From Newsgroup: comp.lang.c

There was a programming-"contest" on comp.lang.c and I wanted to show
the simpler code in C, here it is:

#include <iostream>
#include <sstream>
#include <iomanip>
#include <optional>
#include <algorithm>

using namespace std;

static optional<size_t> parse( const char *str );

int main( int argc, char **argv )
{
if( argc < 3 )
return EXIT_FAILURE;
optional<size_t>
pRows = parse( argv[1] ),
pCols = parse( argv[2] );
if( !pRows || !pCols )
return EXIT_FAILURE;
size_t rows = *pRows, cols = *pCols;
optional<size_t> pClip( rows * cols );
if( argc >= 4 && !(pClip = parse( argv[3] )) )
return EXIT_FAILURE;
size_t clip = min( *pClip, rows * cols );
streamsize width = (ostringstream() << clip).str().length();
for( size_t row = 1; row <= min( rows, clip ); ++row )
{
bool head = true;
for( size_t value = row; value <= clip; value += rows, head = false )
cout << " "sv.substr( head, !head ) << right << setw( width ) << value;
cout << endl;
}
}

static optional<size_t> parse( const char *str )
{
istringstream iss( str );
size_t ret;
iss >> ret;
if( !iss || !iss.eof() )
return nullopt;
return ret;
}

C++ really rocks since you've to deal with much less details than in C.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 12.03.2026 um 07:24 schrieb Bonita Montero:

There was a programming-"contest" on comp.lang.c and I wanted to show
the simpler code in C, here it is:

^^^^^^^^^^^^^^^^^^^^^ C++

#include <iostream>
#include <sstream>
#include <iomanip>
#include <optional>
#include <algorithm>

using namespace std;

static optional<size_t> parse( const char *str );

int main( int argc, char **argv )
{
    if( argc < 3 )
        return EXIT_FAILURE;
    optional<size_t>
        pRows = parse( argv[1] ),
        pCols = parse( argv[2] );
    if( !pRows || !pCols )
        return EXIT_FAILURE;
    size_t rows = *pRows, cols = *pCols;
    optional<size_t> pClip( rows * cols );
    if( argc >= 4 && !(pClip = parse( argv[3] )) )
        return EXIT_FAILURE;
    size_t clip = min( *pClip, rows * cols );
    streamsize width = (ostringstream() << clip).str().length();
    for( size_t row = 1; row <= min( rows, clip ); ++row )
    {
        bool head = true;
        for( size_t value = row; value <= clip; value += rows, head =
false )
            cout << " "sv.substr( head, !head ) << right << setw( width ) << value;
        cout << endl;
    }
}

static optional<size_t> parse( const char *str )
{
    istringstream iss( str );
    size_t ret;
    iss >> ret;
    if( !iss || !iss.eof() )
        return nullopt;
    return ret;
}

C++ really rocks since you've to deal with much less details than in C.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/12/26 07:24, Bonita Montero wrote:

There was a programming-"contest" on comp.lang.c and I wanted to show
the simpler code in C[++], here it is:
[...]

C++ really rocks since you've to deal with much less details than in C.

Concerning your subject; it wouldn't appear to me to use the term
"beauty" in context of C++. C++ inherited so much ugly syntax and
concepts (from "C") and it added yet more syntactic infelicities.
(And I'm saying that as someone who liked to program in C++, also professionally, for many many years.)

You're right in the "details" aspect. C++ adds so much per default
that you just don't have in "C"; that's why I consider it pointless
to compare it in the first place, let alone to argument in a "C"
newsgroup about it. Don't expect missionaries to be welcome here.

C++ has also other benefits beyond the libraries; first of all the
OO concepts and whatnot. But all uninteresting for this group; "C"
won't ever come close to C++. (And "beauty" should be searched in
other places.)

Janis

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 12.03.2026 um 09:32 schrieb Janis Papanagnou:

Concerning your subject; it wouldn't appear to me to use the term
"beauty" in context of C++. C++ inherited so much ugly syntax and
concepts (from "C") and it added yet more syntactic infelicities.
(And I'm saying that as someone who liked to program in C++, also professionally, for many many years.)

My perception of beauty is just as subjective as your perception of the
ugly aspects of C++. I like the span of lowleven means in C++ that are inherited from C to higher level abstractions C++ has inherited from
other more moderrn language.
And if you don't the subjektive part of beauty or ugly parts of this
language C++ is several times more effective than C, i.e. you need
only a fraction of the code size while mostly maintaining the same
performance.
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/12/2026 2:24 AM, Bonita Montero wrote:

There was a programming-"contest" on comp.lang.c and I wanted to show
the simpler code in C, here it is:

#include <iostream>
#include <sstream>
#include <iomanip>
#include <optional>
#include <algorithm>

using namespace std;

static optional<size_t> parse( const char *str );

int main( int argc, char **argv )
{
    if( argc < 3 )
        return EXIT_FAILURE;
    optional<size_t>
        pRows = parse( argv[1] ),
        pCols = parse( argv[2] );
    if( !pRows || !pCols )
        return EXIT_FAILURE;
    size_t rows = *pRows, cols = *pCols;
    optional<size_t> pClip( rows * cols );
    if( argc >= 4 && !(pClip = parse( argv[3] )) )
        return EXIT_FAILURE;
    size_t clip = min( *pClip, rows * cols );
    streamsize width = (ostringstream() << clip).str().length();
    for( size_t row = 1; row <= min( rows, clip ); ++row )
    {
        bool head = true;
        for( size_t value = row; value <= clip; value += rows, head =
false )
            cout << " "sv.substr( head, !head ) << right << setw( width
) << value;
        cout << endl;
    }
}

static optional<size_t> parse( const char *str )
{
    istringstream iss( str );
    size_t ret;
    iss >> ret;
    if( !iss || !iss.eof() )
        return nullopt;
    return ret;
}

C++ really rocks since you've to deal with much less details than in C.

Is your strategy to just ignore reality, and keep making bogus claims
that - for this challenge at least - you can't support?


#include <stdlib.h>
#include <stdio.h>

int main(int argc, char *argv[]) {
if (argc < 3 || argc > 4) {
printf("Enter 2 or 3 arguments:\n$./prog rows columns [stop]\n");
return 0;
}

int rows = atoi(argv[1]);
int cols = atoi(argv[2]);
int max = (argc == 4) ? (atoi(argv[3])) : (rows * cols);
char cw[12];
int colwidth = sprintf(cw,"%d ",rows * cols);

for (int r = 1; r <= rows; r++) {
if (r <= max) {
int nbr = r;
printf("%*d", colwidth, nbr);
for (int i = 0; i < (cols - 1); i++) {
((nbr += rows) <= max) ? (printf("%*d", colwidth, nbr)) : (0) ;
}
printf("\n");
}
}

return 0;

}
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 12.03.2026 um 15:00 schrieb DFS:

On 3/12/2026 2:24 AM, Bonita Montero wrote:

There was a programming-"contest" on comp.lang.c and I wanted to show
the simpler code in C, here it is:

#include <iostream>
#include <sstream>
#include <iomanip>
#include <optional>
#include <algorithm>

using namespace std;

static optional<size_t> parse( const char *str );

int main( int argc, char **argv )
{
     if( argc < 3 )
         return EXIT_FAILURE;
     optional<size_t>
         pRows = parse( argv[1] ),
         pCols = parse( argv[2] );
     if( !pRows || !pCols )
         return EXIT_FAILURE;
     size_t rows = *pRows, cols = *pCols;
     optional<size_t> pClip( rows * cols );
     if( argc >= 4 && !(pClip = parse( argv[3] )) )
         return EXIT_FAILURE;
     size_t clip = min( *pClip, rows * cols );
     streamsize width = (ostringstream() << clip).str().length();
     for( size_t row = 1; row <= min( rows, clip ); ++row )
     {
         bool head = true;
         for( size_t value = row; value <= clip; value += rows, head =
false )
             cout << " "sv.substr( head, !head ) << right <<
setw( width ) << value;
         cout << endl;
     }
}

static optional<size_t> parse( const char *str )
{
     istringstream iss( str );
     size_t ret;
     iss >> ret;
     if( !iss || !iss.eof() )
         return nullopt;
     return ret;
}

C++ really rocks since you've to deal with much less details than in C.

Is your strategy to just ignore reality, and keep making bogus claims
that - for this challenge at least - you can't support?

Where's your check for parsing errors ? Your code would be longer with
that.
But real C++ code is usually multiple times shorter, mostly because of
generic programming. C really sucks since you have to flip every bit
on your own.

#include <stdlib.h>
#include <stdio.h>

int main(int argc, char *argv[]) {
    if (argc < 3 || argc > 4) {
        printf("Enter 2 or 3 arguments:\n$./prog rows columns [stop]\n");
        return 0;
    }

    int rows = atoi(argv[1]);
    int cols = atoi(argv[2]);
    int max  = (argc == 4) ? (atoi(argv[3])) : (rows * cols);
    char cw[12];
    int colwidth = sprintf(cw,"%d ",rows * cols);

    for (int r = 1; r <= rows; r++) {
        if (r <= max) {
            int nbr = r;
            printf("%*d", colwidth, nbr);
            for (int i = 0; i < (cols - 1); i++) {
                ((nbr += rows) <= max) ? (printf("%*d", colwidth,
nbr)) : (0) ;
            }
            printf("\n");
        }
    }

    return 0;

}

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Here, do that in C:

static optional<size_t> parse( const char *str )
{
size_t ret;
if( from_chars_result fcr = from_chars( str, str + strlen( str ), ret ); (bool)fcr.ec || *fcr.ptr )
return nullopt;
return ret;
}

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/12/26 15:03, Bonita Montero wrote:

But real C++ code is usually multiple times shorter, mostly because of generic programming. C really sucks since you have to flip every bit
on your own.

C++ really sucks because you can't know who flip the bits.
--
** **
* tTh des Bourtoulots *
* http://maison.tth.netlib.re/ *
** **
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 12.03.2026 um 15:27 schrieb tTh:

On 3/12/26 15:03, Bonita Montero wrote:

But real C++ code is usually multiple times shorter, mostly because of
generic programming. C really sucks since you have to flip every bit
on your own.

   C++ really sucks because you can't know who flip the bits.

Absolutely not, C++ can anyhing you want if you need that. But in C++
you often doesn't need that. Bad programmers which don't understand
that are rare since C++ has very high skill demands.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/12/2026 10:13 AM, Bonita Montero wrote:

Here, do that in C:

static optional<size_t> parse( const char *str )
{
    size_t ret;
    if( from_chars_result fcr = from_chars( str, str + strlen( str ), ret ); (bool)fcr.ec || *fcr.ptr )
        return nullopt;
    return ret;
}

Explain in detail what it does.


--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 12.03.2026 um 15:43 schrieb DFS:

On 3/12/2026 10:13 AM, Bonita Montero wrote:

Here, do that in C:

static optional<size_t> parse( const char *str )
{
     size_t ret;
     if( from_chars_result fcr = from_chars( str, str + strlen( str ), >> ret ); (bool)fcr.ec || *fcr.ptr )
         return nullopt;
     return ret;
}

Explain in detail what it does.

If that isn't self-explanatory stick with C.

I've got a task for you: Do the same in C:

#include <iostream>
#include <regex>
#include <fstream>
#include <vector>
#include <algorithm>
#include <vector>

using namespace std;

int main( int argc, char **argv )
{
if( argc < 2 )
return EXIT_FAILURE;
ifstream ifs( argv[1] );
static regex rxNameTel( "^\\s*\"([^\"]*)\"\\s*\"([^\"]*)\"\\s*$" );
struct name_tel { string name, tel; };
vector<name_tel> phoneList;
while( !ifs.eof() )
{
string line;
getline( ifs, line );
match_results<string::const_iterator> sm;
if( regex_match( line, sm, rxNameTel ) )
phoneList.emplace_back( string( sm[1].first, sm[1].second ), string(
sm[2].first, sm[2].second ) );
}
sort( phoneList.begin(), phoneList.end(),
[]( const name_tel &left, const name_tel &right ) { return left.name <
right.name; } );
for( name_tel &phone : phoneList )
cout << "\"" << phone.name << "\"\t\"" << phone.tel << "\"" << endl;
}

1. Read a file and parse it with ne mentioned regex-pattern.
2. Split both parts of every line in two strings.
3. Sort the "vector" according to the first string.
4. Print it.

I guess you don't manage to do that with less than five times the work.
Every external lib allowed.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/12/2026 11:10 AM, Bonita Montero wrote:

Am 12.03.2026 um 15:43 schrieb DFS:

On 3/12/2026 10:13 AM, Bonita Montero wrote:

Here, do that in C:

static optional<size_t> parse( const char *str )
{
     size_t ret;
     if( from_chars_result fcr = from_chars( str, str + strlen( str >>> ), ret ); (bool)fcr.ec || *fcr.ptr )
         return nullopt;
     return ret;
}

Explain in detail what it does.

If that isn't self-explanatory stick with C.

I've got a task for you: Do the same in C:

#include <iostream>
#include <regex>
#include <fstream>
#include <vector>
#include <algorithm>
#include <vector>

using namespace std;

int main( int argc, char **argv )
{
    if( argc < 2 )
        return EXIT_FAILURE;
    ifstream ifs( argv[1] );
    static regex rxNameTel( "^\\s*\"([^\"]*)\"\\s*\"([^\"]*)\"\\s*$" );
    struct name_tel { string name, tel; };
    vector<name_tel> phoneList;
    while( !ifs.eof() )
    {
        string line;
        getline( ifs, line );
        match_results<string::const_iterator> sm;
        if( regex_match( line, sm, rxNameTel ) )
            phoneList.emplace_back( string( sm[1].first, sm[1].second
), string( sm[2].first, sm[2].second ) );
    }
    sort( phoneList.begin(), phoneList.end(),
        []( const name_tel &left, const name_tel &right ) { return left.name < right.name; } );
    for( name_tel &phone : phoneList )
        cout << "\"" << phone.name << "\"\t\"" << phone.tel << "\"" <<
endl;
}

1. Read a file and parse it with ne mentioned regex-pattern.
2. Split both parts of every line in two strings.
3. Sort the "vector" according to the first string.
4. Print it.

I guess you don't manage to do that with less than five times the work.
Every external lib allowed.

Give me the file you used.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 12.03.2026 um 16:22 schrieb DFS:

Give me the file you used.

The input file looks like this:

"White House" "001 202 456 1414"
"Mother" "0049 211 151395"
"Mickey Mouse" "001 123 456 7890"

The output shout look sorted:

"White House" "001 202 456 1414"
"Mickey Mouse" "001 123 456 7890"
"Mother" "0049 211 151395"

Input with additional spaces, output without.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 12.03.2026 um 16:25 schrieb Bonita Montero:

Am 12.03.2026 um 16:22 schrieb DFS:

Give me the file you used.

The input file looks like this:

"White House" "001 202 456 1414"
 "Mother" "0049 211 151395"
"Mickey Mouse"   "001 123 456 7890"

The output shout look sorted:

"White House" "001 202 456 1414"
"Mickey Mouse" "001 123 456 7890"
"Mother" "0049 211 151395"

Sorry,

"Mickey Mouse" "001 123 456 7890"
"Mother" "0049 211 151395"
"White House" "001 202 456 1414"

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 12.03.2026 um 16:10 schrieb Bonita Montero:

    static regex rxNameTel( "^\\s*\"([^\"]*)\"\\s*\"([^\"]*)\"\\s*$" );

Easier to read as a C++ raw string: R"~(^\s*"([^"]*)"\s*"([^"]*)"\s*$)~"
The regex is between ~( and )~.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 12.03.2026 um 16:22 schrieb DFS:

Give me the file you used.

Better take this:

"Max Mustermann" "0123-4567890"
"Anna Müller" "0987-6543210"
"Peter Schmidt" "030-1234567"
"Laura Fischer" "040-9876543"
"Tim Becker" "0151-1112223"
"Julia Neumann" "0221-3344556"
"Michael Braun" "0170-9988776"
"Sophie Wagner" "089-2233445"
"Felix Hoffmann" "0711-5566778"
"Lea Richter" "0341-1122334"
"Jonas Klein" "030-4455667"
"Emma Wolf" "040-5566778"
"Lukas König" "069-7788990"
"Clara Hofmann" "0157-2233445"
"Paul Schäfer" "0228-3344556"
"Mia Keller" "089-6677889"
"Leon Zimmermann" "0711-1122445"
"Nina Krause" "0341-4455667"
"David Schulz" "030-9988776"
"Sarah Lehmann" "040-2233445"
"Ben Richter" "069-3344556"
"Hannah Wagner" "0221-5566778"
"Tom Bauer" "0171-1122334"
"Lena Fischer" "0151-6677889"
"Simon Meier" "030-7788990"
"Marie Becker" "040-4455667"
"Jan Hoffmann" "0711-3344556"
"Leonie Klein" "0341-5566778"
"Philipp König" "089-9988776"
"Laura Schulze" "0228-1122334"
"Moritz Wolf" "0157-4455667"
"Jana Zimmer" "030-6677889"
"Felix Neumann" "0170-2233445"
"Sarah Braun" "040-7788990"
"Tim Schäfer" "0711-5566778"
"Anna Fischer" "0341-9988776"
"Maximilian Keller" "030-1122334"
"Lea Wagner" "0151-3344556"
"Lukas Hofmann" "089-6677889"
"Marie Richter" "0221-7788990"
"Jonas Klein" "0171-4455667"
"Clara Zimmermann" "040-5566778"
"Paul Wolf" "030-2233445"
"Sophie Neumann" "0711-3344556"
"Ben Meier" "0341-5566778"
"Emma Bauer" "0157-9988776"
"Leon Krause" "089-1122334"
"Julia Schulz" "0228-4455667"
"Tim Richter" "030-6677889"
"Anna Becker" "0170-2233445"

The output should be this:

"Anna Becker" "0170-2233445"
"Anna Fischer" "0341-9988776"
"Anna M├╝ller" "0987-6543210"
"Ben Meier" "0341-5566778"
"Ben Richter" "069-3344556"
"Clara Hofmann" "0157-2233445"
"Clara Zimmermann" "040-5566778"
"David Schulz" "030-9988776"
"Emma Bauer" "0157-9988776"
"Emma Wolf" "040-5566778"
"Felix Hoffmann" "0711-5566778"
"Felix Neumann" "0170-2233445"
"Hannah Wagner" "0221-5566778"
"Jan Hoffmann" "0711-3344556"
"Jana Zimmer" "030-6677889"
"Jonas Klein" "0171-4455667"
"Jonas Klein" "030-4455667"
"Julia Neumann" "0221-3344556"
"Julia Schulz" "0228-4455667"
"Laura Fischer" "040-9876543"
"Laura Schulze" "0228-1122334"
"Lea Richter" "0341-1122334"
"Lea Wagner" "0151-3344556"
"Lena Fischer" "0151-6677889"
"Leon Krause" "089-1122334"
"Leon Zimmermann" "0711-1122445"
"Leonie Klein" "0341-5566778"
"Lukas Hofmann" "089-6677889"
"Lukas K├Ânig" "069-7788990"
"Marie Becker" "040-4455667"
"Marie Richter" "0221-7788990"
"Max Mustermann" "0123-4567890"
"Maximilian Keller" "030-1122334"
"Mia Keller" "089-6677889"
"Michael Braun" "0170-9988776"
"Moritz Wolf" "0157-4455667"
"Nina Krause" "0341-4455667"
"Paul Schäfer" "0228-3344556"
"Paul Wolf" "030-2233445"
"Peter Schmidt" "030-1234567"
"Philipp K├Ânig" "089-9988776"
"Sarah Braun" "040-7788990"
"Sarah Lehmann" "040-2233445"
"Simon Meier" "030-7788990"
"Sophie Neumann" "0711-3344556"
"Sophie Wagner" "089-2233445"
"Tim Becker" "0151-1112223"
"Tim Richter" "030-6677889"
"Tim Schäfer" "0711-5566778"
"Tom Bauer" "0171-1122334"
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Bonita Montero <Bonita.Montero@gmail.com> writes:

Here, do that in C:

static optional<size_t> parse( const char *str )
{
size_t ret;
if( from_chars_result fcr = from_chars( str, str + strlen( str ), ret
); (bool)fcr.ec || *fcr.ptr )
return nullopt;
return ret;
}

bool parse(const char *str, size_t *rval)
{
char *cp;
*rval = strtoul(str, &cp, 0);
if ((cp == str) || (*cp != '\0')) {
return false;
}
return true;
}

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Bonita Montero <Bonita.Montero@gmail.com> writes:

Am 12.03.2026 um 15:43 schrieb DFS:

On 3/12/2026 10:13 AM, Bonita Montero wrote:

Here, do that in C:

static optional<size_t> parse( const char *str )
{
     size_t ret;
     if( from_chars_result fcr = from_chars( str, str + strlen( str ),
ret ); (bool)fcr.ec || *fcr.ptr )
         return nullopt;
     return ret;
}

Explain in detail what it does.

If that isn't self-explanatory stick with C.

I've got a task for you: Do the same in C:

#include <iostream>
#include <regex>
#include <fstream>
#include <vector>
#include <algorithm>
#include <vector>

using namespace std;

int main( int argc, char **argv )
{
if( argc < 2 )
return EXIT_FAILURE;

A silent failure implies to the interactive user
that the command succeeded. You should print
a responsive message before returning a failure code.

cerr << "Usage: " << argv[0] << " " <phonelist>" << eol;

(although I would prefer fprintf(stderr, ...)).

ifstream ifs( argv[1] );

You didn't bother to check if the file was opened.
You should print a responsive message and exit if the streams
failbit is set.

static regex rxNameTel( "^\\s*\"([^\"]*)\"\\s*\"([^\"]*)\"\\s*$" );
struct name_tel { string name, tel; };
vector<name_tel> phoneList;
while( !ifs.eof() )
{
string line;
getline( ifs, line );
match_results<string::const_iterator> sm;
if( regex_match( line, sm, rxNameTel ) )
phoneList.emplace_back( string( sm[1].first, sm[1].second ), string(
sm[2].first, sm[2].second ) );
}
sort( phoneList.begin(), phoneList.end(),
[]( const name_tel &left, const name_tel &right ) { return left.name <
right.name; } );
for( name_tel &phone : phoneList )
cout << "\"" << phone.name << "\"\t\"" << phone.tel << "\"" << endl;
}

1. Read a file and parse it with ne mentioned regex-pattern.
2. Split both parts of every line in two strings.
3. Sort the "vector" according to the first string.
4. Print it.

This is certainly a task that can be done simply using a shell
script and the standard POSIX utility set. awk(1) would be
a good starting point; it's possible that it can be done with
a single invocation of the posix 'sort' utility.

Of course, the lack of any in-line documentation (e.g. comments) is
a typical defect in your C++ code.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 12.03.26 11:36, Bonita Montero wrote:

Am 12.03.2026 um 09:32 schrieb Janis Papanagnou:

Concerning your subject; it wouldn't appear to me to use the term
"beauty" in context of C++. C++ inherited so much ugly syntax and
concepts (from "C") and it added yet more syntactic infelicities.
(And I'm saying that as someone who liked to program in C++, also
professionally, for many many years.)

My perception of beauty is just as subjective as your perception of the
ugly aspects of C++. I like the span of lowleven means in C++ that are inherited from C to higher level abstractions C++ has inherited from
other more moderrn language.

Or even from older languages, like classes and inheritance from Simula, type-safety (string type checking) as provided by many older langauges, functional principles (cf. STL) from functional languages, and whatnot.

And if you don't the subjektive part of beauty or ugly parts of this
language C++ is several times more effective than C, i.e. you need
only a fraction of the code size while mostly maintaining the same performance.

Yes, I have agreed on that part, what you functionally get with C++.

Concerning the terms beauty and ugliness being subjective is obvious.

I was particularly speaking about syntax. - If you happen to know a
sufficient set of (modern or historic) languages - and I have no
reason to believe you haven't - you can observe that there are many
languages that are _a lot_ clearer designed than C++, and much safer
to use. Languages where the time span from design to a running system
is much shorter than in a language where it's easy to make subtle or
blatant errors.

You mentioned an IMO noteworthy specific aspect of C++; you seem to
like the low-to-high-level option. - Myself I also like both levels
to program in, depending on the application area. - But I consider
it a problem if the syntax and semantics of the inherited low-level
language taints the abstract concepts of a language. Personally I'd
prefer here a readable and powerful high level language with an
*interface* to connect arbitrary code from low-level languages. IMO
there's no need to do all levels in one language. And there are also
existing paragons that support that language design principle.

Janis

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 12.03.26 16:48, Bonita Montero wrote:

Am 12.03.2026 um 16:22 schrieb DFS:

Give me the file you used.

Better take this:
[snip list of names of people and telephone numbers]

The output should be this:
[snip list of names of people and telephone numbers]

I see there are real existing people in that list.

Do you have the consent of these persons to disseminate
their names and telephone numbers?

I'm just curious.

Janis

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/12/2026 11:10 AM, Bonita Montero wrote:

I've got a task for you: Do the same in C:

#include <iostream>
#include <regex>
#include <fstream>
#include <vector>
#include <algorithm>
#include <vector>

using namespace std;

int main( int argc, char **argv )
{
    if( argc < 2 )
        return EXIT_FAILURE;
    ifstream ifs( argv[1] );
    static regex rxNameTel( "^\\s*\"([^\"]*)\"\\s*\"([^\"]*)\"\\s*$" );
    struct name_tel { string name, tel; };
    vector<name_tel> phoneList;
    while( !ifs.eof() )
    {
        string line;
        getline( ifs, line );
        match_results<string::const_iterator> sm;
        if( regex_match( line, sm, rxNameTel ) )
            phoneList.emplace_back( string( sm[1].first, sm[1].second
), string( sm[2].first, sm[2].second ) );
    }
    sort( phoneList.begin(), phoneList.end(),
        []( const name_tel &left, const name_tel &right ) { return left.name < right.name; } );
    for( name_tel &phone : phoneList )
        cout << "\"" << phone.name << "\"\t\"" << phone.tel << "\"" <<
endl;
}

1. Read a file and parse it with ne mentioned regex-pattern.
2. Split both parts of every line in two strings.
3. Sort the "vector" according to the first string.
4. Print it.

I guess you don't manage to do that with less than five times the work.
Every external lib allowed.

I'll try that in C if you commit to trying the following in C++. Deal?

--------------------------------------------------------------------------
* read in a list of words from a file here: https://people.sc.fsu.edu/~jburkardt/datasets/words/words.html

* pick N random words from that list and put them in an array (OK if
there are a few dupes - if you can get no dupes that's better)

* sort and print the array of randoms, adding a blank line each time the
1st letter changes

(note: I just now wrote this. It's not old code.)

usage is
$./randwords filename N
$./randwords special_english.txt 200

-------------------------------------------------------------------------- //public domain code to read a list of words in and print N
//random words in sorted order

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <time.h>
#include <unistd.h>

int compare(const void *a, const void *b) {
const char* ca = *(const char**)a;
const char* cb = *(const char**)b;
return strcmp(ca,cb);
}

int main(int argc, char *argv[]) {

if (argc != 3) {
printf("Enter file name and a number:\n$./prog file.txt 200\n");
return 0;
}

int i = 0;
int lines = 0;
char line[35] = "";
int N = atoi(argv[2]);
int randnbr, randnbr2;

FILE *fin = fopen(argv[1],"r");
while (fgets(line,sizeof line, fin) != NULL) {lines++;}
if (N > lines) {
printf("Enter a value < %d\n", lines);
exit(0);
}

rewind(fin);
char **wordlist = malloc(sizeof(char*) * lines);
while (fgets(line,sizeof line, fin) != NULL) {
int datalen = strlen(line);
wordlist[i] = malloc(datalen + 1);
strncpy(wordlist[i], line, datalen);
wordlist[i][datalen] = '\0';
i++;
}

char **wordrand = malloc(sizeof(char*) * N);
srand(time(NULL) + getpid());
for (i = 0; i < N; i++) {
int randnbr = (rand() % lines) + 1;
int datalen = strlen(wordlist[randnbr]);
wordrand[i] = malloc(datalen + 1);
strncpy(wordrand[i], wordlist[randnbr], datalen);
wordrand[i][datalen] = '\0';
}

qsort(wordrand, N, sizeof(char*), compare);

printf("%d words read in\n", lines);
printf("%d random words extracted\n",N);

int currletter = wordrand[0][0];
for (i = 0; i < N; i++) {
if ((wordrand[i][0] != currletter) && (i > 0)) {
printf("\n");
}
currletter = wordrand[i][0];
printf("%d. %s",i+1, wordrand[i]);
}

printf("\n");

free(wordrand);
free(wordlist);
return 0;
}
--------------------------------------------------------------------------

output is something like:
$ ./randwords special_english.txt 200
1477 words read in
200 random words extracted
1. above
2. accept
3. after
4. against
5. agency
6. ammunition
7. anger
8. anniversary
9. army
10. arrive
11. art
12. artillery
13. automobile
14. autumn

15. bed
16. below
17. bleed
18. blow
19. blue
20. boat
21. boil
22. bread
23. bridge
24. brown
25. business

26. cancer
27. claim
28. clean
29. cloud
30. cloud
31. cloud
32. combine
33. compare
34. conflict
35. consider
36. contain
37. correct
38. credit
39. criticize
40. cross
41. crowd
42. customs

43. decide
44. demand
...
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 12.03.2026 um 18:32 schrieb Scott Lurndal:

You didn't bother to check if the file was opened.

That's not necessary to compare it against a equal solution in C.

This is certainly a task that can be done simply using a shell
script and the standard POSIX utility set. awk(1) would be
a good starting point; it's possible that it can be done with
a single invocation of the posix 'sort' utility.

It's comparison of C against C++.

Of course, the lack of any in-line documentation (e.g. comments) is
a typical defect in your C++ code.

Complete idiot.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 12.03.2026 um 20:25 schrieb Janis Papanagnou:

On 12.03.26 16:48, Bonita Montero wrote:

Am 12.03.2026 um 16:22 schrieb DFS:

Give me the file you used.

Better take this:
[snip list of names of people and telephone numbers]
The output should be this:
[snip list of names of people and telephone numbers]

I see there are real existing people in that list.

Do you have the consent of these persons to disseminate
their names and telephone numbers?

These are no real people.
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Failed ...
You're parsing a completely different type of file.
And with 2.5 times the source length.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 12.03.2026 um 18:32 schrieb Scott Lurndal:

You didn't bother to check if the file was opened.

If the file couldn't been opened nothing happens.
No difference to a check.
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

I shortened my code a bit.
Do that:

#include <iostream>
#include <regex>
#include <fstream>
#include <vector>
#include <algorithm>

using namespace std;

int main( int argc, char **argv )
{
if( argc < 2 )
return EXIT_FAILURE;
ifstream ifs( argv[1] );
static regex rxNameTel( R"~(^\s*"([^"]*)"\s*"([^"]*)"\s*$)~" );
struct name_tel { string name, tel; };
vector<name_tel> phoneList;
smatch sm;
for( string line; getline( ifs, line ); )
if( regex_match( line, sm, rxNameTel ) )
phoneList.emplace_back( sm[1].str(), sm[2].str() );
sort( phoneList.begin(), phoneList.end(),
[]( name_tel &left, name_tel &right ) { return left.name < right.name;
} );
for( name_tel &phone : phoneList )
cout << "\"" << phone.name << "\"\t\"" << phone.tel << "\"" << endl;
}

Have a closer look at the regex.
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

DFS <nospam@dfs.com> writes:

I'll try that in C if you commit to trying the following in C++. Deal?

--------------------------------------------------------------------------
* read in a list of words from a file here: >https://people.sc.fsu.edu/~jburkardt/datasets/words/words.html

* pick N random words from that list and put them in an array (OK if
there are a few dupes - if you can get no dupes that's better)

* sort and print the array of randoms, adding a blank line each time the
1st letter changes

(note: I just now wrote this. It's not old code.)

usage is
$./randwords filename N
$./randwords special_english.txt 200

--------------------------------------------------------------------------

A POSIX version:

#include <errno.h>
#include <fcntl.h>
#include <stddef.h>
#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <time.h>
#include <unistd.h>

#include <sys/mman.h>
#include <sys/stat.h>


int
qsortcompare(const void *a, const void *b)
{
const char* ca = *(const char**)a;
const char* cb = *(const char**)b;
return strcmp(ca,cb);
}

int
main(int argc, const char **argv)
{
int fd;
struct stat st;
char *words;
char *cp;
char *end;
char **wordlist;
char **sorted;
char ch = '\0';
size_t wordcount = 0u;
size_t i;
size_t n;

if (argc < 3) {
fprintf(stderr, "Usage: %s <wordlist> <N>\n", argv[0]);
return 1;
}

n = strtoul(argv[2], &cp, 0);
if ((cp == argv[2]) || (*cp != '\0')) {
fprintf(stderr, "%s: <N> argument '%s' must be fully numeric in base 8, 10 or 16\n", argv[0], argv[2]);
return 1;
}

if (stat(argv[1], &st) == -1) {
fprintf(stderr, "%s: Unable to open '%s': %s\n", argv[0], argv[1], strerror(errno));
return 2;
}

fd = open(argv[1], O_RDONLY, 0);
if (fd == -1) {
fprintf(stderr, "%s: Unable to open '%s': %s\n", argv[0], argv[1], strerror(errno));
return 2;
}

words = mmap(NULL, st.st_size, PROT_READ|PROT_WRITE, MAP_PRIVATE, fd, 0ul);
if (words == MAP_FAILED) {
fprintf(stderr, "%s: Unable to mmap '%s': %s\n", argv[0], argv[1], strerror(errno));
return 2;
}
close(fd);

cp = words;
for(i = 0ul; i < st.st_size; i++, cp++) {
if (*cp == '\n') wordcount++;
}

wordlist = malloc(wordcount * sizeof(const char *));
if (wordlist == NULL) {
fprintf(stderr, "%s: Unable to allocate %zu bytes\n", argv[0], wordcount * sizeof(const char *));
return 3;
}

cp = words;
end = cp + st.st_size;
i = 0u;
while (cp < end) {
wordlist[i++] = cp;
for(; (cp < end) && (*cp != '\n'); cp++) {}
*cp++ = '\0';
}

sorted = malloc(n * sizeof(const char *));
if (sorted == NULL) {
fprintf(stderr, "%s: Unable to allocate %zu bytes\n", argv[0], n * sizeof(const char *));
return 3;
}

srand(time(NULL));
for(i = 0ul; i < n; i++) {
sorted[i] = wordlist[rand() % wordcount];
}
qsort(sorted, n, sizeof(sorted[0]), qsortcompare);

for(size_t w = 0ul; w < n; w++) {
if (ch && (ch != sorted[w][0])) fputc('\n', stdout);
fprintf(stdout, "%s\n", sorted[w]);
ch = sorted[w][0];
}

free(wordlist);
munmap(words, st.st_size);

return 0;
}
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 13.03.26 00:57, Bonita Montero wrote:

Am 12.03.2026 um 20:25 schrieb Janis Papanagnou:

On 12.03.26 16:48, Bonita Montero wrote:

Am 12.03.2026 um 16:22 schrieb DFS:

Give me the file you used.

Better take this:
[snip list of names of people and telephone numbers]
The output should be this:
[snip list of names of people and telephone numbers]

I see there are real existing people in that list.

Do you have the consent of these persons to disseminate
their names and telephone numbers?

These are no real people.

This is an interesting statement. - I first saw the typical
"Max Mustermann" that is in our country the prototype of an
artificial test entry with obvious test number 0123-4567890.
Then I picked an arbitrary entry and looked it up; I found
that person with exactly the associated telephone number. -
Where did you get these real names from?

Janis

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/12/2026 11:48 AM, Bonita Montero wrote:

Am 12.03.2026 um 16:22 schrieb DFS:

Give me the file you used.

Better take this:

That's 50 in and 50 out.

So what's the RegEx for?

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/12/2026 8:54 PM, Scott Lurndal wrote:

DFS <nospam@dfs.com> writes:

I'll try that in C if you commit to trying the following in C++. Deal?

-------------------------------------------------------------------------- >> * read in a list of words from a file here:
https://people.sc.fsu.edu/~jburkardt/datasets/words/words.html

* pick N random words from that list and put them in an array (OK if
there are a few dupes - if you can get no dupes that's better)

* sort and print the array of randoms, adding a blank line each time the
1st letter changes

(note: I just now wrote this. It's not old code.)

usage is
$./randwords filename N
$./randwords special_english.txt 200

--------------------------------------------------------------------------

A POSIX version:

I took out all 7 instances of error trapping, and it throws:
Floating point exception (core dumped)
at line 61 or 62

======================================================================== #include <errno.h>
#include <fcntl.h>
#include <stddef.h>
#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <time.h>
#include <unistd.h>
#include <sys/mman.h>
#include <sys/stat.h>

int
qsortcompare(const void *a, const void *b)
{
const char* ca = *(const char**)a;
const char* cb = *(const char**)b;
return strcmp(ca,cb);
}

int
main(int argc, const char **argv)
{
int fd;
struct stat st;
char *words;
char *cp;
char *end;
char **wordlist;
char **sorted;
char ch = '\0';
size_t wordcount = 0u;
size_t i;
size_t n;

n = strtoul(argv[2], &cp, 0);
fd = open(argv[1], O_RDONLY, 0);
words = mmap(NULL, st.st_size, PROT_READ|PROT_WRITE, MAP_PRIVATE,
fd, 0ul);
close(fd);

cp = words;
for(i = 0ul; i < st.st_size; i++, cp++) {
if (*cp == '\n') wordcount++;
}
wordlist = malloc(wordcount * sizeof(const char *));

cp = words;
end = cp + st.st_size;
i = 0u;
while (cp < end) {
wordlist[i++] = cp;
for(; (cp < end) && (*cp != '\n'); cp++) {}
*cp++ = '\0';
}

sorted = malloc(n * sizeof(const char *));

srand(time(NULL));

//Floating point exception (core dumped)
for(i = 0ul; i < n; i++) {
sorted[i] = wordlist[rand() % wordcount];
}

qsort(sorted, n, sizeof(sorted[0]), qsortcompare);

for(size_t w = 0ul; w < n; w++) {
if (ch && (ch != sorted[w][0])) fputc('\n', stdout);
fprintf(stdout, "%s\n", sorted[w]);
ch = sorted[w][0];
}

free(wordlist);
munmap(words, st.st_size);

return 0;
}
========================================================================

#include <errno.h>
#include <fcntl.h>
#include <stddef.h>
#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <time.h>
#include <unistd.h>

#include <sys/mman.h>
#include <sys/stat.h>

int
qsortcompare(const void *a, const void *b)
{
const char* ca = *(const char**)a;
const char* cb = *(const char**)b;
return strcmp(ca,cb);
}

int
main(int argc, const char **argv)
{
int fd;
struct stat st;
char *words;
char *cp;
char *end;
char **wordlist;
char **sorted;
char ch = '\0';
size_t wordcount = 0u;
size_t i;
size_t n;

if (argc < 3) {
fprintf(stderr, "Usage: %s <wordlist> <N>\n", argv[0]);
return 1;
}

n = strtoul(argv[2], &cp, 0);
if ((cp == argv[2]) || (*cp != '\0')) {
fprintf(stderr, "%s: <N> argument '%s' must be fully numeric in base 8, 10 or 16\n", argv[0], argv[2]);
return 1;
}

if (stat(argv[1], &st) == -1) {
fprintf(stderr, "%s: Unable to open '%s': %s\n", argv[0], argv[1], strerror(errno));
return 2;
}

fd = open(argv[1], O_RDONLY, 0);
if (fd == -1) {
fprintf(stderr, "%s: Unable to open '%s': %s\n", argv[0], argv[1], strerror(errno));
return 2;
}

words = mmap(NULL, st.st_size, PROT_READ|PROT_WRITE, MAP_PRIVATE, fd, 0ul);
if (words == MAP_FAILED) {
fprintf(stderr, "%s: Unable to mmap '%s': %s\n", argv[0], argv[1], strerror(errno));
return 2;
}
close(fd);

cp = words;
for(i = 0ul; i < st.st_size; i++, cp++) {
if (*cp == '\n') wordcount++;
}

wordlist = malloc(wordcount * sizeof(const char *));
if (wordlist == NULL) {
fprintf(stderr, "%s: Unable to allocate %zu bytes\n", argv[0], wordcount * sizeof(const char *));
return 3;
}

cp = words;
end = cp + st.st_size;
i = 0u;
while (cp < end) {
wordlist[i++] = cp;
for(; (cp < end) && (*cp != '\n'); cp++) {}
*cp++ = '\0';
}

sorted = malloc(n * sizeof(const char *));
if (sorted == NULL) {
fprintf(stderr, "%s: Unable to allocate %zu bytes\n", argv[0], n * sizeof(const char *));
return 3;
}

srand(time(NULL));
for(i = 0ul; i < n; i++) {
sorted[i] = wordlist[rand() % wordcount];
}
qsort(sorted, n, sizeof(sorted[0]), qsortcompare);

for(size_t w = 0ul; w < n; w++) {
if (ch && (ch != sorted[w][0])) fputc('\n', stdout);
fprintf(stdout, "%s\n", sorted[w]);
ch = sorted[w][0];
}

free(wordlist);
munmap(words, st.st_size);

return 0;
}

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 13.03.2026 um 02:19 schrieb Janis Papanagnou:

This is an interesting statement. - I first saw the typical
"Max Mustermann" that is in our country the prototype of an
artificial test entry with obvious test number 0123-4567890.
Then I picked an arbitrary entry and looked it up; I found
that person with exactly the associated telephone number. -
Where did you get these real names from?

The list was generated with ChatGpt.
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 13.03.2026 um 02:22 schrieb DFS:

That's 50 in and 50 out.

Yes, but sorted.

So what's the RegEx for?

Learn Regex !
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 13.03.2026 um 05:31 schrieb DFS:

    //Floating point exception (core dumped)
    for(i = 0ul; i < n; i++) {
        sorted[i] = wordlist[rand() % wordcount];
    }

...

    free(wordlist);
    munmap(words, st.st_size);

BIG mistake: If you crash in the first part the memory in
the second part isn't freed, at least under Windows 3.11. ;-)
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/13/2026 12:31 AM, DFS wrote:

On 3/12/2026 8:54 PM, Scott Lurndal wrote:

DFS <nospam@dfs.com> writes:

I took out all 7 instances of error trapping, and it throws:
Floating point exception (core dumped)
at line 61 or 62

Put back in just the stat() call

stat(argv[1], &st);

and the program works again.

I see what happened.
Without the stat() call, st.st_size is 1.
With the stat() call, st.st_size is 9836.

The mmap-ing executes either way. With no stat() call, the words array
still ended up with 606 words (using a file with 1477 words).

But after that wordcount was set to 0:

for(i = 0ul; i < st.st_size;

and then the call to

wordlist[rand() % wordcount]

threw the floating point exception.


Question: wordlist was malloced with a size of 0:

wordlist = malloc(wordcount * sizeof(const char *));

Why are you allowed to malloc size 0?

After malloc, you populate it and it ended up containing only the first
letter of the first word:

wordlist[0] = 'a'
wordlist[1 and up] = (null)

I used file special_english.txt from that link I gave.

======================================================================== #include <errno.h>
#include <fcntl.h>
#include <stddef.h>
#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <time.h>
#include <unistd.h>
#include <sys/mman.h>
#include <sys/stat.h>

int
qsortcompare(const void *a, const void *b)
{
     const char* ca = *(const char**)a;
     const char* cb = *(const char**)b;
     return strcmp(ca,cb);
}

int
main(int argc, const char **argv)
{
    int fd;
    struct stat st;
    char *words;
    char *cp;
    char *end;
    char **wordlist;
    char **sorted;
    char ch = '\0';
    size_t wordcount = 0u;
    size_t i;
    size_t n;

    n = strtoul(argv[2], &cp, 0);
    fd = open(argv[1], O_RDONLY, 0);
    words = mmap(NULL, st.st_size, PROT_READ|PROT_WRITE, MAP_PRIVATE, fd, 0ul);
    close(fd);

    cp = words;
    for(i = 0ul; i < st.st_size; i++, cp++) {
        if (*cp == '\n') wordcount++;
    }
    wordlist = malloc(wordcount * sizeof(const char *));

    cp = words;
    end = cp + st.st_size;
    i = 0u;
    while (cp < end) {
        wordlist[i++] = cp;
    for(; (cp < end) && (*cp != '\n'); cp++) {}
    *cp++ = '\0';
    }

    sorted = malloc(n * sizeof(const char *));

    srand(time(NULL));

    //Floating point exception (core dumped)
    for(i = 0ul; i < n; i++) {
        sorted[i] = wordlist[rand() % wordcount];
    }

    qsort(sorted, n, sizeof(sorted[0]), qsortcompare);

    for(size_t w = 0ul; w < n; w++) {
    if (ch && (ch != sorted[w][0])) fputc('\n', stdout);
        fprintf(stdout, "%s\n", sorted[w]);
    ch = sorted[w][0];
    }

    free(wordlist);
    munmap(words, st.st_size);

    return 0;
}
========================================================================

#include <errno.h>
#include <fcntl.h>
#include <stddef.h>
#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <time.h>
#include <unistd.h>

#include <sys/mman.h>
#include <sys/stat.h>


int
qsortcompare(const void *a, const void *b)
{
      const char* ca = *(const char**)a;
      const char* cb = *(const char**)b;
      return strcmp(ca,cb);
}

int
main(int argc, const char **argv)
{
     int fd;
     struct stat st;
     char *words;
     char *cp;
     char *end;
     char **wordlist;
     char **sorted;
     char ch = '\0';
     size_t wordcount = 0u;
     size_t i;
     size_t n;

     if (argc < 3) {
         fprintf(stderr, "Usage: %s <wordlist> <N>\n", argv[0]);
         return 1;
     }

     n = strtoul(argv[2], &cp, 0);
     if ((cp == argv[2]) || (*cp != '\0')) {
         fprintf(stderr, "%s: <N> argument '%s' must be fully numeric
in base 8, 10 or 16\n", argv[0], argv[2]);
         return 1;
     }

     if (stat(argv[1], &st) == -1) {
         fprintf(stderr, "%s: Unable to open '%s': %s\n", argv[0], >> argv[1], strerror(errno));
         return 2;
     }

     fd = open(argv[1], O_RDONLY, 0);
     if (fd == -1) {
         fprintf(stderr, "%s: Unable to open '%s': %s\n", argv[0], >> argv[1], strerror(errno));
      return 2;
     }

     words = mmap(NULL, st.st_size, PROT_READ|PROT_WRITE, MAP_PRIVATE, >> fd, 0ul);
     if (words == MAP_FAILED) {
         fprintf(stderr, "%s: Unable to mmap '%s': %s\n", argv[0], >> argv[1], strerror(errno));
         return 2;
     }
     close(fd);

     cp = words;
     for(i = 0ul; i < st.st_size; i++, cp++) {
    if (*cp == '\n') wordcount++;
     }

     wordlist = malloc(wordcount * sizeof(const char *));
     if (wordlist == NULL) {
         fprintf(stderr, "%s: Unable to allocate %zu bytes\n",
argv[0], wordcount * sizeof(const char *));
         return 3;
     }

     cp = words;
     end = cp + st.st_size;
     i = 0u;
     while (cp < end) {
         wordlist[i++] = cp;
    for(; (cp < end) && (*cp != '\n'); cp++) {}
    *cp++ = '\0';
     }

     sorted = malloc(n * sizeof(const char *));
     if (sorted == NULL) {
         fprintf(stderr, "%s: Unable to allocate %zu bytes\n",
argv[0], n * sizeof(const char *));
         return 3;
     }

     srand(time(NULL));
     for(i = 0ul; i < n; i++) {
         sorted[i] = wordlist[rand() % wordcount];
     }
     qsort(sorted, n, sizeof(sorted[0]), qsortcompare);

     for(size_t w = 0ul; w < n; w++) {
    if (ch && (ch != sorted[w][0])) fputc('\n', stdout);
         fprintf(stdout, "%s\n", sorted[w]);
    ch = sorted[w][0];
     }

     free(wordlist);
     munmap(words, st.st_size);

     return 0;
}

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/13/2026 1:24 AM, Bonita Montero wrote:

Am 13.03.2026 um 05:31 schrieb DFS:

     //Floating point exception (core dumped)
     for(i = 0ul; i < n; i++) {
         sorted[i] = wordlist[rand() % wordcount];
     }

  ...

     free(wordlist);
     munmap(words, st.st_size);

BIG mistake: If you crash in the first part the memory in
the second part isn't freed, at least under Windows 3.11. ;-)

Lucky I'm running Win95 then...


(actually using Ubuntu on WSL on Win11 Pro 25H2)

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 13.03.2026 um 05:31 schrieb DFS:

I took out all 7 instances of error trapping, and it throws:
Floating point exception (core dumped)
at line 61 or 62

Take my signal_scope<> class, it makes signals thread-specific:

thread_local jmp_buf jb;
signal_scope<SIGFPE> fpeExc( +[]( int, siginfo_t *, void * )
{
static const char PrintThis[] = "caught!\n";
(void)write( 1, PrintThis, sizeof PrintThis - 1 );
siglongjmp( jb, 1 );
} );
if( int ret = sigsetjmp( jb, 1 ); !ret )
...
else
...

Here's the source of the signal_scope class:

#pragma once
#include <cstdlib>
#include <variant>
#include <mutex>
#include <cassert>
#include <unistd.h>
#include <signal.h>
#include <setjmp.h>

template<int SigNo>
struct signal_scope
{
static_assert(SigNo == SIGILL || SigNo == SIGFPE || SigNo == SIGSEGV ||
SigNo == SIGBUS || SigNo == SIGTRAP, "only sychronous signals");
using handler_fn = bool (*)( int );
using siginfo_handler_fn = bool (*)( int, siginfo_t *, void * );
using handler_variant = std::variant<int, handler_fn, siginfo_handler_fn>;
signal_scope( handler_variant handler = handler_variant() ) noexcept;
~signal_scope();
void operator =( handler_variant handler ) noexcept;
static void fallback( handler_variant handler ) noexcept;
static void re_init( const sigset_t *pSet, int flags );
private:
inline static std::mutex g_mtxFallback;
inline static handler_variant g_fallback = handler_variant();
inline static thread_local handler_variant t_handler = handler_variant();
handler_variant m_handlerBefore;
inline static struct init
{
init();
~init();
void reset( const sigset_t *pSet, int flags, bool old );
void dummy() {}
struct sigaction m_saBefore;
} g_init;
static void action( int sig, siginfo_t *info, void *uContext ) noexcept;
static bool callHandler( const handler_variant &handler, int sig, siginfo_t *info, void *uContext );
};

template<int SigNo>
signal_scope<SigNo>::signal_scope( handler_variant handler ) noexcept :
m_handlerBefore( t_handler )
{
(void)g_init;
t_handler = handler;
}

template<int SigNo>
inline signal_scope<SigNo>::~signal_scope()
{
t_handler = m_handlerBefore;
}

template<int SigNo>
inline void signal_scope<SigNo>::operator =( handler_variant handler ) noexcept
{
t_handler = handler;
}

template<int SigNo>
void signal_scope<SigNo>::fallback( handler_variant handler ) noexcept
{
using namespace std;
lock_guard lock( g_mtxFallback );
g_fallback = handler;
}

template<int SigNo>
void signal_scope<SigNo>::re_init( const sigset_t *pSet, int flags )
{
g_init.reset( pSet, flags, false );
}

template<int SigNo>
signal_scope<SigNo>::init::init()
{
reset( nullptr, 0, true );
}

template<int SigNo>
signal_scope<SigNo>::init::~init()
{
sigaction( SigNo, &m_saBefore, nullptr );
}

template<int SigNo>
void signal_scope<SigNo>::init::reset( const sigset_t *pSet, int flags,
bool old )
{
struct sigaction sa;
sa.sa_sigaction = action;
if( pSet )
sa.sa_mask = *pSet;
else
sigfillset( &sa.sa_mask );
sa.sa_flags = flags | SA_SIGINFO;
sigaction( SigNo, &sa, old ? &m_saBefore : nullptr );
}

template<int SigNo>
void signal_scope<SigNo>::action( int sig, siginfo_t *info, void
*uContext ) noexcept
{
using namespace std;
if( callHandler( t_handler, sig, info, uContext ) ) [[likely]]
return;
handler_variant fallback;
{
lock_guard lock( g_mtxFallback );
fallback = g_fallback;
}
callHandler( fallback, sig, info, uContext );
}

template<int SigNo>
bool signal_scope<SigNo>::callHandler( const handler_variant &handler,
int sig, siginfo_t *info, void *uContext )
{
if( holds_alternative<handler_fn>( handler ) ) [[likely]]
return get<handler_fn>( handler )( sig );
if( holds_alternative<siginfo_handler_fn>( handler ) ) [[likely]]
return get<siginfo_handler_fn>( handler )( sig, info, uContext );
return false;
}

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 12.03.2026 um 21:18 schrieb DFS:

output is something like:
$ ./randwords special_english.txt 200
1477 words read in
200 random words extracted
1. above
2. accept
3. after
4. against
5. agency
6. ammunition
7. anger
8. anniversary
9. army
10. arrive
11. art
12. artillery
13. automobile
14. autumn

15. bed
16. below
17. bleed
18. blow
19. blue
20. boat
21. boil
22. bread
23. bridge
24. brown
25. business

26. cancer
27. claim
28. clean
29. cloud
30. cloud
31. cloud
32. combine
33. compare
34. conflict
35. consider
36. contain
37. correct
38. credit
39. criticize
40. cross
41. crowd
42. customs

43. decide
44. demand
...

What's the format of the input file ? Also numbered lines ?
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/13/2026 3:08 AM, Bonita Montero wrote:

Am 12.03.2026 um 21:18 schrieb DFS:

output is something like:
$ ./randwords special_english.txt 200
1477 words read in
200 random words extracted
1. above
2. accept
3. after
...

What's the format of the input file ? Also numbered lines ?

not numbered

* read in a list of words from a file here: https://people.sc.fsu.edu/~jburkardt/datasets/words/words.html

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/12/2026 10:14 PM, Bonita Montero wrote:

Am 13.03.2026 um 02:19 schrieb Janis Papanagnou:

This is an interesting statement. - I first saw the typical
"Max Mustermann" that is in our country the prototype of an
artificial test entry with obvious test number 0123-4567890.
Then I picked an arbitrary entry and looked it up; I found
that person with exactly the associated telephone number. -
Where did you get these real names from?

The list was generated with ChatGpt.

Oh my. Beware, and always double, and triple, check its reams of code.
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 13.03.2026 um 09:48 schrieb Chris M. Thomasson:

On 3/12/2026 10:14 PM, Bonita Montero wrote:

Am 13.03.2026 um 02:19 schrieb Janis Papanagnou:

This is an interesting statement. - I first saw the typical
"Max Mustermann" that is in our country the prototype of an
artificial test entry with obvious test number 0123-4567890.
Then I picked an arbitrary entry and looked it up; I found
that person with exactly the associated telephone number. -
Where did you get these real names from?

The list was generated with ChatGpt.

Oh my. Beware, and always double, and triple, check its reams of code.

It's only word / telephone number tuples as .txt.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/12/2026 11:24 PM, Bonita Montero wrote:

Am 13.03.2026 um 05:31 schrieb DFS:

I took out all 7 instances of error trapping, and it throws:
Floating point exception (core dumped)
at line 61 or 62

Take my signal_scope<> class, it makes signals thread-specific:

thread_local jmp_buf jb;
signal_scope<SIGFPE> fpeExc( +[]( int, siginfo_t *, void * )
    {
        static const char PrintThis[] = "caught!\n";
        (void)write( 1, PrintThis, sizeof PrintThis - 1 );
        siglongjmp( jb, 1 );
    } );
if( int ret = sigsetjmp( jb, 1 ); !ret )
    ...
else
    ...

Here's the source of the signal_scope class:

#pragma once
#include <cstdlib>
#include <variant>
#include <mutex>
#include <cassert>
#include <unistd.h>
#include <signal.h>
#include <setjmp.h>

template<int SigNo>
struct signal_scope
{
    static_assert(SigNo == SIGILL || SigNo == SIGFPE || SigNo == SIGSEGV || SigNo == SIGBUS || SigNo == SIGTRAP, "only sychronous signals");
    using handler_fn = bool (*)( int );
    using siginfo_handler_fn = bool (*)( int, siginfo_t *, void * );
    using handler_variant = std::variant<int, handler_fn, siginfo_handler_fn>;
    signal_scope( handler_variant handler = handler_variant() ) noexcept;
    ~signal_scope();
    void operator =( handler_variant handler ) noexcept;
    static void fallback( handler_variant handler ) noexcept;
    static void re_init( const sigset_t *pSet, int flags );
private:
    inline static std::mutex g_mtxFallback;
    inline static handler_variant g_fallback = handler_variant();
    inline static thread_local handler_variant t_handler = handler_variant();
    handler_variant m_handlerBefore;
    inline static struct init
    {
        init();
        ~init();
        void reset( const sigset_t *pSet, int flags, bool old );
        void dummy() {}
        struct sigaction m_saBefore;
    } g_init;
    static void action( int sig, siginfo_t *info, void *uContext ) noexcept;
    static bool callHandler( const handler_variant &handler, int sig, siginfo_t *info, void *uContext );
};

template<int SigNo>
signal_scope<SigNo>::signal_scope( handler_variant handler ) noexcept :
    m_handlerBefore( t_handler )
{
    (void)g_init;
    t_handler = handler;
}

template<int SigNo>
inline signal_scope<SigNo>::~signal_scope()
{
    t_handler = m_handlerBefore;
}

template<int SigNo>
inline void signal_scope<SigNo>::operator =( handler_variant handler ) noexcept
{
    t_handler = handler;
}

template<int SigNo>
void signal_scope<SigNo>::fallback( handler_variant handler ) noexcept
{
    using namespace std;
    lock_guard lock( g_mtxFallback );
    g_fallback = handler;
}

template<int SigNo>
void signal_scope<SigNo>::re_init( const sigset_t *pSet, int flags )
{
    g_init.reset( pSet, flags, false );
}

template<int SigNo>
signal_scope<SigNo>::init::init()
{
    reset( nullptr, 0, true );
}

template<int SigNo>
signal_scope<SigNo>::init::~init()
{
    sigaction( SigNo, &m_saBefore, nullptr );
}

template<int SigNo>
void signal_scope<SigNo>::init::reset( const sigset_t *pSet, int flags,
bool old )
{
    struct sigaction sa;
    sa.sa_sigaction = action;
    if( pSet )
        sa.sa_mask = *pSet;
    else
        sigfillset( &sa.sa_mask );
    sa.sa_flags = flags | SA_SIGINFO;
    sigaction( SigNo, &sa, old ? &m_saBefore : nullptr );
}

template<int SigNo>
void signal_scope<SigNo>::action( int sig, siginfo_t *info, void
*uContext ) noexcept
{
    using namespace std;
    if( callHandler( t_handler, sig, info, uContext ) ) [[likely]]
        return;
    handler_variant fallback;
    {
        lock_guard lock( g_mtxFallback );
        fallback = g_fallback;
    }
    callHandler( fallback, sig, info, uContext );
}

template<int SigNo>
bool signal_scope<SigNo>::callHandler( const handler_variant &handler,
int sig, siginfo_t *info, void *uContext )
{
    if( holds_alternative<handler_fn>( handler ) ) [[likely]]
        return get<handler_fn>( handler )( sig );
    if( holds_alternative<siginfo_handler_fn>( handler ) ) [[likely]]
        return get<siginfo_handler_fn>( handler )( sig, info, uContext );
    return false;
}

well, is that your code? Or the AI's code? Need to kick that AI to the
curb from time to time. Right?
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 12.03.2026 um 21:18 schrieb DFS:

* read in a list of words from a file here: https://people.sc.fsu.edu/~jburkardt/datasets/words/words.html
* pick N random words from that list and put them in an array (OK if
  there are a few dupes - if you can get no dupes that's better)
* sort and print the array of randoms, adding a blank line each time the
  1st letter changes

#include <iostream>
#include <fstream>
#include <string>
#include <vector>
#include <sstream>
#include <algorithm>
#include <random>

using namespace std;

int main( int argc, char **argv )
{
if( argc < 2 )
return EXIT_FAILURE;
size_t nLines;
if( argc < 3 || !(istringstream( argv[2] ) >> nLines) )
nLines = 200;
ifstream ifs( argv[1] );
vector<string> lines;
size_t iLine = 0;
for( string line; iLine < nLines && !ifs.eof(); ++iLine )
if( getline( ifs, line ) )
lines.emplace_back( line );
vector<string> rndLines;
mt19937_64 mt;
for( size_t n = 0; n < nLines; ++n )
if( size_t i = mt() % nLines; lines[i].size() )
rndLines.emplace_back( move( lines[i] ) );
sort( rndLines.begin(), rndLines.end() );
iLine = 0;
string *pPrev = nullptr;
for( string &rndLine : rndLines )
{
if( pPrev && tolower( rndLine[0] ) != tolower( pPrev->front() ) )
cout << endl;
cout << ++iLine << ". " << rndLine << endl;
pPrev = &rndLine;
}
}

39 lines vs. 72 lines, and much more readability on my side.
And your code may have duplicates in the random list which
is sorted afterwards.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 13.03.2026 um 09:51 schrieb Chris M. Thomasson:

well, is that your code? Or the AI's code? Need to kick that AI to the
curb from time to time. Right?

AI never generates such dodgy ideas.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 13.03.2026 um 09:53 schrieb Bonita Montero:

Am 12.03.2026 um 21:18 schrieb DFS:

* read in a list of words from a file here:
https://people.sc.fsu.edu/~jburkardt/datasets/words/words.html
* pick N random words from that list and put them in an array (OK if
   there are a few dupes - if you can get no dupes that's better)
* sort and print the array of randoms, adding a blank line each time the
   1st letter changes

#include <iostream>
#include <fstream>
#include <string>
#include <vector>
#include <sstream>
#include <algorithm>
#include <random>

using namespace std;

int main( int argc, char **argv )
{
    if( argc < 2 )
        return EXIT_FAILURE;
    size_t nLines;
    if( argc < 3 || !(istringstream( argv[2] ) >> nLines) )
        nLines = 200;
    ifstream ifs( argv[1] );
    vector<string> lines;
    size_t iLine = 0;
    for( string line; iLine < nLines && !ifs.eof(); ++iLine )
        if( getline( ifs, line ) )
            lines.emplace_back( line );
    vector<string> rndLines;
    mt19937_64 mt;
    for( size_t n = 0; n < nLines; ++n )
        if( size_t i = mt() % nLines; lines[i].size() )

if( size_t i = mt() % lines.size(); lines[i].size() )

            rndLines.emplace_back( move( lines[i] ) );
    sort( rndLines.begin(), rndLines.end() );
    iLine = 0;
    string *pPrev = nullptr;
    for( string &rndLine : rndLines )
    {
        if( pPrev && tolower( rndLine[0] ) != tolower( pPrev->front() ) )
            cout << endl;
        cout << ++iLine << ". " << rndLine << endl;
        pPrev = &rndLine;
    }
}

39 lines vs. 72 lines, and much more readability on my side.
And your code may have duplicates in the random list which
is sorted afterwards.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/13/2026 4:53 AM, Bonita Montero wrote:

Am 12.03.2026 um 21:18 schrieb DFS:

* read in a list of words from a file here:
https://people.sc.fsu.edu/~jburkardt/datasets/words/words.html
* pick N random words from that list and put them in an array (OK if
   there are a few dupes - if you can get no dupes that's better)
* sort and print the array of randoms, adding a blank line each time the
   1st letter changes

#include <iostream>
#include <fstream>
#include <string>
#include <vector>
#include <sstream>
#include <algorithm>
#include <random>

using namespace std;

int main( int argc, char **argv )
{
    if( argc < 2 )
        return EXIT_FAILURE;
    size_t nLines;
    if( argc < 3 || !(istringstream( argv[2] ) >> nLines) )
        nLines = 200;
    ifstream ifs( argv[1] );
    vector<string> lines;
    size_t iLine = 0;
    for( string line; iLine < nLines && !ifs.eof(); ++iLine )
        if( getline( ifs, line ) )
            lines.emplace_back( line );
    vector<string> rndLines;
    mt19937_64 mt;
    for( size_t n = 0; n < nLines; ++n )
        if( size_t i = mt() % nLines; lines[i].size() )
            rndLines.emplace_back( move( lines[i] ) );
    sort( rndLines.begin(), rndLines.end() );
    iLine = 0;
    string *pPrev = nullptr;
    for( string &rndLine : rndLines )
    {
        if( pPrev && tolower( rndLine[0] ) != tolower( pPrev->front() ) )
            cout << endl;
        cout << ++iLine << ". " << rndLine << endl;
        pPrev = &rndLine;
    }
}

39 lines vs. 72 lines,

Taking out blank lines
Yours is 37 LOC
Mine is 61 LOC


But your code doesn't work right.
./montero special-english.txt 50

Outputs 35 words, and they're not random. They all begin with a.

and much more readability on my side.

C++ = gag

And your code may have duplicates in the random list which
is sorted afterwards.

My word output definitely does have dupes. C rand() produces tons of
dupe random numbers. I'm working on a method to use only unique randoms.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

I won't test my before code again.

#include <iostream>
#include <fstream>
#include <string>
#include <vector>
#include <sstream>
#include <algorithm>
#include <random>

using namespace std;

int main( int argc, char **argv )
{
if( argc < 2 )
return EXIT_FAILURE;
size_t nLines;
if( argc < 3 || !(istringstream( argv[2] ) >> nLines) )
nLines = 200;
ifstream ifs( argv[1] );
vector<string> lines;
for( string line; !ifs.eof(); )
if( getline( ifs, line ) )
lines.emplace_back( move( line ) );
vector<string> rndLines;
mt19937_64 mt;
for( size_t n = 0; n < nLines; ++n )
{
size_t i = mt() % lines.size();
rndLines.emplace_back( move( lines[i] ) );
lines.erase( lines.begin() + i );
}
sort( rndLines.begin(), rndLines.end() );
size_t iLine = 1;
string *pPrev = nullptr;
for( string &rndLine : rndLines )
{
if( pPrev && tolower( rndLine[0] ) != tolower( pPrev->front() ) )
cout << endl;
cout << iLine++ << ". " << rndLine << endl;
pPrev = &rndLine;
}
}

This is the current version.
If I call it with "test.exe anagram_dictionary.txt 10" it prints:

1. deceived
2. desorption

3. moronic

4. quantitativeness

5. reimpose

6. sleigh

7. tabooing

8. unreconciled

9. waxing
10. woody

C ia a joke, not C++.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/13/26 09:49, Bonita Montero wrote:

Am 13.03.2026 um 09:48 schrieb Chris M. Thomasson:

On 3/12/2026 10:14 PM, Bonita Montero wrote:

The list was generated with ChatGpt.

Oh my. Beware, and always double, and triple, check its reams of code.

It's only word / telephone number tuples as .txt.

The AI strikes back!

First, my apologies; I was mistakenly assuming real people.

What I did was to search for name and number (without area code).
That was answered by the search engine by "missing area code".
So I added it and got a complete record of some doctor of medicine.

But then I noticed too late that I had inadvertently got into some
AI query mode of the search engine.

So I repeated my query in ordinary mode and got nothing as result.

Then I repeated my query once again in that AI mode with the same
data and got again a complete data record, but of *another* doctor.

These inquiries just made-up those entries on the fly. - Scary!

There's a small note at the bottom that AI-results may contain
errors. But these replies are not "errors"; they are completely
made-up statements out of thin air!

Do AI inquiries for *random* test data create not only the list of
test data but also "create" the data entries as *real* in the AI's
memory for later responses to other queries?

Where AI-dreams come true...

You can of course then ask that AI whether that answer is correct,
whether that person is actually existing. - And then the AI will
willingly tell you that there's many indicators that her response
is most likely wrong, and she is enumerating all the issues (that
her original reply had)!

(I thought it's bad. But it's worse than I thought.)

Janis

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/13/26 09:54, Bonita Montero wrote:

Am 13.03.2026 um 09:51 schrieb Chris M. Thomasson:

well, is that your code? Or the AI's code? Need to kick that AI to the
curb from time to time. Right?

AI never generates such dodgy ideas.

It does. Even worse. See my most recent post.

Janis

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 13.03.2026 um 10:29 schrieb Janis Papanagnou:

On 3/13/26 09:54, Bonita Montero wrote:

Am 13.03.2026 um 09:51 schrieb Chris M. Thomasson:

well, is that your code? Or the AI's code? Need to kick that AI to
the curb from time to time. Right?

AI never generates such dodgy ideas.

It does. ...

No, AI doesn't even understand what I did there so that I had to correct
the code while reviewing it through AI. Redirecting signals to threads
is really uncommon.
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

DFS <nospam@dfs.com> writes:

On 3/12/2026 8:54 PM, Scott Lurndal wrote:

DFS <nospam@dfs.com> writes:

I'll try that in C if you commit to trying the following in C++. Deal?

-------------------------------------------------------------------------- >>> * read in a list of words from a file here:
https://people.sc.fsu.edu/~jburkardt/datasets/words/words.html

* pick N random words from that list and put them in an array (OK if
there are a few dupes - if you can get no dupes that's better)

* sort and print the array of randoms, adding a blank line each time the >>> 1st letter changes

(note: I just now wrote this. It's not old code.)

usage is
$./randwords filename N
$./randwords special_english.txt 200

-------------------------------------------------------------------------- >>

A POSIX version:

I took out all 7 instances of error trapping, and it throws:

Why on earth did you remove the error checks? They're
there for a reason.

You don't say how you invoked the executable.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

DFS <nospam@dfs.com> writes:

On 3/13/2026 12:31 AM, DFS wrote:

On 3/12/2026 8:54 PM, Scott Lurndal wrote:

DFS <nospam@dfs.com> writes:

I took out all 7 instances of error trapping, and it throws:
Floating point exception (core dumped)
at line 61 or 62

Put back in just the stat() call

stat(argv[1], &st);

and the program works again.

Yes, you can't willy-nilly remove lines from a program.

Question: wordlist was malloced with a size of 0:

wordlist = malloc(wordcount * sizeof(const char *));

Why are you allowed to malloc size 0?

"If the size of the space requested is 0, the behavior is
implementation-defined: either a null pointer shall be returned,
or the behavior shall be as if the size were some non-zero value,
except that the behavior is undefined if the returned pointer is
used to access an object."

https://pubs.opengroup.org/onlinepubs/9799919799/functions/malloc.html


--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

scott@slp53.sl.home (Scott Lurndal) writes:

DFS <nospam@dfs.com> writes:

On 3/13/2026 12:31 AM, DFS wrote:

On 3/12/2026 8:54 PM, Scott Lurndal wrote:

DFS <nospam@dfs.com> writes:

I took out all 7 instances of error trapping, and it throws:
Floating point exception (core dumped)
at line 61 or 62

Put back in just the stat() call

stat(argv[1], &st);

and the program works again.

Yes, you can't willy-nilly remove lines from a program.

Question: wordlist was malloced with a size of 0:

wordlist = malloc(wordcount * sizeof(const char *));

Why are you allowed to malloc size 0?

"If the size of the space requested is 0, the behavior is
implementation-defined: either a null pointer shall be returned,
or the behavior shall be as if the size were some non-zero value,
except that the behavior is undefined if the returned pointer is
used to access an object."

https://pubs.opengroup.org/onlinepubs/9799919799/functions/malloc.html

IIRC, this caveat was added due to differences in the malloc(3)
implementations for System V Unix and BSD Unix.
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 12/03/2026 15:10, Bonita Montero wrote:

Am 12.03.2026 um 15:43 schrieb DFS:

On 3/12/2026 10:13 AM, Bonita Montero wrote:

Here, do that in C:

static optional<size_t> parse( const char *str )
{
     size_t ret;
     if( from_chars_result fcr = from_chars( str, str +
strlen( str ), ret ); (bool)fcr.ec || *fcr.ptr )
         return nullopt;
     return ret;
}

Explain in detail what it does.

If that isn't self-explanatory stick with C.

I've got a task for you: Do the same in C:

#include <iostream>
#include <regex>
#include <fstream>
#include <vector>
#include <algorithm>
#include <vector>

using namespace std;

int main( int argc, char **argv )
{
    if( argc < 2 )
        return EXIT_FAILURE;
    ifstream ifs( argv[1] );
    static regex rxNameTel( "^\\s*\"([^\"]*)\"\\s*\"([^\"]*)\"\\s*$" );
    struct name_tel { string name, tel; };
    vector<name_tel> phoneList;
    while( !ifs.eof() )
    {
        string line;
        getline( ifs, line );
        match_results<string::const_iterator> sm;
        if( regex_match( line, sm, rxNameTel ) )
            phoneList.emplace_back( string( sm[1].first, sm[1].second ), string( sm[2].first, sm[2].second ) );
    }
    sort( phoneList.begin(), phoneList.end(),
        []( const name_tel &left, const name_tel &right ) { return left.name < right.name; } );
    for( name_tel &phone : phoneList )
        cout << "\"" << phone.name << "\"\t\"" << phone.tel << "\"" <<
endl;
}

1. Read a file and parse it with ne mentioned regex-pattern.
2. Split both parts of every line in two strings.
3. Sort the "vector" according to the first string.
4. Print it.

I guess you don't manage to do that with less than five times the work.

This stuff is child's play with any scripting language. The code will
also be cleaner and simpler.

Here you're trying to write C++ as though it was a scripting language,
by utilising its many bundled libraries, but it fails badly.

There is too much excruciating detail that you still have to write. Even
the easy bit, printing the sorted list, is a mess of punctuation.

In mine, if I use a similar record type and table name, the loop is:

for x in phonelist do
println x.name:"q15jl", x.tel:"q"
end

('q' adds quotes; '15' is field width; 'jl' is left-justified. Note your output uses hard tabs, which can give unpredictable output.)

While your compare function:

[]( const name_tel &left, const name_tel &right ) { return left.name
< right.name; }

would be '{l,r: l.name < r.name}'

I guess you don't manage to do that with less than five times the work.

Every external lib allowed.

If external libraries are allowed then C wouldn't be five times the line count. They just don't come as standard.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 13.03.2026 um 18:32 schrieb Bart:

This stuff is child's play with any scripting language. The code will
also be cleaner and simpler.

Also the C programms here that do similar things. But if your require-
ment is systems programming or performance you won't chose a scripting language.
You're really narrow-minded if you see it so superficially.

Here you're trying to write C++ as though it was a scripting language,
by utilising its many bundled libraries, but it fails badly.

No, the code works correctly according to the requirements.

There is too much excruciating detail that you still have to write.
Even the easy bit, printing the sorted list, is a mess of punctuation.

You're focussed on details without seeing the concept.

While your compare function:

   []( const name_tel &left, const name_tel &right ) { return left.name
< right.name; }
would be '{l,r: l.name < r.name}'

Yes, with one percent of the performance.

If external libraries are allowed then C wouldn't be five times the line count. They just don't come as standard.

The handling of the external libraries is much more complex.
In C++ you won't have to deal with error handling (exceptions)
and there is no explicit memory management if you use containers.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/12/2026 4:56 PM, Bonita Montero wrote:

Am 12.03.2026 um 18:32 schrieb Scott Lurndal:

You didn't bother to check if the file was opened.

That's not necessary to compare it against a equal solution in C.

This is certainly a task that can be done simply using a shell
script and the standard POSIX utility set.  awk(1) would be
a good starting point; it's possible that it can be done with
a single invocation of the posix 'sort' utility.

It's comparison of C against C++.

Of course, the lack of any in-line documentation (e.g. comments) is
a typical defect in your C++ code.

Complete idiot.

Scott is the opposite of idiot.
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/13/2026 2:33 AM, Bonita Montero wrote:

Am 13.03.2026 um 10:29 schrieb Janis Papanagnou:

On 3/13/26 09:54, Bonita Montero wrote:

Am 13.03.2026 um 09:51 schrieb Chris M. Thomasson:

well, is that your code? Or the AI's code? Need to kick that AI to
the curb from time to time. Right?

AI never generates such dodgy ideas.

It does. ...

No, AI doesn't even understand what I did there so that I had to correct
the code while reviewing it through AI. Redirecting signals to threads
is really uncommon.

At least you can take a look at the AI reams o' code, and correct it.
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/13/2026 11:57 AM, Chris M. Thomasson wrote:

On 3/13/2026 2:33 AM, Bonita Montero wrote:

Am 13.03.2026 um 10:29 schrieb Janis Papanagnou:

On 3/13/26 09:54, Bonita Montero wrote:

Am 13.03.2026 um 09:51 schrieb Chris M. Thomasson:

well, is that your code? Or the AI's code? Need to kick that AI to
the curb from time to time. Right?

AI never generates such dodgy ideas.

It does. ...

No, AI doesn't even understand what I did there so that I had to correct
the code while reviewing it through AI. Redirecting signals to threads
is really uncommon.

At least you can take a look at the AI reams o' code, and correct it.

Also, be weary of correcting the AI, the little shit will learn and then
you made it smarter? Uggg...
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/13/2026 1:49 AM, Bonita Montero wrote:

Am 13.03.2026 um 09:48 schrieb Chris M. Thomasson:

On 3/12/2026 10:14 PM, Bonita Montero wrote:

Am 13.03.2026 um 02:19 schrieb Janis Papanagnou:

This is an interesting statement. - I first saw the typical
"Max Mustermann" that is in our country the prototype of an
artificial test entry with obvious test number 0123-4567890.
Then I picked an arbitrary entry and looked it up; I found
that person with exactly the associated telephone number. -
Where did you get these real names from?

The list was generated with ChatGpt.

Oh my. Beware, and always double, and triple, check its reams of code.

It's only word / telephone number tuples as .txt.

Fine. But, think if real names leaked? Oh shit, sorry. But shit can
happen. ;^)
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 13/03/2026 17:44, Bonita Montero wrote:

Am 13.03.2026 um 18:32 schrieb Bart:

This stuff is child's play with any scripting language. The code will
also be cleaner and simpler.

Also the C programms here that do similar things. But if your require-
ment is systems programming or performance you won't chose a scripting language.
You're really narrow-minded if you see it so superficially.

Here you're trying to write C++ as though it was a scripting language,
by utilising its many bundled libraries, but it fails badly.

No, the code works correctly according to the requirements.

There is too much excruciating detail that you still have to write.
Even the easy bit, printing the sorted list, is a mess of punctuation.

You're focussed on details without seeing the concept.

We're not interested in those details! The task is what's important.

While your compare function:

    []( const name_tel &left, const name_tel &right ) { return
left.name < right.name; }
would be '{l,r: l.name < r.name}'

Yes, with one percent of the performance.

It's not that bad. An example of sorting millions of such records by
name, was 13x slower with my scripting language than g++-03.

That's with the sort function itself written in that language. With
CPython where it is built-in, it was only 3x as slow. This is before any acceleration is applied.

There are also languages that are statically typed but have cleaner
syntax with much less boilerplate, and compile to native code.

Obviously that is a far more preferable direction to go in.

BTW that C++ test, based on your small example, took 3 seconds to
compile. It must be pulling in a huge amount of stuff.


--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/12/2026 8:27 PM, Bonita Montero wrote:

I shortened my code a bit.
Do that:

#include <iostream>
#include <regex>
#include <fstream>
#include <vector>
#include <algorithm>

using namespace std;

int main( int argc, char **argv )
{
    if( argc < 2 )
        return EXIT_FAILURE;
    ifstream ifs( argv[1] );
    static regex rxNameTel( R"~(^\s*"([^"]*)"\s*"([^"]*)"\s*$)~" );
    struct name_tel { string name, tel; };
    vector<name_tel> phoneList;
    smatch sm;
    for( string line; getline( ifs, line ); )
        if( regex_match( line, sm, rxNameTel ) )
            phoneList.emplace_back( sm[1].str(), sm[2].str() );
    sort( phoneList.begin(), phoneList.end(),
        []( name_tel &left, name_tel &right ) { return left.name < right.name; } );
    for( name_tel &phone : phoneList )
        cout << "\"" << phone.name << "\"\t\"" << phone.tel << "\"" <<
endl;
}

Have a closer look at the regex.

Got a massive set of compile errors:


$ g++ montero-sort2.cpp -o montero
In file included from /usr/include/x86_64-linux-gnu/c++/11/bits/c++allocator.h:33,
from /usr/include/c++/11/bits/allocator.h:46,
from /usr/include/c++/11/string:41,
from /usr/include/c++/11/bits/locale_classes.h:40,
from /usr/include/c++/11/bits/ios_base.h:41,
from /usr/include/c++/11/ios:42,
from /usr/include/c++/11/ostream:38,
from /usr/include/c++/11/iostream:39,
from montero-sort2.cpp:1: /usr/include/c++/11/ext/new_allocator.h: In instantiation of ‘void __gnu_cxx::new_allocator<_Tp>::construct(_Up*, _Args&& ...) [with _Up = main(int, char**)::name_tel; _Args = {std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::__cxx11::basic_string<char, std::char_traits<char>,
std::allocator<char> >}; _Tp = main(int, char**)::name_tel]’: /usr/include/c++/11/bits/alloc_traits.h:516:17: required from ‘static
void std::allocator_traits<std::allocator<_CharT>

::construct(std::allocator_traits<std::allocator<_CharT>
::allocator_type&, _Up*, _Args&& ...) [with _Up = main(int, char**)::name_tel; _Args = {std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::__cxx11::basic_string<char, std::char_traits<char>,

std::allocator<char> >}; _Tp = main(int, char**)::name_tel; std::allocator_traits<std::allocator<_CharT> >::allocator_type = std::allocator<main(int, char**)::name_tel>]’ /usr/include/c++/11/bits/vector.tcc:115:30: required from ‘std::vector<_Tp, _Alloc>::reference std::vector<_Tp, _Alloc>::emplace_back(_Args&& ...) [with _Args = {std::__cxx11::basic_string<char, std::char_traits<char>,
std::allocator<char> >, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >}; _Tp = main(int, char**)::name_tel; _Alloc = std::allocator<main(int, char**)::name_tel>; std::vector<_Tp, _Alloc>::reference = main(int, char**)::name_tel&]’ montero-sort2.cpp:20:35: required from here /usr/include/c++/11/ext/new_allocator.h:162:11: error: no matching
function for call to ‘main(int, char**)::name_tel::name_tel(std::__cxx11::basic_string<char>, std::__cxx11::basic_string<char>)’
162 | { ::new((void *)__p) _Up(std::forward<_Args>(__args)...); }
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ montero-sort2.cpp:15:16: note: candidate: ‘main(int, char**)::name_tel::name_tel()’
15 | struct name_tel { string name, tel; };
| ^~~~~~~~
montero-sort2.cpp:15:16: note: candidate expects 0 arguments, 2 provided montero-sort2.cpp:15:16: note: candidate: ‘main(int, char**)::name_tel::name_tel(const main(int, char**)::name_tel&)’ montero-sort2.cpp:15:16: note: candidate expects 1 argument, 2 provided montero-sort2.cpp:15:16: note: candidate: ‘main(int, char**)::name_tel::name_tel(main(int, char**)::name_tel&&)’ montero-sort2.cpp:15:16: note: candidate expects 1 argument, 2 provided



--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

DFS <nospam@dfs.com> writes:
...

Question: wordlist was malloced with a size of 0:

wordlist = malloc(wordcount * sizeof(const char *));

Why are you allowed to malloc size 0?

Because, in some contexts, it's convenient to allow a mixture of 0-sized
and non-zero sized allocations, depending upon the value of a variable,
so some pre-standard versions of malloc() supported malloc(0) returning
a unique pointer to memory that could not be accessed. The uniqueness of
the pointer allowed the value of the pointer to be used as an identifier
for the thing that might or might not have been allocated.
This was a sufficiently common feature that the C committee decided to
allow it, but sufficiently rare that the C committee decided not to
mandate it.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 13.03.2026 um 21:11 schrieb DFS:

On 3/12/2026 8:27 PM, Bonita Montero wrote:

I shortened my code a bit.
Do that:

#include <iostream>
#include <regex>
#include <fstream>
#include <vector>
#include <algorithm>

using namespace std;

int main( int argc, char **argv )
{
     if( argc < 2 )
         return EXIT_FAILURE;
     ifstream ifs( argv[1] );
     static regex rxNameTel( R"~(^\s*"([^"]*)"\s*"([^"]*)"\s*$)~" );
     struct name_tel { string name, tel; };
     vector<name_tel> phoneList;
     smatch sm;
     for( string line; getline( ifs, line ); )
         if( regex_match( line, sm, rxNameTel ) )
             phoneList.emplace_back( sm[1].str(), sm[2].str() ); >>      sort( phoneList.begin(), phoneList.end(),
         []( name_tel &left, name_tel &right ) { return left.name < >> right.name; } );
     for( name_tel &phone : phoneList )
         cout << "\"" << phone.name << "\"\t\"" << phone.tel << "\""
<< endl;
}

Have a closer look at the regex.

Got a massive set of compile errors:

= -std=c++20

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 13.03.2026 um 20:36 schrieb Bart:

BTW that C++ test, based on your small example, took 3 seconds
to compile. It must be pulling in a huge amount of stuff.

With C++20 modules that's much faster, but I don't use them since
I've got a fast computer.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 13.03.2026 um 02:22 schrieb DFS:

On 3/12/2026 11:48 AM, Bonita Montero wrote:

Am 12.03.2026 um 16:22 schrieb DFS:

Give me the file you used.

Better take this:

That's 50 in and 50 out.

So what's the RegEx for?

For each line:
1. Skip as many whitespace as possible.
2. Match '"'.
3. Read name until a '"' comes.
4. Match '"'.
5. Skip as many whitespace as possible.
6. Repeat step 1 to 5 for the telephone number.
7. Match line end.
7. Store name and telephone number in a list.
8. Sort the list according to the name.
9. Print each entry.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/14/2026 1:01 AM, Bonita Montero wrote:

Am 13.03.2026 um 21:11 schrieb DFS:

On 3/12/2026 8:27 PM, Bonita Montero wrote:

I shortened my code a bit.
Do that:

#include <iostream>
#include <regex>
#include <fstream>
#include <vector>
#include <algorithm>

using namespace std;

int main( int argc, char **argv )
{
     if( argc < 2 )
         return EXIT_FAILURE;
     ifstream ifs( argv[1] );
     static regex rxNameTel( R"~(^\s*"([^"]*)"\s*"([^"]*)"\s*$)~" ); >>>      struct name_tel { string name, tel; };
     vector<name_tel> phoneList;
     smatch sm;
     for( string line; getline( ifs, line ); )
         if( regex_match( line, sm, rxNameTel ) )
             phoneList.emplace_back( sm[1].str(), sm[2].str() );
     sort( phoneList.begin(), phoneList.end(),
         []( name_tel &left, name_tel &right ) { return left.name <
right.name; } );
     for( name_tel &phone : phoneList )
         cout << "\"" << phone.name << "\"\t\"" << phone.tel << "\""
<< endl;
}

Have a closer look at the regex.

Got a massive set of compile errors:

= -std=c++20

Got it.

Your output is messy. What you want to do is iterate the data and find
the longest name, then pad spaces after the name so the phone numbers
line up.

See at the bottom how it should look.

===========================================================
// read in a Montero file of names and phone numbers
// trim spaces, align data, sort, print

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <ctype.h>
#include <wchar.h> //to handle multibyte characters

//compare function for qsort
int compare(const void *a, const void *b) {
return strcmp(*(const char**)a, *(const char**)b);
}

//replace character in a string
void replace(char *str, char from, char to) {
int i=0;
while(str[i] != '\0'){
if(str[i] == from){
str[i] = to;
} i++;
}
}

//remove leading and trailing spaces
void trim(char *str) {
int i=0, j=0, start=0, end=strlen(str)-1;
for (i = 0; i < strlen(str); i++) {if (isspace(str[i])) {start++;} else {break;}}
for (i = strlen(str) - 1; i > 0; i--) {if (isspace(str[i])) {end--; } else {break;}}
for (i = start; i <= end; i++) {str[j++] = str[i];}
str[j] = '\0';
}

//align name and phone, and handle multibyte characters like ü
char setspacing(char *str, int longname)
{
char *token;
char name[50] = {0};
char spaces[] = " ";
token = strtok(str," ");
int i = 0;
while(token != NULL && i < 3) {
strncat(name, token, strlen(token));
if (i==0) {strncat(name, spaces, 1);}
int namesize = strlen(name);
int mbsize = mbstowcs(NULL, name, 0);
if (mbsize < 0) {namesize += mbsize;}
if (i==1) {strncat(name, spaces, longname - namesize + 3);}
token = strtok(NULL, " ");
i++;
}

strncpy(str,name,strlen(name));
str[strlen(str)] = '\0';
}

int main(int argc, char *argv[]) {
int i = 0;
int names = 0;
char name[100] = "";
int longname = 0;

//count names, and determine longest name
FILE *fin = fopen(argv[1],"r");
while (fgets(name,sizeof name, fin) != NULL) {
names++;
trim(name);
for (i = 1; i < strlen(name); i++) {
if (name[i] == '"') {
if (i > longname) {
longname = i;
}
break;
}
}
}

//add names to array, also trim spaces, remove
//quote marks and align name/phone
rewind(fin);
i = 0;
char **namelist = malloc(sizeof(char*) * names);
while (fgets(name,sizeof name, fin) != NULL) {
trim(name);
replace(name,'"',' ');
char *dupename = strdup(name);
setspacing(dupename, longname);
int datalen = strlen(dupename);
namelist[i] = malloc(datalen + 1);
strncpy(namelist[i], dupename, datalen);
i++;
}
fclose(fin);

//sort and print names
qsort(namelist, names, sizeof(char*), compare);
for (i = 0; i < names; i++) {
printf("%2d. %s\n",i+1, namelist[i]);
}

printf("\n");
free(namelist);
return 0;
}
===========================================================

88 LOC
16.84MB executable
perfect output


$ ./dfs names-numbers-unsorted-montero.txt
1. Anna Becker 0170-2233445
2. Anna Fischer 0341-9988776
3. Anna Müller 0987-6543210
4. Ben Meier 0341-5566778
5. Ben Richter 069-3344556
6. Clara Hofmann 0157-2233445
7. Clara Zimmermann 040-55667788
8. David Schulz 030-9988776
9. Emma Bauer 0157-9988776
10. Emma Wolf 040-5566778
11. Felix Hoffmann 0711-5566778
12. Felix Neumann 0170-2233445
13. Hannah Wagner 0221-5566778
14. Jan Hoffmann 0711-3344556
15. Jana Zimmer 030-6677889
16. Jonas Klein 0171-4455667
17. Jonas Klein 030-4455667
18. Julia Neumann 0221-3344556
19. Julia Schulz 0228-4455667
20. Laura Fischer 040-98765433
21. Laura Schulze 0228-1122334
22. Lea Richter 0341-1122334
23. Lea Wagner 0151-3344556
24. Lena Fischer 0151-6677889
25. Leon Krause 089-1122334
26. Leon Zimmermann 0711-1122445
27. Leonie Klein 0341-5566778
28. Lukas Hofmann 089-6677889
29. Lukas König 069-7788990
30. Marie Becker 040-4455667
31. Marie Richter 0221-7788990
32. Max Mustermann 0123-4567890
33. Maximilian Keller 030-1122334
34. Mia Keller 089-6677889
35. Michael Braun 0170-9988776
36. Moritz Wolf 0157-4455667
37. Nina Krause 0341-4455667
38. Paul Schäfer 0228-3344556
39. Paul Wolf 030-2233445
40. Peter Schmidt 030-1234567
41. Philipp König 089-9988776
42. Sarah Braun 040-7788990
43. Sarah Lehmann 040-2233445
44. Simon Meier 030-7788990
45. Sophie Neumann 0711-3344556
46. Sophie Wagner 089-2233445
47. Tim Becker 0151-1112223
48. Tim Richter 030-6677889
49. Tim Schäfer 0711-5566778
50. Tom Bauer 0171-1122334
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 14.03.2026 um 06:49 schrieb DFS:

Your output is messy.  What you want to do is iterate the data and find
the longest name, then pad spaces after the name so the phone numbers
line up.

No, alphabetically ordered. My code exactly does that.

88 LOC
16.84MB executable
perfect output

$ ./dfs names-numbers-unsorted-montero.txt
 1. Anna Becker          0170-2233445
...

My output is the same without line numbers:

"Anna Becker" "0170-2233445"
"Anna Fischer" "0341-9988776"
"Anna Müller" "0987-6543210"
"Ben Meier" "0341-5566778"
"Ben Richter" "069-3344556"
"Clara Hofmann" "0157-2233445"
"Clara Zimmermann" "040-5566778"
"David Schulz" "030-9988776"
"Emma Bauer" "0157-9988776"
"Emma Wolf" "040-5566778"
"Felix Hoffmann" "0711-5566778"
"Felix Neumann" "0170-2233445"
"Hannah Wagner" "0221-5566778"
"Jan Hoffmann" "0711-3344556"
"Jana Zimmer" "030-6677889"
"Jonas Klein" "0171-4455667"
"Jonas Klein" "030-4455667"
"Julia Neumann" "0221-3344556"
"Julia Schulz" "0228-4455667"
"Laura Fischer" "040-9876543"
"Laura Schulze" "0228-1122334"
"Lea Richter" "0341-1122334"
"Lea Wagner" "0151-3344556"
"Lena Fischer" "0151-6677889"
"Leon Krause" "089-1122334"
"Leon Zimmermann" "0711-1122445"
"Leonie Klein" "0341-5566778"
"Lukas Hofmann" "089-6677889"
"Lukas König" "069-7788990"
"Marie Becker" "040-4455667"
"Marie Richter" "0221-7788990"
"Max Mustermann" "0123-4567890"
"Maximilian Keller" "030-1122334"
"Mia Keller" "089-6677889"
"Michael Braun" "0170-9988776"
"Moritz Wolf" "0157-4455667"
"Nina Krause" "0341-4455667"
"Paul Schäfer" "0228-3344556"
"Paul Wolf" "030-2233445"
"Peter Schmidt" "030-1234567"
"Philipp König" "089-9988776"
"Sarah Braun" "040-7788990"
"Sarah Lehmann" "040-2233445"
"Simon Meier" "030-7788990"
"Sophie Neumann" "0711-3344556"
"Sophie Wagner" "089-2233445"
"Tim Becker" "0151-1112223"
"Tim Richter" "030-6677889"
"Tim Schäfer" "0711-5566778"
"Tom Bauer" "0171-1122334"

But your code is 101 lines of code, my 25. C really sucks.
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Now my output is exactly the same as yours with 38 lines instead of 101.

#include <iostream>
#include <regex>
#include <fstream>
#include <vector>
#include <algorithm>
#include <sstream>
#include <iomanip>

using namespace std;

int main( int argc, char **argv )
{
if( argc < 2 )
return EXIT_FAILURE;
ifstream ifs( argv[1] );
static regex rxNameTel( R"~(^\s*"([^"]*)"\s*"([^"]*)"\s*$)~" );
struct name_tel { string name, tel; };
vector<name_tel> phoneList;
smatch sm;
for( string line; !ifs.eof(); )
if( getline( ifs, line ) && regex_match( line, sm, rxNameTel ) )
phoneList.emplace_back( sm[1].str(), sm[2].str() );
if( !phoneList.size() )
return EXIT_FAILURE;
sort( phoneList.begin(), phoneList.end(),
[]( name_tel &left, name_tel &right ) { return left.name < right.name;
} );
size_t
maxName = max_element( phoneList.begin(), phoneList.end(),
[]( name_tel &a, name_tel &b ) { return a.name.length() <
b.name.length(); } )->name.length(),
maxLineNo = (ostringstream() << phoneList.size()).rdbuf()->view().size();
string totalNamePad( maxName, ' ' );
size_t iLine = 1;
for( name_tel &phone : phoneList )
{
auto namePad = string_view( totalNamePad.begin(), totalNamePad.begin()
+ (maxName - phone.name.length()) );
cout << setw( maxLineNo ) << right << iLine++ << ". " << setw( maxName
) << left << phone.name << " " << phone.tel << endl;
}
}
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Eh, 35 lines. I forget to remove namePad.

#include <iostream>
#include <regex>
#include <fstream>
#include <vector>
#include <algorithm>
#include <sstream>
#include <iomanip>

using namespace std;

int main( int argc, char **argv )
{
if( argc < 2 )
return EXIT_FAILURE;
ifstream ifs( argv[1] );
static regex rxNameTel( R"~(^\s*"([^"]*)"\s*"([^"]*)"\s*$)~" );
struct name_tel { string name, tel; };
vector<name_tel> phoneList;
smatch sm;
for( string line; !ifs.eof(); )
if( getline( ifs, line ) && regex_match( line, sm, rxNameTel ) )
phoneList.emplace_back( sm[1].str(), sm[2].str() );
if( !phoneList.size() )
return EXIT_FAILURE;
sort( phoneList.begin(), phoneList.end(),
[]( name_tel &left, name_tel &right ) { return left.name < right.name;
} );
size_t
maxName = max_element( phoneList.begin(), phoneList.end(),
[]( name_tel &a, name_tel &b ) { return a.name.length() <
b.name.length(); } )->name.length(),
maxLineNo = (ostringstream() << phoneList.size()).rdbuf()->view().size();
string totalNamePad( maxName, ' ' );
size_t iLine = 1;
for( name_tel &phone : phoneList )
cout << setw( maxLineNo ) << right << iLine++ << ". " << setw( maxName
) << left << phone.name << " " << phone.tel << endl;
}
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/14/2026 2:23 AM, Bonita Montero wrote:

Am 14.03.2026 um 06:49 schrieb DFS:

Your output is messy.  What you want to do is iterate the data and
find the longest name, then pad spaces after the name so the phone
numbers line up.

No, alphabetically ordered. My code exactly does that.

88 LOC
16.84MB executable
perfect output


$ ./dfs names-numbers-unsorted-montero.txt
  1. Anna Becker          0170-2233445
...

My output is the same without line numbers:

What? Your output is FUBAR. You don't give anyone a list looking like
this:
> "Anna Becker"    "0170-2233445"

"Anna Fischer"    "0341-9988776"
"Anna Müller"    "0987-6543210"
"Ben Meier"    "0341-5566778"
"Ben Richter"    "069-3344556"
"Clara Hofmann"    "0157-2233445"
"Clara Zimmermann"    "040-5566778"
"Tom Bauer"    "0171-1122334"

....
> But your code is 101 lines of code, my 25. C really sucks.

88 LOC, and C is great.


Did you notice your 25 lines of C++ turns into a ridiculous 665MB
executable. wtf? 2/3 of a GB for that little bit of processing?
Something's wrong with that language.


But C and C++ are the wrong languages for text processing and formatting anyway. You want python (or maybe perl, or other text processing
languages).

-------------------------------------------------------------------
import sys
names = []
with open(sys.argv[1],'r') as f:
    for line in f:
if line != '\n':
line = line.replace('"','')
names.append(line.strip())
for j, name in enumerate(sorted(names)):
    e = [i.strip() for i in name.split(' ') if i]
    print("%2d. %-20s %s" % (j+1, e[0] + ' ' + e[1],e[2])) -------------------------------------------------------------------

$ python3 sortnames.py names-numbers-unsorted.txt
 1. Anna Becker 0170-2233445
 2. Anna Fischer 0341-9988776
 3. Anna Müller 0987-6543210
 4. Ben Meier 0341-5566778
 5. Ben Richter 069-3344556
 6. Clara Hofmann 0157-2233445
 7. Clara Zimmermann 040-5566778
 8. David Schulz 030-9988776
 9. Emma Bauer 0157-9988776
10. Emma Wolf 040-5566778
11. Felix Hoffmann 0711-5566778
12. Felix Neumann 0170-2233445
13. Hannah Wagner 0221-5566778
14. Jan Hoffmann 0711-3344556
15. Jana Zimmer 030-6677889
16. Jonas Klein 030-4455667
17. Jonas Klein 0171-4455667
18. Julia Neumann 0221-3344556
19. Julia Schulz 0228-4455667
20. Laura Fischer 040-9876543
21. Laura Schulze 0228-1122334
22. Lea Richter 0341-1122334
23. Lea Wagner 0151-3344556
24. Lena Fischer 0151-6677889
25. Leon Krause 089-1122334
26. Leon Zimmermann 0711-1122445
27. Leonie Klein 0341-5566778
28. Lukas Hofmann 089-6677889
29. Lukas König 069-7788990
30. Marie Becker 040-4455667
31. Marie Richter 0221-7788990
32. Max Mustermann 0123-4567890
33. Maximilian Keller 030-1122334
34. Mia Keller 089-6677889
35. Michael Braun 0170-9988776
36. Moritz Wolf 0157-4455667
37. Nina Krause 0341-4455667
38. Paul Schäfer 0228-3344556
39. Paul Wolf 030-2233445
40. Peter Schmidt 030-1234567
41. Philipp König 089-9988776
42. Sarah Braun 040-7788990
43. Sarah Lehmann 040-2233445
44. Simon Meier 030-7788990
45. Sophie Neumann 0711-3344556
46. Sophie Wagner 089-2233445
47. Tim Becker 0151-1112223
48. Tim Richter 030-6677889
49. Tim Schäfer 0711-5566778
50. Tom Bauer 0171-1122334



--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/14/2026 2:53 AM, Bonita Montero wrote:

Eh, 35 lines. I forget to remove namePad.

#include <iostream>
#include <regex>
#include <fstream>
#include <vector>
#include <algorithm>
#include <sstream>
#include <iomanip>

using namespace std;

int main( int argc, char **argv )
{
    if( argc < 2 )
        return EXIT_FAILURE;
    ifstream ifs( argv[1] );
    static regex rxNameTel( R"~(^\s*"([^"]*)"\s*"([^"]*)"\s*$)~" );
    struct name_tel { string name, tel; };
    vector<name_tel> phoneList;
    smatch sm;
    for( string line; !ifs.eof(); )
        if( getline( ifs, line ) && regex_match( line, sm, rxNameTel ) )
            phoneList.emplace_back( sm[1].str(), sm[2].str() );
    if( !phoneList.size() )
        return EXIT_FAILURE;
    sort( phoneList.begin(), phoneList.end(),
        []( name_tel &left, name_tel &right ) { return left.name < right.name; } );
    size_t
        maxName = max_element( phoneList.begin(), phoneList.end(),
            []( name_tel &a, name_tel &b ) { return a.name.length() <
b.name.length(); } )->name.length(),
        maxLineNo = (ostringstream() << phoneList.size()).rdbuf()->view().size();
    string totalNamePad( maxName, ' ' );
    size_t iLine = 1;
    for( name_tel &phone : phoneList )
        cout << setw( maxLineNo ) << right << iLine++ << ". " << setw(
maxName ) << left << phone.name << " " << phone.tel << endl;
}

You didn't handle multibyte characters like ä and ö and ü, so the phone numbers aren't lined up.

1. Anna Becker 0170-2233445
2. Anna Fischer 0341-9988776
3. Anna Müller 0987-6543210
4. Ben Meier 0341-5566778
5. Ben Richter 069-3344556
6. Clara Hofmann 0157-2233445
7. Clara Zimmermann 040-5566778
8. David Schulz 030-9988776
9. Emma Bauer 0157-9988776
10. Emma Wolf 040-5566778
11. Felix Hoffmann 0711-5566778
12. Felix Neumann 0170-2233445
13. Hannah Wagner 0221-5566778
14. Jan Hoffmann 0711-3344556
15. Jana Zimmer 030-6677889
16. Jonas Klein 030-4455667
17. Jonas Klein 0171-4455667
18. Julia Neumann 0221-3344556
19. Julia Schulz 0228-4455667
20. Laura Fischer 040-9876543
21. Laura Schulze 0228-1122334
22. Lea Richter 0341-1122334
23. Lea Wagner 0151-3344556
24. Lena Fischer 0151-6677889
25. Leon Krause 089-1122334
26. Leon Zimmermann 0711-1122445
27. Leonie Klein 0341-5566778
28. Lukas Hofmann 089-6677889
29. Lukas König 069-7788990
30. Marie Becker 040-4455667
31. Marie Richter 0221-7788990
32. Max Mustermann 0123-4567890
33. Maximilian Keller 030-1122334
34. Mia Keller 089-6677889
35. Michael Braun 0170-9988776
36. Moritz Wolf 0157-4455667
37. Nina Krause 0341-4455667
38. Paul Schäfer 0228-3344556
39. Paul Wolf 030-2233445
40. Peter Schmidt 030-1234567
41. Philipp König 089-9988776
42. Sarah Braun 040-7788990
43. Sarah Lehmann 040-2233445
44. Simon Meier 030-7788990
45. Sophie Neumann 0711-3344556
46. Sophie Wagner 089-2233445
47. Tim Becker 0151-1112223
48. Tim Richter 030-6677889
49. Tim Schäfer 0711-5566778
50. Tom Bauer 0171-1122334

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 14.03.2026 um 08:05 schrieb DFS:

You didn't handle multibyte characters like ä and ö and ü, so the phone numbers aren't lined up.

No, my output is correct. Don't trust the console,
print everyhting into a file with "xxx yyy > filename".

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/14/2026 3:10 AM, Bonita Montero wrote:

Am 14.03.2026 um 08:05 schrieb DFS:

You didn't handle multibyte characters like ä and ö and ü, so the
phone numbers aren't lined up.

No, my output is correct. Don't trust the console,
print everyhting into a file with "xxx yyy > filename".

Your output is most definitely NOT correct.

Mine is, though. See the setspacing() function to see how I did it.






--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/14/2026 1:48 AM, Bonita Montero wrote:

Am 13.03.2026 um 02:22 schrieb DFS:

On 3/12/2026 11:48 AM, Bonita Montero wrote:

Am 12.03.2026 um 16:22 schrieb DFS:

Give me the file you used.

Better take this:

That's 50 in and 50 out.

So what's the RegEx for?

For each line:
1. Skip as many whitespace as possible.
2. Match '"'.
3. Read name until a '"' comes.
4. Match '"'.
5. Skip as many whitespace as possible.
6. Repeat step 1 to 5 for the telephone number.
7. Match line end.
7. Store name and telephone number in a list.
8. Sort the list according to the name.
9. Print each entry.

Thanks.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 14.03.2026 um 08:17 schrieb DFS:

Your output is most definitely NOT correct.

The problem is that setw( xxx ) on cout doesn't support UTF-8 strings
with Visual C++ even if I imbue cout to "de_DE.utf8". I'm looking for
a solution without counting the characters myself.
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 14.03.2026 um 08:17 schrieb DFS:

On 3/14/2026 3:10 AM, Bonita Montero wrote:

Am 14.03.2026 um 08:05 schrieb DFS:

You didn't handle multibyte characters like ä and ö and ü, so the
phone numbers aren't lined up.

No, my output is correct. Don't trust the console,
print everyhting into a file with "xxx yyy > filename".

Your output is most definitely NOT correct.

Mine is, though.  See the setspacing() function to see how I did it.

Now it works without imbuing cout to de_DE.utf8:

#include <iostream>
#include <regex>
#include <fstream>
#include <vector>
#include <algorithm>
#include <sstream>
#include <iomanip>
#include <locale>

using namespace std;

int main( int argc, char **argv )
{
if( argc < 2 )
return EXIT_FAILURE;
ifstream ifs( argv[1] );
static regex rxNameTel( R"~(^\s*"([^"]*)"\s*"([^"]*)"\s*$)~" );
struct name_tel { string name, tel; };
vector<name_tel> phoneList;
smatch sm;
for( string line; !ifs.eof(); )
if( getline( ifs, line ) && regex_match( line, sm, rxNameTel ) )
phoneList.emplace_back( sm[1].str(), sm[2].str() );
if( !phoneList.size() )
return EXIT_FAILURE;
sort( phoneList.begin(), phoneList.end(),
[]( name_tel &left, name_tel &right ) { return left.name < right.name;
} );
static auto length = []( string_view sv ) { return count_if( sv.begin(), sv.end(), []( char c ) { return (c & 0xC0) != 0x80; } ); };
static auto extents = []( string_view sv ) { return count_if( sv.begin(), sv.end(), []( char c ) { return (c & 0xC0) == 0x80; } ); };
size_t
maxName = length( max_element( phoneList.begin(), phoneList.end(),
[&]( name_tel &a, name_tel &b ) { return length( a.name ) < length(
b.name ); } )->name ),
maxLineNo = (ostringstream() << phoneList.size()).rdbuf()->view().size();
size_t iLine = 1;
for( name_tel &phone : phoneList )
cout << setw( maxLineNo ) << right << iLine++ << ". " << setw( maxName
+ extents( phone.name ) ) << left << phone.name << " " << phone.tel << endl;
}
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 14/03/2026 06:53, Bonita Montero wrote:

Eh, 35 lines. I forget to remove namePad.

#include <iostream>
#include <regex>
#include <fstream>
#include <vector>
#include <algorithm>
#include <sstream>
#include <iomanip>

using namespace std;

int main( int argc, char **argv )
{
    if( argc < 2 )
        return EXIT_FAILURE;
    ifstream ifs( argv[1] );
    static regex rxNameTel( R"~(^\s*"([^"]*)"\s*"([^"]*)"\s*$)~" );
    struct name_tel { string name, tel; };
    vector<name_tel> phoneList;
    smatch sm;
    for( string line; !ifs.eof(); )
        if( getline( ifs, line ) && regex_match( line, sm, rxNameTel ) )
            phoneList.emplace_back( sm[1].str(), sm[2].str() );
    if( !phoneList.size() )
        return EXIT_FAILURE;
    sort( phoneList.begin(), phoneList.end(),
        []( name_tel &left, name_tel &right ) { return left.name < right.name; } );
    size_t
        maxName = max_element( phoneList.begin(), phoneList.end(),
            []( name_tel &a, name_tel &b ) { return a.name.length() <
b.name.length(); } )->name.length(),
        maxLineNo = (ostringstream() << phoneList.size()).rdbuf()-

view().size();

    string totalNamePad( maxName, ' ' );
    size_t iLine = 1;
    for( name_tel &phone : phoneList )
        cout << setw( maxLineNo ) << right << iLine++ << ". " << setw( maxName ) << left << phone.name << " " << phone.tel << endl;
}

I got it down to 32 lines (you left in 2 blank lines, and there was one
{ on its own line).

Doing the same to DFS's 88 line solution (that is, 88 lines without
comments), it was down to around 80 lines.

In terms of file size, it's about 1.9:1. Both use spaced indents. If I
get rid of leading white space (and some trailing white space in DFS
version), then difference is 1.7:1.

However, even at 1.9:1, the readability of DFS version is far superior.
It's also easier to maintain, and to port.

Your C++ always looks like total gobbledygook. It's also impossible to
port, even to C++ (eg. to older versions).

As for binary sizes, those aren't so interesting: using -Os -s:

BM: 87KB
DFS: 50KB

Compilation time however was very different:

BM g++-Os -s 4.9 seconds, or 6.5 lps
DFS gcc-Os -s 0.3 seconds, or 270 lps

Both are poor frankly, but the C++ was still significantly slower.

When I don't care about size or performance and need a quick build-cycle:

BM g++ -s 4.0 seconds (producing a 280KB file)
DFS tcc 0.05 seconds (producing a 5KB file)

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 14.03.2026 um 13:15 schrieb Bart:

In terms of file size, it's about 1.9:1. Both use spaced indents. If I
get rid of leading white space (and some trailing white space in DFS version), then difference is 1.7:1.

Ridiculous, a rough estimate of the relation is sufficient.

Your C++ always looks like total gobbledygook.

If you can handle C++ it's easier.

As for binary sizes, those aren't so interesting: using -Os -s:
BM:    87KB
DFS:   50KB

My PC has 128GiB of memory.

BM  g++-Os -s    4.9 seconds,  or 6.5 lps
DFS gcc-Os -s    0.3 seconds,  or 270 lps

This is worth the less development times.

Both are poor frankly, but the C++ was still significantly slower.

It's for sure as slow as most of the time is taken by the kernel
flushes to cout / stdout.
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 14/03/2026 06:53, Bonita Montero wrote:

Eh, 35 lines. I forget to remove namePad.

#include <iostream>
#include <regex>
#include <fstream>
#include <vector>
#include <algorithm>
#include <sstream>
#include <iomanip>

using namespace std;

int main( int argc, char **argv )
{
    if( argc < 2 )
        return EXIT_FAILURE;
    ifstream ifs( argv[1] );
    static regex rxNameTel( R"~(^\s*"([^"]*)"\s*"([^"]*)"\s*$)~" );
    struct name_tel { string name, tel; };
    vector<name_tel> phoneList;
    smatch sm;
    for( string line; !ifs.eof(); )
        if( getline( ifs, line ) && regex_match( line, sm, rxNameTel ) )
            phoneList.emplace_back( sm[1].str(), sm[2].str() );
    if( !phoneList.size() )
        return EXIT_FAILURE;
    sort( phoneList.begin(), phoneList.end(),
        []( name_tel &left, name_tel &right ) { return left.name < right.name; } );
    size_t
        maxName = max_element( phoneList.begin(), phoneList.end(),
            []( name_tel &a, name_tel &b ) { return a.name.length() <
b.name.length(); } )->name.length(),
        maxLineNo = (ostringstream() << phoneList.size()).rdbuf()-

view().size();

    string totalNamePad( maxName, ' ' );
    size_t iLine = 1;

    for( name_tel &phone : phoneList )
        cout << setw( maxLineNo ) << right << iLine++ << ". " << setw( maxName ) << left << phone.name << " " << phone.tel << endl;

In C it would be:

printf("%*d. %-*s %s\n", maxLineNo, iLine++, maxName, phone.name, phone.tel);

(I assume left/right set justification; a weird way of doing it.)

For that matter, this can be done in C++ too, but you chose the most long-winded syntax possible.

You seem incapable of using a simple approach when a more complicated
exists! You don't /have/ to use every possible feature you know.
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 14.03.2026 um 17:22 schrieb Bart:

cout << setw( maxLineNo ) << right << iLine++ << ". " <<
setw( maxName ) << left << phone.name << " " << phone.tel << endl;

In C it would be:
   printf("%*d. %-*s %s\n", maxLineNo, iLine++, maxName, phone.name, phone.tel);

With dynamic widths according to the maximum length of name and tel ?

For that matter, this can be done in C++ too, but you chose the most long-winded syntax possible.

With additional flexibility as I said, and type-safe;
printf() is sick in that sense.

You seem incapable of using a simple approach when a more complicated exists! You don't /have/ to use every possible feature you know.

My approach is simpler since you didn't check the constraints.
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 14/03/2026 17:04, Bonita Montero wrote:

Am 14.03.2026 um 17:22 schrieb Bart:

cout << setw( maxLineNo ) << right << iLine++ << ". " <<
setw( maxName ) << left << phone.name << " " << phone.tel << endl;

In C it would be:
    printf("%*d. %-*s %s\n", maxLineNo, iLine++, maxName, phone.name,
phone.tel);

With dynamic widths according to the maximum length of name and tel ?

That's the purpose of those '*' characters; each is replaced by the next
print item. While that '-' sign will left-justify the field.

For that matter, this can be done in C++ too, but you chose the most
long-winded syntax possible.

With additional flexibility as I said, and type-safe;
printf() is sick in that sense.

Yes that's one of my many criticisms of it: you have to get those %d and
%s formats right. But given that, here it does the job.

(In my languages, 'print' is generally simpler and also type-safe. With complex formatting needs, it becomes fiddly too. But it still uses conventional syntax. For this it would be:

println maxlineno:"v", iline++:"*", maxname:"v", phone.name:"*jl", phone.tel

However formatting info can be set up beforehand, so it can also be:

println iline++:lfmt, phone.name:namefmt, phone.tel

(You'd set the 'lfmt' variable instead of 'maxLineNo' for example.)

That 'iline++' would be different, as even a loop like this has an
accessible in index:

for i, phone in phonelist

Does 'for (name_tel &phone : phoneList)' not expose the internal index?)

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Am 14.03.2026 um 18:39 schrieb Bart:

Yes that's one of my many criticisms of it: you have to get those %d and
%s formats right. But given that, here it does the job.

Maybe some things are shorter in C, but that's mostly not typical, and
the not typesafe C code you've shown is a no-go. The C++ in this case
is more verbose and I think it's much more redable.
Most of the higher productivity in C++ comes from generic components.
Imagine you would write the containers C++ has for every template type yourself; with C++ that's a one-liner. RAII is another productivity
-enhancer and it's much easier to write safe code with that.
You've chosen a small part of my code and you're losing the overall
view. The whole program is shorter at last and doesn't have that much
lowlevel details which are hard to control in larger programs.
C is an outdated language, but it is the choice for kernel-hackers since
they don't fit with abstractions. Rust would be an improvement the Rust integration in the Linux-kernel with several thousand lines. It's just
because the kernel hackers want to flip every bit themselfes, even if
the abstractions in Rust or C++ have mostly no additional cost over a
manual implementation.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

scott@slp53.sl.home (Scott Lurndal) writes:

scott@slp53.sl.home (Scott Lurndal) writes:

DFS <nospam@dfs.com> writes:

On 3/13/2026 12:31 AM, DFS wrote:

On 3/12/2026 8:54 PM, Scott Lurndal wrote:

DFS <nospam@dfs.com> writes:

I took out all 7 instances of error trapping, and it throws:
Floating point exception (core dumped)
at line 61 or 62

Put back in just the stat() call

stat(argv[1], &st);

and the program works again.

Yes, you can't willy-nilly remove lines from a program.

Question: wordlist was malloced with a size of 0:

wordlist = malloc(wordcount * sizeof(const char *));

Why are you allowed to malloc size 0?

"If the size of the space requested is 0, the behavior is
implementation-defined: either a null pointer shall be returned,
or the behavior shall be as if the size were some non-zero value,
except that the behavior is undefined if the returned pointer is
used to access an object."

https://pubs.opengroup.org/onlinepubs/9799919799/functions/malloc.html

IIRC, this caveat was added due to differences in the malloc(3) implementations for System V Unix and BSD Unix.

That sounds right, except I might say "put in" rather than "added"
since as best I can determine that allowance was present in the
earliest drafts of the C standard and POSIX discussions.

Note that malloc() is not mentioned in K&R, and apparently was
added to AT&T Unix in Unix V7. The timing on that was about the
same time as the first BSD Unix.
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

James Kuyper <jameskuyper@alumni.caltech.edu> writes:

DFS <nospam@dfs.com> writes:
...

Question: wordlist was malloced with a size of 0:

wordlist = malloc(wordcount * sizeof(const char *));

Why are you allowed to malloc size 0?

Because, in some contexts, it's convenient to allow a mixture of
0-sized and non-zero sized allocations, depending upon the value of
a variable, so some pre-standard versions of malloc() supported
malloc(0) returning a unique pointer to memory that could not be
accessed. The uniqueness of the pointer allowed the value of the
pointer to be used as an identifier for the thing that might or
might not have been allocated.
This was a sufficiently common feature that the C committee decided
to allow it, but sufficiently rare that the C committee decided not
to mandate it.

The explanation offered by Scott Lurndal is more convincing. The
rule was put in to accommodate both AT&T Unix and BSD Unix. There
is no reason in evidence to suggest there was any evaluation by
the C committee; they allowed both behaviors because the two main
branches of Unix at the time had made different decisions.
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

Tim Rentsch <tr.17687@z991.linuxsc.com> writes:

scott@slp53.sl.home (Scott Lurndal) writes:

scott@slp53.sl.home (Scott Lurndal) writes:

DFS <nospam@dfs.com> writes:

On 3/13/2026 12:31 AM, DFS wrote:

On 3/12/2026 8:54 PM, Scott Lurndal wrote:

DFS <nospam@dfs.com> writes:

I took out all 7 instances of error trapping, and it throws:
Floating point exception (core dumped)
at line 61 or 62

Put back in just the stat() call

stat(argv[1], &st);

and the program works again.

Yes, you can't willy-nilly remove lines from a program.

Question: wordlist was malloced with a size of 0:

wordlist = malloc(wordcount * sizeof(const char *));

Why are you allowed to malloc size 0?

"If the size of the space requested is 0, the behavior is
implementation-defined: either a null pointer shall be returned,
or the behavior shall be as if the size were some non-zero value,
except that the behavior is undefined if the returned pointer is
used to access an object."

https://pubs.opengroup.org/onlinepubs/9799919799/functions/malloc.html

IIRC, this caveat was added due to differences in the malloc(3)
implementations for System V Unix and BSD Unix.

That sounds right, except I might say "put in" rather than "added"
since as best I can determine that allowance was present in the
earliest drafts of the C standard and POSIX discussions.

Note that malloc() is not mentioned in K&R, and apparently was
added to AT&T Unix in Unix V7. The timing on that was about the
same time as the first BSD Unix.

Right. K&R 1st Ed. defined a 'morecore' function based on sbrk.

K&R 2nd Ed. (ANSI) added malloc.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/14/2026 8:15 AM, Bart wrote:

On 14/03/2026 06:53, Bonita Montero wrote:

Eh, 35 lines. I forget to remove namePad.

<snip>

Your C++ always looks like total gobbledygook.

Not sure you can fix it - it's just how C++ looks.

Without using or looking back at Montero's code, I did my own research
and replicated this little bit of functionality using recommended C++
style and libraries and functions:

read file in
clean data (remove whitespace, remove quote marks)
sort
print

and this is what it looks like:

=================================================================================
#include <iostream>
#include <fstream>
#include <string>
#include <vector>
#include <regex>
#include <iomanip>

using namespace std;

int main(int argc, char* argv[]) {
vector<string> lines; // store all lines
string line; // store one line

regex re_strip ("^\\s+|\\s+$"); // regex to strip whitespace
regex re_quotes ("\""); // regex to remove quote marks

ifstream file(argv[1]); // Open the file
while (getline(file, line)) { // read lines, clean, add to array
line = regex_replace(line, re_strip , "");
line = regex_replace(line, re_quotes, "");
lines.push_back(line);
}
file.close();

sort(lines.begin(), lines.end());
int i = 0;
for (const string& stored_line : lines) {
cout << right << setw(2) << ++i << ". " << stored_line << endl;
}
return 0;
} =================================================================================

fairly hideous to look at, and 6 includes required?

Even with the -Os compile flag you mentioned, the executable is 101MB.
That's crazy for that tiny bit of functionality.

But for this little code, C++ does some nice things for you: memory management, easier regex usage, and easier sorting.

But for my money, python is the best tradeoff of ease of use,
readability, speed and functionality.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

DFS <nospam@dfs.com> writes:

On 3/14/2026 8:15 AM, Bart wrote:

On 14/03/2026 06:53, Bonita Montero wrote:

Eh, 35 lines. I forget to remove namePad.

<snip>

Your C++ always looks like total gobbledygook.

Not sure you can fix it - it's just how C++ looks.

Without using or looking back at Montero's code, I did my own research
and replicated this little bit of functionality using recommended C++
style and libraries and functions:

read file in
clean data (remove whitespace, remove quote marks)
sort
print

Technically, your example removes _leading_ and _trailing_
whitespace, no?

$ sed -e 's/^[ \t]*//;s/[ \t]*$//' -e 's/"//' < inputfile | sort

But for my money, python is the best tradeoff of ease of use,
readability, speed and functionality.

For functionality that can be composed from standard command
line utilities, even python loses.
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 16/03/2026 20:43, DFS wrote:

read file in
clean data (remove whitespace, remove quote marks)
sort
print

---
#!/bin/ksh

while read
do
echo ${REPLY//[ \"]/}
done | sort
---

[After Scott's sed]


--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 16/03/2026 20:43, DFS wrote:

On 3/14/2026 8:15 AM, Bart wrote:

On 14/03/2026 06:53, Bonita Montero wrote:

Eh, 35 lines. I forget to remove namePad.

<snip>

Your C++ always looks like total gobbledygook.

Not sure you can fix it - it's just how C++ looks.

Without using or looking back at Montero's code, I did my own research
and replicated this little bit of functionality using recommended C++
style and libraries and functions:

read file in
clean data (remove whitespace, remove quote marks)
sort
print

and this is what it looks like:

=================================================================================
#include <iostream>
#include <fstream>
#include <string>
#include <vector>
#include <regex>
#include <iomanip>

using namespace std;

int main(int argc, char* argv[]) {
    vector<string> lines;             // store all lines
    string line;                      // store one line

    regex re_strip  ("^\\s+|\\s+$"); // regex to strip whitespace
    regex re_quotes ("\"");             // regex to remove quote marks

    ifstream file(argv[1]);            // Open the file
    while (getline(file, line)) {     // read lines, clean, add to array
        line = regex_replace(line, re_strip , "");
        line = regex_replace(line, re_quotes, "");
        lines.push_back(line);
    }
    file.close();

    sort(lines.begin(), lines.end());
    int i = 0;
    for (const string& stored_line : lines) {
        cout << right << setw(2) << ++i << ". " << stored_line << endl;
    }
    return 0;
}

This works very differently:

* The two bits of info on each line (name and telno) are combined into
one string

* Extra white space between name and teleno, even outside the quotes, is
retained, and contributes to sorting

* Sorting is done on both name and telephone number

* The output doesn't put the telephone number into its own tabulated
column

* It doesn't ignore blank lines in the input


With this test input:

---------------------
"White House" "001 202 456 1414"
"White House" "001 202 456 1413"
"White House" "001 202 456 1415"
"Mother" "0049 211 151395"
"Mickey Mouse" "001 123 456 7890"

---------------------

(Note this has an extra blank line I hadn't noticed.)

BM's version produces:

1. Mickey Mouse 001 123 456 7890
2. Mother 0049 211 151395
3. White House 001 202 456 1414
4. White House 001 202 456 1413
5. White House 001 202 456 1415

Notice the three White House numbers stay in the same order. With yours:

1.
2. Mickey Mouse 001 123 456 7890
3. Mother 0049 211 151395
4. White House 001 202 456 1415
5. White House 001 202 456 1413
6. White House 001 202 456 1414

The blank line has not been ignored. The numbers aren't lined up. The
internal white space has been retained. The WH entries have moved
around, even the two without the extra spacing.

Perhaps it's not as simple as it seemed! Did your Python version do the
same as this?

It's a rather fiddly spec for me to bother doing it properly, but I
threw something together in my scripting language which is shown below.
It shows the equivalent of BM's output.

But it cuts some corners: the column widths of the output are hardcoded
(easy to fix but untidy), and it uses a custom sort routine (not shown,
but is a bubble sort), as I don't have a ready-made library routine that
takes a compare function.

=================================================================================

fairly hideous to look at, and 6 includes required?

Even with the -Os compile flag you mentioned, the executable is 101MB. That's crazy for that tiny bit of functionality.

But for this little code, C++ does some nice things for you: memory management,

Yeah. But in a language like C (or my static one), I'd just use a
slightly different approach. Maybe an extra pass over the data to
establish the size of the table, then you just allocate it all at once.

(My compiler project never actually free any memory!)

easier regex usage, and easier sorting.

I've never used regex. As you can see from my example, decent i/o
routines can eliminate the need.

Sorting though can get complex. You'd need a lot more than Sort with a
compare function to work with real phone data.

-------------------------
record rec =
var name, tel
end

f:=openfile(sread("n"))

phonelist::=()

while not eof(f) do
readln @f, name:"s", tel:"s"
nextloop when name=""

phonelist &:= rec(name, tel)
end

closefile(f)

sort(phonelist)

for i,x in phonelist do
fprintln "#. # #", i:"3", x.name:"15jl", x.tel
end

------------------

(The "s" input format reads case-preserved names or files. Inputs can be quoted to allow embedded separators within context. Quotes are
discarded. White space around items is skipped anyway.

Here also, names can have embedded quotes, but they need to be doubled up:

"White ""House""" -> White "House")


--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 16/03/2026 22:26, Richard Harnden wrote:

On 16/03/2026 20:43, DFS wrote:

read file in
clean data (remove whitespace, remove quote marks)
sort
print

---
#!/bin/ksh

while read
do
    echo ${REPLY//[ \"]/}
done | sort
---

[After Scott's sed]

If I change 'ksh' to 'bash' then this gives me this output (for the
input shown in my last post):

-----------------

MickeyMouse0011234567890
Mother0049211151395
WhiteHouse0012024561413
WhiteHouse0012024561414
WhiteHouse0012024561415
-----------------

(As I explained there, the sorting needs to be on the name field only,
and the two fields displayed in separate columns.)
--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/16/2026 4:57 PM, Scott Lurndal wrote:

DFS <nospam@dfs.com> writes:

On 3/14/2026 8:15 AM, Bart wrote:

On 14/03/2026 06:53, Bonita Montero wrote:

Eh, 35 lines. I forget to remove namePad.

<snip>

Your C++ always looks like total gobbledygook.

Not sure you can fix it - it's just how C++ looks.

Without using or looking back at Montero's code, I did my own research
and replicated this little bit of functionality using recommended C++
style and libraries and functions:

read file in
clean data (remove whitespace, remove quote marks)
sort
print

Technically, your example removes _leading_ and _trailing_
whitespace, no?

Yes.

$ sed -e 's/^[ \t]*//;s/[ \t]*$//' -e 's/"//' < inputfile | sort

This is what I got from that:

Anna Becker" "0170-2233445"
Anna Fischer" "0341-9988776"
Anna Müller" "0987-6543210"
Ben Meier" "0341-5566778"
Ben Richter" "069-3344556"
...

Not what we're looking for.

But for my money, python is the best tradeoff of ease of use,
readability, speed and functionality.

For functionality that can be composed from standard command
line utilities, even python loses.

Sure, but sed and regex are inscrutable.



original Montero file

"Max Mustermann" "0123-4567890"
"Anna Müller" "0987-6543210"
"Peter Schmidt" "030-1234567"
"Laura Fischer" "040-9876543"
"Tim Becker" "0151-1112223"
"Julia Neumann" "0221-3344556"
"Michael Braun" "0170-9988776"
"Sophie Wagner" "089-2233445"
"Felix Hoffmann" "0711-5566778"
"Lea Richter" "0341-1122334"
"Jonas Klein" "030-4455667"
"Emma Wolf" "040-5566778"
"Lukas König" "069-7788990"
"Clara Hofmann" "0157-2233445"
"Paul Schäfer" "0228-3344556"
"Mia Keller" "089-6677889"
"Leon Zimmermann" "0711-1122445"
"Nina Krause" "0341-4455667"
"David Schulz" "030-9988776"
"Sarah Lehmann" "040-2233445"
"Ben Richter" "069-3344556"
"Hannah Wagner" "0221-5566778"
"Tom Bauer" "0171-1122334"
"Lena Fischer" "0151-6677889"
"Simon Meier" "030-7788990"
"Marie Becker" "040-4455667"
"Jan Hoffmann" "0711-3344556"
"Leonie Klein" "0341-5566778"
"Philipp König" "089-9988776"
"Laura Schulze" "0228-1122334"
"Moritz Wolf" "0157-4455667"
"Jana Zimmer" "030-6677889"
"Felix Neumann" "0170-2233445"
"Sarah Braun" "040-7788990"
"Tim Schäfer" "0711-5566778"
"Anna Fischer" "0341-9988776"
"Maximilian Keller" "030-1122334"
"Lea Wagner" "0151-3344556"
"Lukas Hofmann" "089-6677889"
"Marie Richter" "0221-7788990"
"Jonas Klein" "0171-4455667"
"Clara Zimmermann" "040-5566778"
"Paul Wolf" "030-2233445"
"Sophie Neumann" "0711-3344556"
"Ben Meier" "0341-5566778"
"Emma Bauer" "0157-9988776"
"Leon Krause" "089-1122334"
"Julia Schulz" "0228-4455667"
"Tim Richter" "030-6677889"
"Anna Becker" "0170-2233445"


This 12-line python reads the file, strips leading/trailing spaces,
removes quote marks, determines the longest name for spacing (if new
longer names are added it still pretty-prints), and outputs
first-last-phone numbered and aligned and sorted by last-first.

import sys
names = []
longname = 0
with open(sys.argv[1],'r') as f:
for line in f:
line = line.replace('"','')
e = [i.strip() for i in line.split(' ') if i]
ln = len(e[0]) + len(e[1])
if ln > longname: longname = ln
names.append((e[1], e[0], e[2]))
for i,n in enumerate(sorted(names)):
print("%2d. %-*s %s" % (i+1, longname+2, n[1]+' '+n[0], n[2]))

1. Emma Bauer 0157-9988776
2. Tom Bauer 0171-1122334
3. Anna Becker 0170-2233445
4. Marie Becker 040-4455667
5. Tim Becker 0151-1112223
6. Michael Braun 0170-9988776
7. Sarah Braun 040-7788990
8. Anna Fischer 0341-9988776
9. Laura Fischer 040-9876543
10. Lena Fischer 0151-6677889
11. Felix Hoffmann 0711-5566778
12. Jan Hoffmann 0711-3344556
13. Clara Hofmann 0157-2233445
14. Lukas Hofmann 089-6677889
15. Maximilian Keller 030-1122334
16. Mia Keller 089-6677889
17. Jonas Klein 0171-4455667
18. Jonas Klein 030-4455667
19. Leonie Klein 0341-5566778
20. Leon Krause 089-1122334
21. Nina Krause 0341-4455667
22. Lukas König 069-7788990
23. Philipp König 089-9988776
24. Sarah Lehmann 040-2233445
25. Ben Meier 0341-5566778
26. Simon Meier 030-7788990
27. Max Mustermann 0123-4567890
28. Anna Müller 0987-6543210
29. Felix Neumann 0170-2233445
30. Julia Neumann 0221-3344556
31. Sophie Neumann 0711-3344556
32. Ben Richter 069-3344556
33. Lea Richter 0341-1122334
34. Marie Richter 0221-7788990
35. Tim Richter 030-6677889
36. Peter Schmidt 030-1234567
37. David Schulz 030-9988776
38. Julia Schulz 0228-4455667
39. Laura Schulze 0228-1122334
40. Paul Schäfer 0228-3344556
41. Tim Schäfer 0711-5566778
42. Hannah Wagner 0221-5566778
43. Lea Wagner 0151-3344556
44. Sophie Wagner 089-2233445
45. Emma Wolf 040-5566778
46. Moritz Wolf 0157-4455667
47. Paul Wolf 030-2233445
48. Jana Zimmer 030-6677889
49. Clara Zimmermann 040-5566778
50. Leon Zimmermann 0711-1122445


If you can do that with sed and regex and pipes in 1 or 2 lines, I'm
gonna hurl.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/16/2026 6:26 PM, Richard Harnden wrote:

On 16/03/2026 20:43, DFS wrote:

read file in
clean data (remove whitespace, remove quote marks)
sort
print

---
#!/bin/ksh

while read
do
    echo ${REPLY//[ \"]/}
done | sort
---

[After Scott's sed]

output:

AnnaBecker0170-2233445
AnnaFischer0341-9988776
AnnaMüller0987-6543210
BenMeier0341-5566778
BenRichter069-3344556
ClaraHofmann0157-2233445


Me know that not right!




--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 16/03/2026 23:09, DFS wrote:

On 3/16/2026 6:26 PM, Richard Harnden wrote:

On 16/03/2026 20:43, DFS wrote:

read file in
clean data (remove whitespace, remove quote marks)
sort
print

---
#!/bin/ksh

while read
do
     echo ${REPLY//[ \"]/}
done | sort
---

[After Scott's sed]

output:

AnnaBecker0170-2233445
AnnaFischer0341-9988776
AnnaMüller0987-6543210
BenMeier0341-5566778
BenRichter069-3344556
ClaraHofmann0157-2233445

Me know that not right!

Yes. You and Bart are right.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/16/2026 7:17 PM, Richard Harnden wrote:

On 16/03/2026 23:09, DFS wrote:

On 3/16/2026 6:26 PM, Richard Harnden wrote:

On 16/03/2026 20:43, DFS wrote:

read file in
clean data (remove whitespace, remove quote marks)
sort
print

---
#!/bin/ksh

while read
do
     echo ${REPLY//[ \"]/}
done | sort
---

[After Scott's sed]

output:

AnnaBecker0170-2233445
AnnaFischer0341-9988776
AnnaMüller0987-6543210
BenMeier0341-5566778
BenRichter069-3344556
ClaraHofmann0157-2233445


Me know that not right!

Yes.  You and Bart are right.

Can you make it look like:

Anna Becker 0170-2233445
Anna Fischer 0341-9988776
Anna Müller 0987-6543210
Ben Meier 0341-5566778
Ben Richter 069-3344556
Clara Hofmann 0157-2233445



--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 16/03/2026 23:21, DFS wrote:

On 3/16/2026 7:17 PM, Richard Harnden wrote:

On 16/03/2026 23:09, DFS wrote:

On 3/16/2026 6:26 PM, Richard Harnden wrote:

On 16/03/2026 20:43, DFS wrote:

read file in
clean data (remove whitespace, remove quote marks)
sort
print

---
#!/bin/ksh

while read
do
     echo ${REPLY//[ \"]/}
done | sort
---

[After Scott's sed]

output:

AnnaBecker0170-2233445
AnnaFischer0341-9988776
AnnaMüller0987-6543210
BenMeier0341-5566778
BenRichter069-3344556
ClaraHofmann0157-2233445


Me know that not right!

Yes.  You and Bart are right.

Can you make it look like:

Anna Becker     0170-2233445
Anna Fischer    0341-9988776
Anna Müller     0987-6543210
Ben Meier       0341-5566778
Ben Richter     069-3344556
Clara Hofmann   0157-2233445

Kinda ...

$ awk -F\" '{printf("%-20s|%s\n", $2, $4)}' input.txt \ |
sort -t\| -k1 \ |
tr -d "|" \ |
head
Anna Becker 0170-2233445
Anna Fischer 0341-9988776
Anna Müller 0987-6543210
Ben Meier 0341-5566778
Ben Richter 069-3344556
Clara Hofmann 0157-2233445
Clara Zimmermann 040-5566778
David Schulz 030-9988776
Emma Bauer 0157-9988776
Emma Wolf 040-5566778

... it's not pretty.


--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 3/16/2026 6:26 PM, Bart wrote:

On 16/03/2026 20:43, DFS wrote:

On 3/14/2026 8:15 AM, Bart wrote:

On 14/03/2026 06:53, Bonita Montero wrote:

Eh, 35 lines. I forget to remove namePad.

<snip>

Your C++ always looks like total gobbledygook.

Not sure you can fix it - it's just how C++ looks.

Without using or looking back at Montero's code, I did my own research
and replicated this little bit of functionality using recommended C++
style and libraries and functions:

read file in
clean data (remove whitespace, remove quote marks)
sort
print

and this is what it looks like:

=================================================================================
#include <iostream>
#include <fstream>
#include <string>
#include <vector>
#include <regex>
#include <iomanip>

using namespace std;

int main(int argc, char* argv[]) {
     vector<string> lines;             // store all lines >>      string line;                      // store one line

     regex re_strip  ("^\\s+|\\s+$"); // regex to strip whitespace
     regex re_quotes ("\"");             // regex to remove quote marks

     ifstream file(argv[1]);            // Open the file
     while (getline(file, line)) {     // read lines, clean, add to array
         line = regex_replace(line, re_strip , "");
         line = regex_replace(line, re_quotes, "");
         lines.push_back(line);
     }
     file.close();

     sort(lines.begin(), lines.end());
     int i = 0;
     for (const string& stored_line : lines) {
         cout << right << setw(2) << ++i << ". " << stored_line << endl;
     }
     return 0;
}

This works very differently:

It wasn't meant to duplicate Montero's.

It was meant as a demo of how C++ looks by default.

Perhaps it's not as simple as it seemed! Did your Python version do the
same as this?

My Python version:

1. Anna Becker 0170-2233445
2. Anna Fischer 0341-9988776
3. Anna Müller 0987-6543210
4. Ben Meier 0341-5566778
5. Ben Richter 069-3344556
6. Clara Hofmann 0157-2233445
7. Clara Zimmermann 040-5566778
...

it handled blank lines, but no other malformed data checks.

It's a rather fiddly spec for me to bother doing it properly, but I
threw something together in my scripting language which is shown below.
It shows the equivalent of BM's output.

But it cuts some corners: the column widths of the output are hardcoded (easy to fix but untidy), and it uses a custom sort routine (not shown,
but is a bubble sort), as I don't have a ready-made library routine that takes a compare function.

This is a python implementation of the 'quicksort' algorithm I found online.

# Simplified in-place QuickSort in Python
def quick_sort(array, low, high):
if low < high:
pi = partition(array, low, high)
quick_sort(array, low, pi - 1)
quick_sort(array, pi + 1, high)

def partition(array, low, high):
pivot = array[high]
i = low - 1
for j in range(low, high):
if array[j] <= pivot:
i += 1
array[i], array[j] = array[j], array[i]
array[i + 1], array[high] = array[high], array[i + 1]
return i + 1


I assume you could easily port it to BartScript?


Note: Python now uses PowerSort - a recent change from the long-time
TimSort.

=================================================================================

fairly hideous to look at, and 6 includes required?

Even with the -Os compile flag you mentioned, the executable is 101MB.
That's crazy for that tiny bit of functionality.

But for this little code, C++ does some nice things for you: memory
management,

Yeah. But in a language like C (or my static one), I'd just use a
slightly different approach. Maybe an extra pass over the data to
establish the size of the table, then you just allocate it all at once.

(My compiler project never actually free any memory!)

easier regex usage, and easier sorting.

I've never used regex. As you can see from my example, decent i/o
routines can eliminate the need.

Sorting though can get complex. You'd need a lot more than Sort with a compare function to work with real phone data.

-------------------------
record rec =
    var name, tel
end

f:=openfile(sread("n"))

phonelist::=()

while not eof(f) do
    readln @f, name:"s", tel:"s"
    nextloop when name=""

    phonelist &:= rec(name, tel)
end

closefile(f)

sort(phonelist)

for i,x in phonelist do
    fprintln "#. # #", i:"3", x.name:"15jl", x.tel
end

------------------

(The "s" input format reads case-preserved names or files. Inputs can be quoted to allow embedded separators within context. Quotes are
discarded. White space around items is skipped anyway.

Here also, names can have embedded quotes, but they need to be doubled up:

  "White ""House""" -> White "House")

How can we run a BartScript program?

I assume you rarely reply with C code because of the coding time?

Python takes me 1/5th to 1/3rd the time of C to write. Fantastic
language in many ways.


--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 16/03/2026 23:34, Richard Harnden wrote:

On 16/03/2026 23:21, DFS wrote:

On 3/16/2026 7:17 PM, Richard Harnden wrote:

On 16/03/2026 23:09, DFS wrote:

On 3/16/2026 6:26 PM, Richard Harnden wrote:

On 16/03/2026 20:43, DFS wrote:

read file in
clean data (remove whitespace, remove quote marks)
sort
print

---
#!/bin/ksh

while read
do
     echo ${REPLY//[ \"]/}
done | sort
---

[After Scott's sed]

output:

AnnaBecker0170-2233445
AnnaFischer0341-9988776
AnnaMüller0987-6543210
BenMeier0341-5566778
BenRichter069-3344556
ClaraHofmann0157-2233445


Me know that not right!

Yes.  You and Bart are right.

Can you make it look like:

Anna Becker     0170-2233445
Anna Fischer    0341-9988776
Anna Müller     0987-6543210
Ben Meier       0341-5566778
Ben Richter     069-3344556
Clara Hofmann   0157-2233445

Kinda ...

$ awk -F\" '{printf("%-20s|%s\n", $2, $4)}' input.txt \ |
sort -t\| -k1 \ |
tr -d "|" \ |
head
Anna Becker         0170-2233445
Anna Fischer        0341-9988776
Anna Müller         0987-6543210
Ben Meier           0341-5566778
Ben Richter         069-3344556
Clara Hofmann       0157-2233445
Clara Zimmermann    040-5566778
David Schulz        030-9988776
Emma Bauer          0157-9988776
Emma Wolf           040-5566778

... it's not pretty.

I split the long line and got the '| \' backwards :(

Second attempt, this time with line numbers ...

$ awk -F\" '{printf("%-20s|%s\n", $2, $4)}' input.txt | \
sort -t\| -k1 | \
awk -F\| 'BEGIN {i=1} {printf("%2d. %-20s|%s\n", i,$1,$2); i++}' | \
tr -d "|"

1. Anna Becker 0170-2233445
2. Anna Fischer 0341-9988776
3. Anna Müller 0987-6543210
4. Ben Meier 0341-5566778
5. Ben Richter 069-3344556
[...]
47. Tim Becker 0151-1112223
48. Tim Richter 030-6677889
49. Tim Schäfer 0711-5566778
50. Tom Bauer 0171-1122334


--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 16/03/2026 23:41, DFS wrote:

On 3/16/2026 6:26 PM, Bart wrote:

This works very differently:

It wasn't meant to duplicate Montero's.

It was meant as a demo of how C++ looks by default.

OK.

This is a python implementation of the 'quicksort' algorithm I found
online.

# Simplified in-place QuickSort in Python
def quick_sort(array, low, high):
    if low < high:
        pi = partition(array, low, high)
        quick_sort(array, low, pi - 1)
        quick_sort(array, pi + 1, high)

def partition(array, low, high):
    pivot = array[high]
    i = low - 1
    for j in range(low, high):
        if array[j] <= pivot:
            i += 1
            array[i], array[j] = array[j], array[i]
    array[i + 1], array[high] = array[high], array[i + 1]
    return i + 1

I assume you could easily port it to BartScript?

Well, I have quicksort routines, they could have been adapted, but for
this task the tricky bits were input, string processing, and formatted
output.

(Also I quite like using bubble sort because so many deride it!)

   "White ""House""" -> White "House")

How can we run a BartScript program?

I assume you're on Linux? Then download this C file:

https://github.com/sal55/langs/blob/master/qq.c

(Note: this is 2MB and is some 75Kloc of machine-generated C.)

Build it like this (instructions are at the top too):

gcc -O2 qq.c -o qq -lm -ldl -fno-strict-aliasing

At the same link you will also see hello.q and test.q (the full version
of the program that I'd posted). Note the language is actually called
'Q' and source files have .q extensions.

Test like this (file extensions are optional):

./qq hello

Run the telno program 'test.q' like this:

./qq test input

(Replace 'input' with the name of the data-file.)

(The qq.c file was generated on Windows by a special version of my
systems compiler, like this:

c:\qx>mc -linux qq
Compiling qq.m to qq.c

This translates the intermediate representation of this 35-module 40Kloc application into a single C source file, but it is very poor quality
code. The -linux option makes it include the right OS-specific module.)

I assume you rarely reply with C code because of the coding time?

I've been posting here on and off for some decades. I've posted plenty
of C code! But I'm interested in PL design and ergonomics, so take a
wider view of things than most here.

Python takes me 1/5th to 1/3rd the time of C to write.  Fantastic
language in many ways.

Yeah, but it is rather on the big scale compared to how my own tools
work. Here is how I've just run that same 'test.q' file, but with a
couple of twists:

c:\qapps>tim mm8 -r \mx\mm -r \qx\qq test abc
Compiling \mx\mm.m to \mx\mm.(run)
Compiling \qx\qq.m to \qx\qq.(run)
1. Mickey Mouse 001 123 456 7890
2. Mother 0049 211 151395
3. White House 001 202 456 1414
4. White House 001 202 456 1413
5. White House 001 202 456 1415
Time: 0.157

'tim' is a timing tool. 'mm8.exe' is a production compiler for v8 of my systems language. It has the ability to run programs from source (-r).
'qq' is the interpreter for 'Bartscript'.

Here, I built a new version of this compiler from source, ran it
immediately to build and immediately run the Q interpreter from source,
which then runs this test. It took 1/6th second in total.

This is the equivalent of first using gcc to build gcc from source; then
using that to build CPython from source; then using that to run test.py.

That's not going to happen in 1/6th of a second! Not unless you have a supercomputer.

--- Synchronet 3.21d-Linux NewsLink 1.2

From Newsgroup: comp.lang.c

On 16/03/2026 23:07, DFS wrote:

On 3/16/2026 4:57 PM, Scott Lurndal wrote:

DFS <nospam@dfs.com> writes:

On 3/14/2026 8:15 AM, Bart wrote:

On 14/03/2026 06:53, Bonita Montero wrote:

Eh, 35 lines. I forget to remove namePad.

<snip>

Your C++ always looks like total gobbledygook.

Not sure you can fix it - it's just how C++ looks.

Without using or looking back at Montero's code, I did my own research
and replicated this little bit of functionality using recommended C++
style and libraries and functions:

read file in
clean data (remove whitespace, remove quote marks)
sort
print

Technically, your example removes _leading_ and _trailing_
whitespace, no?

Yes.

$ sed -e 's/^[ \t]*//;s/[ \t]*$//' -e 's/"//' < inputfile | sort

This is what I got from that:

Anna Becker" "0170-2233445"
Anna Fischer"   "0341-9988776"
Anna Müller"   "0987-6543210"
Ben Meier" "0341-5566778"
Ben Richter" "069-3344556"
...

Not what we're looking for.

But for my money, python is the best tradeoff of ease of use,
readability, speed and functionality.

For functionality that can be composed from standard command
line utilities, even python loses.

Sure, but sed and regex are inscrutable.

original Montero file

   "Max Mustermann"   "0123-4567890"
"Anna Müller"   "0987-6543210"
    "Peter Schmidt" "030-1234567"
"Laura Fischer"      "040-9876543"
"Tim Becker" "0151-1112223"
    "Julia Neumann" "0221-3344556"
"Michael Braun"   "0170-9988776"
  "Sophie Wagner"   "089-2233445"
"Felix Hoffmann" "0711-5566778"
    "Lea Richter"   "0341-1122334"
"Jonas Klein"   "030-4455667"
"Emma Wolf" "040-5566778"
   "Lukas König" "069-7788990"
"Clara Hofmann"     "0157-2233445"
"Paul Schäfer" "0228-3344556"
    "Mia Keller" "089-6677889"
"Leon Zimmermann"   "0711-1122445"
"Nina Krause" "0341-4455667"
"David Schulz" "030-9988776"
    "Sarah Lehmann"   "040-2233445"
"Ben Richter" "069-3344556"
"Hannah Wagner" "0221-5566778"
   "Tom Bauer"   "0171-1122334"
"Lena Fischer" "0151-6677889"
   "Simon Meier" "030-7788990"
"Marie Becker"   "040-4455667"
"Jan Hoffmann"   "0711-3344556"
"Leonie Klein" "0341-5566778"
    "Philipp König" "089-9988776"
"Laura Schulze" "0228-1122334"
"Moritz Wolf" "0157-4455667"
"Jana Zimmer" "030-6677889"
    "Felix Neumann" "0170-2233445"
"Sarah Braun" "040-7788990"
"Tim Schäfer" "0711-5566778"
  "Anna Fischer"   "0341-9988776"
"Maximilian Keller" "030-1122334"
"Lea Wagner" "0151-3344556"
   "Lukas Hofmann" "089-6677889"
"Marie Richter" "0221-7788990"
     "Jonas Klein" "0171-4455667"
"Clara Zimmermann"   "040-5566778"
"Paul Wolf" "030-2233445"
"Sophie Neumann" "0711-3344556"
    "Ben Meier" "0341-5566778"
"Emma Bauer" "0157-9988776"
"Leon Krause" "089-1122334"
"Julia Schulz"   "0228-4455667"
"Tim Richter" "030-6677889"
"Anna Becker" "0170-2233445"

This 12-line python reads the file, strips leading/trailing spaces,
removes quote marks, determines the longest name for spacing (if new
longer names are added it still pretty-prints), and outputs
first-last-phone numbered and aligned and sorted by last-first.

import sys
names = []
longname = 0
with open(sys.argv[1],'r') as f:
    for line in f:
        line = line.replace('"','')
        e = [i.strip() for i in line.split(' ') if i]
        ln = len(e[0]) + len(e[1])
        if ln > longname: longname = ln
        names.append((e[1], e[0], e[2]))
for i,n in enumerate(sorted(names)):
    print("%2d. %-*s %s" % (i+1, longname+2, n[1]+' '+n[0], n[2]))

 1. Emma Bauer         0157-9988776
 2. Tom Bauer          0171-1122334
 3. Anna Becker        0170-2233445
 4. Marie Becker       040-4455667
 5. Tim Becker         0151-1112223
 6. Michael Braun      0170-9988776
 7. Sarah Braun        040-7788990
 8. Anna Fischer       0341-9988776
 9. Laura Fischer      040-9876543
10. Lena Fischer       0151-6677889
11. Felix Hoffmann     0711-5566778
12. Jan Hoffmann       0711-3344556
13. Clara Hofmann      0157-2233445
14. Lukas Hofmann      089-6677889
15. Maximilian Keller  030-1122334
16. Mia Keller         089-6677889
17. Jonas Klein        0171-4455667
18. Jonas Klein        030-4455667
19. Leonie Klein       0341-5566778
20. Leon Krause        089-1122334
21. Nina Krause        0341-4455667
22. Lukas König        069-7788990
23. Philipp König      089-9988776
24. Sarah Lehmann      040-2233445
25. Ben Meier          0341-5566778
26. Simon Meier        030-7788990
27. Max Mustermann     0123-4567890
28. Anna Müller        0987-6543210
29. Felix Neumann      0170-2233445
30. Julia Neumann      0221-3344556
31. Sophie Neumann     0711-3344556
32. Ben Richter        069-3344556
33. Lea Richter        0341-1122334
34. Marie Richter      0221-7788990
35. Tim Richter        030-6677889
36. Peter Schmidt      030-1234567
37. David Schulz       030-9988776
38. Julia Schulz       0228-4455667
39. Laura Schulze      0228-1122334
40. Paul Schäfer       0228-3344556
41. Tim Schäfer        0711-5566778
42. Hannah Wagner      0221-5566778
43. Lea Wagner         0151-3344556
44. Sophie Wagner      089-2233445
45. Emma Wolf          040-5566778
46. Moritz Wolf        0157-4455667
47. Paul Wolf          030-2233445
48. Jana Zimmer        030-6677889
49. Clara Zimmermann   040-5566778
50. Leon Zimmermann    0711-1122445

If you can do that with sed and regex and pipes in 1 or 2 lines, I'm
gonna hurl.

I can do something that works with the provided data with util-linux,
sed, awk, and coreutils:

{ sed 's/"//g' | awk '{print $1 " " $2 ":" $3}' | sort -k2 | column -t
-s: | nl -s". " ; } < data

where the file "data" contains the original list. Although there will be
some differences in behaviour for some other data and I think your
python and the above will do unintuitive things with other data.

The pipeline goes as follows:
sed strips quotes
awk makes two delimited columns
sort sorts on surnames, and where surnames are identical on telno
column presents columns with a visual model
nl adds number prefixes

If I really try and use tee and fifos I can use many fewer tools.

None of these are really good languages/toolsets for the task. A DSL
with a data schema would be best.
--
Tristan Wibberley

The message body is Copyright (C) 2026 Tristan Wibberley except
citations and quotations noted. All Rights Reserved except that you may,
of course, cite it academically giving credit to me, distribute it
verbatim as part of a usenet system or its archives, and use it to
promote my greatness and general superiority without misrepresentation
of my opinions other than my opinion of my greatness and general
superiority which you _may_ misrepresent. You definitely MAY NOT train
any production AI system with it but you may train experimental AI that
will only be used for evaluation of the AI methods it implements.

--- Synchronet 3.21d-Linux NewsLink 1.2