| 1 | =head1 NAME |
| 2 | |
| 3 | a2p - Awk to Perl translator |
| 4 | |
| 5 | =head1 SYNOPSIS |
| 6 | |
| 7 | B<a2p> [I<options>] [I<filename>] |
| 8 | |
| 9 | =head1 DESCRIPTION |
| 10 | |
| 11 | I<A2p> takes an awk script specified on the command line (or from |
| 12 | standard input) and produces a comparable I<perl> script on the |
| 13 | standard output. |
| 14 | |
| 15 | =head2 OPTIONS |
| 16 | |
| 17 | Options include: |
| 18 | |
| 19 | =over 5 |
| 20 | |
| 21 | =item B<-DE<lt>numberE<gt>> |
| 22 | |
| 23 | sets debugging flags. |
| 24 | |
| 25 | =item B<-FE<lt>characterE<gt>> |
| 26 | |
| 27 | tells a2p that this awk script is always invoked with this B<-F> |
| 28 | switch. |
| 29 | |
| 30 | =item B<-nE<lt>fieldlistE<gt>> |
| 31 | |
| 32 | specifies the names of the input fields if input does not have to be |
| 33 | split into an array. If you were translating an awk script that |
| 34 | processes the password file, you might say: |
| 35 | |
| 36 | a2p -7 -nlogin.password.uid.gid.gcos.shell.home |
| 37 | |
| 38 | Any delimiter can be used to separate the field names. |
| 39 | |
| 40 | =item B<-E<lt>numberE<gt>> |
| 41 | |
| 42 | causes a2p to assume that input will always have that many fields. |
| 43 | |
| 44 | =item B<-o> |
| 45 | |
| 46 | tells a2p to use old awk behavior. The only current differences are: |
| 47 | |
| 48 | =over 5 |
| 49 | |
| 50 | =item * |
| 51 | |
| 52 | Old awk always has a line loop, even if there are no line |
| 53 | actions, whereas new awk does not. |
| 54 | |
| 55 | =item * |
| 56 | |
| 57 | In old awk, sprintf is extremely greedy about its arguments. |
| 58 | For example, given the statement |
| 59 | |
| 60 | print sprintf(some_args), extra_args; |
| 61 | |
| 62 | old awk considers I<extra_args> to be arguments to C<sprintf>; new awk |
| 63 | considers them arguments to C<print>. |
| 64 | |
| 65 | =back |
| 66 | |
| 67 | =back |
| 68 | |
| 69 | =head2 "Considerations" |
| 70 | |
| 71 | A2p cannot do as good a job translating as a human would, but it |
| 72 | usually does pretty well. There are some areas where you may want to |
| 73 | examine the perl script produced and tweak it some. Here are some of |
| 74 | them, in no particular order. |
| 75 | |
| 76 | There is an awk idiom of putting int() around a string expression to |
| 77 | force numeric interpretation, even though the argument is always |
| 78 | integer anyway. This is generally unneeded in perl, but a2p can't |
| 79 | tell if the argument is always going to be integer, so it leaves it |
| 80 | in. You may wish to remove it. |
| 81 | |
| 82 | Perl differentiates numeric comparison from string comparison. Awk |
| 83 | has one operator for both that decides at run time which comparison to |
| 84 | do. A2p does not try to do a complete job of awk emulation at this |
| 85 | point. Instead it guesses which one you want. It's almost always |
| 86 | right, but it can be spoofed. All such guesses are marked with the |
| 87 | comment "C<#???>". You should go through and check them. You might |
| 88 | want to run at least once with the B<-w> switch to perl, which will |
| 89 | warn you if you use == where you should have used eq. |
| 90 | |
| 91 | Perl does not attempt to emulate the behavior of awk in which |
| 92 | nonexistent array elements spring into existence simply by being |
| 93 | referenced. If somehow you are relying on this mechanism to create |
| 94 | null entries for a subsequent for...in, they won't be there in perl. |
| 95 | |
| 96 | If a2p makes a split line that assigns to a list of variables that |
| 97 | looks like (Fld1, Fld2, Fld3...) you may want to rerun a2p using the |
| 98 | B<-n> option mentioned above. This will let you name the fields |
| 99 | throughout the script. If it splits to an array instead, the script |
| 100 | is probably referring to the number of fields somewhere. |
| 101 | |
| 102 | The exit statement in awk doesn't necessarily exit; it goes to the END |
| 103 | block if there is one. Awk scripts that do contortions within the END |
| 104 | block to bypass the block under such circumstances can be simplified |
| 105 | by removing the conditional in the END block and just exiting directly |
| 106 | from the perl script. |
| 107 | |
| 108 | Perl has two kinds of array, numerically-indexed and associative. |
| 109 | Perl associative arrays are called "hashes". Awk arrays are usually |
| 110 | translated to hashes, but if you happen to know that the index is |
| 111 | always going to be numeric you could change the {...} to [...]. |
| 112 | Iteration over a hash is done using the keys() function, but iteration |
| 113 | over an array is NOT. You might need to modify any loop that iterates |
| 114 | over such an array. |
| 115 | |
| 116 | Awk starts by assuming OFMT has the value %.6g. Perl starts by |
| 117 | assuming its equivalent, $#, to have the value %.20g. You'll want to |
| 118 | set $# explicitly if you use the default value of OFMT. |
| 119 | |
| 120 | Near the top of the line loop will be the split operation that is |
| 121 | implicit in the awk script. There are times when you can move this |
| 122 | down past some conditionals that test the entire record so that the |
| 123 | split is not done as often. |
| 124 | |
| 125 | For aesthetic reasons you may wish to change the array base $[ from 1 |
| 126 | back to perl's default of 0, but remember to change all array |
| 127 | subscripts AND all substr() and index() operations to match. |
| 128 | |
| 129 | Cute comments that say "# Here is a workaround because awk is dumb" |
| 130 | are passed through unmodified. |
| 131 | |
| 132 | Awk scripts are often embedded in a shell script that pipes stuff into |
| 133 | and out of awk. Often the shell script wrapper can be incorporated |
| 134 | into the perl script, since perl can start up pipes into and out of |
| 135 | itself, and can do other things that awk can't do by itself. |
| 136 | |
| 137 | Scripts that refer to the special variables RSTART and RLENGTH can |
| 138 | often be simplified by referring to the variables $`, $& and $', as |
| 139 | long as they are within the scope of the pattern match that sets them. |
| 140 | |
| 141 | The produced perl script may have subroutines defined to deal with |
| 142 | awk's semantics regarding getline and print. Since a2p usually picks |
| 143 | correctness over efficiency. it is almost always possible to rewrite |
| 144 | such code to be more efficient by discarding the semantic sugar. |
| 145 | |
| 146 | For efficiency, you may wish to remove the keyword from any return |
| 147 | statement that is the last statement executed in a subroutine. A2p |
| 148 | catches the most common case, but doesn't analyze embedded blocks for |
| 149 | subtler cases. |
| 150 | |
| 151 | ARGV[0] translates to $ARGV0, but ARGV[n] translates to $ARGV[$n]. A |
| 152 | loop that tries to iterate over ARGV[0] won't find it. |
| 153 | |
| 154 | =head1 ENVIRONMENT |
| 155 | |
| 156 | A2p uses no environment variables. |
| 157 | |
| 158 | =head1 AUTHOR |
| 159 | |
| 160 | Larry Wall E<lt>F<larry@wall.org>E<gt> |
| 161 | |
| 162 | =head1 FILES |
| 163 | |
| 164 | =head1 SEE ALSO |
| 165 | |
| 166 | perl The perl compiler/interpreter |
| 167 | |
| 168 | s2p sed to perl translator |
| 169 | |
| 170 | =head1 DIAGNOSTICS |
| 171 | |
| 172 | =head1 BUGS |
| 173 | |
| 174 | It would be possible to emulate awk's behavior in selecting string |
| 175 | versus numeric operations at run time by inspection of the operands, |
| 176 | but it would be gross and inefficient. Besides, a2p almost always |
| 177 | guesses right. |
| 178 | |
| 179 | Storage for the awk syntax tree is currently static, and can run out. |