Read a configuration file
You are encouraged to solve this task according to the task description, using any language you may know.
The task is to read a configuration file in standard configuration file, and set variables accordingly. For this task, we have a configuration file as follows:
# This is a configuration file in standard configuration file format # # Lines beginning with a hash or a semicolon are ignored by the application # program. Blank lines are also ignored by the application program. # This is the fullname parameter FULLNAME Foo Barber # This is a favourite fruit FAVOURITEFRUIT banana # This is a boolean that should be set NEEDSPEELING # This boolean is commented out ; SEEDSREMOVED # Configuration option names are not case sensitive, but configuration parameter # data is case sensitive and may be preserved by the application program. # An optional equals sign can be used to separate configuration parameter data # from the option name. This is dropped by the parser. # A configuration option may take multiple parameters separated by commas. # Leading and trailing whitespace around parameter names and parameter data fields # are ignored by the application program. OTHERFAMILY Rhu Barber, Harry Barber
For the task we need to set four variables according to the configuration entries as follows:
- fullname = Foo Barber
- favouritefruit = banana
- needspeeling = true
- seedsremoved = false
We also have an option that contains multiple parameters. These may be stored in an array.
- otherfamily(1) = Rhu Barber
- otherfamily(2) = Harry Barber
See also:
Ada
Uses package config_file_parser described in Ada in Denmak Wiki http://wiki.ada-dk.org/simple_configuration_file_reader_-_using_a_generic <lang Ada>with Ada.Strings.Unbounded; with Config_File_Parser; pragma Elaborate_All (Config_File_Parser);
package Config is
function TUS (S : String) return Ada.Strings.Unbounded.Unbounded_String renames Ada.Strings.Unbounded.To_Unbounded_String; -- Convenience rename. TUS is much shorter than To_Unbounded_String.
type Keys is ( FULLNAME, FAVOURITEFRUIT, NEEDSPEELING, SEEDSREMOVED, OTHERFAMILY); -- These are the valid configuration keys.
type Defaults_Array is array (Keys) of Ada.Strings.Unbounded.Unbounded_String; -- The array type we'll use to hold our default configuration settings.
Defaults_Conf : Defaults_Array := (FULLNAME => TUS ("John Doe"), FAVOURITEFRUIT => TUS ("blackberry"), NEEDSPEELING => TUS ("False"), SEEDSREMOVED => TUS ("False"), OTHERFAMILY => TUS ("Daniel Defoe, Ada Byron")); -- Default values for the Program object. These can be overwritten by -- the contents of the rosetta.cfg file(see below).
package Rosetta_Config is new Config_File_Parser ( Keys => Keys, Defaults_Array => Defaults_Array, Defaults => Defaults_Conf, Config_File => "rosetta.cfg"); -- Instantiate the Config configuration object.
end Config;</lang> Main program : <lang Ada> with Ada.Text_IO; with Config; use Config;
procedure Read_Config is
use Ada.Text_IO; use Rosetta_Config;
begin
New_Line; Put_Line ("Reading Configuration File."); Put_Line ("Fullname := " & Get (Key => FULLNAME)); Put_Line ("Favorite Fruit := " & Get (Key => FAVOURITEFRUIT)); Put_Line ("Other Family := " & Get (Key => OTHERFAMILY)); if Has_Value (Key => NEEDSPEELING) then Put_Line ("NEEDSPEELLING := " & Get (Key => NEEDSPEELING)); else Put_Line ("NEEDSPEELLING := True"); end if; if Has_Value (Key => SEEDSREMOVED) then Put_Line ("SEEDSREMOVED := " & Get (Key => SEEDSREMOVED)); else Put_Line ("SEEDSREMOVED := True"); end if;
end Read_Config;</lang>
output
Reading Configuration File. Fullname := Foo Barber Favorite Fruit := banana Other Family := Rhu Barber, Harry Barber NEEDSPEELLING := True SEEDSREMOVED := False
AutoHotkey
<lang AutoHotkey>
- Author
- AlephX, Aug 18 2011
data = %A_scriptdir%\rosettaconfig.txt comma := ","
Loop, Read, %data% { if NOT (instr(A_LoopReadLine, "#") == 1 OR A_LoopReadLine == "")
{ if instr(A_LoopReadLine, ";") == 1 { parameter := RegExReplace(Substr(A_LoopReadLine,2), "^[ \s]+|[ \s]+$", "") %parameter% = "1" } else { parameter := RegExReplace(A_LoopReadLine, "^[ \s]+|[ \s]+$", "")
if instr(parameter, A_Space) { value := substr(parameter, instr(parameter, A_Space)+1,999) parameter := substr(parameter, 1, instr(parameter, A_Space)-1)
if (instr(value, ",") <> 0) { Loop, Parse, value, %comma% ,%A_Space% %parameter%%A_Index% := A_Loopfield } else %parameter% = %value% } else %parameter% = "0" } } } msgbox, FULLNAME %fullname%`nFAVOURITEFRUIT %FAVOURITEFRUIT%`nNEEDSPEELING %NEEDSPEELING%`nSEEDSREMOVED %SEEDSREMOVED%`nOTHERFAMILY %OTHERFAMILY1% + %OTHERFAMILY2% </lang>
AWK
<lang AWK>
- syntax: GAWK -f READ_A_CONFIGURATION_FILE.AWK
BEGIN {
fullname = favouritefruit = "" needspeeling = seedsremoved = "false" fn = "READ_A_CONFIGURATION_FILE.INI" while (getline rec <fn > 0) { tmp = tolower(rec) if (tmp ~ /^ *fullname/) { fullname = extract(rec) } else if (tmp ~ /^ *favouritefruit/) { favouritefruit = extract(rec) } else if (tmp ~ /^ *needspeeling/) { needspeeling = "true" } else if (tmp ~ /^ *seedsremoved/) { seedsremoved = "true" } else if (tmp ~ /^ *otherfamily/) { split(extract(rec),otherfamily,",") } } close(fn) printf("fullname=%s\n",fullname) printf("favouritefruit=%s\n",favouritefruit) printf("needspeeling=%s\n",needspeeling) printf("seedsremoved=%s\n",seedsremoved) for (i=1; i<=length(otherfamily); i++) { sub(/^ +/,"",otherfamily[i]) # remove leading spaces sub(/ +$/,"",otherfamily[i]) # remove trailing spaces printf("otherfamily(%d)=%s\n",i,otherfamily[i]) } exit(0)
} function extract(rec, pos,str) {
sub(/^ +/,"",rec) # remove leading spaces before parameter name pos = match(rec,/[= ]/) # determine where data begins str = substr(rec,pos) # extract the data gsub(/^[= ]+/,"",str) # remove leading "=" and spaces sub(/ +$/,"",str) # remove trailing spaces return(str)
} </lang>
output:
fullname=Foo Barber favouritefruit=banana needspeeling=true seedsremoved=false otherfamily(1)=Rhu Barber otherfamily(2)=Harry Barber
BBC BASIC
<lang bbcbasic> BOOL = 1
NAME = 2 ARRAY = 3 optfile$ = "options.cfg" fullname$ = FNoption(optfile$, "FULLNAME", NAME) favouritefruit$ = FNoption(optfile$, "FAVOURITEFRUIT", NAME) needspeeling% = FNoption(optfile$, "NEEDSPEELING", BOOL) seedsremoved% = FNoption(optfile$, "SEEDSREMOVED", BOOL) !^otherfamily$() = FNoption(optfile$, "OTHERFAMILY", ARRAY) PRINT "fullname = " fullname$ PRINT "favouritefruit = " favouritefruit$ PRINT "needspeeling = "; : IF needspeeling% PRINT "true" ELSE PRINT "false" PRINT "seedsremoved = "; : IF seedsremoved% PRINT "true" ELSE PRINT "false" PRINT "otherfamily(1) = " otherfamily$(1) PRINT "otherfamily(2) = " otherfamily$(2) END DEF FNoption(file$, key$, type%) LOCAL file%, opt$, comma%, bool%, name$, size%, !^array$() file% = OPENIN(file$) IF file% = 0 THEN = 0 WHILE NOT EOF#file% opt$ = GET$#file% WHILE RIGHT$(opt$) = " " opt$ = LEFT$(opt$) : ENDWHILE IF opt$ = key$ OR LEFT$(opt$, LEN(key$)+1) = key$ + " " THEN opt$ = MID$(opt$, LEN(key$) + 1) WHILE LEFT$(opt$,1) = " " opt$ = MID$(opt$,2) : ENDWHILE CASE type% OF WHEN BOOL: bool% = TRUE : EXIT WHILE WHEN NAME: name$ = opt$ : EXIT WHILE WHEN ARRAY: REPEAT comma% = INSTR(opt$, ",", comma%+1) IF comma% size% += 1 UNTIL comma% = 0 DIM array$(size% + 1) size% = 0 REPEAT comma% = INSTR(opt$, ",") IF comma% THEN size% += 1 array$(size%) = LEFT$(opt$, comma%-1) opt$ = MID$(opt$, comma%+1) WHILE LEFT$(opt$,1) = " " opt$ = MID$(opt$,2) : ENDWHILE ENDIF UNTIL comma% = 0 array$(size% + 1) = opt$ EXIT WHILE ENDCASE ENDIF ENDWHILE CLOSE #file% CASE type% OF WHEN BOOL: = bool% WHEN NAME: = name$ WHEN ARRAY: = !^array$() ENDCASE = 0</lang>
Output:
fullname = Foo Barber favouritefruit = banana needspeeling = true seedsremoved = false otherfamily(1) = Rhu Barber otherfamily(2) = Harry Barber
C++
unoptimized
<lang cpp>#include "stdafx.h"
- include <iostream>
- include <fstream>
- include <vector>
- include <string>
- include <boost/tokenizer.hpp>
- include <boost/algorithm/string/case_conv.hpp>
using namespace std; using namespace boost;
typedef boost::tokenizer<boost::char_separator<char> > Tokenizer; static const char_separator<char> sep(" ","#;,");
//Assume that the config file represent a struct containing all the parameters to load struct configs{ string fullname; string favoritefruit; bool needspelling; bool seedsremoved; vector<string> otherfamily; } conf;
void parseLine(const string &line, configs &conf) { if (line[0] == '#' || line.empty()) return; Tokenizer tokenizer(line, sep); vector<string> tokens; for (Tokenizer::iterator iter = tokenizer.begin(); iter != tokenizer.end(); iter++) tokens.push_back(*iter); if (tokens[0] == ";"){ algorithm::to_lower(tokens[1]); if (tokens[1] == "needspeeling") conf.needspelling = false; if (tokens[1] == "seedsremoved") conf.seedsremoved = false; } algorithm::to_lower(tokens[0]); if (tokens[0] == "needspeeling") conf.needspelling = true; if (tokens[0] == "seedsremoved") conf.seedsremoved = true; if (tokens[0] == "fullname"){ for (unsigned int i=1; i<tokens.size(); i++) conf.fullname += tokens[i] + " "; conf.fullname.erase(conf.fullname.size() -1, 1); } if (tokens[0] == "favouritefruit") for (unsigned int i=1; i<tokens.size(); i++) conf.favoritefruit += tokens[i]; if (tokens[0] == "otherfamily"){ unsigned int i=1; string tmp; while (i<=tokens.size()){ if ( i == tokens.size() || tokens[i] ==","){ tmp.erase(tmp.size()-1, 1); conf.otherfamily.push_back(tmp); tmp = ""; i++; } else{ tmp += tokens[i]; tmp += " "; i++; } } } }
int _tmain(int argc, TCHAR* argv[]) { if (argc != 2) { wstring tmp = argv[0]; wcout << L"Usage: " << tmp << L" <configfile.ini>" << endl; return -1; } ifstream file (argv[1]);
if (file.is_open()) while(file.good()) { char line[255]; file.getline(line, 255); string linestring(line); parseLine(linestring, conf); } else { cout << "Unable to open the file" << endl; return -2; }
cout << "Fullname= " << conf.fullname << endl; cout << "Favorite Fruit= " << conf.favoritefruit << endl; cout << "Need Spelling= " << (conf.needspelling?"True":"False") << endl; cout << "Seed Removed= " << (conf.seedsremoved?"True":"False") << endl; string otherFamily; for (unsigned int i = 0; i < conf.otherfamily.size(); i++) otherFamily += conf.otherfamily[i] + ", "; otherFamily.erase(otherFamily.size()-2, 2); cout << "Other Family= " << otherFamily << endl;
return 0; } </lang>
output
Fullname= Foo Barber Favorite Fruit= banana Need Spelling= True Seed Removed= False Other Family= Rhu Barber, Harry Barber
Solution without Boost libraries. No optimisation. <lang cpp>#include <iostream>
- include <iomanip>
- include <string>
- include <exception>
- include <fstream>
- include <vector>
- include <algorithm>
struct confi { std::string fullname; std::string favouritefruit; bool needspeeling; bool seedsremoved; std::vector<std::string> otherfamily; };
void read_config(std::ifstream& in, confi& out) { in.open("Config.txt"); std::string str; out.needspeeling = false; out.seedsremoved = false; while(!in.eof()) { while(getline(in,str)) { std::string::size_type begin = str.find_first_not_of(" \f\t\v"); //Skips blank lines if(begin == std::string::npos) continue; //Skips # if(std::string("#").find(str[begin]) != std::string::npos) continue; std::string firstWord; try { firstWord = str.substr(0,str.find(" ")); } catch(std::exception& e) { firstWord = str.erase(str.find_first_of(" "),str.find_first_not_of(" ")); } std::transform(firstWord.begin(),firstWord.end(),firstWord.begin(), ::toupper); if(firstWord == "FULLNAME") out.fullname = str.substr(str.find(" ")+1,str.length()); if(firstWord == "FAVOURITEFRUIT") out.favouritefruit = str.substr(str.find(" ")+1,str.length()); if(firstWord == "NEEDSPEELING") out.needspeeling = true; if(firstWord == "SEEDSREMOVED") out.seedsremoved = true; if(firstWord == "OTHERFAMILY") { size_t found = str.find(","); if(found != std::string::npos) { out.otherfamily.push_back(str.substr(str.find_first_of(" ")+1,found-str.find_first_of(" ")-1)); out.otherfamily.push_back(str.substr(found+2,str.length())); } }
} } std::cout << "Full Name: " << out.fullname << std::endl; std::cout << "Favourite Fruit: " << out.favouritefruit << std::endl; std::cout << "Needs peeling?: "; if(out.needspeeling == true) std::cout << "True" << std::endl; else std::cout << "False" << std::endl; std::cout << "Seeds removed?: "; if(out.seedsremoved == true) std::cout << "True" << std::endl; else std::cout << "False" << std::endl; std::cout << "Other family members: " << out.otherfamily[0] << ", " << out.otherfamily[1] << std::endl; } int main() { std::ifstream inp; confi outp; read_config(inp,outp); } </lang>
output
Full Name: Foo Barber Favourite Fruit: banana Needs peeling?: True Seeds removed?: False Other family members: Rhu Barber, Harry Barber
D
<lang d>import std.stdio, std.string, std.conv, std.regex, std.getopt;
enum VarName(alias var) = var.stringof.toUpper;
void setOpt(alias Var)(in string line) {
auto m = match(line, regex(`^` ~ VarName!Var ~ `(\s+(.*))?`)); if (!m.empty) { static if (is(typeof(Var) == string)) Var = m.captures.length > 2 ? m.captures[2] : ""; static if (is(typeof(Var) == bool)) Var = true; static if (is(typeof(Var) == int)) Var = m.captures.length > 2 ? to!int(m.captures[2]) : 0; }
}
void main(in string[] args) {
string fullName, favouriteFruit, otherFamily; bool needsPeeling, seedsRemoved; // Default false. auto f = "readcfg.txt".File;
foreach (line; f.byLine) { auto opt = line.strip.idup; setOpt!fullName(opt); setOpt!favouriteFruit(opt); setOpt!needsPeeling(opt); setOpt!seedsRemoved(opt); setOpt!otherFamily(opt); }
writefln("%14s = %s", VarName!fullName, fullName); writefln("%14s = %s", VarName!favouriteFruit, favouriteFruit); writefln("%14s = %s", VarName!needsPeeling, needsPeeling); writefln("%14s = %s", VarName!seedsRemoved, seedsRemoved); writefln("%14s = %s", VarName!otherFamily, otherFamily);
}</lang>
- Output:
FULLNAME = Foo Barber AVOURITEFRUIT = banana NEEDSPEELING = true SEEDSREMOVED = false OTHERFAMILY = Rhu Barber, Harry Barber
Erlang
<lang Erlang> -module( configuration_file ).
-export( [read/1, task/0] ).
read( File ) ->
{ok, Binary} = file:read_file( File ), Lines = [X || <<First:8, _T/binary>> = X <- binary:split(Binary, <<"\n">>, [global]), First =/= $#, First =/= $;], [option_from_binaries(binary:split(X, <<" ">>)) || X <- Lines].
task() ->
Defaults = [{fullname, "Kalle"}, {favouritefruit, "apple"}, {needspeeling, false}, {seedsremoved, false}], Options = read( "configuration_file" ) ++ Defaults, [io:fwrite("~p = ~p~n", [X, proplists:get_value(X, Options)]) || X <- [fullname, favouritefruit, needspeeling, seedsremoved, otherfamily]].
option_from_binaries( [Option] ) -> {erlang:list_to_atom(string:to_lower(erlang:binary_to_list(Option))), true}; option_from_binaries( [Option, Values] ) -> {erlang:list_to_atom(string:to_lower(erlang:binary_to_list(Option))), option_from_binaries_value(binary:split(Values, <<", ">>))}.
option_from_binaries_value( [Value] ) -> erlang:binary_to_list(Value); option_from_binaries_value( Values ) -> [erlang:binary_to_list(X) || X <- Values]. </lang>
- Output:
50> configuration_file:task(). fullname = "Foo Barber" favouritefruit = "banana" needspeeling = true seedsremoved = false otherfamily = ["Rhu Barber","Harry Barber"]
Fantom
<lang fantom> class Main {
// remove the given key and an optional '=' from start of line Str removeKey (Str key, Str line) { remainder := line[key.size..-1].trim if (remainder.startsWith("=")) { remainder = remainder.replace("=", "").trim } return remainder }
Void main () { // define the variables which need configuring fullname := "" favouritefruit := "" needspeeling := false seedsremoved := false Str[] otherfamily := [,]
// loop through the file, setting variables as needed File(`config.dat`).eachLine |Str line| { line = line.trim if (line.isEmpty || line.startsWith("#") || line.startsWith(";")) { // do nothing for empty and comment lines } else if (line.upper.startsWith("FULLNAME")) { fullname = removeKey("FULLNAME", line) } else if (line.upper.startsWith("FAVOURITEFRUIT")) { favouritefruit = removeKey("FAVOURITEFRUIT", line) } else if (line.upper.startsWith("NEEDSPEELING")) { needspeeling = true } else if (line.upper.startsWith("SEEDSREMOVED")) { seedsremoved = true } else if (line.upper.startsWith("OTHERFAMILY")) { otherfamily = removeKey("OTHERFAMILY", line).split(',') } }
// report results echo ("Full name is $fullname") echo ("Favourite fruit is $favouritefruit") echo ("Needs peeling is $needspeeling") echo ("Seeds removed is $seedsremoved") echo ("Other family is " + otherfamily.join(", ")) }
} </lang>
Fortran
<lang Fortran> program readconfig
implicit none integer, parameter :: strlen = 100 logical :: needspeeling = .false., seedsremoved =.false. character(len=strlen) :: favouritefruit = "", fullname = "", fst, snd character(len=strlen), allocatable :: otherfamily(:), tmp(:) character(len=1000) :: line integer :: lun, stat, j, j0, j1, ii = 1, z integer, parameter :: state_begin=1, state_in_fst=2, state_in_sep=3
open(newunit=lun, file="config.ini", status="old") do read(lun, "(a)", iostat=stat) line if (stat<0) exit if ((line(1:1) == "#") .or. & (line(1:1) == ";") .or. & (len_trim(line)==0)) then cycle end if z = state_begin do j = 1, len_trim(line) if (z == state_begin) then if (line(j:j)/=" ") then j0 = j z = state_in_fst end if elseif (z == state_in_fst) then if (index("= ",line(j:j))>0) then fst = lower(line(j0:j-1)) z = state_in_sep end if elseif (z == state_in_sep) then if (index(" =",line(j:j)) == 0) then snd = line(j:) exit end if else stop "not possible to be here" end if end do if (z == state_in_fst) then fst = lower(line(j0:)) elseif (z == state_begin) then cycle end if
if (fst=="fullname") then read(snd,"(a)") fullname elseif (fst=="favouritefruit") then read(snd,"(a)") favouritefruit elseif (fst=="seedsremoved") then seedsremoved = .true. elseif (fst=="needspeeling") then needspeeling = .true. elseif (fst=="otherfamily") then j = 1; ii = 1 do while (len_trim(snd(j:)) >0) j1 = index(snd(j:),",") if (j1==0) then j1 = len_trim(snd) else j1 = j + j1 - 2 end if do if (j>len_trim(snd)) exit if (snd(j:j) /= " ") exit j = j +1 end do allocate(tmp(ii)) tmp(1:ii-1) = otherfamily call move_alloc(tmp, otherfamily) read(snd(j:j1),"(a)"), otherfamily(ii) j = j1 + 2 ii = ii + 1 end do else print *, "unknown option '"//trim(fst)//"'"; stop end if end do close(lun)
print "(a,a)","fullname = ", trim(fullname) print "(a,a)","favouritefruit = ", trim(favouritefruit) print "(a,l)","needspeeling = ", needspeeling print "(a,l)","seedsremoved = ", seedsremoved print "(a,*(a,:,', '))", "otherfamily = ", & (trim(otherfamily(j)), j=1,size(otherfamily))
contains
pure function lower (str) result (string)
implicit none character(*), intent(In) :: str character(len(str)) :: string Integer :: ic, i
character(26), parameter :: cap = 'ABCDEFGHIJKLMNOPQRSTUVWXYZ' character(26), parameter :: low = 'abcdefghijklmnopqrstuvwxyz'
string = str do i = 1, len_trim(str) ic = index(cap, str(i:i)) if (ic > 0) string(i:i) = low(ic:ic) end do
end function
end program </lang>
Go
This make assumptions about the way the config file is supposed to be structured similar to the ones made by the Python solution. <lang go>package config
import ( "errors" "io" "fmt" "bytes" "strings" "io/ioutil" )
var ( ENONE = errors.New("Requested value does not exist") EBADTYPE = errors.New("Requested type and actual type do not match") EBADVAL = errors.New("Value and type do not match") )
type varError struct { err error n string t VarType }
func (err *varError) Error() string { return fmt.Sprintf("%v: (%q, %v)", err.err, err.n, err.t) }
type VarType int
const ( Bool VarType = 1 + iota Array String )
func (t VarType) String() string { switch t { case Bool: return "Bool" case Array: return "Array" case String: return "String" }
panic("Unknown VarType") }
type confvar struct { Type VarType Val interface{} }
type Config struct { m map[string]confvar }
func Parse(r io.Reader) (c *Config, err error) { c = new(Config) c.m = make(map[string]confvar)
buf, err := ioutil.ReadAll(r) if err != nil { return }
lines := bytes.Split(buf, []byte{'\n'})
for _, line := range lines { line = bytes.TrimSpace(line) if len(line) == 0 { continue } switch line[0] { case '#', ';': continue }
parts := bytes.SplitN(line, []byte{' '}, 2) nam := string(bytes.ToLower(parts[0]))
if len(parts) == 1 { c.m[nam] = confvar{Bool, true} continue }
if strings.Contains(string(parts[1]), ",") { tmpB := bytes.Split(parts[1], []byte{','}) for i := range tmpB { tmpB[i] = bytes.TrimSpace(tmpB[i]) } tmpS := make([]string, 0, len(tmpB)) for i := range tmpB { tmpS = append(tmpS, string(tmpB[i])) }
c.m[nam] = confvar{Array, tmpS} continue }
c.m[nam] = confvar{String, string(bytes.TrimSpace(parts[1]))} }
return }
func (c *Config) Bool(name string) (bool, error) { name = strings.ToLower(name)
if _, ok := c.m[name]; !ok { return false, nil }
if c.m[name].Type != Bool { return false, &varError{EBADTYPE, name, Bool} }
v, ok := c.m[name].Val.(bool) if !ok { return false, &varError{EBADVAL, name, Bool} } return v, nil }
func (c *Config) Array(name string) ([]string, error) { name = strings.ToLower(name)
if _, ok := c.m[name]; !ok { return nil, &varError{ENONE, name, Array} }
if c.m[name].Type != Array { return nil, &varError{EBADTYPE, name, Array} }
v, ok := c.m[name].Val.([]string) if !ok { return nil, &varError{EBADVAL, name, Array} } return v, nil }
func (c *Config) String(name string) (string, error) { name = strings.ToLower(name)
if _, ok := c.m[name]; !ok { return "", &varError{ENONE, name, String} }
if c.m[name].Type != String { return "", &varError{EBADTYPE, name, String} }
v, ok := c.m[name].Val.(string) if !ok { return "", &varError{EBADVAL, name, String} }
return v, nil }</lang>
Usage example: <lang go>package main
import ( "os" "fmt" "config" )
func main() { if len(os.Args) != 2 { fmt.Printf("Usage: %v <configfile>\n", os.Args[0]) os.Exit(1) }
file, err := os.Open(os.Args[1]) if err != nil { fmt.Println(err) os.Exit(1) } defer file.Close()
conf, err := config.Parse(file) if err != nil { fmt.Println(err) os.Exit(1) }
fullname, err := conf.String("fullname") if err != nil { fmt.Println(err) os.Exit(1) }
favouritefruit, err := conf.String("favouritefruit") if err != nil { fmt.Println(err) os.Exit(1) }
needspeeling, err := conf.Bool("needspeeling") if err != nil { fmt.Println(err) os.Exit(1) }
seedsremoved, err := conf.Bool("seedsremoved") if err != nil { fmt.Println(err) os.Exit(1) }
otherfamily, err := conf.Array("otherfamily") if err != nil { fmt.Println(err) os.Exit(1) }
fmt.Printf("FULLNAME: %q\n", fullname) fmt.Printf("FAVOURITEFRUIT: %q\n", favouritefruit) fmt.Printf("NEEDSPEELING: %q\n", needspeeling) fmt.Printf("SEEDSREMOVED: %q\n", seedsremoved) fmt.Printf("OTHERFAMILY: %q\n", otherfamily) }</lang>
Groovy
<lang groovy>def config = [:] def loadConfig = { File file ->
String regex = /^(;{0,1})\s*(\S+)\s*(.*)$/ file.eachLine { line -> (line =~ regex).each { matcher, invert, key, value -> if (key == || key.startsWith("#")) return parts = value ? value.split(/\s*,\s*/) : (invert ? [false] : [true]) if (parts.size() > 1) { parts.eachWithIndex{ part, int i -> config["$key(${i + 1})"] = part} } else { config[key] = parts[0] } } }
}</lang> Testing: <lang groovy>loadConfig new File('config.ini') config.each { println it }</lang> Output:
FULLNAME=Foo Barber FAVOURITEFRUIT=banana NEEDSPEELING=true SEEDSREMOVED=false OTHERFAMILY(1)=Rhu Barber OTHERFAMILY(2)=Harry Barber
Haskell
<lang haskell> import Data.Char import Data.List import Data.List.Split
main :: IO () main = readFile "config" >>= (print . parseConfig)
parseConfig :: String -> Config parseConfig = foldr addConfigValue defaultConfig . clean . lines
where clean = filter (not . flip any ["#", ";", "", " "] . (==) . take 1)
addConfigValue :: String -> Config -> Config addConfigValue raw config = case key of
"fullname" -> config {fullName = values} "favouritefruit" -> config {favoriteFruit = values} "needspeeling" -> config {needsPeeling = True} "seedsremoved" -> config {seedsRemoved = True} "otherfamily" -> config {otherFamily = splitOn "," values} _ -> config where (k, vs) = span (/= ' ') raw key = map toLower k values = tail vs
data Config = Config
{ fullName :: String , favoriteFruit :: String , needsPeeling :: Bool , seedsRemoved :: Bool , otherFamily :: [String] } deriving (Show)
defaultConfig :: Config defaultConfig = Config "" "" False False [] </lang>
Or, use Data.Configfile:
<lang haskell> import Data.ConfigFile import Data.Either.Utils
getSetting cp x = forceEither $ get cp "Default" x
cp <- return . forceEither =<< readfile emptyCP "name_of_configuration_file" let username = getSetting cp "username"
password = getSetting cp "password"
</lang> This works with configuration files in standard format, i.e.,
# this is a comment username = myname
Icon and Unicon
The following works in both languages. However boolean values don't exist in either language. So a variable whose value in other languages is false simply has no value in Icon and Unicon. If it contains any value then it can considered as being equivalent to true:
<lang unicon>procedure main(A)
ws := ' \t' vars := table() every line := !&input do { line ? { tab(many(ws)) if any('#;') | pos(0) then next vars[map(tab(upto(ws)\1|0))] := getValue() } } show(vars)
end
procedure getValue()
ws := ' \t' a := [] while not pos(0) do { tab(many(ws)) put(a, trim(tab(upto(',')|0))) move(1) } return a
end
procedure show(t)
every pair := !sort(t) do { every (s := pair[1]||" = ") ||:= !pair[2] || ", " write(s[1:-2]) }
end</lang>
Sample run on above input:
->rcf <rcf.in favouritefruit = banana fullname = Foo Barber needspeeling otherfamily = Rhu Barber, Harry Barber ->
Note that seedsremoved doesn't exist.
==J==http://rosettacode.org/mw/skins/common/images/button_nowiki.png
<lang j>require'regex' set=:4 :'(x)=:y'
cfgString=:4 :0
y set (1;&,~'(?i:',y,')\s*(.*)') y&set rxapply x
)
cfgBoolean=:4 :0
y set 0 (1;&,~'(?i:',y,')\s*(.*)') y&set rxapply x if.-.0-:y do.y set 1 end.
)
taskCfg=:3 :0
cfg=: ('[#;].*';) rxrplc 1!:1<y cfg cfgString 'fullname' cfg cfgString 'favouritefruit' cfg cfgBoolean 'needspeeling' cfg cfgBoolean 'seedsremoved' i.0 0
)</lang>
Example use:
<lang j> taskCfg 'fruit.conf'
(,' = ',]&.do)&>;: 'fullname favouritefruit needspeeling seedsremoved'
fullname = Foo Barber favouritefruit = banana needspeeling = 1 seedsremoved = 0 </lang>
JavaScript
In JavaScript using an object makes more sense than local variables. This function takes our config file in plain text as the parameter.
<lang javascript>function parseConfig(config) {
// this expression matches a line starting with an all capital word, // and anything after it var regex = /^([A-Z]+)(.*)$/mg; var configObject = {}; // loop until regex.exec returns null var match; while (match = regex.exec(config)) { // values will typically be an array with one element // unless we want an array // match[0] is the whole match, match[1] is the first group (all caps word), // and match[2] is the second (everything through the end of line) var key = match[1], values = match[2].split(","); if (values.length === 1) { configObject[key] = values[0]; } else { configObject[key] = values.map(function(value){ return value.trim(); }); } } return configObject;
} </lang>
The result is an object, which can be represented with this JSON.
<lang javascript>{
"FULLNAME": " Foo Barber", "FAVOURITEFRUIT": " banana", "NEEDSPEELING": "", "OTHERFAMILY": [ "Rhu Barber", "Harry Barber" ]
} </lang>
Java
A more natural way to do this in Java would be Properties.load(InputStream) but the example data is not in the format expected by that method (no equals signs). <lang Java>import java.io.BufferedReader; import java.io.FileReader; import java.io.IOException; import java.util.Arrays; import java.util.HashMap; import java.util.Map; import java.util.regex.Matcher; import java.util.regex.Pattern;
public class ConfigReader {
private static final Pattern LINE_PATTERN = Pattern.compile( "([^ ]+) ?(.*)" );
public static void main( final String[] args ) { System.out.println( parseFile( args[ 0 ] ) ); }
public static Map<String, ? extends Object> parseFile( final String fileName ) { final Map<String, Object> result = new HashMap<String, Object>(); /*v*/ BufferedReader reader = null;
try { reader = new BufferedReader( new FileReader( fileName ) ); for ( String line; null != ( line = reader.readLine() ); ) { parseLine( line, result ); } } catch ( final IOException x ) { throw new RuntimeException( "Oops: " + x, x ); } finally { if ( null != reader ) try { reader.close(); } catch ( final IOException x2 ) { System.err.println( "Could not close " + fileName + " - " + x2 ); } }
return result; }
private static void parseLine( final String line, final Map<String, Object> map ) { if ( "".equals( line.trim() ) || line.startsWith( "#" ) || line.startsWith( ";" ) ) return;
final Matcher matcher = LINE_PATTERN.matcher( line );
if ( ! matcher.matches() ) { System.err.println( "Bad config line: " + line ); return; }
final String key = matcher.group( 1 ).trim().toUpperCase(); final String value = matcher.group( 2 ).trim();
if ( "".equals( value ) ) { map.put( key, true ); } else if ( -1 == value.indexOf( ',' ) ) { map.put( key, value ); } else { final String[] values = value.split( "," );
for ( int i = 0; i < values.length; i++ ) { values[ i ] = values[ i ].trim(); } map.put( key, Arrays.asList( values ) ); } }
}</lang>
The output is:
{NEEDSPEELING=true, FULLNAME=Foo Barber, FAVOURITEFRUIT=banana, OTHERFAMILY=[Rhu Barber, Harry Barber]}
Lasso
<lang Lasso>local(config = '# This is a configuration file in standard configuration file format
- Lines beginning with a hash or a semicolon are ignored by the application
- program. Blank lines are also ignored by the application program.
- This is the fullname parameter
FULLNAME Foo Barber
- This is a favourite fruit
FAVOURITEFRUIT = banana
- This is a boolean that should be set
NEEDSPEELING
- This boolean is commented out
- SEEDSREMOVED
- Configuration option names are not case sensitive, but configuration parameter
- data is case sensitive and may be preserved by the application program.
- An optional equals sign can be used to separate configuration parameter data
- from the option name. This is dropped by the parser.
- A configuration option may take multiple parameters separated by commas.
- Leading and trailing whitespace around parameter names and parameter data fields
- are ignored by the application program.
OTHERFAMILY Rhu Barber, Harry Barber ') // if config is in a file collect it like this //local(config = file('path/and/file.name') -> readstring)
define getconfig(term::string, config::string) => {
local( regexp = regexp(-find = `(?m)^` + #term + `($|\s*=\s*|\s+)(.*)$`, -input = #config, -ignorecase), result )
while(#regexp -> find) => { #result = (#regexp -> groupcount > 1 ? (#regexp -> matchString(2) -> trim& || true)) if(#result -> asstring >> ',') => { #result = #result -> split(',') #result -> foreach => {#1 -> trim} } return #result } return false
}
local( fullname = getconfig('FULLNAME', #config), favorite = getconfig('FAVOURITEFRUIT', #config), sedsremoved = getconfig('SEEDSREMOVED', #config), needspeel = getconfig('NEEDSPEELING', #config), otherfamily = getconfig('OTHERFAMILY', #config) )
- fullname
'
'
- favorite
'
'
- sedsremoved
'
'
- needspeel
'
'
- otherfamily
'
'</lang>
Output:
Foo Barber banana false true array(Rhu Barber, Harry Barber)
Liberty BASIC
<lang lb> dim confKeys$(100) dim confValues$(100)
optionCount = ParseConfiguration("a.txt")
fullName$ = GetOption$( "FULLNAME", optionCount) favouriteFruit$ = GetOption$( "FAVOURITEFRUIT", optionCount) needsPeeling = HasOption("NEEDSPEELING", optionCount) seedsRemoved = HasOption("SEEDSREMOVED", optionCount) otherFamily$ = GetOption$( "OTHERFAMILY", optionCount) 'it's easier to keep the comma-separated list as a string
print "Full name: "; fullName$ print "likes: "; favouriteFruit$ print "needs peeling: "; needsPeeling print "seeds removed: "; seedsRemoved
print "other family:" otherFamily$ = GetOption$( "OTHERFAMILY", optionCount) counter = 1 while word$(otherFamily$, counter, ",") <> ""
print counter; ". "; trim$(word$(otherFamily$, counter, ",")) counter = counter + 1
wend end
'parses the configuration file, stores the uppercase keys in array confKeys$ and corresponding values in confValues$ 'returns the number of key-value pairs found function ParseConfiguration(fileName$)
count = 0 open fileName$ for input as #f while not(eof(#f)) line input #f, s$ if not(Left$(s$,1) = "#" or Left$( s$,1) = ";" or trim$(s$) = "") then 'ignore empty and comment lines s$ = trim$(s$) key$ = ParseKey$(s$) value$ = trim$(Mid$(s$,len(key$) + 1)) if Left$( value$,1) = "=" then value$ = trim$(Mid$(value$,2)) 'optional = count = count + 1 confKeys$(count) = upper$(key$) confValues$(count) = value$ end if wend close #f ParseConfiguration = count
end function
function ParseKey$(s$)
'key is the first word in s$, delimited by whitespace or = s$ = word$(s$, 1) ParseKey$ = trim$(word$(s$, 1, "="))
end function
function GetOption$( key$, optionCount)
index = Find.confKeys( 1, optionCount, key$) if index > 0 then GetOption$ =(confValues$(index))
end function
function HasOption(key$, optionCount)
HasOption = Find.confKeys( 1, optionCount, key$) > 0
end function
function Find.confKeys( Start, Finish, value$)
Find.confKeys = -1 for i = Start to Finish if confKeys$(i) = value$ then Find.confKeys = i : exit for next i
end function </lang>
Output:
Full name: Foo Barber likes: banana needs peeling: 1 seeds removed: 0 other family: 1. Rhu Barber 2. Harry Barber
Lua
<lang lua>conf = {}
fp = io.open( "conf.txt", "r" )
for line in fp:lines() do
line = line:match( "%s*(.+)" ) if line and line:sub( 1, 1 ) ~= "#" and line:sub( 1, 1 ) ~= ";" then option = line:match( "%S+" ):lower()
value = line:match( "%S*%s*(.*)" )
if not value then
conf[option] = true
else if not value:find( "," ) then conf[option] = value else value = value .. "," conf[option] = {} for entry in value:gmatch( "%s*(.-)," ) do conf[option][#conf[option]+1] = entry end end end
end
end
fp:close()
print( "fullname = ", conf["fullname"] )
print( "favouritefruit = ", conf["favouritefruit"] )
if conf["needspeeling"] then print( "needspeeling = true" ) else print( "needspeeling = false" ) end
if conf["seedsremoved"] then print( "seedsremoved = true" ) else print( "seedsremoved = false" ) end
if conf["otherfamily"] then
print "otherfamily:" for _, entry in pairs( conf["otherfamily"] ) do
print( "", entry )
end
end</lang>
MATLAB / Octave
This is defined as a function, parameters are returned as part of a struct. When the first line, and the assignment to return values are removed, it is a script that stores the parameters in the local workspace.
<lang MATLAB>function R = readconf(configfile) % READCONF reads configuration file. % % The value of boolean parameters can be tested with % exist(parameter,'var')
if nargin<1,
configfile = 'q.conf';
end;
fid = fopen(configfile); if fid<0, error('cannot open file %s\n',a); end;
while ~feof(fid)
line = strtrim(fgetl(fid)); if isempty(line) || all(isspace(line)) || strncmp(line,'#',1) || strncmp(line,';',1),
; % no operation
else
[var,tok] = strtok(line,' \t='); var = upper(var); if any(tok==','), k = 1; while (1) [val, tok]=strtok(tok,','); R.(var){k} = strtrim(val); % return value of function eval(sprintf('%s{%i}=%s;',var,k,strtrim(val))); % stores variable in local workspace if isempty(tok), break; end; k=k+1; end; else tok = strtrim(tok); R.(var) = tok; % return value of function eval(sprintf('%s=%s; ',var,tok)); % stores variable in local workspace end;
end;
end; fclose(fid); whos, % shows the parameter in the local workspace
</lang>
R=readconf('file.conf') Variables in the current scope: Attr Name Size Bytes Class ==== ==== ==== ===== ===== FAVOURITEFRUIT 1x6 6 char FULLNAME 1x10 10 char NEEDSPEELING 0x0 0 char OTHERFAMILY 1x2 22 cell f R 1x1 38 struct f configfile 1x6 6 char fid 1x1 8 double k 1x1 8 double line 0x0 0 char tok 0x0 0 char val 1x13 13 char var 1x11 11 char Total is 51 elements using 122 bytes R = scalar structure containing the fields: FULLNAME = Foo Barber FAVOURITEFRUIT = banana NEEDSPEELING = OTHERFAMILY = { [1,1] = Rhu Barber [1,2] = Harry Barber }
OCaml
Using the library ocaml-inifiles:
<lang ocaml>#use "topfind"
- require "inifiles"
open Inifiles
let print_field ini (label, field) =
try let v = ini#getval "params" field in Printf.printf "%s: %s\n" label v with Invalid_element _ -> Printf.printf "%s: not defined\n" label
let () =
let ini = new inifile "./conf.ini" in let lst = [ "Full name", "FULLNAME"; "likes", "FAVOURITEFRUIT"; "needs peeling", "NEEDSPEELING"; "seeds removed", "SEEDSREMOVED"; ] in List.iter (print_field ini) lst;
let v = ini#getaval "params" "OTHERFAMILY" in print_endline "other family:"; List.iter (Printf.printf "- %s\n") v;
- </lang>
The file "conf.ini":
# This is a configuration file # # Lines begininning with a hash are ignored # Blank lines are also ignored # Leading and trailing whitespace around parameter names and # parameter data fields are ignored # ocaml-inifiles needs at least one section name [params] # This is the fullname parameter FULLNAME = Foo Barber # This is a favourite fruit FAVOURITEFRUIT = banana # This is a boolean that should be set NEEDSPEELING = true # This boolean is commented out #SEEDSREMOVED = false # A configuration option may take multiple values # these values will be returned as a list OTHERFAMILY = Rhu Barber OTHERFAMILY = Harry Barber
Output:
$ ocaml conf.ml Full name: Foo Barber likes: banana needs peeling: true seeds removed: not defined other family: - Harry Barber - Rhu Barber
ooRexx
Here's another way of doing this, which stores the values in a Rexx stem (array), and stores each value of a multivalued variable as a separate item: <lang ooRexx>
- !/usr/bin/rexx
/*.----------------------------------------------------------------------.*/ /*|readconfig: Read keyword value pairs from a configuration file into |*/ /*| Rexx variables. |*/ /*| |*/ /*|Usage: |*/ /*| .-~/rosetta.conf-. |*/ /*|>>-readconfig-+----------------+------------------------------------><|*/ /*| |-configfilename-| |*/ /*| |-- -? ----------| |*/ /*| |-- -h ----------| |*/ /*| '- --help -------' |*/ /*| |*/ /*|where |*/ /*| configfilename |*/ /*| is the name of the configuration file to be processed. if not|*/ /*| specified, ~/rosetta.conf is used. |*/ /*| |*/ /*|All values retrieved from the configuration file are stored in |*/ /*|compound variables with the stem config. Variables with multiple |*/ /*|values have a numeric index appended, and the highest index number |*/ /*|is stored in the variable with index 0; e.g. if CONFIG.OTHERFAMILY.1 |*/ /*|and CONFIG.OTHERFAMILY.2 have values assigned, CONFIG.OTHERFAMILY.0 = |*/ /*|2. |*/ /*|-?, -h or --help all cause this documentation to be displayed. |*/ /*| |*/ /*|This program was tested using Open Object Rexx 4.1.1. It should work |*/ /*|with most other dialects as well. |*/ /*'----------------------------------------------------------------------'*/
call usage arg(1) trace normal signal on any name error
/* Prepare for processing the configuration file. */
keywords = 'FULLNAME FAVOURITEFRUIT NEEDSPEELING SEEDSREMOVED OTHERFAMILY'
/* Set default values for configuration variables here */
config_single?. = 1 config. = config.NEEDSPEELING = 0 config.SEEDSREMOVED = 1
/* Validate command line inputs. */
parse arg configfilename
if length(configfilename) = 0 then configfilename = '~/rosetta.conf'
configfile = stream(configfilename,'COMMAND','QUERY EXISTS')
if length(configfile) = 0 then do say configfilename 'was not found.' exit 28 end
signal on notready /* if an I/O error occurs. */
/* Open the configuration file. */
response = stream(configfile,'COMMAND','OPEN READ SHARED')
/* Parse the contents of the configuration file into variables. */
do while lines(configfile) > 0 statement = linein(configfile)
select when left(statement,1) = '#', | left(statement,1) = ';', | length(strip(statement)) = 0, then /* a comment or a blank line. */ nop /* skip it. */
otherwise do if pos('=',word(statement,1)) > 0, | left(word(statement,2),1) = '=', then /* a keyword/value pair with = between. */ parse var statement keyword '=' value
else /* a keyword/value pair with no =. */ parse var statement keyword value
keyword = translate(strip(keyword)) /* make it uppercase. */ single? = pos(',',value) = 0 /* a single value, or multiple values? */ call value 'CONFIG_single?.'keyword,single? /* remember. */
if single? then do if length(value) > 0 then call value 'CONFIG.'keyword,strip(value) end /* strip keeps internal whitespace only. */
else /* store each value with its index. */ do v = 1 by 1 while length(value) > 0 parse var value value1 ',' value
if length(value1) > 0 then do call value 'CONFIG.'keyword'.'v,strip(value1) call value 'CONFIG.'keyword'.0',v /* remember this. */ end end end end end
/* Display the values of the configuration variables. */
say 'Values associated with configuration file' configfilename':' say
do while words(keywords) > 0 parse var keywords keyword keywords
if value('CONFIG_single?.'keyword) then say right(keyword,20) '= "'value('CONFIG.'keyword)'"'
else do lastv = value('CONFIG.'keyword'.0')
do v = 1 to lastv say right(keyword,20-(length(v)+2))'['v'] = "'value('CONFIG.'keyword'.'v)'"' end end end
say
notready: /* I/O errors come here. */
filestatus = stream(configfile,'STATE')
if filestatus \= 'READY' then say 'An I/O error occurred; the file status is' filestatus'.'
response = stream(configfile,'COMMAND','CLOSE')
error: /*? = sysdumpvariables() */ /* see everything Rexx used. */ exit
usage: procedure
trace normal
if arg(1) = '-h', | arg(1) = '-?', | arg(1) = '--help' then do line = '/*|' say
do l = 3 by 1 while line~left(3) = '/*|' line = sourceline(l) parse var line . '/*|' text '|*/' . say text end
say exit 0 end
return
</lang>
It produces the following output:
$ readconfig.rex Values associated with configuration file ~/rosetta.conf: FULLNAME = "Foo Barber" FAVOURITEFRUIT = "banana" NEEDSPEELING = "0" SEEDSREMOVED = "1" OTHERFAMILY[1] = "Rhu Barber" OTHERFAMILY[2] = "Harry Barber"
Leslie
Pascal
This solution makes use of FCL-STL package shipped with FPC >= 2.6.0, moreover it can be run directly like a script (just chmod +x) using the new instantfpc feature.
<lang Pascal>#!/usr/bin/instantfpc
{$if not defined(fpc) or (fpc_fullversion < 20600)}
{$error FPC 2.6.0 or greater required}
{$endif}
{$mode objfpc}{$H+}
uses
Classes,SysUtils,gvector,ghashmap;
type
TStrHashCaseInsensitive = class class function hash(s: String; n: Integer): Integer; end;
class function TStrHashCaseInsensitive.hash(s: String; n: Integer): Integer; var
x: Integer; c: Char;
begin
x := 0; for c in UpCase(s) do Inc(x,Ord(c)); Result := x mod n;
end;
type
TConfigValues = specialize TVector<String>; TConfigStorage = class(specialize THashMap<String,TConfigValues,TStrHashCaseInsensitive>) destructor Destroy; override; end;
destructor TConfigStorage.Destroy; var
It: TIterator;
begin
if Size > 0 then begin It := Iterator; repeat It.Value.Free; until not It.Next; It.Free; end; inherited Destroy;
end;
var
ConfigStrings,ConfigValues: TStrings; ConfigStorage: TConfigStorage; ConfigLine,ConfigName,ConfigValue: String; SeparatorPos: Integer;
begin
ConfigStrings := TStringList.Create; ConfigValues := TStringList.Create; ConfigValues.Delimiter := ','; ConfigValues.StrictDelimiter := true; ConfigStorage := TConfigStorage.Create;
ConfigStrings.LoadFromFile('config.test'); for ConfigLine in ConfigStrings do begin if Length(ConfigLine) > 0 then begin case ConfigLine[1] of '#',';': ; // ignore else begin // look for = first SeparatorPos := Pos('=',ConfigLine); // if not found, then look for space if SeparatorPos = 0 then begin SeparatorPos := Pos(' ',ConfigLine); end; // found space if SeparatorPos <> 0 then begin ConfigName := UpCase(Copy(ConfigLine,1,SeparatorPos - 1)); ConfigValues.DelimitedText := Copy(ConfigLine,SeparatorPos + 1,Length(ConfigLine) - SeparatorPos); // no = or space found, take the whole line as a key name end else begin ConfigName := UpCase(Trim(ConfigLine)); end; if not ConfigStorage.Contains(ConfigName) then begin ConfigStorage[ConfigName] := TConfigValues.Create; end; for ConfigValue in ConfigValues do begin ConfigStorage[ConfigName].PushBack(Trim(ConfigValue)); end; end; end; end; end;
WriteLn('FULLNAME = ' + ConfigStorage['FULLNAME'][0]); WriteLn('FAVOURITEFRUIT = ' + ConfigStorage['FAVOURITEFRUIT'][0]); WriteLn('NEEDSPEELING = ' + BoolToStr(ConfigStorage.Contains('NEEDSPEELING'),true)); WriteLn('SEEDSREMOVED = ' + BoolToStr(ConfigStorage.Contains('SEEDSREMOVED'),true)); WriteLn('OTHERFAMILY(1) = ' + ConfigStorage['OTHERFAMILY'][0]); WriteLn('OTHERFAMILY(2) = ' + ConfigStorage['OTHERFAMILY'][1]);
ConfigStorage.Free; ConfigValues.Free; ConfigStrings.Free;
end.</lang>
Output:
FULLNAME = Foo Barber FAVOURITEFRUIT = banana NEEDSPEELING = True SEEDSREMOVED = False OTHERFAMILY(1) = Rhu Barber OTHERFAMILY(2) = Harry Barber
Perl
This is an all-singing, all-dancing version that checks the configuration file syntax and contents and raises exceptions if it fails. (It is intentionally over-commented for pedagogical purposes.)
<lang perl>my $fullname; my $favouritefruit; my $needspeeling; my $seedsremoved; my @otherfamily;
- configuration file definition. See read_conf_file below for explanation.
my $conf_definition = {
'fullname' => [ 'string', \$fullname ], 'favouritefruit' => [ 'string', \$favouritefruit ], 'needspeeling' => [ 'boolean', \$needspeeling ], 'seedsremoved' => [ 'boolean', \$seedsremoved ], 'otherfamily' => [ 'array', \@otherfamily ],
};
my $arg = shift; # take the configuration file name from the command line
# (or first subroutine argument if this were in a sub)
my $file; # this is going to be a file handle reference open $file, $arg or die "Can't open configuration file '$arg': $!";
read_conf_file($file, $conf_definition);
print "fullname = $fullname\n"; print "favouritefruit = $favouritefruit\n"; print "needspeeling = ", ($needspeeling ? 'true' : 'false'), "\n"; print "seedsremoved = ", ($seedsremoved ? 'true' : 'false'), "\n"; for (my $i = 0; $i < @otherfamily; ++$i) {
print "otherfamily(", $i + 1, ") = ", $otherfamily[$i], "\n";
}
- read_conf_file: Given a file handle opened for reading and a configuration definition,
- read the file.
- If the configuration file doesn't match the definition, raise an exception with "die".
- The configuration definition is (a reference to) an associative array
- where the keys are the configuration variable names in all lower case
- and the values are references to arrays.
- The first element of each of these arrays is the expected type: 'boolean', 'string', or 'array';
- the second element is a reference to the variable that should be assigned the data.
sub read_conf_file {
my ($fh, $def) = @_; # copy parameters
local $_; # avoid interfering with use of $_ in main program while (<$fh>) { # read a line from $fh into $_ until end of file next if /^#/; # skip "#" comments next if /^;/; # skip ";" comments next if /^$/; # skip blank lines chomp; # strip final newline
$_ =~ /^\s*(\w+)\s*(.*)$/i or die "Syntax error"; my $key = $1; my $rest = $2; $key =~ tr/[A-Z]/[a-z]/; # convert keyword to lower case
if (!exists $def->{$key}) { die "Unknown keyword: '$key'"; }
if ($def->{$key}[0] eq 'boolean') { if ($rest) { die "Syntax error: extra data following boolean '$key'"; } ${$def->{$key}[1]} = 1; next; # done with this line, go back to "while" }
$rest =~ s/\s*$//; # drop trailing whitespace $rest =~ s/^=\s*//; # drop equals sign if present
if ($def->{$key}[0] eq 'string') { ${$def->{$key}[1]} = $rest; } elsif ($def->{$key}[0] eq 'array') { @{$def->{$key}[1]} = split /\s*,\s*/, $rest; } else { die "Internal error (unknown type in configuration definition)"; } }
} </lang>
Perl 6
This demonstrates several interesting features of Perl 6, including full grammar support, derived grammars, alternation split across derivations, and longest-token matching that works across derivations. It also shows off Perl 6's greatly cleaned up regex syntax. <lang perl6>my $fullname; my $favouritefruit; my $needspeeling = False; my $seedsremoved = False; my @otherfamily;
grammar ConfFile {
token TOP {
:my $*linenum = 0; ^ <fullline>* [$ || (\N*) { die "Parse failed at $0" } ]
}
token fullline {
<?before .> { ++$*linenum } <line> [ \n || { die "Parse failed at line $*linenum" } ]
}
proto token line()
token line:misc { {} (\S+) { die "Unrecognized word: $0" } }
token line:sym<comment> { ^^ [ ';' | '#' ] \N* } token line:sym<blank> { ^^ \h* $$ }
token line:sym<fullname> {:i fullname» <rest> { $fullname = $<rest>[0].trim } } token line:sym<favouritefruit> {:i favouritefruit» <rest> { $favouritefruit = $<rest>[0].trim } } token line:sym<needspeeling> {:i needspeeling» <yes> { $needspeeling = defined $<yes>[0] } } token rest { \h* '='? (\N*) } token yes { :i \h* '='? \h* [
|| ([yes|true|1]) || [no|false|0] || (<?>) ] \h*
}
}
grammar MyConfFile is ConfFile {
token line:sym<otherfamily> {:i otherfamily» <many> { @otherfamily = $<many>[0]».trim} } token many { \h*'='? ([ <![,]> \N ]*) ** ',' }
}
MyConfFile.parsefile('file.cfg');
.perl.say for
:$fullname, :$favouritefruit, :$needspeeling, :$seedsremoved, :@otherfamily;</lang>
Output:
"fullname" => "Foo Barber" "favouritefruit" => "banana" "needspeeling" => Bool::True "seedsremoved" => Bool::False "otherfamily" => ["Rhu Barber", "Harry Barber"]
PicoLisp
'read' supports only a single comment character. Therefore, we use a pipe to filter the comments. <lang PicoLisp>(de rdConf (File)
(pipe (in File (while (echo "#" ";") (till "^J"))) (while (read) (skip) (set @ (or (line T) T)) ) ) )</lang>
Test: <lang PicoLisp>(off FULLNAME FAVOURITEFRUIT NEEDSPEELING SEEDSREMOVED OTHERFAMILY) (rdConf "conf.txt")</lang> Output:
: (list FULLNAME FAVOURITEFRUIT NEEDSPEELING SEEDSREMOVED OTHERFAMILY) -> ("Foo Barber" "banana" T NIL "Rhu Barber, Harry Barber")
PL/I
<lang PL/I> set: procedure options (main);
declare text character (100) varying; declare (fullname, favouritefruit) character (100) varying initial (); declare needspeeling bit (1) static initial ('0'b); declare seedsremoved bit (1) static initial ('0'b); declare otherfamily(10) character (100) varying; declare (i, j) fixed binary; declare in file;
open file (in) title ( '/RD-CON.DAT,TYPE(TEXT),RECSIZE(200)' );
on endfile (in) go to variables;
otherfamily = ; j = 0;
do forever; get file (in) edit (text) (L); text = trim(text);
if length(text) = 0 then iterate; if substr(text, 1, 1) = ';' then iterate; if substr(text, 1, 1) = '#' then iterate; if length(text) >= 9 then if substr(text, 1, 9) = 'FULLNAME ' then fullname = trim( substr(text, 9) ); if length(text) >= 15 then if substr(text, 1, 15) = 'FAVOURITEFRUIT ' then favouritefruit = trim( substr(text, 15) ); if length(text) >= 12 then if text = 'NEEDSPEELING' then needspeeling = '1'b; if length(text) >= 12 then if text = 'SEEDSREMOVED' then seedsremoved = '1'b; if length(text) >= 12 then if substr(text, 1, 12) = 'OTHERFAMILY ' then do; text = trim(substr(text, 12) ); i = index(text, ','); do while (i > 0); j = j + 1; otherfamily(j) = substr(text, 1, i-1); text = trim(substr(text, i+1)); i = index(text, ','); end; j = j + 1; otherfamily(j) = trim(text); end; end;
variables:
put skip data (fullname); put skip data (favouritefruit); put skip data (needspeeling); put skip data (seedsremoved); do i = 1 to j; put skip data (otherfamily(i)); end;
end set; </lang> Output, using the given text file as input:-
FULLNAME='Foo Barber'; FAVOURITEFRUIT='banana'; NEEDSPEELING='1'B; SEEDSREMOVED='0'B; OTHERFAMILY(1)='Rhu Barber'; OTHERFAMILY(2)='Harry Barber';
Protium
Despite the discussion, this task is still a bit ambiguous. I've taken it that
* blank lines and lines beginning with # and ; should be ignored. * an all uppercase word defines a configuration symbol * no tail means a boolean true * a tail including a comma means a list * any other kind of tail is a string
What we end up with after processing rosetta.config are three VARs and a LST, named FAVOURITEFRUIT, FULLNAME, NEEDSPEELING and OTHERFAMILY respectively.
<lang html><@ DEFUDRLIT>__ReadConfigurationFile| <@ LETSCPPNTPARSRC>Data|1</@><@ OMT> read file into locally scope variable</@> <@ LETCGDLSTLLOSCP>List|Data</@><@ OMT> split Data into a list of lines </@> <@ OMT> Remove comment lines, and blank lines </@> <@ ACTOVRBEFLSTLIT>List|;</@> <@ ACTOVRBEFLSTLIT>List|#</@> <@ ACTRMELST>List</@> <@ OMT> Iterate over the lines of the list </@> <@ ITEENULSTLit>List| <@ LETVARUPTVALLSTLIT>key|...| </@> <@ LETVARAFTVALLSTLIT>val|...| </@> <@ OMT> test for an empty key (in the case of a boolean) </@> <@ TSTVARLIT>key|</@> <@ IFE><@ LETPNTVARVARLIT>val|__True</@></@> <@ ELS> <@ TSTGT0ATBVARLIT>val|,</@> <@ IFE><@ ACTEXEEMMCAP><&prot; LETCNDLSTLITLIT>&key;&pipe;&val;&pipe;, </&prot;></@></@> <@ ELS><@ LETPNTVARVARVAR>key|val</@></@> </@> </@> </@>
<@ ACTUDRLIT>__ReadConfigurationFile|c:\rosetta.config</@> <@ SAYVAR>FAVOURITEFRUIT</@> <@ SAYVAR>FULLNAME</@> <@ SAYVAR>NEEDSPEELING</@> <@ SAYDMPLST>OTHERFAMILY</@> </lang>
Python
This task is not well-defined, so we have to make some assumptions (see comments in code). <lang python>def readconf(fn):
ret = {} with file(fn) as fp: for line in fp: # Assume whitespace is ignorable line = line.strip() if not line or line.startswith('#'): continue boolval = True # Assume leading ";" means a false boolean if line.startswith(';'): # Remove one or more leading semicolons line = line.lstrip(';') # If more than just one word, not a valid boolean if len(line.split()) != 1: continue boolval = False bits = line.split(None, 1) if len(bits) == 1: # Assume booleans are just one standalone word k = bits[0] v = boolval else: # Assume more than one word is a string value k, v = bits ret[k.lower()] = v return ret
if __name__ == '__main__':
import sys conf = readconf(sys.argv[1]) for k, v in sorted(conf.items()): print k, '=', v</lang>
Racket
Use the shared Racket/Options code.
<lang Racket>
- lang racket
(require "options.rkt")
(read-options "options-file") (define-options fullname favouritefruit needspeeling seedsremoved otherfamily) (printf "fullname = ~s\n" fullname) (printf "favouritefruit = ~s\n" favouritefruit) (printf "needspeeling = ~s\n" needspeeling) (printf "seedsremoved = ~s\n" seedsremoved) (printf "otherfamily = ~s\n" otherfamily) </lang>
Output:
fullname = "Foo Barber" favouritefruit = "banana" needspeeling = #t seedsremoved = #f otherfamily = ("Rhu Barber" "Harry Barber")
REXX
No assumptions were made about what variables are (or aren't) in the configuration file.
Code was written to make the postmortem report as readable as possible.
<lang rexx>/*REXX program to read a config file and assign VARs as found within. */
signal on syntax; signal on novalue /*handle REXX program errors. */
parse arg cFID _ . /*cFID = config file to be read. */
if cFID== then cFID='CONFIG.DAT' /*Not specified? Use the default*/
bad= /*this will contain all bad VARs.*/
varList= /*this will contain all the VARs.*/
maxLenV=0; blanks=0; hashes=0; semics=0; badVar=0 /*zero 'em.*/
do j=0 while lines(cFID)\==0 /*J count's the file's lines. */ txt=strip(linein(cFID)) /*read a line (record) from file,*/ /*& strip leading/trailing blanks*/ if txt = then do; blanks=blanks+1; iterate; end if left(txt,1)=='#' then do; hashes=hashes+1; iterate; end if left(txt,1)==';' then do; semics=semics+1; iterate; end eqS=pos('=',txt) /*can't use the TRANSLATE bif. */ if eqS\==0 then txt=overlay(' ',txt,eqS) /*replace 1st '=' with blank*/ parse var txt xxx value; upper xxx /*get the variableName and value.*/ value=strip(value) /*strip leading & trailing blanks*/ if value= then value='true' /*if no value, then use "true". */ if symbol(xxx)=='BAD' then do /*can REXX use the variable name?*/ badVar=badVar+1; bad=bad xxx; iterate end varList=varList xxx /*add it to the list of variables*/ call value xxx,value /*now, use VALUE to set the var. */ maxLenV=max(maxLenV,length(value)) /*maxLen of varNames, pretty disp*/ end /*j*/
vars=words(varList)
say #(j) 'record's(j) "were read from file: " cFID
if blanks\==0 then say #(blanks) 'blank record's(blanks) "were read." if hashes\==0 then say #(hashes) 'record's(hashes) "ignored that began with a # (hash)." if semics\==0 then say #(semics) 'record's(semics) "ignored that began with a ; (semicolon)." if badVar\==0 then say #(badVar) 'bad variable name's(badVar) 'detected:' bad say; say 'The list of' vars "variable"s(vars) 'and' s(vars,'their',"it's") "value"s(vars) 'follows:'; say
do k=1 for vars v=word(varList,k) say right(v,maxLenV) '=' value(v) end /*k*/
exit /*stick a fork in it, we're done.*/ /*───────────────────────────────error handling subroutines and others.─*/ s: if arg(1)==1 then return arg(3); return word(arg(2) 's',1)
- return right(arg(1),length(j)+11) /*right justify a number +indent*/
err: say; say; say center(' error! ', max(40, linesize()%2), "*"); say
do j=1 for arg(); say arg(j); say; end; say; exit 13
novalue: syntax: call err 'REXX program' condition('C') "error",,
condition('D'),'REXX source statement (line' sigl"):",sourceline(sigl)</lang>
output when using the input (file) as specified in the task description:
28 records were read from file: CONFIG.DAT 8 blank records were read. 15 records ignored that began with a # (hash). 1 record ignored that began with a ; (semicolon). The list of 4 variables and their values follows: FULLNAME = Foo Barber FAVOURITEFRUIT = banana NEEDSPEELING = true OTHERFAMILY = Rhu Barber, Harry Barber
Ruby
<lang ruby>fullname = favouritefruit = "" needspeeling = seedsremoved = false otherfamily = []
IO.foreach("config.file") do |line|
line.chomp! key, value = line.split(nil, 2) case key when /^([#;]|$)/; # ignore line when "FULLNAME"; fullname = value when "FAVOURITEFRUIT"; favouritefruit = value when "NEEDSPEELING"; needspeeling = true when "SEEDSREMOVED"; seedsremoved = true when "OTHERFAMILY"; otherfamily = value.split(",").map(&:strip) when /^./; puts "#{key}: unknown key" end
end
puts "fullname = #{fullname}" puts "favouritefruit = #{favouritefruit}" puts "needspeeling = #{needspeeling}" puts "seedsremoved = #{seedsremoved}" otherfamily.each_with_index do |name, i|
puts "otherfamily(#{i+1}) = #{name}"
end</lang>
- Output:
fullname = Foo Barber favouritefruit = banana needspeeling = true seedsremoved = false otherfamily(1) = Rhu Barber otherfamily(2) = Harry Barber
Run BASIC
<lang Runbasic>dim param$(6) dim paramVal$(6) param$(1) = "fullname" param$(2) = "favouritefruit" param$(3) = "needspeeling" param$(4) = "seedsremoved" param$(5) = "otherfamily" for i = 1 to 6
paramVal$(i) = "false"
next i
open DefaultDir$ + "\public\a.txt" for binary as #f while not(eof(#f)) line input #f, a$ a$ = trim$(a$) if instr("#;",left$(a$,1)) = 0 and a$ <> "" then thisParam$ = lower$(word$(a$,1," ")) for i = 1 to 5 if param$(i) = thisParam$ then paramVal$(i) = "true" aa$ = trim$(mid$(a$,len(thisParam$)+2)) if aa$ <> "" then paramVal$(i) = aa$ end if next i end if wend close #f for i = 1 to 5
if instr(paramVal$(i),",") > 0 then for j = 1 to 2 print param$(i);"(";j;")";chr$(9);trim$(word$(paramVal$(i),j,",")) next j else print param$(i);chr$(9);paramVal$(i) end if
next i</lang>Output
fullname Foo Barber favouritefruit banana needspeeling true seedsremoved false otherfamily(1) Rhu Barber otherfamily(2) Harry Barber
Scala
A "one liner" version which:
- Filters out empty and comment lines
- Splits field name and value(s)
- Adds "true" to value-less fields
- Converts values to (k, List(values)
- Converts the entire collection to a Map
<lang Scala>val conf = scala.io.Source.fromFile("config.file").
getLines. toList. filter(_.trim.size > 0). filterNot("#;" contains _(0)). map(_ split(" ", 2) toList). map(_ :+ "true" take 2). map { s:List[String] => (s(0).toLowerCase, s(1).split(",").map(_.trim).toList) }.toMap</lang>
Test code: <lang Scala> for ((k,v) <- conf) {
if (v.size == 1) println("%s: %s" format (k, v(0))) else for (i <- 0 until v.size) println("%s(%s): %s" format (k, i+1, v(i)))
} </lang> Output
fullname: Foo Barber favouritefruit: banana needspeeling: true otherfamily(1): Rhu Barber otherfamily(2): Harry Barber
Seed7
The Seed7 library scanfile.s7i, can be used to scan a file with functions like getWord, skipSpace and getLine.
<lang seed7>$ include "seed7_05.s7i";
include "scanfile.s7i";
var string: fullname is ""; var string: favouritefruit is ""; var boolean: needspeeling is FALSE; var boolean: seedsremoved is FALSE; var array string: otherfamily is 0 times "";
const proc: main is func
local var file: configFile is STD_NULL; var string: symbol is ""; var integer: index is 0; begin configFile := open("readcfg.txt", "r"); configFile.bufferChar := getc(configFile); symbol := lower(getWord(configFile)); while symbol <> "" do skipSpace(configFile); if symbol = "#" or symbol = ";" then skipLine(configFile); elsif symbol = "fullname" then fullname := getLine(configFile); elsif symbol = "favouritefruit" then favouritefruit := getLine(configFile); elsif symbol = "needspeeling" then needspeeling := TRUE; elsif symbol = "seedsremoved" then seedsremoved := TRUE; elsif symbol = "otherfamily" then otherfamily := split(getLine(configFile), ","); for key index range otherfamily do otherfamily[index] := trim(otherfamily[index]); end for; else writeln(" *** Illegal line " <& literal(getLine(configFile))); end if; symbol := lower(getWord(configFile)); end while; close(configFile); writeln("fullname: " <& fullname); writeln("favouritefruit: " <& favouritefruit); writeln("needspeeling: " <& needspeeling); writeln("seedsremoved: " <& seedsremoved); for key index range otherfamily do writeln(("otherfamily[" <& index <& "]:") rpad 16 <& otherfamily[index]); end for; end func;</lang>
Output:
fullname: Foo Barber favouritefruit: banana needspeeling: TRUE seedsremoved: FALSE otherfamily[1]: Rhu Barber otherfamily[2]: Harry Barber
Smalltalk
This code retrieves the configuration values as a Dictionary; code to set an object's instance variables follows below: <lang smalltalk>dict := Dictionary new. configFile asFilename readingLinesDo:[:line |
(line isEmpty or:[ line startsWithAnyOf:#('#' ';') ]) ifFalse:[ s := line readStream. (s skipSeparators; atEnd) ifFalse:[ |optionName values| optionName := s upToSeparator. values := (s upToEnd asCollectionOfSubstringsSeparatedBy:$,) collect:[:each | each withoutSeparators] thenSelect:[:vals | vals notEmpty]. dict at:optionName asLowercase put:(values isEmpty ifTrue:[true] ifFalse:[ values size == 1 ifTrue:[values first] ifFalse:[values]]). ] ].
]</lang> gives us in dict Dictionary ('fullname'->'Foo Barber' 'needspeeling'->true 'favouritefruit'->'banana' 'otherfamily'->OrderedCollection('Rhu Barber' 'Harry Barber'))
assuming that the target object has setters for each option name, we could write: <lang smalltalk>dict keysAndValuesDo:[:eachOption :eachValue |
someObject perform:(eachOption,':') asSymbol with:eachValue ifNotUnderstood: [ self error: 'unknown option: ', eachOption ]
]</lang> or assign variables with: <lang smalltalk>fullname := dict at: 'fullname' ifAbsent:false. needspeeling := dict at: 'needspeeling' ifAbsent:false. favouritefruit := dict at: 'favouritefruit' ifAbsent:false. otherfamily := dict at: 'otherfamily' ifAbsent:false. seedsremoved := dict at: 'seedsremoved' ifAbsent:false. </lang>
Tcl
This code stores the configuration values in a global array (cfg
) so they can't pollute the global namespace in unexpected ways.
<lang tcl>proc readConfig {filename {defaults {}}} {
global cfg # Read the file in set f [open $filename] set contents [read $f] close $f # Set up the defaults, if supplied foreach {var defaultValue} $defaults {
set cfg($var) $defaultValue
} # Parse the file's contents foreach line [split $contents "\n"] {
set line [string trim $line] # Skip comments if {[string match "#*" $line] || [string match ";*" $line]} continue # Skip blanks if {$line eq ""} continue
if {[regexp {^\w+$} $line]} { # Boolean case set cfg([string tolower $line]) true } elseif {[regexp {^(\w+)\s+([^,]+)$} $line -> var value]} { # Simple value case set cfg([string tolower $var]) $value } elseif {[regexp {^(\w+)\s+(.+)$} $line -> var listValue]} { # List value case set cfg([string tolower $var]) {} foreach value [split $listValue ","] { lappend cfg([string tolower $var]) [string trim $value] } } else { error "malformatted config file: $filename" }
}
}
- Need to supply some default values due to config file grammar ambiguities
readConfig "fruit.cfg" {
needspeeling false seedsremoved false
} puts "Full name: $cfg(fullname)" puts "Favourite: $cfg(favouritefruit)" puts "Peeling? $cfg(needspeeling)" puts "Unseeded? $cfg(seedsremoved)" puts "Family: $cfg(otherfamily)"</lang>
TXR
Prove the logic by transliterating to a different syntax: <lang txr>@(collect) @ (cases)
- @/.*/
@ (or)
- @/.*/
@ (or) @{IDENT /[A-Z_][A-Z_0-9]+/}@/ */ @(bind VAL ("true")) @ (or) @{IDENT /[A-Z_][A-Z_0-9]+/}@(coll)@/ */@{VAL /[^,]+/}@/ */@(end) @ (or) @{IDENT /[A-Z_][A-Z_0-9]+/}@(coll)@/ */@{VAL /[^, ]+/}@/,? */@(end) @(flatten VAL) @ (or) @/ */ @ (or) @ (throw error "bad configuration syntax") @ (end) @(end) @(output) @ (repeat) @IDENT = @(rep)@VAL, @(first){ @VAL, @(last)@VAL };@(single)@VAL;@(end) @ (end) @(end) </lang>
Sample run:
$ txr configfile.txr configfile FULLNAME = Foo Barber; FAVOURITEFRUIT = banana; NEEDSPEELING = true; OTHERFAMILY = { Rhu Barber, Harry Barber };
Vedit macro language
<lang vedit>#11 = 0 // needspeeling = FALSE
- 12 = 0 // seedsremoved = FALSE
Reg_Empty(21) // fullname Reg_Empty(22) // favouritefruit Reg_Empty(23) // otherfamily
File_Open("|(PATH_ONLY)\example.cfg")
if (Search("|<FULLNAME|s", BEGIN+ADVANCE+NOERR)) {
Match("=", ADVANCE) // skip optional '=' Reg_Copy_Block(21, CP, EOL_pos)
} if (Search("|<FAVOURITEFRUIT|s", BEGIN+ADVANCE+NOERR)) {
Match("=", ADVANCE) Reg_Copy_Block(22, CP, EOL_pos)
} if (Search("|<OTHERFAMILY|s", BEGIN+ADVANCE+NOERR)) {
Match("=", ADVANCE) Reg_Copy_Block(23, CP, EOL_pos)
} if (Search("|<NEEDSPEELING|s", BEGIN+ADVANCE+NOERR)) {
#11 = 1
} if (Search("|<SEEDSREMOVED|s", BEGIN+ADVANCE+NOERR)) {
#12 = 1
}
Buf_Quit(OK) // close .cfg file
// Display the variables Message("needspeeling = ") Num_Type(#11, LEFT) Message("seedsremoved = ") Num_Type(#12, LEFT) Message("fullname = ") Reg_Type(21) TN Message("favouritefruit = ") Reg_Type(22) TN Message("otherfamily = ") Reg_Type(23) TN</lang>
Output:
needspeeling = 1 seedsremoved = 0 fullname = Foo Barber favouritefruit = banana otherfamily = Rhu Barber, Harry Barber
Visual Basic
<lang vb>' Configuration file parser routines. ' ' (c) Copyright 1993 - 2011 Mark Hobley ' ' This configuration parser contains code ported from an application program ' written in Microsoft Quickbasic ' ' This code can be redistributed or modified under the terms of version 1.2 of ' the GNU Free Documentation Licence as published by the Free Software Foundation.
Sub readini()
var.filename = btrim$(var.winpath) & ini.inifile var.filebuffersize = ini.inimaxlinelength Call openfileread If flg.error = "Y" Then flg.abort = "Y" Exit Sub End If If flg.exists <> "Y" Then flg.abort = "Y" Exit Sub End If var.inistream = var.stream
readinilabela:
Call readlinefromfile If flg.error = "Y" Then flg.abort = "Y" Call closestream flg.error = "Y" Exit Sub End If If flg.endoffile <> "Y" Then iniline$ = message$ If iniline$ <> "" Then If Left$(iniline$, 1) <> ini.commentchar AND Left$(iniline$, 1) <> ini.ignorechar Then endofinicommand% = 0 For l% = 1 To Len(iniline$) If Mid$(iniline$, l%, 1) < " " Then endofinicommand% = l% End If If Not (endofinicommand%) Then If Mid$(iniline$, l%, 1) = " " Then endofinicommand% = l% End If End If If endofinicommand% Then l% = Len(iniline$) End If Next l% iniarg$ = "" If endofinicommand% Then If endofinicommand% <> Len(iniline$) Then iniarg$ = btrim$(Mid$(iniline$, endofinicommand% + 1)) If iniarg$ = "" Then GoTo readinilabelb End If inicommand$ = Left$(iniline$, endofinicommand% - 1) End If Else inicommand$ = btrim$(iniline$) End If
readinilabelb:
'interpret command inicommand$ = UCase$(inicommand$) Select Case inicommand$ Case "FULLNAME" If iniarg$ <> "" Then ini.fullname = iniarg$ End If Case "FAVOURITEFRUIT" If iniarg$ <> "" Then ini.favouritefruit = iniarg$ End If Case "NEEDSPEELING" ini.needspeeling = "Y" Case "SEEDSREMOVED" ini.seedsremoved = "Y" Case "OTHERFAMILY" If iniarg$ <> "" Then ini.otherfamily = iniarg$ CALL familyparser End If Case Else '!! error handling required End Select End If End If GoTo readinilabela End If Call closestream Exit Sub
readinierror:
End Sub
Sub openfileread()
flg.streamopen = "N" Call checkfileexists If flg.error = "Y" Then Exit Sub If flg.exists <> "Y" Then Exit Sub Call getfreestream If flg.error = "Y" Then Exit Sub var.errorsection = "Opening File" var.errordevice = var.filename If ini.errortrap = "Y" Then On Local Error GoTo openfilereaderror End If flg.endoffile = "N" Open var.filename For Input As #var.stream Len = var.filebuffersize flg.streamopen = "Y" Exit Sub
openfilereaderror:
var.errorcode = Err Call errorhandler resume '!!
End Sub
Public Sub checkfileexists()
var.errorsection = "Checking File Exists" var.errordevice = var.filename If ini.errortrap = "Y" Then On Local Error GoTo checkfileexistserror End If flg.exists = "N" If Dir$(var.filename, 0) <> "" Then flg.exists = "Y" End If Exit Sub
checkfileexistserror:
var.errorcode = Err Call errorhandler
End Sub
Public Sub getfreestream()
var.errorsection = "Opening Free Data Stream" var.errordevice = "" If ini.errortrap = "Y" Then On Local Error GoTo getfreestreamerror End If var.stream = FreeFile Exit Sub
getfreestreamerror:
var.errorcode = Err Call errorhandler resume '!!
End Sub
Sub closestream()
If ini.errortrap = "Y" Then On Local Error GoTo closestreamerror End If var.errorsection = "Closing Stream" var.errordevice = "" flg.resumenext = "Y" Close #var.stream If flg.error = "Y" Then flg.error = "N" '!! Call unexpectederror End If flg.streamopen = "N" Exit Sub
closestreamerror:
var.errorcode = Err Call errorhandler resume next
End Sub
Sub readlinefromfile()
If ini.errortrap = "Y" Then On Local Error GoTo readlinefromfileerror End If If EOF(var.stream) Then flg.endoffile = "Y" Exit Sub End If Line Input #var.stream, tmp$ message$ = tmp$ Exit Sub
readlinefromfileerror:
var.errorcode = Err Call errorhandler resume '!!
End Sub
Public Sub errorhandler()
tmp$ = btrim$(var.errorsection) tmp2$ = btrim$(var.errordevice) If tmp2$ <> "" Then tmp$ = tmp$ + " (" + tmp2$ + ")" End If tmp$ = tmp$ + " : " + Str$(var.errorcode) tmp1% = MsgBox(tmp$, 0, "Error!") flg.error = "Y" If flg.resumenext = "Y" Then flg.resumenext = "N"
' Resume Next
Else flg.error = "N"
' Resume
End If
End Sub
Public Function btrim$(arg$)
btrim$ = LTrim$(RTrim$(arg$))
End Function</lang>
- Programming Tasks
- File handling
- Ada
- AutoHotkey
- AWK
- BBC BASIC
- C++
- Boost
- D
- Erlang
- Fantom
- Fortran
- Go
- Groovy
- Haskell
- Icon
- Unicon
- J
- JavaScript
- Java
- Lasso
- Liberty BASIC
- Lua
- MATLAB
- Octave
- OCaml
- OoRexx
- Pascal
- SysUtils
- Perl
- Perl 6
- PicoLisp
- PL/I
- Protium
- Python
- Racket
- REXX
- Ruby
- Run BASIC
- Scala
- Seed7
- Smalltalk
- Tcl
- TXR
- Vedit macro language
- Visual Basic
- GUISS/Omit
- Lilypond/Omit
- Openscad/Omit
- Initialization