Rosetta Code/Run examples

From Rosetta Code
Rosetta Code/Run examples is a draft programming task. It is not yet considered ready to be promoted as a complete task, for reasons that should be found in its talk page.

This task is based on an idea hatched from this C1R Implementation.

Write a program that will accept as input the name of a task from Rosetta Code and the name of a language. The program should then download the solution for the specified task in the specified language, present the source to the user and prompt the user to confirm running the example.

The program should verify that the tools needed to compile or run the solution are present before running it. If the solution can not be run, a graceful exit should happen. (i.e. the program should not crash)

Besides its own language, the program should support at least two other languages. (Ideally it would support most of the languages available, but that is too much to ask. Though some languages are simple, e.g. python, pike, perl, bash and several more only need the solution saved in a file and given to the language as argument to run, so it should be easy to support many languages like that).

If you know what is needed to support a particular language, please help to add support for that language to implementations in other languages.

Extra credit: add a function to get a list of all solutions of a given language and run them to create a report on which solutions failed to run.

More credit: also test if the output of a solution compares to a given result. The expected output should be loaded from a file with the name of the task. (This way all implementations can share the same set of files, and anyone can add more files. In the future the content of these files could be stored directly in a section of the task on Rosetta Code itself.)

Go

This is a fairly basic program which is designed to work for Go, Perl or Python entries. The latter two have been chosen as they are often included in Linux distros by default.

However, in the case of Python, it assumes a Python 3 compatible program and so some earlier programs written for Python 2 will not work.

Although all necessary resources should still be available for Go programs, I have no idea about the other languages and would have to leave that to the user to judge.

Having checked that the task exists, the program downloads its 'Edit' page and looks for the first runnable program after the language header. In practice, of course, there may be several runnable programs for a given language though this strategy should at least ensure that programs for different languages which use the same syntax highlighting (such as Nim which uses Python S/H) are not extracted by mistake.

No attempt has been at 'extra credit' which would take all day for Go alone and 'more credit' seems pointless as program output is seldom, if ever, exactly the same for all languages.

package main

import (
    "bufio"
    "fmt"
    "html"
    "io/ioutil"
    "log"
    "net/http"
    "os"
    "os/exec"
    "regexp"
    "strings"
)

func getAllTasks() map[string]bool {
    ex := `<li><a href="/wiki/(.*?)"`
    re := regexp.MustCompile(ex)
    url1 := "http://rosettacode.org/wiki/Category:Programming_Tasks"
    url2 := "http://rosettacode.org/wiki/Category:Draft_Programming_Tasks"
    urls := []string{url1, url2}
    tasks := make(map[string]bool)
    for _, url := range urls {
        resp, _ := http.Get(url)
        body, _ := ioutil.ReadAll(resp.Body)
        // find all tasks
        matches := re.FindAllStringSubmatch(string(body), -1)
        resp.Body.Close()
        for _, match := range matches {
            // exclude any 'category' references
            if !strings.HasPrefix(match[1], "Category:") {
                tasks[match[1]] = true
            }
        }
    }
    return tasks
}

func check(err error) {
    if err != nil {
        log.Fatal(err)
    }
}

func main() {
    tasks := getAllTasks()
    for {
        fmt.Print("Enter the exact name of the task  : ")
        in := bufio.NewReader(os.Stdin)
        task, err := in.ReadString('\n')
        check(err)
        task = strings.TrimSpace(task)
        task = strings.ReplaceAll(task, " ", "_")
        if !tasks[task] {
            fmt.Println("Sorry a task with that name doesn't exist.")
        } else {
            url := "https://rosettacode.org/mw/index.php?title=" + task + "&action=edit"
            resp, err := http.Get(url)
            check(err)
            body, _ := ioutil.ReadAll(resp.Body)
            resp.Body.Close()
            var lang string
            for {
                fmt.Print("Enter the language Go/Perl/Python : ")
                lang, err = in.ReadString('\n')
                check(err)
                lang = strings.TrimSpace(lang)
                lang = strings.ToLower(lang)
                if lang == "go" || lang == "perl" || lang == "python" {
                    break
                }
                fmt.Println("Sorry that language is not supported.")
            }
            var lang2, lang3, ext string
            switch lang {
            case "go":
                lang2 = "Go"
                lang3 = "(go|Go|GO)"
                ext = "go"
            case "perl":
                lang2 = "Perl"
                lang3 = "(perl|Perl)"
                ext = "pl"
            case "python":
                lang2 = "Python"
                lang3 = "(python|Python)"
                ext = "py"
            }
            fileName := "rc_temp." + ext
            header := fmt.Sprintf(`(?s)==\{\{header\|%s\}\}==.*?&lt;lang %s>`, lang2, lang3)
            exp := header + `(.*?)&lt;/lang>`
            re := regexp.MustCompile(exp)
            page := string(body)
            matches := re.FindStringSubmatch(page)
            if matches == nil {
                fmt.Println("No runnable task entry for that language was detected.")
            } else {
                source := html.UnescapeString(matches[2])
                fmt.Println("\nThis is the source code for the first or only runnable program:\n")
                fmt.Println(source)
                fmt.Print("\nDo you want to run it y/n : ")
                yn, err := in.ReadString('\n')
                check(err)
                if yn[0] == 'y' || yn[0] == 'Y' {
                    err = ioutil.WriteFile(fileName, []byte(source), 0666)
                    check(err)
                    var cmd *exec.Cmd
                    switch lang {
                    case "go":
                        cmd = exec.Command("go", "run", fileName)
                    case "perl":
                        cmd = exec.Command("perl", fileName)
                    case "python":
                        cmd = exec.Command("python3", fileName)
                    }
                    cmd.Stdout = os.Stdout
                    cmd.Stderr = os.Stderr
                    cmd.Run() // allow any error(s) to go to StdEerr
                    check(os.Remove(fileName))
                }
            }
            fmt.Print("\nDo another one y/n : ")
            yn, err := in.ReadString('\n')
            check(err)
            if yn[0] != 'y' &&  yn[0] != 'Y' {
                break
            }
        }
    }
}
Output:

Sample output (same shortish task used for all 3 languages):

Enter the exact name of the task  : Palindrome dates
Enter the language Go/Perl/Python : Go

This is the source code for the first or only runnable program:

package main

import (
    "fmt"
    "time"
)

func reverse(s string) string {
    chars := []rune(s)
    for i, j := 0, len(chars)-1; i < j; i, j = i+1, j-1 {
        chars[i], chars[j] = chars[j], chars[i]
    }
    return string(chars)
}

func main() {
    const (
        layout  = "20060102"
        layout2 = "2006-01-02"
    )
    fmt.Println("The next 15 palindromic dates in yyyymmdd format after 20200202 are:")
    date := time.Date(2020, 2, 2, 0, 0, 0, 0, time.UTC)
    count := 0
    for count < 15 {
        date = date.AddDate(0, 0, 1)
        s := date.Format(layout)
        r := reverse(s)
        if r == s {
            fmt.Println(date.Format(layout2))
            count++
        }
    }
}

Do you want to run it y/n : y
The next 15 palindromic dates in yyyymmdd format after 20200202 are:
2021-12-02
2030-03-02
2040-04-02
2050-05-02
2060-06-02
2070-07-02
2080-08-02
2090-09-02
2101-10-12
2110-01-12
2111-11-12
2120-02-12
2121-12-12
2130-03-12
2140-04-12

Do another one y/n : y
Enter the exact name of the task  : Palindrome dates
Enter the language Go/Perl/Python : Perl

This is the source code for the first or only runnable program:

use Time::Piece;
my $d = Time::Piece->strptime("2020-02-02", "%Y-%m-%d");

for (my $k = 1 ; $k <= 15 ; $d += Time::Piece::ONE_DAY) {
    my $s = $d->strftime("%Y%m%d");
    if ($s eq reverse($s) and ++$k) {
        print $d->strftime("%Y-%m-%d\n");
    }
}

Do you want to run it y/n : y
2020-02-02
2021-12-02
2030-03-02
2040-04-02
2050-05-02
2060-06-02
2070-07-02
2080-08-02
2090-09-02
2101-10-12
2110-01-12
2111-11-12
2120-02-12
2121-12-12
2130-03-12

Do another one y/n : y
Enter the exact name of the task  : Palindrome dates
Enter the language Go/Perl/Python : Python

This is the source code for the first or only runnable program:

'''Palindrome dates'''

from datetime import datetime
from itertools import chain


# palinDay :: Int -> [ISO Date]
def palinDay(y):
    '''A possibly empty list containing the palindromic
       date for the given year, if such a date exists.
    '''
    s = str(y)
    r = s[::-1]
    iso = '-'.join([s, r[0:2], r[2:]])
    try:
        datetime.strptime(iso, '%Y-%m-%d')
        return [iso]
    except ValueError:
        return []


# --------------------------TEST---------------------------
# main :: IO ()
def main():
    '''Count and samples of palindromic dates [2021..9999]
    '''
    palinDates = list(chain.from_iterable(
        map(palinDay, range(2021, 10000))
    ))
    for x in [
            'Count of palindromic dates [2021..9999]:',
            len(palinDates),
            '\nFirst 15:',
            '\n'.join(palinDates[0:15]),
            '\nLast 15:',
            '\n'.join(palinDates[-15:])
    ]:
        print(x)


# MAIN ---
if __name__ == '__main__':
    main()

Do you want to run it y/n : y
Count of palindromic dates [2021..9999]:
284

First 15:
2021-12-02
2030-03-02
2040-04-02
2050-05-02
2060-06-02
2070-07-02
2080-08-02
2090-09-02
2101-10-12
2110-01-12
2111-11-12
2120-02-12
2121-12-12
2130-03-12
2140-04-12

Last 15:
9170-07-19
9180-08-19
9190-09-19
9201-10-29
9210-01-29
9211-11-29
9220-02-29
9221-12-29
9230-03-29
9240-04-29
9250-05-29
9260-06-29
9270-07-29
9280-08-29
9290-09-29

Do another one y/n : n

Liberty BASIC

'   ********************************************************************
'   **                                                                **
'   **    parseAndRun.bas  v26b  tenochtitlanuk    November 2012      **
'   **                                                                **
'   **      select a LB solution from RC site & run it locally        **
'   **                                                                **
'   ********************************************************************
'retrieve proper temporary path and filename to save downloaded HTML:
Source$ = GetTempFileName$("htm")
'nomainwin

'   Download main RC LB page which has current tasks on it. Save as 'source.html'
'   run "C:\Program Files\Mozilla Firefox\firefox.exe http://rosettacode.org/wiki/Category:Liberty_BASIC"   'testing routine
print " Fetching current RC page of completed Liberty BASIC RC solutions."
'result = DownloadToFile( "http://rosettacode.org/wiki/Category:Liberty_BASIC", "E:\source.html")
result = DownloadToFile( "http://rosettacode.org/wiki/Category:Liberty_BASIC", Source$)

if result <>0 then  print "Error downloading LB solved tasks.": end else print: print " Displaying solved tasks.": print

'   Load source into a string. Go through and save in a 2D array all topic titles
'       and the appropriate web addresses to find them.
'open "E:\source.html" for input as #f
open Source$ for input as #f
    html$ = input$( #f, lof( #f))
close #f
kill Source$   'remove temp file

dim solutions$( 500, 2)

global count
count   =1
first   =0
last    =0
reading =0

'   The first topic is the '100 doors' so skip all html jump ref's earlier than this.
do
    r$ =getHtmlSection$( html$, first, last)
    if instr( r$, "/rosettacode.org/mw/index.php") then exit do '   We've read all LB solved tasks.
    if r$ ="wiki/100_doors" then reading =1

    if reading =1 then   '   we can start recording path & name
        solutions$( count, 1) ="http://rosettacode.org/" +r$ +"#Liberty_BASIC"

        special =instr( r$, "%2B"):    if special <>0 then r$ =left$( r$, special -1) +"+" +mid$( r$, special +3)
        special =instr( r$, "%27"):    if special <>0 then r$ =left$( r$, special -1) +"'" +mid$( r$, special +3)
        special =instr( r$, "%C3%A8"): if special <>0 then r$ =left$( r$, special -1) +chr$( 232) +mid$( r$, special +6)
        solutions$( count, 0) =mid$( r$, 6)  '   we want the bit beyond '/wiki/'
        if instr(  solutions$( count, 0), "/") then
            newName$ =""
            for ii =1 to len(  solutions$( count, 0) )
                n$ =mid$( solutions$( count, 0), ii, 1)
                if n$ ="/" then n$ ="_"
                newName$ =newName$ +n$
            next ii
            solutions$( count, 0) =newName$
        end if
        print count, solutions$( count, 0)'; tab( 60); solutions$( count, 1)
        count =count +1
    end if
loop until 0
print: print count -1; " tasks solved in LB."

'input " Choose task # "; R                                              '   Choose a page to try.
for R =1 to 283
print
print " Choosing a task at random viz #"; R; " out of "; count -1; " completed in LB."
print " Task is "; chr$( 34); solutions$( R, 0); chr$( 34)
print

'********************run "C:\Program Files\Mozilla Firefox\firefox.exe " +solutions$( R, 1)

'   Fetch the RC task page with all the sol'ns including LB one.
print " Downloading the page for this task."
result = DownloadToFile( solutions$( R, 1), "rx.html")

if result <>0 then  print "Error downloading.": end

print " Now finding the LB section of the html code."   '   Now finding the appropriate LB section on this topic.

open "rx.html" for input as #r
    L =lof( #r)
    print " Length of source html of this topic's page is "; L
    t$ =input$( #r, L)
close #r

preamble$ =">Liberty BASIC</a></span></h2>" +chr$( 10)

lP =len( preamble$)
print " Finding the preamble string at ";
beg =instr( t$, preamble$)' +len( preamble$)
print beg

lookFor$ ="source" +chr$( 34) +">"
beg =instr( t$, lookFor$, beg)    '   get to start of BASIC code.
beg =beg +len( lookFor$)

print " Found LB section at "; beg;

fin =instr( t$, "</pre>", beg)
print " and ending at "; fin

print " Chopping off unwanted earlier & later sections of html source."
t$ =mid$( t$, beg, fin -beg)    '   discard earlier & later parts of html code.

open solutions$( R, 0) +".txt" for output as #LbText
    #LbText t$;
close #LbText

L =len( t$)

print " Relevant html code LB section being parsed for LB BASIC code."

'   Read the rest of the LB code section to </pre> section ..
LB$  =""
j    =1

print " Dropping html tags & translating html entities."
print
print " LB code follows."
print

do
    nxtChr$  =mid$( t$, j, 1)
    select case '   _______________________________________________________________________________________________________
        case ( nxtChr$ =chr$( 10)) or ( nxtChr$ =chr$( 13))
            j =L
            print "End reached- CRLF"

        case nxtChr$ ="<"                                                           '   we've found a html tag. Omit.
            'print " Starting a tag with a <";
            item$ ="<"
            do                                                                      '   keep looking until find a '>' or finish...
                j =j +1
                nxtChr$  =mid$( t$, j, 1)
                item$ =item$ +nxtChr$
            loop until nxtChr$ =">"
            'print " Closing a tag with a >."
            if item$ ="</pre>" then j =L                                            '   end reached
            if item$ ="<br />" then LB$ =LB$ +chr$( 10)                             '   code for CRLF
            if item$ ="<br/>" then LB$ =LB$ +chr$( 10)                             '   code for CRLF, now
            if j <>L then j =j +1

        case nxtChr$ ="&"                                                           '   we've found an html entity.
                                                                                    '   replace with plain-text equivalents.
            'print " html entity starting with & ";
            select case '   ..............................................................................
                case mid$( t$, j+1, 5) ="quot;"
                    LB$ =LB$ +chr$( 34): j =j +6                                    '   &guot;    "
                case mid$( t$, j+1, 3) ="gt;"
                    LB$ =LB$ +">": j =j +4                                          '   >      >
                case mid$( t$, j+1, 3) ="lt;"
                    LB$ =LB$ +"<": j =j +4                                          '   <      <
                case right$( mid$( t$, j, 5), 1) =";"
                    v =val( mid$( t$, j +2, 2)): j =j +5                            '   2-digit character-code
                    if v =39 then LB$ =LB$ +chr$( 39) else LB$ =LB$ +chr$( v)       '       eg '  ( 40,41)   '()
                case right$( mid$( t$, j, 6), 1) =";"                               '   3-digit character-code
                    v =val( mid$( t$, j +2, 3))
                    if v =160 then v =32    'print "Hard space!"                    '   convert   hard- to soft-space.
                    j =j +6: LB$ =LB$ +chr$( v)
            end select  '   ..............................................................................
            'print " and finishing with ;"
        case else  '   not an html entity nor a tag. Use as-is unless it's the final hard-space plus semi-colon..
            if mid$( t$, j +1, 5) ="#160;" and mid$( t$, j +5, 6) ="</pre>" then j =L else LB$ =LB$ +nxtChr$: j =j +1

    end select  '   _________________________________________________________________________________________________________
    scan
loop until j >= fin -beg -4

print: print LB$

open solutions$( R, 0) +".bas" for output as #LB
    #LB LB$;
close #LB

print
print " Done"

timer 5000, [on2]
wait
[on2]
timer 0

'   Run with LB.
'   *************************************run chr$( 34) +"C:\Program Files\Liberty BASIC v4.04\liberty.exe" +chr$( 34) +" -R E:\" +solutions$( R, 0) +".bas"
next R
end
    '   **************************************************************
Function DownloadToFile( urlfile$, localfile$)
    open "URLmon" for dll as #url
    calldll #url, "URLDownloadToFileA",_
    0 as long,_         'null
    urlfile$ as ptr,_   'url to download
    localfile$ as ptr,_ 'save file name
    0 as long,_         'reserved, must be 0
    0 as long,_         'callback address, can be 0
    DownloadToFile as ulong  '0=success
    close #url
end function
end

function getHtmlSection$( string$, byref first, last)
    a                    =instr(  string$, "<a href=" +chr$( 34), first)
        if a =0 then getHtmlSection$ ="  Sorry! html link not found": exit function
    b                    =instr(  string$, chr$( 34), a +9)
    getHtmlSection$      =mid$(   string$, a +10, b -a -10)
    first                =b +1
            '   Reset value of "first" so that in the next call to
            '   getHtmlSection$( the next html link can be found
end function

function GetTempFileName$(prefix$)
    TempPath$=GetTempPath$()
    TempFile$ = space$(256)+chr$(0)

    calldll #kernel32, "GetTempFileNameA",_
    TempPath$ as ptr,_  'directory for temp file
    prefix$ as ptr,_    'desired prefix for temp filename
    0 as ulong,_        '0=file created,nonzero=you must create file
    TempFile$ as ptr,_  'string buffer to hold qualified path and filename
    result as ulong     'nonzero=success

    'TempFile$ holds complete path and filename info
    GetTempFileName$ = TempFile$
    end function

Function GetTempPath$()
    CallDLL #kernel32, "GetTempPathA",_
    0 as long,_
    _NULL as long,_
    length as long

    buf$ = space$(length)

    CallDLL #kernel32, "GetTempPathA",_
    length as long,_
    buf$ as ptr,_
    ret as long

    GetTempPath$ = buf$
End Function


Nim

Translation of: Go

This is a translation of the Go solution. The accepted languages are Go, Nim, Perl and Python (version 3 only). There is no tentative to check if the requirements to run a program are fulfilled. If they are not, the execution of the external program will fail but without consequences on the main program.

On Linux, Perl and Python are installed by default but third party modules may be not. For Go and Nim, an installation is required.

Note that when running a Nim program a compilation is done and messages are displayed on the terminal before the messages displayed when executing the compiled program. To avoid compiling issues and execution issues, programs are compiled with options -d:release -d:ssl --threads:on. The tow last options are not needed for most programs.

The present program must be compiled with option -d:ssl

import htmlparser, httpclient, os, osproc, re, sets, strutils, xmltree

const
  Url1 = "http://rosettacode.org/wiki/Category:Programming_Tasks"
  Url2 = "http://rosettacode.org/wiki/Category:Draft_Programming_Tasks"
  Urls = [Url1, Url2]

proc getAllTasks(client: HttpClient): HashSet[string] =
  let regex = re("""<li><a href="/wiki/(.*?)"""")
  var matches: array[1, string]
  var start = 0
  for url in Urls:
    let body = client.getContent(url)
    # Find all tasks.
    while true:
      start = body.find(regex, matches, start) + 1
      if start == 0: break
      if not matches[0].startsWith("Category:"):
        result.incl matches[0]

let client = newHttpClient()
let tasks = client.getAllTasks()

while true:
  stdout.write "Enter the exact name of the task: "
  stdout.flushFile()
  let task = stdin.readLine().strip().replace(' ', '_')
  if task notin tasks:
    echo "Sorry a task with that name doesn't exist."
    continue

  let url = "https://rosettacode.org/w/index.php?title=" & task & "&action=edit"
  let body = client.getContent(url)
  var lang: string
  while true:
    stdout.write "Enter the language Go/Nim/Perl/Python : "
    stdout.flushFile()
    lang = stdin.readLine().strip().toLowerAscii
    if lang in ["go", "nim", "perl", "python"]:
      break
    echo "Sorry that language is not supported."

  var lang2, lang3, ext: string
  case lang
  of "go":
    lang2 = "Go"
    lang3 = """("go"|"Go"|"GO")"""
    ext = "go"
  of "nim":
    lang2 = "Nim"
    lang3 = """("nim"|"Nim")"""
    ext = "nim"
  of "perl":
    lang2 = "Perl"
    lang3 = """("perl"|"Perl")"""
    ext = "pl"
  of "python":
    lang2 = "Python"
    lang3 = """("python"|"Python")"""
    ext = "py"

  let fileName = "rc_temp." & ext
  let header = r"(?s)==\{\{header\|$#\}\}".format(lang2)
  let language = r"lt;syntaxhighlight lang=$#>".format(lang3)
  let regex = re(header & r"==.*?&" & language & r"(.*?)&lt;/syntaxhighlight>")
  var matches: array[2, string]
  let idx = body.find(regex, matches)
  if idx < 0:
    echo "No runnable task entry for that language was detected."
    continue

  let source = parseHtml(matches[1]).innerText()
  echo "\nThis is the source code for the first or only runnable program:\n"
  echo source
  stdout.write "\nDo you want to run it (y/n)? "
  stdout.flushFile()
  var answer = stdin.readLine()
  if answer in ["y", "Y"]:
    fileName.writeFile(source)
    let cmd = case lang
              of "go": "go run " & fileName
              of "nim": "nim r -d:release --threads:on -d:ssl " & fileName
              of "perl": "perl " & fileName
              of "python": "python " & fileName
              else: ""
    discard execCmd cmd
    removeFile fileName

  stdout.write "\nDo another one (y/n)? "
  stdout.flushFile()
  answer = stdin.readLine()
  if answer notin ["y", "Y"]:
    break

Phix

Screenshot here

--
-- demo\rosetta\Run_examples.exw
-- =============================
--
-- Full GUI, supports Go, Julia, Phix, Python, and Wren (those I've installed).
--
-- Uses much of the same code as Count_Examples.exw, along with 130MB+ of ".raw"
--  downloaded files, thankfully cached, and not surprising given that program
--  reports there are 83,050 entires, which averages out at 1662 bytes per...
-- It can of course take some time to download 1440 tasks at not much under 100K
--  each, but once in the cache it only takes a few seconds to load everything.
--
-- Should you only want to run Phix programs, demo/pGUI/pdemo.exw offers many
--  other non-rosettacode demo programs, and many but not all 1440 rc tasks.
--
--  No attempt has been made for the "Extra credit" (report which tasks failed
--  to run) or "More credit" (save output and compare with expected) - for the
--  latter you might want to look at test/terror[X].exw and if I'm having that
--  much trouble just on Phix with only things that I actually wrote, well...
--
-- To do: Entries such as Boids just get "See Boids/Go" with that link not
--        even being downloaded into rc_cache at all..... Exactly the same o/c 
--        for both the Julia and Phix entries for Boids, that is not counting
--        the additional screenshot link on the Phix entry.
--
--        The table and codetext are not shrinking as I would like (help?).
--
-- sug: Add buttons "Run (F5)", "Copy cmd (F6)", "Edit rcre.xxx"
--      Save some IupConfig() settings...
--
without js -- (obviously this will never ever run in a browser!)
include pGUI.e
requires(WINDOWS) -- currently windows-only, unless you can munge the following

-- Without any doubt these will need editing, or perhaps just commenting out
--  There are no hard-coded lengths and buttons etc are generated from this
constant gopath = `C:\Program Files\Go\bin\`,
         jspath = getenv("TEMP"), 
      juliapath = `E:\downloads\misc\wren\julia-1.6.2-win64\julia-1.6.2\bin\`,
       phixpath = `C:\Program Files (x86)\Phix\`, 
         pypath = `C:\Program Files (x86)\Python37-32\`,
       wrenpath = `E:\downloads\misc\wren\wren_cli-windows-0.3.0\`,
        gopause = "var s string\nfmt.Printf(\"done\")\nfmt.Scan(&s)\n",
      phixpause = "?\"done\"\n{}=wait_key()\n",
--        jsrun = {`pw.exe`, `runjs.exw`}
      cliordome = {`wren_cli-0.3.0.exe`,`dome.exe`},
      -- it should be pretty straightforward to comment out any that you don't want,
      -- and/or insert similar entries for other programming languages that you do:
      languages = {{`Go`,         gopath, `go.exe`,    `.go`, `"%s" run "%s"`, gopause},
                   {`JavaScript`, jspath, `open`,    `.html`, `"%s"`,``},
                   {`Julia`,   juliapath, `julia.exe`, `.jl`, `"%s" "%s"`, ``},
                   {`Phix`,     phixpath, `pw.exe`,   `.exw`, `"%s" "%s"`, phixpause},
                   {`Python`,     pypath, `python.exe`,`.py`, `"%s" "%s"`, ``},
                   {`Wren`,     wrenpath,  cliordome,`.wren`, `"%s" "%s"`, ``}},
          langs = vslice(languages,1)

Ihandle langsel, taskfilter, textsearch, table, add_pause, codetext, statusbar, clipboard
sequence langchks = {},
         radioset = {},
         disabled = repeat(false,length(languages))

constant help_text = """
Note you will have to edit the language table in the source code with directories of
executables you have previously installed. The status bar will show a list of any 
entries that need editing, when the program is first run, and the corresponding 
checkboxes and radio buttons will be permanently disabled if the executables cannot 
be found.

Selecting a language via the dropdown and/or the checkboxes filters the list of tasks
to those with implementations in the selected language(s). You can actually achieve
much the same effects just by the checkboxes and ignoring the dropdown, or even by 
sorting the table by column, apart from the full text search described next.

Text entered into the Filter box excludes all task names that do not match. The search
text option is only enabled when a language has been selected in the main dropdown, and
will also search all entries for that language for the specified string as well. The
latter can be a bit slow, hence the main table and the radios under it are deliberately 
disabled while filtering to avoid strange crashes caused by clicking on things changing
underneath your feet, such as selecting a table entry that will be removed in about ten
seconds.

Selecting an entry in the main task table populates the lower half of the screen and
enables/disables the radio butons appropriately.

Selecting a language via the radio buttons shows the text of any entry or entries found.

When the first line shows "<1 of n>" use left and right arrows to select.

Press F5 to run the code shown, which you can edit as needed though it will not be saved,
except in an often-overwritten rcre.<ext> in the executable directory. Note that if you
delete the "<n of m>" then left and right will no longer cycle through them until it has
been restored, manually or via selecting table entries or clicking radio buttons.

Note that JavaScript runs html in the browser, making no attempt to run anything under
node.js (not my thing), and Wren automatically switches to dome.exe when it spots an 
"import dome", as well as checking that all other imports can be found.

Also note that I am assuming you have write permissions to all the bin directories, for
said rcre.<ext> files.

The add pause checkbox appends a "done", pause to the end of the code (for some only).

Alternatively F6 copies the command to the clipboard so it can be run in a terminal
(you may need to cd <bin directory> for that to work).

It should not be spectacularly difficult to run this on Linux, he says.

Lastly, this was written with a fully-populated rc_cache (from previously running
Count_examples.exw) so I have no real handle on how long that might take to re-
populate, and hopefully pretty obviously, should you leave this running and something 
else updates rc_cache, then it will all go horribly wrong.
"""
--DEV/SUG: ^^ I suppose you could always put date/time/size in the main task_table
--         and then obviously check and reload the individual entries as needed...

function help()
    IupMessage("Run examples",help_text)
    return IUP_IGNORE
end function

function get_selected_language()
    for i=1 to length(radioset) do
        if IupGetInt(radioset[i],"VALUE") then
            return i
        end if
    end for
    -- oops?
end function

function run(atom c)
    string text = IupGetAttribute(codetext,"VALUE"), cmd
    if length(text) then
        if text[1]='<' then
            -- remove any `<n of m>` at the start
            integer k = find('>',text)
            string nofm = text[1..k]
            if match(` of `,nofm) then
                text = trim_head(text[k+1..$])
            end if
        end if
        integer ldx = get_selected_language(), 
                mode = 8 -- (no result/wait, no redirect)
        sequence {name,d,exe,ext,fmt} = languages[ldx]
        string filename = join_path({d,"rcre"&ext})
        if name="Wren" then
            exe = exe[iff(match(`import "dome"`,text)?2:1)]
            sequence lines = split(text,"\n"),
                   missing = {}
            for i=1 to length(lines) do
                string li = lines[i]
                if length(li)>7 and li[1..7]="import " then
                    string lib = trim(li[8..find(' ',li,9)-1],". \"/")
                    lib &= iff(lib="dome"?".exe":".wren")
                    string libpath = join_path({d,lib})
                    if not file_exists(libpath) then
                        missing = append(missing,lib)
                    end if
                end if
            end for
            if length(missing) then
                missing = "NOT INSTALLED:"&join(missing,",")
                IupSetStrAttribute(statusbar,"TITLE",missing)
                return IUP_IGNORE
            end if
        elsif name="JavaScript" then
            if not match("<html>",text) then
                IupSetStrAttribute(statusbar,"TITLE","<html> only")
                return IUP_IGNORE
            end if
            cmd = sprintf(fmt,{filename})
            mode = 4
        end if
        if mode=8 then
            string execname = join_path({d,exe})
            cmd = sprintf(fmt,{execname,filename})
        end if
-- erm, does not work, but you do have to be in the right directory.
--      if c=K_F6 and name="Wren" then
--          cmd = sprintf("cd \"%s\"; %s",{directory,cmd})
--      end if
        integer fn = open(filename,"w")
        -- you may need to add write permissons for this...
        if fn=-1 then crash("error opening "&filename) end if
        puts(fn,text)
        if IupGetInt(add_pause,"VALUE") then
            -- I suppose you could check that Go imports fmt,
            -- and maybe append this thing much earlier on,
            -- and maybe disable add_pause when length==0.
            sequence pauses = vslice(languages,6)
            puts(fn,pauses[ldx])
        end if
        close(fn)
        -- (clear CF_UNICODETEXT etc first, as per IupClipboard() docs)
        IupSetAttribute(clipboard,"TEXT",NULL)
        IupSetAttribute(clipboard,"TEXT",cmd)
        IupSetStrAttribute(statusbar, "TITLE", cmd)
        if c=K_F5 then
            string prevd = current_dir()
            if not chdir(d) then
                crash("cannot chdir to "&d)
            end if
            {} = system_exec(cmd,mode) 
            if not chdir(prevd) then
                crash("cannot chdir back to "&prevd)
            end if
        end if
    end if
    return IUP_IGNORE -- (for key_cb)
end function

include rosettacode_cache.e -- see Rosetta_Code/Count_examples#Phix

procedure set_title(string title)
    IupSetStrAttribute(statusbar, "TITLE", title)
end procedure
show_title = set_title

--  (We use a few '&' here, fairly obviously for everyone's sanity..)
constant cleanups = {{`<!--<l`&`ang Phix>(phixonline)-->`,`<l`&`ang Phix>`},
                     {`<!--<l`&`ang Phix>(notonline)-->`,`<l`&`ang Phix>`},
                     {`<!--<l`&`ang Phix>-->`,`<l`&`ang Phix>`},
                     {`<!--</l`&`ang>-->`,`</l`&`ang>`},
                     {`<span style="color: #008080;">`,``},
                     {`<span style="color: #000000;">`,``},
                     {`<span style="color: #0000FF;">`,``},
                     {`<span style="color: #7060A8;">`,``},
                     {`<span style="color: #008000;">`,``},
                     {`<span style="color: #004080;">`,``},
                     {`<span style="color: #004600;">`,``},
                     {`<span style="color: #000080;font-style:italic;">`,``},
                     {`</span>`,``},
                     {`&`&`gt;`,`>`},
                     {`&`&`lt;`,`<`}},
         {cleanstrs,cleanreps} = columnize(cleanups)

function cleanup(string s)
    if match(`<!--<l`&`ang Phix>`,s) then
        -- (Phix "manual" syntax colouring inserts a
        --  space at the start of every single line,
        --  otherwise blank lines break code blocks.)
        s = substitute(s,"\n ","\n")
    end if
    return substitute_all(s,cleanstrs,cleanreps)
end function

sequence task_table -- [[<task_name>,...langYN,"file.raw",starts,finishes]]
-- eg [["100 doors","Y","Y","Y","Y","Y","100_doors.raw",
--      {176247,209655,257786,287434,358111},
--      {177928,210102,262527,288564,358644}],...]
-- Each entry is length(languages)+4 long, starts and finishes are
--  nested tables of length(languages) each. In theory the Y/N could
--  be removed and starts[idx]!=0 relied on instead, except that we
--  use/need them for the IupTable() form, as next.

sequence data,  -- task_table in IupTable() form, and filtered
         blocks -- for the selected language
integer blockn, -- <1 of n> handling
        data_idx = 0 -- selected table entry

procedure set_blocks(sequence di, integer i)
    sequence starts = di[$-1],
             finishes = di[$]
    integer start = starts[i],
           finish = finishes[i]
    string filename = join_path({"rc_cache",di[$-2]})
    string text = trim(get_text(filename))
    -- (above is full text, below is language-only)
    -- (assumes rc_cache not edited since load_tasks)
    text = cleanup(text[start..finish])
    blocks = {}
    start = 1
    while true do
        start = match(`<l`&`ang `,text,start)
        if start=0 then exit end if
        start = find('>',text,start)+1
        if start=1 then ?9/0 end if
        finish = match(`</l`&`ang>`,text,start)
        blocks = append(blocks,trim_head(text[start..finish-1]))
        start = finish+7
    end while
    -- eg "See Boids/Go" cases (flag as un-runnable?)
    if blocks = {} then blocks = {text} end if
end procedure

function set_codetext(integer n=0)
    if data_idx!=0 then
        string text
        if n=0 then
            sequence di = data[1][data_idx]
            integer ldx = get_selected_language()
            set_blocks(data[1][data_idx],ldx)
            n = 1
        else
            text = IupGetAttribute(codetext,"VALUE")
            if length(text) 
            and text[1]!='<' then   -- manually removed?
                return IUP_DEFAULT  -- allow manual edits then
            end if
        end if
        blockn = n
        if length(blocks)=1 then
            text = blocks[1]
        else
            text = sprintf("<%d of %d>\n%s",{n,length(blocks),blocks[n]})
        end if
        IupSetAttribute(codetext,"VALUE",text)
    end if
    return IUP_IGNORE -- (for key_cb)
end function

procedure disable_radios()
    for i=1 to length(radioset) do
        IupSetInt(radioset[i],"ACTIVE",false)
    end for
end procedure

bool inIdle = false  -- [so we can IupLoopStep()/quit on esc]
bool bFilter = false -- (once set, load_tasks() performs filtering)

integer ttdx = 1    -- re-entrant filter index

function load_tasks()
    if not bFilter then
        if inIdle then return IUP_DEFAULT end if
        inIdle = true
        -- don't clobber any "NEED EDITING" message:
        wastitle = IupGetAttribute(statusbar, "TITLE")
        if get_file_type("rc_cache")!=FILETYPE_DIRECTORY then
            if not create_directory("rc_cache") then
                crash("cannot create rc_cache directory")
            end if
        end if
        sequence tasks = sort(dewiki(open_category("Programming_Tasks"))&
                              dewiki(open_category("Draft_Programming_Tasks")))
        task_table = {}
        atom t1 = time()+0.2
        for i=1 to length(tasks) do
            string ti = tasks[i],
                   url = sprintf("http://rosettacode.org/mw/index.php?title=%s&action=raw",{ti}),
                   contents = open_download(ti&".raw",url,i,length(tasks)),
                   prev = "", curr
            integer count = 0, start = 1, finishk = 0
            sequence found = repeat("N",length(langs)),
                     starts = repeat(0,length(langs)),
                     finishes = repeat(length(contents),length(langs))
            while true do
                -- Note this must handle (from Animation.raw):
                -- ==JavaScript + HTML==
                -- ==JavaScript + SVG==
                start = match(`=={`&`{he`&`ader|`,contents,start)
                if start=0 then exit end if
                integer finish = match(`}`&`}`,contents,start+1)
                curr = contents[start+11..finish-1]
                if curr!=prev then
                    if finishk then
                        finishes[finishk] = start-3
                        finishk = 0
                    end if
                    integer k = find(curr,langs)
                    if k then
                        found[k] = "Y"
                        starts[k] = finish+5
                        finishk = k
                    end if
                    count += 1
                end if
                prev = curr
                start += length(`{`&`{he`&`ader|`)
            end while
            if find("Y",found) then
                assert(find(true,sq_lt(finishes,starts))=0)
                found = prepend(found,substitute(html_clean(tasks[i]),'_',' '))
                found = append(found,ti&".raw")
                found = append(found,starts)
                found = append(found,finishes)
                task_table = append(task_table,found)
            end if
            if platform()!=JS             -- (fat chance! - but code consistently)
            and IupLoopStep()=IUP_CLOSE then -- (nb requires that inIdle handling)
                return IUP_CLOSE
            end if
            if time()>t1 and wastitle="" then
                IupSetStrAttribute(statusbar, "TITLE", "Processing %d/%d (%.1f%%)\r",
                                                {i,length(tasks),i/length(tasks)*100})
                t1 = time()+0.2
            end if
        end for
        curl_cleanup()
        if wastitle="" then
            IupSetStrAttribute(statusbar, "TITLE", "%d tasks loaded",{length(task_table)})
        end if
        ttdx = 1
        bFilter = true
        return IUP_DEFAULT  -- (brb)
    end if
    -- filtering:
    integer ls = IupGetInt(langsel,"VALUE")
    sequence cs = repeat(-1,length(langchks))
    for i=1 to length(langchks) do
        cs[i] = IupGetInt(langchks[i],"VALUE")
    end for
    string tf = IupGetAttribute(taskfilter,"VALUE")
    integer ts = IupGetInt(textsearch,"VALUE"),
            bcount = 0
    if ttdx=1 then data = {} end if
    data_idx = 0
    IupTableClearSelected(table)
    IupSetInt(table,"ACTIVE",false)
    disable_radios()
    for i=ttdx to length(task_table) do
        sequence ti = task_table[i],
                 starts = ti[$-1],
                 finishes = ti[$]
        if ls=1 or (ls>1 and starts[ls-1]!=0) then
            bool bHas = true
            for c=1 to length(langchks) do
                bHas = (starts[c]!=0 or not cs[c])
                if not bHas then exit end if
            end for
            if bHas then
                bool bText = (tf=="")
                if not bText then
                    bText = match(tf,ti[1],case_sensitive:=false)!=0
                    if not bText and ls>1 and ts then
                        bcount += 1
--                      if bcount>=10 then exit end if
                        if bcount>=25 then exit end if -- (marginally better)
--                      if bcount>=50 then exit end if
                        set_blocks(ti,ls-1)
                        for b=1 to length(blocks) do
                            bText = match(tf,blocks[b],case_sensitive:=false)
                            if bText then exit end if
                        end for
                    end if
                end if  
                if bText then
                    data = append(data,ti)
                end if
            end if
        end if
        ttdx += 1
    end for
    string title
    integer res = IUP_DEFAULT,
            ltt = length(task_table)
    if ttdx>ltt then
        title = sprintf("%d/%d tasks filtered",{length(data),ltt})
        data = {data,{}}
        IupTableSetData(table, data)
        IupSetInt(table,"ACTIVE",true)
--      enable_radios(true) -- no, when user (re-) selects an entry
        res = IUP_IGNORE    -- remove callback
    else
        title = sprintf("filtering %d/%d",{ttdx,ltt})
    end if
    IupSetStrAttribute(statusbar, "TITLE", title)
    return res
end function

procedure apply_filters()
    ttdx = 1
    IupSetGlobalFunction("IDLE_ACTION", Icallback("load_tasks"))
end procedure

function valuechanged_cb(Ihandle ih)
    integer v = IupGetInt(ih,"VALUE")
    if ih=langsel then
        bool bActive = v>1 and not disabled[v-1]
        IupSetInt(textsearch,"ACTIVE",bActive)
        if bActive then
            -- mark/unmark and disable/enable
            -- eg if the dropdown is(/was) Go then that checkbox
            --    would be checked and disabled; if I then change
            --    it to Julia, then enable Go and disable Julia, 
            --    and also uncheck Go (why not) and check Julia.
            v -= 1
            integer sanity_count = 0
            for i=1 to length(langchks) do
                if not disabled[i]
                and IupGetInt(langchks[i],"ACTIVE")=(i=v) then
                    IupSetInt(langchks[i],"VALUE",(i=v))
                    IupSetInt(langchks[i],"ACTIVE",(i!=v))
                    sanity_count += 1
                end if
            end for
            if sanity_count>2 then ?9/0 end if -- what?!
        elsif v=1 then
            -- re-enable and uncheck, when "<any>" re-selected.
            for i=1 to length(langchks) do
                if not disabled[i]
                and not IupGetInt(langchks[i],"ACTIVE") then
                    IupSetInt(langchks[i],"ACTIVE",true)
                    IupSetInt(langchks[i],"VALUE",false)
                end if
            end for
        end if
    end if
    apply_filters()
    return IUP_DEFAULT
end function
constant cb_valuechanged = Icallback("valuechanged_cb")

function enteritem_cb(Ihandle table, integer lin, col)
    {} = IupTableEnterItem_cb(table,lin,col) -- as per docs
    data_idx = IupTableGetSelected(table)
    if data_idx then
        -- enable/disable radio buttons and if necessary
        -- transfer selected to something which is legal.
        sequence di = data[1][data_idx],
                 starts = di[$-1]
        integer badx = 0, validx = 0
        for i=1 to length(radioset) do
            if not disabled[i] then
                bool bActive = (starts[i]!=0)
--              bool bActive = (di[i+1]=="Y") -- (same)
                IupSetInt(radioset[i],"ACTIVE",bActive)
                if bActive then
                    if badx!=0 then
                        -- unset that by setting this
                        IupSetInt(radioset[i],"VALUE",1)
                        badx = 0 -- (not really needed)
                    elsif validx=0 then
                        -- unset future by setting this
                        validx = i
                    end if
                elsif IupGetInt(radioset[i],"VALUE") then
                    if validx!=0 then
                        -- unset by setting prior
                        IupSetInt(radioset[validx],"VALUE",1)
                    else
                        -- set future to unset this
                        badx = i
                    end if
                end if
            end if
        end for
        {} = set_codetext()
    end if
    return IUP_DEFAULT
end function

function radiochanged_cb(Ihandle ih)
    if IupGetInt(ih,"VALUE") then -- (ignore the automatic "unsets")
        {} = set_codetext()
    end if
    return IUP_DEFAULT
end function
constant cb_radiochanged = Icallback("radiochanged_cb")

function filter_action_cb(Ihandle /*ih*/)
    apply_filters()
    return IUP_DEFAULT
end function
constant cb_filter_action = Icallback(routine_id("filter_action_cb"))

function key_cb(Ihandle /*ih*/, atom c)
    if    c=K_F1 then return help()
    elsif c=K_F5 or c=K_F6 then return run(c)
    elsif c=K_LEFT then return set_codetext(max(1,blockn-1))
    elsif c=K_RIGHT then return set_codetext(min(blockn+1,length(blocks)))
    end if
    return iff(c=K_ESC?IUP_CLOSE:IUP_CONTINUE)
end function

procedure main()
    IupOpen()
    IupSetGlobal("UTF8MODE","YES")
    langsel = IupList("DROPDOWN=YES, 1=<any>, VALUE=1")
    IupSetCallback(langsel,"VALUECHANGED_CB",cb_valuechanged)
    sequence status_msgs = {}
    for i=1 to length(langs) do
        {string lang, string d, sequence cmd} = languages[i]
        IupSetStrAttributeId(langsel,"",i+1,lang)
        Ihandle checkbox = IupToggle(lang),
                radiobtn = IupToggle(lang)
        bool bOK = true
        if cmd!="open" then
            bOK = (get_file_type(d)==FILETYPE_DIRECTORY)
            if bOK then
                if string(cmd) then
                    bOK = file_exists(join_path({d,cmd}))
                else
                    for c=1 to length(cmd) do
                        bOK = file_exists(join_path({d,cmd[c]}))
                        if not bOK then exit end if
                    end for
                end if
            end if
            if not bOK then
                disabled[i] = true
                status_msgs = append(status_msgs,lang)
            end if
        end if
        IupSetInt({checkbox,radiobtn},"ACTIVE",bOK)
        langchks = append(langchks,checkbox)
        radioset = append(radioset,radiobtn)
    end for
    IupSetCallback(langchks,"VALUECHANGED_CB",cb_valuechanged)
    IupSetCallback(radioset,"VALUECHANGED_CB",cb_radiochanged)
    IupSetInt(langsel,"VISIBLEITEMS",length(langs)+2)
    sequence langset = {IupLabel("Language:"),langsel,
                        IupLabel("and:")} & langchks
    codetext = IupMultiLine(`VISIBLELINES=20, VISIBLECOLUMNS=80`)
    IupSetAttributes(codetext,`EXPAND=YES, FONT="Courier, 10"`)
    taskfilter = IupText("EXPAND=HORIZONTAL")
    textsearch = IupToggle("search text","ACTIVE=NO")
    IupSetCallback(taskfilter,"VALUECHANGED_CB",cb_filter_action)
    IupSetCallback(textsearch,"VALUECHANGED_CB",cb_valuechanged)
    add_pause = IupToggle("add pause")
    sequence columns = {{"Task",200,"ALEFT"}}
    for i=1 to length(langs) do
        columns = append(columns,{langs[i],40,"ACENTER"})
    end for
    table = IupTable(columns,{{},{}})
    statusbar = IupLabel("","EXPAND=HORIZONTAL, PADDING=10x5")
    Ihandle hbfilt = IupHbox({IupLabel("Filter:"),taskfilter,textsearch},
                             "NORMALIZESIZE=VERTICAL"),
            hlangs = IupHbox(langset,"GAP=20, NORMALIZESIZE=VERTICAL"),
            radios = IupRadio(IupHbox(radioset)),
            hradio = IupHbox({radios,IupFill(),add_pause},
                             "NORMALIZESIZE=VERTICAL"),
            dlg = IupDialog(IupVbox({hlangs,
                                     hbfilt,
                                     table,
                                     hradio,
                                     codetext,
                                     statusbar}),
                            "MARGIN=10x10, GAP=5, SHRINK=YES") 
    IupSetCallback(table,"ENTERITEM_CB",Icallback("enteritem_cb"))
    IupSetAttribute(dlg,"TITLE","Run examples")
    IupSetCallback(dlg,"KEY_CB",Icallback("key_cb"))
    IupSetAttributeHandle(NULL,"PARENTDIALOG",dlg)
    IupSetInt(radioset,"ACTIVE",false)
    for i=1 to length(radioset) do
        if not disabled[i] then
            IupSetInt(radioset[i],"VALUE",1)
            exit
        end if
    end for
    clipboard = IupClipboard()
    IupShow(dlg)
    if length(status_msgs) then
        string msg = "***NEED EDITING***: "&join(status_msgs,",")
        IupSetStrAttribute(statusbar, "TITLE", msg)
    end if
    apply_filters() -- (set load_tasks() as the idle action)
    if platform()!=JS then -- (no chance, but code consistently)
        IupMainLoop()
        IupClose()
    end if
end procedure
main()

Raku

(formerly Perl 6)

Works with: Rakudo version 2018.03

This is a fairly comprehensive task code runner. It is set up to work for Raku by default, but has basic configurations to run Perl, Python, Tcl and Go tasks as well. It can be easily tweaked to work with other languages by adding a load-lang('whatever'){} routine similar to the Raku, Perl, Python, Tcl and Go ones. (And ensuring that the appropriate compiler is installed and accessible.) There is so much variation to the task requirements and calling conventions that it would be problematic to make a general purpose, language agnostic code runner so some configuration is necessary to make it work with other languages.

(Note that there is no dependency download code nor resource hash for Python and Go, though they can run a remarkable number of tasks without.)

By default, this will download the Raku section of any (every) task that has a Raku example, extract the code blocks and attempt to run them. Many tasks require files or user interaction to proceed, others are not complete runnable code blocks (example code fragments), some tasks run forever. To try to deal with and compensate for this, this implementation can load a %resource hash that will: supply input files where necessary, skip unrunnable code fragments, limit long and/or infinite running blocks, supply user interaction code where possible, and skip blocks where user interaction is unavoidable.

There are several command line options to control its actions. See the README in the repository for details.

The complete implementation is too large and cumbersome to post in it's entirety here, only the main task retrieval and execution code is included.

For the whole ball of wax see the rc-run github repository.

Run with no parameters to run every implemented task on Rosetta Code. Feed it a task name to only download / run that task. Give it command line switches to adjust its behaviour.

Note: This is set up to run under Linux. It could be adapted for Windows (or OSX I suppose) fairly easily but I don't have access to those OSs, nor do I care to seek it.

use HTTP::UserAgent;
use URI::Escape;
use JSON::Fast;
use Text::Levenshtein::Damerau;
use MONKEY-SEE-NO-EVAL;

#####################################
say "Version = 2020-03-15T12:15:31";
#####################################

sleep 1;

my %*SUB-MAIN-OPTS = :named-anywhere;

unit sub MAIN(
    Str $run = '',        #= Task or file name
    Str :$lang = 'raku',  #= Language, default raku - used to load configuration settings
    Int :$skip = 0,       #= Skip # to continue partially into a list
    Bool :f(:$force),     #= Override any task skip parameter in %resource hash
    Bool :l(:$local),     #= Only use code from local cache
    Bool :r(:$remote),    #= Only use code from remote server (refresh local cache)
    Bool :q(:$quiet),     #= Less verbose, don't display source code
    Bool :d(:$deps),      #= Load dependencies
    Bool :p(:$pause),     #= pause after each task
    Bool :b(:$broken),    #= pause after each task which is broken or fails in some way
    Int  :$sleep = 0,     #= sleep for $sleep after each task
    Bool :t(:$timer),     #= save timing data for each task
);

die 'You can select local or remote, but not both...' if $local && $remote;

## INITIALIZATION

my $client   = HTTP::UserAgent.new;
my $url      = 'https://rosettacode.org/w';

my %c = ( # text colors
    code  => "\e[0;92m", # green
    delim => "\e[0;93m", # yellow
    cmd   => "\e[1;96m", # cyan
    bad   => "\e[0;91m", # red
    warn  => "\e[38;2;255;155;0m", # orange
    dep   => "\e[38;2;248;24;148m", # pink
    clr   => "\e[0m",    # clear formatting
);

my $view      = 'xdg-open';       # image viewer, this will open default under Linux
my %l         = load-lang($lang); # load language parameters
my %resource  = load-resources($lang);
my $get-tasks = True;

my @tasks;

run('clear');

## FIGURE OUT WHICH TASKS TO RUN

if $run {
    if $run.IO.e and $run.IO.f {# is it a file?
        @tasks = $run.IO.lines; # yep, treat each line as a task name
    } else {                    # must be a single task name
        @tasks = ($run);        # treat it so
    }
    $get-tasks = False;         # don't need to retrieve task names from web
}

if $get-tasks { # load tasks from web if cache is not found, older than one day or forced
    if !"%l<dir>.tasks".IO.e or (now - "%l<dir>.tasks".IO.modified) > 86400 or $remote {
        note 'Retrieving task list from site.';
        @tasks = mediawiki-query( # get tasks from web
        $url, 'pages',
        :generator<categorymembers>,
        :gcmtitle("Category:%l<language>"),
        :gcmlimit<350>,
        :rawcontinue(),
        :prop<title>
<title>.grep( * !~~ /^'Category:'/ ).sort;
        "%l<dir>.tasks".IO.spurt: @tasks.sort.join("\n");
    } else {
        note 'Using cached task list.';
        @tasks = "%l<dir>.tasks".IO.slurp.lines; # load tasks from file
    }
}

my $tfile;
if $timer {
    $tfile = open :w, "{$lang}-time.txt";
    $tfile.close;
}

note "Skipping first $skip tasks..." if $skip;
my $redo;

## MAIN LOOP

for @tasks -> $title {
    $redo = False;
    next if $++ < $skip;
    next unless $title ~~ /\S/; # filter blank lines (from files)
    say my $tasknum = $skip + ++$, ")  $title";

    my $name = $title.subst(/<-[-0..9A..Za..z]>/, '_', :g);
    my $taskdir = "./rc/%l<dir>/$name";

    my $modified = "$taskdir/$name.txt".IO.e ?? "$taskdir/$name.txt".IO.modified !! 0;

    my $entry;
    if $remote or !"$taskdir/$name.txt".IO.e or ((now - $modified) > 86400 * 7) {
        my $page = $client.get("{ $url }/index.php?title={ uri-escape $title }&action=raw").content;

        uh-oh("Whoops, can't find page: $url/$title :check spelling.\n\n{fuzzy-search($title)}", 'warn')
            and next if $page.elems == 0;
        say "Getting code from: https://rosettacode.org/wiki/{ $title.subst(' ', '_', :g) }#%l<language>";

        $entry = $page.comb(rx:i/'=={{header|' $(%l<header>) '}}==' .+? [<?before \n'=='<-[={]>*'{{header'> || $] /).Str //
          uh-oh("No code found\nMay be bad markup", 'warn');

        if $entry ~~ /^^ 'See [[' (.+?) '/' $(%l<language>) / { # no code on main page, check sub page
            $entry = $client.get("{ $url }/index.php?title={ uri-escape $/[0].Str ~ '/' ~ %l<language> }&action=raw").content;
        }
        mkdir $taskdir unless $taskdir.IO.d;
        spurt( "$taskdir/$name.txt", $entry );
    } else {
        if "$taskdir/$name.txt".IO.e {
            $entry = "$taskdir/$name.txt".IO.slurp;
            say "Loading code from: $taskdir/$name.txt";
        } else {
            uh-oh("Task code $taskdir/$name.txt not found, check spelling or run remote.", 'warn');
            next;
        }
    }

    my @blocks = $entry.comb: %l<tag>;

    unless @blocks {
        uh-oh("No code found\nMay be bad markup", 'warn') unless %resource{"$name"}<skip> ~~ /'ok to skip'/;
        say "Skipping $name: ", %resource{"$name"}<skip>, "\n" if %resource{"$name"}<skip>
    }

    for @blocks.kv -> $k, $v {
        my $n = +@blocks == 1 ?? '' !! $k;
        spurt( "$taskdir/$name$n%l<ext>", $v );
        if %resource{"$name$n"}<skip> && !$force {
            dump-code ("$taskdir/$name$n%l<ext>");
            if %resource{"$name$n"}<skip> ~~ /'broken'/ {
                uh-oh(%resource{"$name$n"}<skip>, 'bad');
                pause if $broken;
            } else {
                say "{%c<warn>}Skipping $name$n: ", %resource{"$name$n"}<skip>, "{%c<clr>}\n";
            }
            next;
        }
        say "\nTesting $name$n";
        run-it($taskdir, "$name$n", $tasknum);
    }
    say  %c<delim>, '=' x 79, %c<clr>;
    redo if $redo;
    sleep $sleep if $sleep;
    pause if $pause;
}

## SUBROUTINES

sub mediawiki-query ($site, $type, *%query) {
    my $url = "$site/api.php?" ~ uri-query-string(
        :action<query>, :format<json>, :formatversion<2>, |%query);
    my $continue = '';

    gather loop {
        my $response = $client.get("$url&$continue");
        my $data = from-json($response.content);
        take $_ for $data.<query>.{$type}.values;
        $continue = uri-query-string |($data.<query-continue>{*}».hash.hash or last);
    }
}

sub run-it ($dir, $code, $tasknum) {
    my $current = $*CWD;
    chdir $dir;
    if %resource{$code}<file> -> $fn {
        copy "$current/rc/resources/{$_}", "./{$_}" for $fn[]
    }
    dump-code ("$code%l<ext>") unless $quiet;
    check-dependencies("$code%l<ext>", $lang) if $deps;
    my @cmd = %resource{$code}<cmd> ?? |%resource{$code}<cmd> !! "%l<exe> $code%l<ext>\n";
    if $timer {
        $tfile = open :a, "{$current}/{$lang}-time.txt";
    }
    my $time = 'NA: not run or killed before completion';
    for @cmd -> $cmd {
        say "\nCommand line: {%c<cmd>}$cmd",%c<clr>;
        if $timer { $tfile.say: "Command line: $cmd".chomp }
        my $start = now;
        try shell $cmd;
        $time = (now - $start).round(.001);
        CATCH {
            when /'exit code: 137'/ { }
            default {
                .resume unless $broken;
                uh-oh($_, 'bad');
                if %resource{$code}<fail-by-design> {
                    say %c<warn>, 'Fails by design, (or at least, it\'s not unexpected).', %c<clr>;
                } else {
                    if pause.lc eq 'r' {
                       unlink "$code.txt";
                       $redo = True;
                    }
                }
             }
        }
    if $timer { $tfile.say("#$tasknum - Wallclock seconds: $time\n") }
    }
    chdir $current;
    say "\nDone task #$tasknum: $code - wallclock seconds: $time\e[?25h";
    $tfile.close if $timer;
}

sub pause {
    prompt "Press enter to procede:> ";
    # or
    # sleep 5;
}

sub dump-code ($fn) {
    say "\n", %c<delim>, ('vvvvvvvv' xx 7).join(' CODE '), %c<clr>, "\n", %c<code>;
    print $fn.IO.slurp;
    say %c<clr>,"\n\n",%c<delim>,('^^^^^^^^' xx 7).join(' CODE '),%c<clr>;
}

sub uri-query-string (*%fields) { %fields.map({ "{.key}={uri-escape .value}" }).join('&') }

sub clear { "\r" ~ ' ' x 100 ~ "\r" }

sub uh-oh ($err, $class='warn') { put %c{$class}, "{'#' x 79}\n\n $err \n\n{'#' x 79}", %c<clr> }

sub fuzzy-search ($title) {
    my @tasknames;
    if "%l<dir>.tasks".IO.e {
        @tasknames = "%l<dir>.tasks".IO.slurp.lines;
    }
    return '' unless @tasknames.elems;
    " Did you perhaps mean:\n\n\t" ~
    @tasknames.grep( {.lc.contains($title.lc) or dld($_, $title) < (5 min $title.chars)} ).join("\n\t");
}                # Damerau Levenshtein distance  ^^^

multi check-dependencies ($fn, 'raku') {
    my @use = $fn.IO.slurp.comb(/<?after ^^ \h* 'use '> \N+? <?before \h* ';'>/);
    if +@use {
        say %c<dep>, 'Checking dependencies...', %c<clr>;
        for @use -> $module {
            if $module eq any('v6', 'v6.c', 'v6.d', 'nqp', 'NativeCall', 'Test') or $module.contains('MONKEY')
              or $module.contains('experimental') or $module.starts-with('lib') or $module.contains('from<Perl5>') {
                print %c<dep>;
                say 'ok, no installation necessary: ', $module;
                print %c<clr>;
                next;
            }
            my $installed = $*REPO.resolve(CompUnit::DependencySpecification.new(:short-name($module)));
            my @mods = $module;
            if './../../../raku-modules.txt'.IO.e {
                my $fh = open( './../../../perl6-modules.txt', :r ) or die $fh;
                @mods.append: $fh.lines;
                $fh.close;
            }
            my $fh = open( './../../../raku-modules.txt', :w ) or die $fh;
            $fh.spurt: @mods.Bag.keys.sort.join: "\n";
            $fh.close;
            print %c<dep>;
            if $installed {
                say 'ok, installed: ', $module
            } else {
                say 'not installed: ', $module;
                shell("zef install $module");
            }
            print %c<clr>;
        }
    }
}

multi check-dependencies ($fn, 'perl') {
    my @use = $fn.IO.slurp.comb(/<?after ^^ \h* 'use '> \N+? <?before \h* ';'>/);
    if +@use {
        for @use -> $module {
            next if $module eq $module.lc;
            next if $module.starts-with(any('constant','bignum'));
            my $installed = shell( "%l<exe> -e 'eval \"use {$module}\"; exit 1 if \$@'" );
            print %c<dep>;
            if $installed {
                say 'ok:            ', $module
            } else {
                say 'not installed: ', $module;
                try shell("sudo cpan $module");
            }
            print %c<clr>;
        }
    }
}

multi check-dependencies  ($fn, $unknown) {
    note "Sorry, don't know how to handle dependencies for $unknown language."
};

multi load-lang ('raku') { ( # Language specific variables. Adjust to suit.
    language => 'Raku',  # language category name
    exe      => 'raku',  # executable name to run perl6 in a shell
    ext      => '.raku', # file extension for perl6 code (optional, but nice to have)
    dir      => 'raku',  # directory to save tasks to
    header   => 'Raku',  # header text (=={{header|Raku}}==)
    # tags marking blocks of code - spaced out to placate wiki formatter
    # and to avoid getting tripped up when trying to run _this_ task.
    # note that this tag only selects the syntax highlighting, continue to
    # leave 'perl6' as an option, 'raku' is now live on the site.
    tag => rx/:i <?after '<syntaxhighlight lang="' ['perl6'|'raku'] '"' ' line'? '>' > .*? <?before '</' 'syntaxhighlight>'>/,
) }

multi load-lang ('perl') { (
    language => 'Perl',
    exe      => 'perl',
    ext      => '.pl',
    dir      => 'perl',
    header   => 'Perl',
    tag => rx/:i <?after '<syntaxhighlight lang="' 'perl' '"' ' line'? '>' > .*? <?before '</' 'lang>'>/,
) }

multi load-lang ('python') { (
    language => 'Python',
    exe      => 'python',
    ext      => '.py',
    dir      => 'python',
    header   => 'Python',
    tag => rx/:i <?after '<syntaxhighlight lang="' 'python' '"' ' line'? '>' > .*? <?before '</' 'lang>'>/,
) }

multi load-lang ('go') { (
    language => 'Go',
    exe      => 'go run',
    ext      => '.go',
    dir      => 'go',
    header   => 'Go',
    tag => rx/:i <?after '<syntaxhighlight lang="' 'go'  '"' ' line'? '>' > .*? <?before '</' 'lang>'>/,
) }

multi load-lang ('tcl') { (
    language => 'Tcl',
    exe      => 'tclsh',
    ext      => '.tcl',
    dir      => 'tcl',
    header   => 'Tcl',
    tag => rx/:i <?after '<syntaxhighlight lang="' 'tcl'  '"' ' line'? '>' > .*? <?before '</' 'lang>'>/,
) }

multi load-lang ($unknown) { die "Sorry, don't know how to handle $unknown language." };

multi load-resources ($unknown) { () };
Output:
with command line
raku RC-run.p6 -q "Determine if a string is numeric"
Retrieving tasks
1)  Determine if a string is numeric
Getting code from: http://rosettacode.org/wiki/Determine_if_a_string_is_numeric#Raku

Testing Determine_if_a_string_is_numeric

Command line: raku Determine_if_a_string_is_numeric.p6

               Coerce     Don't coerce
    String   whitespace    whitespace
       <1>      True         True
     <1.2>      True         True
   <1.2.3>     False        False
      <-6>      True         True
     <1/2>      True         True
     <12e>     False        False
     <B17>     False        False
 <1.3e+12>      True         True
  <1.3e12>      True         True
 <-2.6e-3>      True         True
    <zero>     False        False
      <0x>     False        False
   <0xA10>      True         True
  <0b1001>      True         True
    <0o16>      True         True
    <0o18>     False        False
    <2+5i>      True         True
    <True>     False        False
   <False>     False        False
     <Inf>      True         True
     <NaN>      True         True
 <0x10.50>      True         True
   <0b102>     False        False
  <0o_5_3>      True         True
      <௫௯>      True         True
  <  12  >      True         True
   <1 1 1>     False        False
        <>      True        False
       < >      True        False

Done task #1: Determine_if_a_string_is_numeric - wallclock seconds: 0.171
===============================================================================

Or, if the full %resource hash is loaded it will automatically feed input parameters to tasks that require them:
raku RC-run.p6 Lucky_and_even_lucky_numbers -q

Retrieving tasks
1 Lucky_and_even_lucky_numbers
Getting code from: http://rosettacode.org/wiki/Lucky_and_even_lucky_numbers#Raku

Testing Lucky_and_even_lucky_numbers

Command line: raku Lucky_and_even_lucky_numbers.p6 20 , lucky

79

Command line: raku Lucky_and_even_lucky_numbers.p6 1 20

(1 3 7 9 13 15 21 25 31 33 37 43 49 51 63 67 69 73 75 79)

Command line: raku Lucky_and_even_lucky_numbers.p6 1 20 evenlucky

(2 4 6 10 12 18 20 22 26 34 36 42 44 50 52 54 58 68 70 76)

Command line: raku Lucky_and_even_lucky_numbers.p6 6000 -6100

(6009 6019 6031 6049 6055 6061 6079 6093)

Command line: raku Lucky_and_even_lucky_numbers.p6 6000 -6100 evenlucky

(6018 6020 6022 6026 6036 6038 6050 6058 6074 6090 6092)

Done Lucky_and_even_lucky_numbers
===============================================================================

Try running a Go task - Command line: raku RC-run.p6 -q --lang=go "Determine if a string is numeric"

Finds two task entries, downloads both and runs each:

1)  Determine if a string is numeric
Getting code from: http://rosettacode.org/wiki/Determine_if_a_string_is_numeric#Go

Testing Determine_if_a_string_is_numeric0

Command line: go run Determine_if_a_string_is_numeric0.go

Are these strings numeric?
     1 -> true
  3.14 -> true
  -100 -> true
   1e2 -> true
   NaN -> true
  rose -> false

Done task #1: Determine_if_a_string_is_numeric0 - wallclock seconds: 0.16

Testing Determine_if_a_string_is_numeric1

Command line: go run Determine_if_a_string_is_numeric1.go

Are these strings integers?
    1 -> true
  one -> false

Done task #1: Determine_if_a_string_is_numeric1 - wallclock seconds: 0.147
===============================================================================

Run BASIC

bf$	= "<SPAN STYLE='font-family:Arial; font-weight:700; font-size:12pt'>"
a$	= httpGet$("http://rosettacode.org/wiki/Category:Run_BASIC") ' get RB tasks from [RC]
a1$	= word$(a$,2,"Pages in category ""Run BASIC")
a1$	= word$(a1$,1,"</tr></table>")
i	= 2
b$	= word$(a1$,i,"<li><a href=""/wiki/")
'
' Create a drop down window for selection of a task
'
html bf$;"<center><TABLE BORDER=1 CELLPADDING=0 CELLSPACING=0 bgcolor=wheat>"
html "<TR align=center><TD colspan=2>Tasks</TD></TR><TR>"
html "<TD align=right>Task</TD><TD>"

html "<select size=10 id='runProg' name='runProg'>" 
while b$ <> ""
  b$	= left$(b$,instr(b$,"""")-1)
  b$	= strRep$(b$,"%2B","+")
  b$	= strRep$(b$,"%27","'")
  html "<option>"+b$+"</option>"
  i 	= i + 1
  b$	= word$(a1$,i,"<li><a href=""/wiki/")
wend
html "</select></TD></TR><TR><TD colspan=2 ALIGN=CENTER>"

' BUTTON options to Run It or Exit
    button #run, "Run It", [runProg]
    button #ex, "Exit", [quit]

html "</TD></TR></TABLE></center>"  ' close the drop down table and wait
wait

[runProg]
progName$	= #request get$("runProg")
print progName$
a$	= httpGet$("http://rosettacode.org/wiki/"+progName$)

i	= instr(a$,"<a href=""#Run_BASIC"">")

a$	= mid$(a$,i-6,6)
a$	= word$(a$,2,"-")
a$	= word$(a$,1,"""")
cls                            ' clear screen
'print a$                      ' this is the program number used in the [RC] editor

a$ = httpGet$("http://rosettacode.org/mw/index.php?title="+progName$+"&action=edit&section="+a$)

a$	= word$(a$,2,"{header|Run BASIC}")
i	= instr(a$,">")
a$	= mid$(a$,i+1)
i 	= instr(a$,"/lang>")
a$	= left$(a$,i-5)
a$	= strRep$(a$,"&lt;","<")               ' this is the good program code
' place the code in the rb$ file
rb$     = DefaultDir$ + "\projects\a_project\rcCode.bas" ' RC program
open rb$ for output as #f
print #f,a$
close #f

print "================== Run Basic Solution ==========================="
run rb$,#handle      ' point RunBasic to the file with the program
render #handle       ' render the runned code
[quit]               ' that's it folks
end

' --------------------------------
' string replace rep str with
' --------------------------------
FUNCTION strRep$(str$,rep$,with$)
ln  = len(rep$)
ln1 = ln - 1
i = 1
while i <= len(str$)
    if mid$(str$,i,ln) = rep$ then
        strRep$ = strRep$ + with$
        i = i + ln1
    else
        strRep$ = strRep$ + mid$(str$,i,1)
    end if
i = i + 1
WEND
END FUNCTION

Tcl

This code only includes support for running Tcl task solutions, but it can download any language's; it assumes that the first <lang…> is sufficient when it comes to task extraction (definitely not true universally, but mostly good enough).

Library: Tcllib (Package: uri)
# Code to download task contents from find-bare-lang-tags task
package require Tcl 8.5
package require http
package require uri

proc getUrlWithRedirect {base args} {
    set url $base?[http::formatQuery {*}$args]
    while 1 {
	set t [http::geturl $url]
	if {[http::status $t] ne "ok"} {
	    error "Oops: url=$url\nstatus=$s\nhttp code=[http::code $token]"
	}
	if {[string match 2?? [http::ncode $t]]} {
	    return $t
	}
	# OK, but not 200? Must be a redirect...
	set url [uri::resolve $url [dict get [http::meta $t] Location]]
	http::cleanup $t
    }
}
proc getTaskContent {task} {
    set token [getUrlWithRedirect http://rosettacode.org/mw/index.php \
	    title $task action raw]
    set content [http::data $token]
    http::cleanup $token
    return $content
}

# Code to extract the first <syntaxhighlight lang="text"> section for a language
proc getTaskCodeForLanguage {task language} {
    set content [getTaskContent $task]
    set startRE {==\s*\{\{header\|@LANG@(?:\|[^{}]+)?\}\}\s*==}
    set startRE [string map [list @LANG@ $language] $startRE]
    if {![regexp -indices $startRE $content start]} {
	error "$language does not implement task \"$task\""
    }
    if {![regexp -indices -start [lindex $start end] \
	      "==\\s*\\\{\\\{header" $content end]} {
	set end {end end}
    }
    set content [string range $content [lindex $start 1] [lindex $end 0]]
    # Extended format RE used to allow embedding within _this_ task's <syntaxhighlight lang="text">!
    if {![regexp {(?x)<syntaxhighlight lang=".*?">(.*?)</ lang>} $content -> solution]} {
	error "$language solution of task \"$task\" has no useful code"
    }
    return "$solution\n"
}

# How to download and run a Tcl task
proc runTclTaskForLanguage {task} {
    puts "Fetching task solution..."
    set solution [getTaskCodeForLanguage $task Tcl]
    set filename rcsoln_[string map {/ _ " " _} $task].tcl
    set f [open $filename w]
    puts $f $solution
    close $f
    puts "Executing task solution with: tclsh $filename"
    exec [info nameofexecutable] $filename <@stdin >@stdout 2>@stderr
}
runTclTaskForLanguage {*}$argv

UNIX Shell

See C1R Implementation for an incomplete implementation. (only supports C)

Wren

Library: WrenGo
Library: Wren-pattern
Library: Wren-str

An embedded program with a Go host as Wren-cli currently has no way to download web pages.

This is designed to work for Go, Perl, Python and Wren itself though see the remarks in the Go entry about using the first three languages and other more general points.

As far as Wren is concerned, in addition to Wren-cli programs, this should be able to run DOME, Wren-gmp, Wren-sql, Wren-i64, Wren-linear, Wren-regex and Wren-psieve programs as it is easy to detect when these are being used and the filename is always given as a command line argument. All Wren modules are assumed to be present in the current working directory.

However, no attempt has been made - at least for now - to run other embedded programs (which can be identified by the presence of the 'foreign' keyword) due to a number of technical difficulties in doing so.

/* Rosetta_Code_Run_examples.wren */

import "./pattern" for Pattern
import "./str" for Str

class Http {
    // gets the response body, copies it to a string and automatically closes it
    foreign static getBodyText(url)
}

class Html {
    foreign static unescapeString(s)
}

class Stdin {
    foreign static readLine()
}

class IOUtil {
    foreign static writeFile(fileName, text)

    foreign static removeFile(fileName)
}

class Exec {
    foreign static run2(lang, fileName)
    foreign static run3(lang, param, fileName)
}

var p1 = Pattern.new("title/=\"[+1^\"]\"")
var p2 = Pattern.new("cmcontinue/=\"[+1^\"]\"")

var findTasks = Fn.new { |category|
    var url = "https://www.rosettacode.org/w/api.php?action=query&list=categorymembers&cmtitle=Category:%(category)&cmlimit=500&format=xml"
    var cmcontinue = ""
    var tasks = []
    while (true) {
        var content = Http.getBodyText(url + cmcontinue)
        var matches1 = p1.findAll(content)
        for (m in matches1) {
            var title = m.capsText[0].replace("&#039;", "'").replace("&quot;", "\"")
            tasks.add(title)
        }
        var m2 = p2.find(content)
        if (m2) cmcontinue = "&cmcontinue=%(m2.capsText[0])" else break
    }
    return tasks
}

var tasks = findTasks.call("Programming_Tasks") // 'full' tasks only
tasks.addAll(findTasks.call("Draft_Programming_Tasks"))
var langs = ["go", "perl", "python", "wren"]
while (true) {
    System.write("Enter the exact name of the task : ")
    var task = Stdin.readLine().trim()
    if (!tasks.contains(task)) {
        System.print("Sorry a task with that name doesn't exist.")
    } else {
        task = task.replace(" ", "_")
        var url = "https://rosettacode.org/w/index.php?title=" + task + "&action=edit"
        var page = Http.getBodyText(url).replace("&lt;", "<")
        var lang
        while (true) {
            System.write("Enter the language Go/Perl/Python/Wren : ")
            lang = Str.lower(Stdin.readLine().trim())
            if (langs.contains(lang)) break
            System.print("Sorry that language is not supported.")
        }
        var lang2
        var lang3
        var ext
        if (lang == "go") {
            lang2 = "Go"
            lang3 = "[go|Go|GO]"
            ext   = "go"
        } else if (lang == "perl") {
            lang2 = "Perl"
            lang3 = "[perl|Perl]"
            ext   = "pl"
        } else if (lang == "python") {
            lang2 = "Python"
            lang3 = "[python|Python]"
            ext   = "py"
        } else if (lang == "wren") {
            lang2 = "Wren"
            lang3 = "[wren|Wren]"
            ext   = "wren"
        }
        var fileName = "rc_temp." + ext
        var p1 = Pattern.new("/=/={{header/|%(lang2)}}/=/=")
        var p2 = Pattern.new("<syntaxhighlight lang/=\"%(lang3)\">")
        var p3 = Pattern.new("<//syntaxhighlight>")
        var s = p1.split(page, 1, 0)
        if (s.count > 1) s = p2.split(s[1], 1, 0)
        var preamble = s[0]
        if (s.count > 1) s = p3.split(s[1], 1, 0)
        if (s.count == 1) {
            System.print("No runnable task entry for that language was detected.")
        } else if (lang == "wren" && s[0].contains(" foreign ")) {
            System.print("This is an embedded script which cannot be run automatically at present.")
        } else {
            var source = Html.unescapeString(s[0])
            System.print("\nThis is the source code for the first or only runnable program:\n")
            System.print(source)
            System.write("\nDo you want to run it y/n : ")
            var yn = Stdin.readLine().trim()[0]
            // note that the executable names may differ from the ones I'm currently using
            if (yn == "y" || yn == "Y") {
                IOUtil.writeFile(fileName, source)
                if (lang == "go") {
                    Exec.run3("go", "run", fileName)
                } else if (lang == "perl") {
                    Exec.run2("perl", fileName)
                } else if (lang == "python") {
                    Exec.run2("python3", fileName)
                } else if (lang == "wren") {
                    if (preamble.contains("{{libheader|DOME}}")) {
                        Exec.run2("dome171", fileName)
                    } else if (preamble.contains("{{libheader|Wren-gmp}}")) {
                        Exec.run2("./wren-gmp", fileName)
                    } else if (preamble.contains("{{libheader|Wren-sql}}")) {
                        Exec.run2("./wren-sql", fileName)
                    } else if (preamble.contains("{{libheader|Wren-i64}}")) {
                        Exec.run2("./wren-i64", fileName)
                    } else if (preamble.contains("{{libheader|Wren-linear}}")) {
                        Exec.run2("./wren-linear", fileName)
                    } else if (preamble.contains("{{libheader|Wren-regex}}")) {
                        Exec.run2("./wren-regex", fileName)
                    } else if (preamble.contains("{{libheader|Wren-psieve}}")) {
                        Exec.run2("./wren-psieve", fileName) 
                    } else { // Wren-cli
                        Exec.run2("wren4", fileName)
                    }
                }
                IOUtil.removeFile(fileName)
            }
        }
        System.write("\nDo another one y/n : ")
        var yn = Stdin.readLine().trim()[0]
        if (yn != "y" && yn != "Y") return
    }
}

We now embed this script in the following Go program and run it:

/* go run Rosetta_Code_Run_examples.go  */

package main

import (
    "bufio"
    wren "github.com/crazyinfin8/WrenGo"
    "html"
    "io/ioutil"
    "log"
    "net/http"
    "os"
    "os/exec"
    "strings"
)

type any = interface{}

var in = bufio.NewReader(os.Stdin)

func check(err error) {
    if err != nil {
        log.Fatal(err)
    }
}

func getBodyText(vm *wren.VM, parameters []any) (any, error) {
    url := parameters[1].(string)
    resp, _ := http.Get(url)
    body, _ := ioutil.ReadAll(resp.Body)
    resp.Body.Close()
    return string(body), nil
}

func unescapeString(vm *wren.VM, parameters []any) (any, error) {
    s := parameters[1].(string)
    return html.UnescapeString(s), nil
}

func readLine(vm *wren.VM, parameters []any) (any, error) {
    l, err := in.ReadString('\n')
    check(err)
    return l, nil
}

func writeFile(vm *wren.VM, parameters []any) (any, error) {
    fileName := parameters[1].(string)
    text := parameters[2].(string)
    err := ioutil.WriteFile(fileName, []byte(text), 0666)
    check(err)
    return nil, nil
}

func removeFile(vm *wren.VM, parameters []any) (any, error) {
    fileName := parameters[1].(string)
    err := os.Remove(fileName)
    check(err)
    return nil, nil
}

func run2(vm *wren.VM, parameters []any) (any, error) {
    lang := parameters[1].(string)
    fileName := parameters[2].(string)
    cmd := exec.Command(lang, fileName)
    cmd.Stdout = os.Stdout
    cmd.Stderr = os.Stderr
    cmd.Run()
    return nil, nil
}

func run3(vm *wren.VM, parameters []any) (any, error) {
    lang := parameters[1].(string)
    param := parameters[2].(string)
    fileName := parameters[3].(string)
    cmd := exec.Command(lang, param, fileName)
    cmd.Stdout = os.Stdout
    cmd.Stderr = os.Stderr
    cmd.Run()
    return nil, nil
}

func moduleFn(vm *wren.VM, name string) (string, bool) {
    if name != "meta" && name != "random" && !strings.HasSuffix(name, ".wren") {
        name += ".wren"
    }
    return wren.DefaultModuleLoader(vm, name)
}

func main() {
    cfg := wren.NewConfig()
    cfg.LoadModuleFn = moduleFn
    vm := cfg.NewVM()

    httpMethodMap := wren.MethodMap{
        "static getBodyText(_)": getBodyText,
    }

    htmlMethodMap := wren.MethodMap{
        "static unescapeString(_)": unescapeString,
    }

    stdinMethodMap := wren.MethodMap{
        "static readLine()": readLine,
    }

    ioutilMethodMap := wren.MethodMap{
        "static writeFile(_,_)": writeFile,
        "static removeFile(_)":  removeFile,
    }

    execMethodMap := wren.MethodMap{
        "static run2(_,_)":   run2,
        "static run3(_,_,_)": run3,
    }

    classMap := wren.ClassMap{
        "Http":   wren.NewClass(nil, nil, httpMethodMap),
        "Html":   wren.NewClass(nil, nil, htmlMethodMap),
        "Stdin":  wren.NewClass(nil, nil, stdinMethodMap),
        "IOUtil": wren.NewClass(nil, nil, ioutilMethodMap),
        "Exec":   wren.NewClass(nil, nil, execMethodMap),
    }

    module := wren.NewModule(classMap)
    fileName := "Rosetta_Code_Run_examples.wren"
    vm.SetModule(fileName, module)
    vm.InterpretFile(fileName)
    vm.Free()
}
Output:

Sample run:

Enter the exact name of the task : Palindrome dates
Enter the language Go/Perl/Python/Wren : Go

This is the source code for the first or only runnable program:

package main

import (
    "fmt"
    "time"
)

func reverse(s string) string {
    chars := []rune(s)
    for i, j := 0, len(chars)-1; i < j; i, j = i+1, j-1 {
        chars[i], chars[j] = chars[j], chars[i]
    }
    return string(chars)
}

func main() {
    const (
        layout  = "20060102"
        layout2 = "2006-01-02"
    )
    fmt.Println("The next 15 palindromic dates in yyyymmdd format after 20200202 are:")
    date := time.Date(2020, 2, 2, 0, 0, 0, 0, time.UTC)
    count := 0
    for count < 15 {
        date = date.AddDate(0, 0, 1)
        s := date.Format(layout)
        r := reverse(s)
        if r == s {
            fmt.Println(date.Format(layout2))
            count++
        }
    }
}

Do you want to run it y/n : y
The next 15 palindromic dates in yyyymmdd format after 20200202 are:
2021-12-02
2030-03-02
2040-04-02
2050-05-02
2060-06-02
2070-07-02
2080-08-02
2090-09-02
2101-10-12
2110-01-12
2111-11-12
2120-02-12
2121-12-12
2130-03-12
2140-04-12

Do another one y/n : y
Enter the exact name of the task : Palindrome dates
Enter the language Go/Perl/Python/Wren : Perl

This is the source code for the first or only runnable program:

use Time::Piece;
my $d = Time::Piece->strptime("2020-02-02", "%Y-%m-%d");

for (my $k = 1 ; $k <= 15 ; $d += Time::Piece::ONE_DAY) {
    my $s = $d->strftime("%Y%m%d");
    if ($s eq reverse($s) and ++$k) {
        print $d->strftime("%Y-%m-%d\n");
    }
}

Do you want to run it y/n : y
2020-02-02
2021-12-02
2030-03-02
2040-04-02
2050-05-02
2060-06-02
2070-07-02
2080-08-02
2090-09-02
2101-10-12
2110-01-12
2111-11-12
2120-02-12
2121-12-12
2130-03-12

Do another one y/n : y
Enter the exact name of the task : Palindrome dates
Enter the language Go/Perl/Python/Wren : Python

This is the source code for the first or only runnable program:

'''Palindrome dates'''

from datetime import datetime
from itertools import chain


# palinDay :: Int -> [ISO Date]
def palinDay(y):
    '''A possibly empty list containing the palindromic
       date for the given year, if such a date exists.
    '''
    s = str(y)
    r = s[::-1]
    iso = '-'.join([s, r[0:2], r[2:]])
    try:
        datetime.strptime(iso, '%Y-%m-%d')
        return [iso]
    except ValueError:
        return []


# --------------------------TEST---------------------------
# main :: IO ()
def main():
    '''Count and samples of palindromic dates [2021..9999]
    '''
    palinDates = list(chain.from_iterable(
        map(palinDay, range(2021, 10000))
    ))
    for x in [
            'Count of palindromic dates [2021..9999]:',
            len(palinDates),
            '\nFirst 15:',
            '\n'.join(palinDates[0:15]),
            '\nLast 15:',
            '\n'.join(palinDates[-15:])
    ]:
        print(x)


# MAIN ---
if __name__ == '__main__':
    main()

Do you want to run it y/n : y
Count of palindromic dates [2021..9999]:
284

First 15:
2021-12-02
2030-03-02
2040-04-02
2050-05-02
2060-06-02
2070-07-02
2080-08-02
2090-09-02
2101-10-12
2110-01-12
2111-11-12
2120-02-12
2121-12-12
2130-03-12
2140-04-12

Last 15:
9170-07-19
9180-08-19
9190-09-19
9201-10-29
9210-01-29
9211-11-29
9220-02-29
9221-12-29
9230-03-29
9240-04-29
9250-05-29
9260-06-29
9270-07-29
9280-08-29
9290-09-29

Do another one y/n : y
Enter the exact name of the task : Palindrome dates
Enter the language Go/Perl/Python/Wren : Wren

This is the source code for the first or only runnable program:

import "/fmt" for Fmt
import "/date" for Date

var isPalDate = Fn.new { |date|
    date = date.format(Date.rawDate)
    return date == date[-1..0]
}

Date.default = Date.isoDate
System.print("The next 15 palindromic dates in yyyy-mm-dd format after 2020-02-02 are:")
var date = Date.new(2020, 2, 2)
var count = 0
while (count < 15) {
    date = date.addDays(1)
    if (isPalDate.call(date)) {
        System.print(date)
        count = count + 1
    }
}

Do you want to run it y/n : y
The next 15 palindromic dates in yyyy-mm-dd format after 2020-02-02 are:
2021-12-02
2030-03-02
2040-04-02
2050-05-02
2060-06-02
2070-07-02
2080-08-02
2090-09-02
2101-10-12
2110-01-12
2111-11-12
2120-02-12
2121-12-12
2130-03-12
2140-04-12

Do another one y/n : y
Enter the exact name of the task : Deceptive numbers
Enter the language Go/Perl/Python/Wren : Wren

This is the source code for the first or only runnable program:

/* deceptive_numbers.wren */

import "./gmp" for Mpz
import "./math" for Int

var count = 0
var limit = 25
var n = 17
var repunit = Mpz.from(1111111111111111)
var deceptive = []
while (count < limit) {
    if (!Int.isPrime(n) && n % 3 != 0 && n % 5 != 0) {
        if (repunit.isDivisibleUi(n)) {
            deceptive.add(n)
            count = count + 1
        }
    }
    n = n + 2
    repunit.mul(100).add(11)
}
System.print("The first %(limit) deceptive numbers are:")
System.print(deceptive)

Do you want to run it y/n : y
The first 25 deceptive numbers are:
[91, 259, 451, 481, 703, 1729, 2821, 2981, 3367, 4141, 4187, 5461, 6533, 6541, 6601, 7471, 7777, 8149, 8401, 8911, 10001, 11111, 12403, 13981, 14701]

Do another one y/n : n