src/env_parallel.bash

   1 #!/usr/bin/env bash
   2
   3 # This file must be sourced in bash:
   4 #
   5 #   source `which env_parallel.bash`
   6 #
   7 # after which 'env_parallel' works
   8 #
   9 #
  10 # Copyright (C) 2016-2020 Ole Tange, http://ole.tange.dk and Free
  11 # Software Foundation, Inc.
  12 #
  13 # This program is free software; you can redistribute it and/or modify
  14 # it under the terms of the GNU General Public License as published by
  15 # the Free Software Foundation; either version 3 of the License, or
  16 # (at your option) any later version.
  17 #
  18 # This program is distributed in the hope that it will be useful, but
  19 # WITHOUT ANY WARRANTY; without even the implied warranty of
  20 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  21 # General Public License for more details.
  22 #
  23 # You should have received a copy of the GNU General Public License
  24 # along with this program; if not, see <http://www.gnu.org/licenses/>
  25 # or write to the Free Software Foundation, Inc., 51 Franklin St,
  26 # Fifth Floor, Boston, MA 02110-1301 USA
  27
  28 env_parallel() {
  29     # env_parallel.bash
  30
  31     _names_of_ALIASES() {
  32         compgen -a
  33     }
  34     _bodies_of_ALIASES() {
  35         local _i
  36         for _i in $@; do
  37             if [ $(alias $_i | wc -l) == 1 ] ; then
  38                 true Alias is a single line. Good.
  39             else
  40                 _warning_PAR "Alias '$_i' contains newline."
  41                 _warning_PAR "Make sure the command has at least one newline after '$_i'."
  42                 _warning_PAR "See BUGS in 'man env_parallel'."
  43             fi
  44         done
  45         alias "$@"
  46     }
  47     _names_of_FUNCTIONS() {
  48         compgen -A function
  49     }
  50     _bodies_of_FUNCTIONS() {
  51         typeset -f "$@"
  52     }
  53     _names_of_VARIABLES() {
  54         compgen -A variable
  55     }
  56     _bodies_of_VARIABLES() {
  57         typeset -p "$@"
  58     }
  59     _ignore_HARDCODED() {
  60         # These names cannot be detected
  61         echo '(_|TIMEOUT|GROUPS|FUNCNAME|DIRSTACK|PIPESTATUS|USERNAME|BASHPID|BASH_[A-Z_]+)'
  62     }
  63     _ignore_READONLY() {
  64         readonly | perl -e '@r = map {
  65                 chomp;
  66                 # sh on UnixWare: readonly TIMEOUT
  67                 # ash: readonly var='val'
  68                 # ksh: var='val'
  69                 s/^(readonly )?([^= ]*)(=.*|)$/$2/ or
  70                 # bash: declare -ar BASH_VERSINFO=([0]="4" [1]="4")
  71                 # zsh: typeset -r var='val'
  72                   s/^\S+\s+\S+\s+(\S[^=]*)(=.*|$)/$1/;
  73                 $_ } <>;
  74             $vars = join "|",map { quotemeta $_ } @r;
  75             print $vars ? "($vars)" : "(,,nO,,VaRs,,)";
  76             '
  77     }
  78     _remove_bad_NAMES() {
  79         # Do not transfer vars and funcs from env_parallel
  80         _ignore_RO="`_ignore_READONLY`"
  81         _ignore_HARD="`_ignore_HARDCODED`"
  82         # Macos-grep does not like long patterns
  83         # Old Solaris grep does not support -E
  84         # Perl Version of:
  85         # grep -Ev '^(...)$' |
  86         perl -ne '/^(
  87                      PARALLEL_ENV|
  88                      PARALLEL_TMP|
  89                      _alias_NAMES|
  90                      _bodies_of_ALIASES|
  91                      _bodies_of_FUNCTIONS|
  92                      _bodies_of_VARIABLES|
  93                      _error_PAR|
  94                      _function_NAMES|
  95                      _get_ignored_VARS|
  96                      _grep_REGEXP|
  97                      _ignore_HARD|
  98                      _ignore_HARDCODED|
  99                      _ignore_READONLY|
 100                      _ignore_RO|
 101                      _ignore_UNDERSCORE|
 102                      _list_alias_BODIES|
 103                      _list_function_BODIES|
 104                      _list_variable_VALUES|
 105                      _make_grep_REGEXP|
 106                      _names_of_ALIASES|
 107                      _names_of_FUNCTIONS|
 108                      _names_of_VARIABLES|
 109                      _names_of_maybe_FUNCTIONS|
 110                      _parallel_exit_CODE|
 111                      _prefix_PARALLEL_ENV|
 112                      _prefix_PARALLEL_ENV|
 113                      _remove_bad_NAMES|
 114                      _remove_readonly|
 115                      _variable_NAMES|
 116                      _warning_PAR|
 117                      _which_PAR)$/x and next;
 118             # Filter names matching --env
 119             /^'"$_grep_REGEXP"'$/ or next;
 120             /^'"$_ignore_UNDERSCORE"'$/ and next;
 121             # Remove readonly variables
 122             /^'"$_ignore_RO"'$/ and next;
 123             /^'"$_ignore_HARD"'$/ and next;
 124             print;'
 125     }
 126     _prefix_PARALLEL_ENV() {
 127         shopt 2>/dev/null |
 128         perl -pe 's:\s+off:;: and s/^/shopt -u /;
 129                   s:\s+on:;: and s/^/shopt -s /;
 130                   s:;$:&>/dev/null;:';
 131         echo 'shopt -s expand_aliases &>/dev/null';
 132     }
 133
 134     _get_ignored_VARS() {
 135         perl -e '
 136             for(@ARGV){
 137                 $next_is_env and push @envvar, split/,/, $_;
 138                 $next_is_env=/^--env$/;
 139             }
 140             if(grep { /^_$/ } @envvar) {
 141                 if(not open(IN, "<", "$ENV{HOME}/.parallel/ignored_vars")) {
 142                     print STDERR "parallel: Error: ",
 143                     "Run \"parallel --record-env\" in a clean environment first.\n";
 144                 } else {
 145                     chomp(@ignored_vars = <IN>);
 146                 }
 147             }
 148             if($ENV{PARALLEL_IGNORED_NAMES}) {
 149                 push @ignored_vars, split/\s+/, $ENV{PARALLEL_IGNORED_NAMES};
 150                 chomp @ignored_vars;
 151             }
 152             $vars = join "|",map { quotemeta $_ } @ignored_vars;
 153             print $vars ? "($vars)" : "(,,nO,,VaRs,,)";
 154             ' -- "$@"
 155     }
 156
 157     # Get the --env variables if set
 158     # --env _ should be ignored
 159     # and convert  a b c  to (a|b|c)
 160     # If --env not set: Match everything (.*)
 161     _make_grep_REGEXP() {
 162         perl -e '
 163             for(@ARGV){
 164                 /^_$/ and $next_is_env = 0;
 165                 $next_is_env and push @envvar, split/,/, $_;
 166                 $next_is_env = /^--env$/;
 167             }
 168             $vars = join "|",map { quotemeta $_ } @envvar;
 169             print $vars ? "($vars)" : "(.*)";
 170             ' -- "$@"
 171     }
 172     _which_PAR() {
 173         # type returns:
 174         #   ll is an alias for ls -l (in ash)
 175         #   bash is a tracked alias for /bin/bash
 176         #   true is a shell builtin (in bash)
 177         #   myfunc is a function (in bash)
 178         #   myfunc is a shell function (in zsh)
 179         #   which is /usr/bin/which (in sh, bash)
 180         #   which is hashed (/usr/bin/which)
 181         #   gi is aliased to `grep -i' (in bash)
 182         #   aliased to `alias | /usr/bin/which --tty-only --read-alias --show-dot --show-tilde'
 183         # Return 0 if found, 1 otherwise
 184         LANG=C type "$@" |
 185             perl -pe '$exit += (s/ is an alias for .*// ||
 186                                 s/ is aliased to .*// ||
 187                                 s/ is a function// ||
 188                                 s/ is a shell function// ||
 189                                 s/ is a shell builtin// ||
 190                                 s/.* is hashed .(\S+).$/$1/ ||
 191                                 s/.* is (a tracked alias for )?//);
 192                       END { exit not $exit }'
 193     }
 194     _warning_PAR() {
 195         echo "env_parallel: Warning: $@" >&2
 196     }
 197     _error_PAR() {
 198         echo "env_parallel: Error: $@" >&2
 199     }
 200
 201     # Bash is broken in version 3.2.25 and 4.2.39
 202     # The crazy '[ "`...`" == "" ]' is needed for the same reason
 203     if [ "`_which_PAR parallel`" == "" ]; then
 204         _error_PAR 'parallel must be in $PATH.'
 205         return 255
 206     fi
 207
 208     # Grep regexp for vars given by --env
 209     _grep_REGEXP="`_make_grep_REGEXP \"$@\"`"
 210     unset _make_grep_REGEXP
 211
 212     # Deal with --env _
 213     _ignore_UNDERSCORE="`_get_ignored_VARS \"$@\"`"
 214     unset _get_ignored_VARS
 215
 216     # --record-env
 217     # Bash is broken in version 3.2.25 and 4.2.39
 218     # The crazy '[ "`...`" == 0 ]' is needed for the same reason
 219     if [ "`perl -e 'exit grep { /^--record-env$/ } @ARGV' -- "$@"; echo $?`" == 0 ] ; then
 220         true skip
 221     else
 222         (_names_of_ALIASES;
 223          _names_of_FUNCTIONS;
 224          _names_of_VARIABLES) |
 225             cat > $HOME/.parallel/ignored_vars
 226         return 0
 227     fi
 228
 229     # --session
 230     # Bash is broken in version 3.2.25 and 4.2.39
 231     # The crazy '[ "`...`" == 0 ]' is needed for the same reason
 232     if [ "`perl -e 'exit grep { /^--session$/ } @ARGV' -- "$@"; echo $?`" == 0 ] ; then
 233         true skip
 234     else
 235         # Insert ::: between each level of session
 236         # so you can pop off the last ::: at --end-session
 237         PARALLEL_IGNORED_NAMES="`echo \"$PARALLEL_IGNORED_NAMES\";
 238           echo :::;
 239           (_names_of_ALIASES;
 240            _names_of_FUNCTIONS;
 241            _names_of_VARIABLES) | perl -ne '
 242             BEGIN{
 243               map { $ignored_vars{$_}++ }
 244                 split/\s+/, $ENV{PARALLEL_IGNORED_NAMES};
 245             }
 246             chomp;
 247             for(split/\s+/) {
 248               if(not $ignored_vars{$_}) {
 249                 print $_,\"\\n\";
 250               }
 251             }
 252             '`"
 253         export PARALLEL_IGNORED_NAMES
 254         return 0
 255     fi
 256     # Bash is broken in version 3.2.25 and 4.2.39
 257     # The crazy '[ "`...`" == 0 ]' is needed for the same reason
 258     if [ "`perl -e 'exit grep { /^--end.?session$/ } @ARGV' -- "$@"; echo $?`" == 0 ] ; then
 259         true skip
 260     else
 261         # Pop off last ::: from PARALLEL_IGNORED_NAMES
 262         PARALLEL_IGNORED_NAMES="`perl -e '
 263           $ENV{PARALLEL_IGNORED_NAMES} =~ s/(.*):::.*?$/$1/s;
 264           print $ENV{PARALLEL_IGNORED_NAMES}
 265         '`"
 266         return 0
 267     fi
 268     # Grep alias names
 269     _alias_NAMES="`_names_of_ALIASES | _remove_bad_NAMES | xargs echo`"
 270     _list_alias_BODIES="_bodies_of_ALIASES $_alias_NAMES"
 271     if [ "$_alias_NAMES" = "" ] ; then
 272         # no aliases selected
 273         _list_alias_BODIES="true"
 274     fi
 275     unset _alias_NAMES
 276
 277     # Grep function names
 278     _function_NAMES="`_names_of_FUNCTIONS | _remove_bad_NAMES | xargs echo`"
 279     _list_function_BODIES="_bodies_of_FUNCTIONS $_function_NAMES"
 280     if [ "$_function_NAMES" = "" ] ; then
 281         # no functions selected
 282         _list_function_BODIES="true"
 283     fi
 284     unset _function_NAMES
 285
 286     # Grep variable names
 287     _variable_NAMES="`_names_of_VARIABLES | _remove_bad_NAMES | xargs echo`"
 288     _list_variable_VALUES="_bodies_of_VARIABLES $_variable_NAMES"
 289     if [ "$_variable_NAMES" = "" ] ; then
 290         # no variables selected
 291         _list_variable_VALUES="true"
 292     fi
 293     unset _variable_NAMES
 294
 295     _which_TRUE="`_which_PAR true`"
 296     # Copy shopt (so e.g. extended globbing works)
 297     # But force expand_aliases as aliases otherwise do not work
 298     PARALLEL_ENV="`
 299         _prefix_PARALLEL_ENV
 300         $_list_alias_BODIES;
 301         $_list_function_BODIES;
 302         $_list_variable_VALUES;
 303     `"
 304     export PARALLEL_ENV
 305     unset _list_alias_BODIES _list_variable_VALUES _list_function_BODIES
 306     unset _bodies_of_ALIASES _bodies_of_VARIABLES _bodies_of_FUNCTIONS
 307     unset _names_of_ALIASES _names_of_VARIABLES _names_of_FUNCTIONS
 308     unset _ignore_HARDCODED _ignore_READONLY _ignore_UNDERSCORE
 309     unset _remove_bad_NAMES _grep_REGEXP
 310     unset _prefix_PARALLEL_ENV
 311     # Test if environment is too big
 312     if [ "`_which_PAR true`" == "$_which_TRUE" ] ; then
 313         parallel "$@"
 314         _parallel_exit_CODE=$?
 315         # Clean up variables/functions
 316         unset PARALLEL_ENV
 317         unset _which_PAR _which_TRUE
 318         unset _warning_PAR _error_PAR
 319         # Unset _parallel_exit_CODE before return
 320         eval "unset _parallel_exit_CODE; return $_parallel_exit_CODE"
 321     else
 322         unset PARALLEL_ENV;
 323         _error_PAR "Your environment is too big."
 324         _error_PAR "You can try 3 different approaches:"
 325         _error_PAR "1. Run 'env_parallel --session' before you set"
 326         _error_PAR "   variables or define functions."
 327         _error_PAR "2. Use --env and only mention the names to copy."
 328         _error_PAR "3. Try running this in a clean environment once:"
 329         _error_PAR "     env_parallel --record-env"
 330         _error_PAR "   And then use '--env _'"
 331         _error_PAR "For details see: man env_parallel"
 332         return 255
 333     fi
 334 }
 335
 336 parset() {
 337     _parset_PARALLEL_PRG=parallel
 338     _parset_main "$@"
 339 }
 340
 341 env_parset() {
 342     _parset_PARALLEL_PRG=env_parallel
 343     _parset_main "$@"
 344 }
 345
 346 _parset_main() {
 347     # If $1 contains ',' or space:
 348     #   Split on , to get the destination variable names
 349     # If $1 is a single destination variable name:
 350     #   Treat it as the name of an array
 351     #
 352     #   # Create array named myvar
 353     #   parset myvar echo ::: {1..10}
 354     #   echo ${myvar[5]}
 355     #
 356     #   # Put output into $var_a $var_b $var_c
 357     #   varnames=(var_a var_b var_c)
 358     #   parset "${varnames[*]}" echo ::: {1..3}
 359     #   echo $var_c
 360     #
 361     #   # Put output into $var_a4 $var_b4 $var_c4
 362     #   parset "var_a4 var_b4 var_c4" echo ::: {1..3}
 363     #   echo $var_c4
 364
 365     _make_TEMP() {
 366         # mktemp does not exist on some OS
 367         perl -e 'use File::Temp qw(tempfile);
 368                  $ENV{"TMPDIR"} ||= "/tmp";
 369                  print((tempfile(DIR=>$ENV{"TMPDIR"}, TEMPLATE => "parXXXXX"))[1])'
 370     }
 371
 372     _parset_NAME="$1"
 373     if [ "$_parset_NAME" = "" ] ; then
 374         echo parset: Error: No destination variable given. >&2
 375         echo parset: Error: Try: >&2
 376         echo parset: Error: ' ' parset myarray echo ::: foo bar >&2
 377         return 255
 378     fi
 379     shift
 380     echo "$_parset_NAME" |
 381         perl -ne 'chomp;for (split /[, ]/) {
 382             # Allow: var_32 var[3]
 383             if(not /^[a-zA-Z_][a-zA-Z_0-9]*(\[\d+\])?$/) {
 384                 print STDERR "parset: Error: $_ is an invalid variable name.\n";
 385                 print STDERR "parset: Error: Variable names must be letter followed by letters or digits.\n";
 386                 $exitval = 255;
 387             }
 388         }
 389         exit $exitval;
 390         ' || return 255
 391     _exit_FILE=`_make_TEMP`
 392     if perl -e 'exit not grep /,| /, @ARGV' "$_parset_NAME" ; then
 393         # $_parset_NAME contains , or space
 394         # Split on , or space to get the names
 395         eval "$(
 396             # Compute results into files
 397             ($_parset_PARALLEL_PRG --files -k "$@"; echo $? > "$_exit_FILE") |
 398                 # var1=`cat tmpfile1; rm tmpfile1`
 399                 # var2=`cat tmpfile2; rm tmpfile2`
 400                 parallel --plain -q echo {2}='`cat {1}; rm {1}`' :::: - :::+ $(
 401                     echo "$_parset_NAME" | perl -pe 's/,/ /g'
 402                          )
 403             );
 404         "
 405     else
 406         # $_parset_NAME does not contain , or space
 407         # => $_parset_NAME is the name of the array to put data into
 408         # Supported in: bash zsh ksh mksh
 409         # Arrays do not work in: sh ash dash
 410         eval "$_parset_NAME=( $(
 411             # Compute results into files. Save exit value
 412             ($_parset_PARALLEL_PRG --files -k "$@"; echo $? > "$_exit_FILE") |
 413                 perl -pe 'chop;$_="\"\`cat $_; rm $_\`\" "'
 414             ) )"
 415     fi
 416     unset _parset_NAME _parset_PARALLEL_PRG _parallel_exit_CODE
 417     # Unset _exit_FILE before return
 418     eval "unset _exit_FILE; return \`cat $_exit_FILE; rm $_exit_FILE\`"
 419 }