#!/usr/bin/wish # This file provides many valuable data processing procedures like lappend_unique, get_var, etc. proc lappend_unique { args } { # Add the each entry to a list if and only if they do not already exist in the list. # Description of argument(s): # args The first argument should be the list name. All other arguments are # items to be added to the list. set list_name [lindex $args 0] # Remove first entry from args list. set args [lreplace $args 0 0] upvar 1 $list_name list if { ! [info exists list] } { set list {} } foreach arg $args { if { [lsearch -exact $list "${arg}"] != -1 } { continue } lappend list $arg } } proc lsubtract {main_list_name removal_list} { upvar $main_list_name ref_main_list # Remove any entry from the main list that is contained in removal list. # Description of argument(s): # main_list_name The name of your master list. # removal_list The list of items to be removed from master list. # For each element in the removal list, find the corresponding entry in the master list and remove it. for {set removal_ix 0} {$removal_ix < [llength $removal_list ]}\ {incr removal_ix} { set value [lindex $removal_list $removal_ix] set master_ix [lsearch $ref_main_list $value] set ref_main_list [lreplace $ref_main_list $master_ix $master_ix] } } proc list_map { list expression } { # Create and return a new list where each element of the new list is a result of running the given # expression on the corresponding entry from the original list. # Description of argument(s): # list A list to be operated on. # expression A command expression to be run on each element in the list (e.g. '[string # range $x 1 end]'). foreach x $list { set cmd_buf "lappend new_list ${expression}" eval $cmd_buf } return $new_list } proc list_filter { list expression } { # Create and return a new list consisting of all elements of the original list that do NOT pass the # expression. # Description of argument(s): # list A list to be operated on. # expression A command expression to be run on each element in the list (e.g. 'regexp # -expanded {^[[:blank:]]*\#|^[[:blank:]]*$} $x', 'string equal $x ""', # etc.). set new_list {} foreach x $list { set cmd_buf "set result \[${expression}\]" eval $cmd_buf if { ! $result } { lappend new_list $x } } return $new_list } proc list_filter_comments { list } { # Filter comments from list and return new_list as a result. # Description of argument(s): # list A list to be operated on. set comment_regexp {^[[:blank:]]*\#|^[[:blank:]]*$} set new_list [list_filter $list "regexp -expanded {$comment_regexp} \$x"] return $new_list } proc get_var { var_var { default ""} } { upvar 1 $var_var var_ref # Return the value of the variable expression or the value of default if the variable is not defined. # Example use: # set PATH [get_var ::env(PATH) "/usr/bin"] # Description of argument(s): # var_var The name of a variable (e.g. "::env(NANOSECOND)" or "var1"). # default The default value to return if the variable named in var_var does not # exist. expr { [info exists var_ref] ? [return $var_ref] : [return $default] } } proc set_var_default { var_name { default ""} } { upvar 1 $var_name var_ref # If the variable named in var_name is either blank or non-existent, set its value to the default. # Example use: # set_var_default indent 0 # Description of argument(s): # var_name The name of a variable. # default The default value to assign to the variable if the variable named in # var_name is blank or non-existent. if { ! ([info exists var_ref] && $var_ref != "") } { set var_ref $default } } proc split_path {path dir_path base_name} { upvar $dir_path ref_dir_path upvar $base_name ref_base_name # Split a path into it's dir_path and base_name. The dir_path variable will include a trailing slash. # Description of argument(s): # path The directory or file path. # dir_path The variable to contain the resulting directory path which will include a # trailing slash. # base_name The variable to contain the resulting base directory or file name. set ref_dir_path "[file dirname ${path}]/" set ref_base_name "[file tail $path]" } proc read_properties_file {parm_file_path} { # Read properties files and return key/value pairs as a list. # Description of argument(s): # parm_file_path The path to the properties file. # The properties file must have the following format: # var_name=var_value # Comment lines (those beginning with a "#") and blank lines are allowed and will be ignored. Leading and # trailing single or double quotes will be stripped from the value. E.g. # var1="This one" # Quotes are stripped so the resulting value for var1 is: # This one # Suggestion: The caller can then process the result as an array or a dictionary. # Example usage: # array set properties [read_properties_file $file_path] # print_var properties # With the following result... # properties: # properties(command): string # Or... # set properties [read_properties_file $file_path] # print_dict properties # With the following result... # properties: # properties[command]: string # Initialize properties array. set properties [list] # Read the entire file into a list, filtering comments out. set file_descriptor [open $parm_file_path r] set file_data [list_filter_comments [split [read $file_descriptor] "\n"]] close $file_descriptor foreach line $file_data { # Split = into component parts. set pair [split $line =] lappend properties [lindex ${pair} 0] lappend properties [string trim [lindex ${pair} 1] {"}] } return $properties } proc convert_array_key {key {convert_commands} {prefix ""} } { # Convert the key according to the caller's convert_commands and return the result. # This is designed as a helper procedure to be called by convert_array_keys. # See convert_array_keys for description of arguments. set new_key $key foreach command $convert_commands { if { $command == "prefix" } { regsub -all "^$prefix" $new_key {} new_key set new_key "$prefix$new_key" } elseif { $command == "rm_prefix" } { regsub -all "^$prefix" $new_key {} new_key set new_key "$new_key" } if { $command == "upper" } { set new_key [string toupper $new_key] } elseif { $command == "lower" } { set new_key [string tolower $new_key] } } return $new_key } proc convert_array_keys {source_arr target_arr {convert_commands}\ {prefix ""} } { upvar $source_arr source_arr_ref upvar $target_arr target_arr_ref # Convert the keys of source_arr according to the caller's convert_commands and put the resulting array in # target_arr. If this procedure fails for any reason, it will return non-zero. # Note that despite the name of this procedure, it will also work on a dictionary. In other words, if # source_arr is NOT an array, it will be processed as a dictionary and target_arr will be created as a # dictionary as well. # Description of argument(s): # source_arr The source array that is to be converted. # target_arr The target array that results from the conversion. # convert_commands A list of custom commands that indicate the type of conversion(s) the # caller wishes to see. Currently the accepted values are as follows: # upper Convert key value to uppercase. # lower Convert key value to lowercase. # prefix Prepend prefix to the key, provided that it does not already exist. If # upper or lower is included in convert_commands list, the prefix will be # converted to the specified case as well. # rm_prefix Remove a prefix that is prepended, provided that it exists. # prefix The prefix to be used for "prefix" and "rm_prefix" commands (see # convert_commands text above). # Validate arguments. if { [lsearch $convert_commands lower] != -1 } { if { [lsearch $convert_commands upper] != -1 } { return -code error "Cannot convert to both upper and lower cases." } } if { [lsearch $convert_commands rm_prefix] != -1} { if { [lsearch $convert_commands prefix] != -1} { return -code error "Cannot add and remove a prefix." } } if { [lsearch $convert_commands prefix] != -1 ||\ [lsearch $convert_commands rm_prefix] != -1 } { if { [lsearch $convert_commands upper] != -1 } { set prefix [string toupper $prefix] } elseif { [lsearch $convert_commands lower] != -1 } { set prefix [string tolower $prefix] } } if { [array exists source_arr_ref] } { # Initialize targ array. array set target_arr_ref {} # Walk the source array doing the conversion specified in convert_commands. set search_token [array startsearch source_arr_ref] while {[array anymore source_arr_ref $search_token]} { set key [array nextelement source_arr_ref $search_token] set value $source_arr_ref($key) set new_key [convert_array_key $key $convert_commands $prefix] set cmd_buf "set target_arr_ref($new_key) $value" eval $cmd_buf } array donesearch source_arr_ref $search_token } else { # Initialize targ dictionary. set target_arr_ref [list] # Walk the source dictionary doing the conversion specified in convert_commands. foreach {key value} $source_arr_ref { set new_key [convert_array_key $key $convert_commands $prefix] set cmd_buf "dict append target_arr_ref $new_key \$value" eval $cmd_buf } } } proc expand_shell_string {buffer} { upvar $buffer ref_buffer # Call upon the shell to expand the string in "buffer", i.e. the shell will make substitutions for # environment variables and glob expressions. # Description of argument(s): # buffer The buffer to be expanded. # This is done to keep echo from interpreting all of the double quotes away. regsub -all {\"} $ref_buffer "\\\"" ref_buffer # Bash will compress extra space delimiters if you don't quote the string. So, we quote the argument to # echo. if {[catch {set ref_buffer [exec bash -c "echo \"$ref_buffer\""]} result]} { puts stderr $result exit 1 } } proc add_trailing_string { buffer { add_string "/" } } { upvar $buffer ref_buffer # Add the add string to the end of the buffer if and only if it doesn't already end with the add_string. # Description of argument(s): # buffer The buffer to be modified. # add_string The string to conditionally append to the buffer. regsub -all "${add_string}$" $ref_buffer {} ref_buffer set ref_buffer "${ref_buffer}${add_string}" }