Fix smells

CLAUDE.md

@@ -0,0 +1,34 @@
+# CLAUDE.md
+
+This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
+
+## What this is
+
+`chicken` is a chicken-ipsum text generator (a Perl port/extension of https://github.com/sebastiancarlos/chicken-ipsum). Plain Perl 5 with only core modules (`Getopt::Long`); no build system, no test suite, no dependencies to install.
+
+## Files
+
+- `chicken.pl` — the whole program. Installed as `/usr/bin/chicken`.
+- `man_chicken.pl` — generator for the man page. Its `__DATA__` section is a roff template; every `$(args)` is replaced by the output of `./chicken.pl --seed <s> args` (anything after the args, e.g. `| tr ...`, is passed through the shell as part of the `qx()` call).
+- `chicken.1` — **generated** man page. Do not edit by hand; edit the template in `man_chicken.pl` and regenerate.
+- `install.sh` — copies the script and man page into `${DESTDIR}/usr/bin` and `${DESTDIR}/usr/share/man/man1` (used for packaging with `bee`).
+
+## Commands
+
+```sh
+./chicken.pl                        # one paragraph, 15-75 words
+./chicken.pl -nw 10 --seed 42       # deterministic output
+perl -c chicken.pl                  # syntax check
+./man_chicken.pl > chicken.1        # regenerate the man page (must run from repo root; it calls ./chicken.pl)
+man ./chicken.1                     # view it
+DESTDIR=/some/root ./install.sh     # staged install
+```
+
+There are no tests. To verify a change, compare seeded output before/after (e.g. `./man_chicken.pl | diff - chicken.1`); the man page regeneration is fully deterministic and currently reproduces the committed `chicken.1` byte-for-byte.
+
+## Things that are easy to break
+
+- **Man page seeds derive from `$VERSION`.** `man_chicken.pl` starts at `seed = VERSION * 1000` and increments once per `$(...)` substitution, in template order. So bumping `$VERSION`, inserting/removing a `$(...)` in the template, or changing the order/number of `rand()` calls in `chicken.pl` changes the text of the whole (rest of the) man page. That's expected — just regenerate `chicken.1` and commit it together with the change.
+- **When adding an option**, update three places: `GetOptions`/`USAGE` in `chicken.pl`, and the SYNOPSIS + Options sections of the template in `man_chicken.pl`; then regenerate `chicken.1`.
+- `--maxlength` works by skipping words that don't fit. Skipped words must not affect capitalization, and the paragraph is closed off afterwards: the trailing space is removed and a `.` is added, with room for it held back during the loop. Keep the default (no `-ml`) path free of extra `rand()` calls so seeded output and the man page stay stable.
+- `--namethischicken` builds its name pool from `@WORDS` (only words without a hyphen), so changing `@WORDS` changes the set of possible names.

chicken.pl

@@ -10,14 +10,14 @@ sub USAGE {
 usage: $0 [options]
 
   --numberofwords -nw <n[-m]> : number of words, ranged from n to m
-  --maxlength -ml <n>         : maximun length
+  --maxlength -ml <n>         : maximum length
   --enumerate -en     <n>     : create a list of arguments
   --namethischicken           : tell the chickens name
   --printchickenversion       : print out "$VERSION"
   --seed <n>                  : feed the random chicken
 
 eof
-    exit;
+    exit 1;
 }
 
 our %OPT;
@@ -29,7 +29,7 @@ sub USAGE {
 my $opt_maxlength               = 0;
 my $numberOfWords               = 0;
 my $opt_namethischicken         = 0;
-my $opt_seed                    = 0;
+my $opt_seed;
 
 my ( $f, $t ) = ( 0, 0 );
 my $opt_enumerate = 0;
@@ -48,7 +48,7 @@ sub USAGE {
     exit;
 }
 
-srand( $opt_seed ? $opt_seed : time() );
+srand($opt_seed) if defined $opt_seed;
 
 my @WORDS = ( "puk", "pukaak", "cluck", "cluckity", "bwak", "waak", "bok", "bwok", "cluck-a-buh-gawk", "cock-a-doodle-doo", "bwwwaak", "gobble", "honk" );
 
@@ -57,6 +57,7 @@ sub USAGE {
 if ($opt_namethischicken) {
 
     my %ChickenPopulation;
+    my @MarriedChickens;
     my @ChickenNames = grep { !/-/ } @WORDS;
 
     for my $pn (@ChickenNames) {
@@ -68,11 +69,11 @@ sub USAGE {
     # "Waak Bok" could also be married to "Cluck" which would end up in "Waak Bok-Cluck"
     for my $pn (@ChickenNames) {
         for ( keys %ChickenPopulation ) {
-            push @_, ucfirst($pn) . ' ' . $_ unless $ChickenPopulation{$_} == 1;
+            push @MarriedChickens, ucfirst($pn) . ' ' . $_ unless $ChickenPopulation{$_} == 1;
         }
     }
 
-    for (@_) { s/( \w+) /$1-/; $ChickenPopulation{$_} = 3 }
+    for (@MarriedChickens) { s/( \w+) /$1-/; $ChickenPopulation{$_} = 3 }
     my @CN = keys %ChickenPopulation;
 
     my $chickensname = $CN[ rand( scalar @CN ) ];
@@ -89,7 +90,7 @@ sub USAGE {
 else {
     print STDERR "wrong format for --numberofwords\n";
     print STDERR "should be either 'n' or 'n-m'. ( '-nw 50' or '-nw 30-90''\n";
-    exit;
+    exit 1;
 }
 
 my $enumerate = ($opt_enumerate) ? $opt_enumerate : 1;
@@ -127,14 +128,17 @@ sub USAGE {
 
         # optionally capitalize the word
         # note: always capitalize if it's the first words, or if preceded by punctuation
-        my $shouldCapitalize = $i == 0 || $previousWordHadPunctuation || rand() > 0.7;
+        my $shouldCapitalize = $paragraph eq '' || $previousWordHadPunctuation || rand() > 0.7;
 
         $word = $shouldCapitalize ? uc( substr( $word, 0, 1 ) ) . substr( $word, 1 ) : $word;
 
-        # preserve info for next iteration
-        $previousWordHadPunctuation = $shouldAddPunctuation;
+        # keep room for a final '.' in case this ends up being the last word that fits
+        my $reserve = $shouldAddPunctuation ? 0 : 1;
+
+        if ( $opt_maxlength == 0 or length($paragraph) + length($word) + $reserve <= $opt_maxlength ) {
+            # preserve info for next iteration
+            $previousWordHadPunctuation = $shouldAddPunctuation;
 
-        if ( $opt_maxlength == 0 or length($paragraph) + length($word) - 1 < $opt_maxlength ) {
             # add word to paragraph
             $paragraph = $paragraph . $word;
 
@@ -143,6 +147,12 @@ sub USAGE {
         }
     }
 
+    # words may have been skipped by --maxlength: drop the trailing space and terminate the sentence
+    if ($opt_maxlength) {
+        $paragraph =~ s/ $//;
+        $paragraph .= '.' if $paragraph ne '' and $paragraph !~ /[.!?]$/;
+    }
+
     if ($opt_enumerate) {
         printf "%d. ", $e + 1;
     }

man_chicken.pl

@@ -4,12 +4,13 @@
 my $seed = $_ * 1000;
 
 sub bokbokbok {
+    my ($args) = @_;
     $seed++;
-    return qx(./chicken.pl --seed $seed $1);
+    return qx(./chicken.pl --seed $seed $args);
 };
 
 while(<DATA>) {
-    s/\$\((.*?)\)/&bokbokbok($&)/eg;
+    s/\$\((.*?)\)/bokbokbok($1)/eg;
     print;
 }