src/parser/prescc.sh

   1 # This file is part of NIT ( http://www.nitlanguage.org ).
   2 #
   3 # Copyright 2009 Jean Privat <jean@pryen.org>
   4 #
   5 # Licensed under the Apache License, Version 2.0 (the "License");
   6 # you may not use this file except in compliance with the License.
   7 # You may obtain a copy of the License at
   8 #
   9 #     http://www.apache.org/licenses/LICENSE-2.0
  10 #
  11 # Unless required by applicable law or agreed to in writing, software
  12 # distributed under the License is distributed on an "AS IS" BASIS,
  13 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  14 # See the License for the specific language governing permissions and
  15 # limitations under the License.
  16
  17 # prescc, a Sablecc preprocessor.
  18 #
  19 # Synopsis
  20 #
  21 #   Extends a sablecc grammar with parametrized productions.
  22 #
  23 # Description
  24 #
  25 #   A production named foo~bar~baz semantically correspond to a production foo with two boolean parameters bar and baz
  26 #   In fact foo is a family of 4 distinct productions: foo, foo_bar, foo_baz and foo_bar_baz
  27 #   In a parametrized production with a parameter ~xxx:
  28 #    * parameters (~xxx) are substituted with _xxx if the parameter is true and removed if the parameter is false
  29 #    * guarded alternatives (!xxx) are disabled if the parameter is true
  30 #
  31 # Limitations
  32 #
  33 #   prescc is badly implemented with shell, sed and perl and is not robust.
  34 #   Users must remember the following:
  35 #    * parametrized productions MUST be terminated with a line containing only a single semicolon (;)
  36 #    * parameters (~) and guards (!) in alternatives MUST correspond to a parameter of the enclosing production
  37 #    * if required, names in transformations MUST contain the full invocation name (with all parameters)
  38 #        foo bar_x~y~z_t baz {-> New p(foo, bar_x~y~z_t.q)}
  39 #    * guards do not understand grammar, they just remove the whole line
  40 #    * The AST MUST start with a line containing only "Abstract Syntax Tree"
  41 #
  42 # Example of the dangling else implementation:
  43 #
  44 #   stmt~withelse =
  45 #                'if' expr 'then' stmt_withelse 'else' stmt~withelse |
  46 #     !withelse  'if' expr 'then' stmt |
  47 #                nop
  48 #                ;
  49
  50
  51 case $# in
  52         2);;
  53         *) echo "Usage: prescc infile outfile"; exit
  54 esac
  55
  56
  57 infile=$1
  58 outfile=$2
  59 tmpfile=`mktemp "$2.XXXXXX"`
  60
  61 echo -n "/* This file is autogenerated, do not modify it */" > "$outfile"
  62 cat "$infile" >> "$outfile"
  63
  64 # The perl code is used to list all the available parameters in the extended grammar
  65 for token in `perl -ne 'if (/\~(\w+)/ && !$found{$1}) {print "$1\n"; $found{$1}=1}' "$infile"`
  66 do
  67         echo "Parameter ~$token"
  68         # first, sed starts from first line to the AST line and removes ~xxx and !xxx
  69         sed -n -e "1,/^Abstract Syntax Tree/{/^Abstract Syntax Tree/b;s/[\~!]$token//g;p}" "$outfile" > "$tmpfile"
  70         # second, sed clones ~xxx parametrized productions, substitute ~xxx with _xxx and delete !xxx lines
  71         sed -n -e "/\~$token/,/;/{s/\~$token/_$token/g;/!$token/d;p}" "$outfile" >> "$tmpfile"
  72         # third, sed continues fron AST line to last line and remove ~xxx and !xxx
  73         sed -n -e "/^Abstract Syntax Tree/,\${s/[\~!]$token//g;p}" "$outfile" >> "$tmpfile"
  74         mv "$tmpfile" "$outfile"
  75 done
  76