version) = ($RCSversion =~ /Revision:\s+([\d]+\.[\d]+)/);
$pr = $0; # name of this program
$pr =~ s|.*\/||; # basename of this program
$ilen = 1; # characters per indent
$ichar = "\t"; # indent character (typically <SPACE> or <TAB>)
$ilevel = 0; # indent level
$caselevel = 0; # case level for sh/ksh scripts
$type = ''; # unknown type

# usage message
$usage = "usage:
$pr -help
$pr [-t|-s#] [-ksh|-sh] [infile [outfile]]
-t indent is one tab character (default)
-s# indent is <#> space characters
-sh script is Bourne shell
-ksh script is Korn shell
infile input file name (default STDIN)
outfile output file name (default STDOUT)
notes:
* you can use '-' for STDIN/STDOUT
* $pr gives outfile execute permissions when possible
example:
* to convert an ugly script into a pretty script, using a 4space indent:
$pr -s4 script.ugly script.pretty
";

# process command options
ARGS: while ($#ARGV >= 0) {
$_ = $ARGV[0];
ARG0: {
/^-h/ && do { print "$usage"; exit; }; # print usage
/^-t$/ && do { $ichar="\t"; $ilen=1 ; last ARG0 }; # one tab
/^-s(\d+)$/ && do { $ichar=" " ; $ilen=$1; last ARG0 }; # $1 spaces
/^-sh$/ && do { $type='sh'; last ARG0 }; # Bourne shell
/^-ksh$/ && do { $type='ksh'; last ARG0 }; # Korn shell
/./ && last ARGS; # file to fmt
}
shift;
}

# check remaining args
if ($#ARGV > 1 || $ARGV[0] =~ /^-./) {
die "$usage";
}

#============================================================================
# regular expression stuff
#
# list of regex to increase next line
$flow = '^(if|for|foreach|while|until)\s'; # flow control statements
$caseelem = '^[^\(\)]+\)(\s|$)'; # sh case element (no internal parens!)
$function = '\(\s*\)(\s*\{\s*)?$'; # sh lists, perl functions, code blocks
$list = '[\{\(]\s*$'; # list executed, possibly in subshell
$hereis = '<<\s*(\S+)\s*$'; # hereis documents

# list of regex to decrease this line
$endflow = '^(done|end|endif|fi)\b'; # fi/done/end/endif
$endfn = '^\}'; # end of function declaration or block
$endlist = '^[\}\)]'; # end of list or block execution
$endcaseelem = '^(;;|breaksw)'; # end of sh/csh case elements

# list of regex to decrease this line *and* increase next line
$else = '^(\}\s*)?(else|elif|elsif|else\s+if)\b'; # else/elsif variants
@endhereis = (); # words used to begin/end here-is docs

# list of regex to *postpone* indentation until *next* line
$postponeinc = '^(then|do)\b'; # sh if/for/while/until ... \nthen/do

# list of useful regex - watch out for oneliners like
# "for ... ; do ... ; done"
# "case ... in ; ...) ... ;; ...) ... ;; esac"
# "...) ... ;; # part of condensed form case statement
# we could even have something like
# for ... ; do ... ; done | whatever ...
$linecaseelem = "$caseelem.*;;"; # handle condensed form case element
$endinline = ';\s*(done|end|esac|fi)\b'; # inline flow statement end
$falseincnext = "$linecaseelem|$endinline"; # false alarm

# regex arrays @decthis and @incnext notes:
# 1) $caseelem regex is pushed onto @incnext only while inside a sh/ksh case
# 2) $hereis regex is pushed onto @incnext only for sh/ksh/csh scripts
# 3) $else has to go *before* the other regex in @decthis, because of $endfn
@decthis = ($else, $endflow, $endfn, $endlist, $endcaseelem);
@incnext = ($else, $flow, $function, $list);
if ($type =~ /sh$/) {
push(@incnext, $hereis);
}

#============================================================================


# start our work - open files, etc.
$istr = $ichar x $ilen; # indent string (typically " " or "\t")
$input = shift(@ARGV) || '-';
$output = shift(@ARGV) || '-';
open( INPUT, "$input") || die "$pr: Unable to open $input: $!\n";
open(OUTPUT, ">$output") || die "$pr: Unable to create $output: $!\n";

# process input
while ($line = <INPUT>) {

# initial processing for every line
$doincnext = 0; # increase indent on next line?
$dodecthis = 0; # decrease indent on this line?
$dopostponeinc = 0; # postpone increase indent on this line?
$badcaseelem = 0; # do we have a bad case element?
$comment = ''; # assume this line has no inline comment
chop($line); # strip newline
$line =~ s/\s+$//; # strip trailing whitespace
$line =~ s/^\s+//; # strip indentation

# blank lines and comment lines can be passed straight through, with
# no effect on indentation
if ($line =~ /^$/) {
print OUTPUT "\n";
next;
}
if ($line =~ /^#/) {
# figure out what type of script we are processing
if ($. == 1) {
if ($line =~ m,#!\s*/\S+/(\w+)(\s|$), ) {
$type = $1;
}
}
print OUTPUT $istr x $ilevel, $line, "\n";
next;
}

# inline comments can be stripped to protect them from other substitutions;
# the tricky part is deciding what is a comment, since hashmarks can
# appear inside quotes;
# we will err on the safe side; meaning we will not strip the comment
# unless we are pretty sure it is safe
if ($line =~ s/(\s+#[^'"`]+$)//) {
$comment = $1;
}
# inline substitutions # discouraged preferred
$line =~ s/\$\{(\w+)\}(\W|$)/\$$1$2/g; # ${var} $var
if ($type =~ /^k?sh$/) {
$line =~ s,>([^\s\&>]),> $1,; # x>file x> file
$line =~ s,([^\s\d>])>,$1 >,; # x> file x > file
# # [ ... ] test ...
$line =~ s/^((if|while)(\s+))?(\[)(.*[^\\"'])(\])/$1test$5/;
}

# track case level - use this to see if we should check for case elements
# there is no point in checking for case elements if we aren't in a
# ksh/sh script and actually inside a case statement
if ($type =~ /^k?sh$/) {
if ($line =~ /^case\b/ && $line !~ /$linecaseelem/) {
($caselevel == 0) && push(@incnext, $caseelem);
$caselevel++;
} elsif ($line =~ /^esac\b/) {
$caselevel--;
($caselevel == 0) && pop(@incnext);
}
}

# see if we are going to decrease this line's indentation
foreach $regex (@decthis) {
($line =~ /$regex/) && ($dodecthis=1) && last;
}
if ($dodecthis == 1) {
$ilevel--;
}

# handle "command ;;" problems, but make sure we don't trip over
# condensed form case
if ($type =~ /^k?sh$/ && $line =~ /.;;/ && $line !~ /$linecaseelem/) {
$badcaseelem = 1;
$line =~ s/\s*;;//;
}

# see if we are going to postpone increasing this line's indentation
if ($line =~ /^(then|do)\b/) {
$dopostponeinc = 1;
$ilevel--;
}

# print this line (but don't indent blank lines)
if (! length($line)) {
print OUTPUT "\n";
} else {
print OUTPUT $istr x $ilevel, $line, $comment, "\n";
}

# if we found a bad case element print the closing ";;" which we stripped
if ($badcaseelem) {
$ilevel--;
print OUTPUT $istr x $ilevel, ";;\n";
}

# we postponed it above, but make sure we do it now
if ($dopostponeinc == 1) {
$ilevel++;
} else {
# see if we are going to increase the next line's indentation
foreach $regex (@incnext) {
($line =~ /$regex/) && ($doincnext=1) && last;
}
if ($doincnext == 1) {
if ($regex == $flow) {
# if we are bourne shell, we can do some advanced checking
if ($type =~ /^k?sh$/) {
# for "for" and "case" we can do further checking, but
# for "if", "elif", "else", and "while" we usually can't
# prove anything; however there is one small exception
if ($line =~ /^for/) {
if ($line !~ /^for\s+\w+($|\s+[;#]|\s+in\s+\S+)/) {
$doincnext = 0;
}
} elsif ($line =~ /^case/) {
if ($line !~ /^case\s+\S+\s+in\b/) {
$doincnext = 0;
}
} else {
# flow keywords can be in quoted multi-line text, like:
# usage="$pr [opts]\nif you like chocolate, honk!"\n
# this can sometimes be reliably detected (assuming the
# code actually runs)
# if the line has no comments and exactly 1 quote char,
# it is probably like the usage example above
if ($line !~ /#/) {
@chars = split(//, $line);
$nquote = grep(/[\'\"\`]/, @chars);
if ($nquote == 1) {
$doincnext = 0;
}
}
}
}
}
if ($doincnext == 1 && $line !~ /$falseincnext/) {
$ilevel++;
}
}
}
#print STDERR $pr, ": $doincnext/$dodecthis=$ilevel:", $line, "\n";
}
close(INPUT);
close(STDOUT);

# fix permissions on output file
if ($output ne '-') {
if ($input ne '-') {
($dev,$ino,$mode,@junk) = stat($input);
chmod($mode,$output);
} else {
system("chmod +x $output");
}
}
exit;

Comments

Popular posts from this blog

Data Guard - Changing IP Addresses

Install Oracle Internet Directory (OID) in Standalone mode

Fixing & Registering ORACLE_HOMES in Central Inventory