Update fuzzing seed-corpus generation

fuzzing/Jamfile

@@ -1,27 +1,52 @@
 #
 # Copyright (c) 2019-2023 Ruben Perez Hidalgo (rubenperez038 at gmail dot com)
 # Copyright (c) 2024 Matt Borland
+# Copyright (c) 2025 Alexander Grund
 #
-# Distributed under the Boost Software License, Version 1.0. (See accompanying
-# file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
+# Distributed under the Boost Software License, Version 1.0.
+# https://www.boost.org/LICENSE_1_0.txt.
 #
 
 import common ;
+import path ;
+import python ;
 import regex ;
+import toolset ;
+
+path-constant HERE : . ;
 
 local all_fuzzers = [ regex.replace-list
     [ glob "fuzz_*.cpp" ] : ".cpp" : ""
 ] ;
 
+if ! [ python.configured ]
+{
+    using python ;
+}
+
+.make-corpus-script = $(HERE)/make-corpus.py ;
+
+rule make-corpus ( target : sources + : properties * )
+{
+    RUNNER on $(target) = [ path.native $(.make-corpus-script) ] ;
+}
+actions make-corpus
+{
+    "$(PYTHON:E=python)" "$(RUNNER)" "$(<)" "$(>)"
+}
+toolset.flags $(__name__).make-corpus PYTHON <python.interpreter> ;
+
 for local fuzzer in $(all_fuzzers)
 {
-    # These two fuzzers are the most complex ones. The rest are really
-    # simple, so less time is enough
     local fuzz_time = 30 ;
+    local corpus = /tmp/corpus/$(fuzzer) ;
+    local min_corpus = /tmp/mincorpus/$(fuzzer) ;
+    local seed_corpus = $(HERE)/seedcorpus/$(fuzzer) ;
+    local seed_files = [ glob "$(seed_corpus)/*" ] ;
 
     # Create the output corpus directories
-    make /tmp/corpus/$(fuzzer) : : common.MkDir ;
-    make /tmp/mincorpus/$(fuzzer) : : common.MkDir ;
+    make $(corpus) : $(seed_files) : make-corpus : <dependency>$(.make-corpus-script) ;
+    make $(min_corpus) : : common.MkDir ;
 
     # Build the fuzzer
     exe $(fuzzer)
@@ -37,31 +62,21 @@ for local fuzzer in $(all_fuzzers)
             <library>/boost/charconv//boost_charconv
     ;
 
-    # Make sure that any old crashes are run without problems
-    local old_crashes = [ glob-tree-ex old_crashes/$(fuzzer) : * ] ;
-    if $(old_crashes)
-    {
-        run $(fuzzer)
-            : target-name $(fuzzer)-old-crashes
-            : input-files [ SORT $(old_crashes) ]
-        ;
-    }
-
     # Run the fuzzer for a short while
     run $(fuzzer)
-        : <testing.arg>"seedcorpus/$(fuzzer) -max_total_time=$(fuzz_time)"
+        : <testing.arg>"$(corpus) -max_total_time=$(fuzz_time)"
         : target-name $(fuzzer)-fuzzing
         : requirements
-            <dependency>/tmp/corpus/$(fuzzer)
+            <dependency>$(corpus)
     ;
 
     # Minimize the corpus
     run $(fuzzer)
-        : <testing.arg>"/tmp/mincorpus/$(fuzzer) /tmp/corpus/$(fuzzer) -merge=1"
+        : <testing.arg>"$(min_corpus) $(corpus) -merge=1"
         : target-name $(fuzzer)-minimize-corpus
         : requirements
             <dependency>$(fuzzer)-fuzzing
-            <dependency>/tmp/corpus/$(fuzzer)
-            <dependency>/tmp/mincorpus/$(fuzzer)
+            <dependency>$(corpus)
+            <dependency>$(min_corpus)
     ;
 }

fuzzing/make-corpus.py

@@ -0,0 +1,32 @@
+#!/bin/env python
+
+# Copyright (c) 2025 Alexander Grund
+# Distributed under the Boost Software License, Version 1.0.
+# https://www.boost.org/LICENSE_1_0.txt.
+
+import os
+import sys
+
+def get_samples(input_files):
+    for file_name in input_files:
+        if not os.path.isfile(file_name):
+            raise RuntimeError("Not a file: " + file_name)
+        with open(file_name, 'r') as input_file:
+            yield from input_file
+
+
+def process_files(output_folder, input_files):
+    if not os.path.exists(output_folder):
+        os.makedirs(output_folder)
+
+    for i, sample in enumerate(get_samples(input_files)):
+        with open(os.path.join(output_folder, str(i) + ".txt"), 'w') as output_file:
+            output_file.write(sample)
+
+
+if __name__ == "__main__":
+    if len(sys.argv) < 3:
+        print("Usage: python script.py <output_folder> <input_file1> [<input_file2> ...]")
+        sys.exit(1)
+
+    process_files(output_folder=sys.argv[1], input_files=sys.argv[2:])