From b7e09dc7f4cad2f30d929ec6fa2887c58dfb8a5f Mon Sep 17 00:00:00 2001
From: DavidKorczynski <david@adalogics.com>
Date: Thu, 2 Jan 2025 18:21:23 +0000
Subject: [PATCH] core: avoid needing to specify language for analysis (#1931)

* core: avoid needing to specify language for analysis

Signed-off-by: David Korczynski <david@adalogics.com>

* nit

Signed-off-by: David Korczynski <david@adalogics.com>

* nit

Signed-off-by: David Korczynski <david@adalogics.com>

---------

Signed-off-by: David Korczynski <david@adalogics.com>
---
 src/fuzz_introspector/commands.py           |  3 ++
 src/fuzz_introspector/frontends/oss_fuzz.py |  8 ++++-
 src/fuzz_introspector/utils.py              | 33 +++++++++++++++++++++
 3 files changed, 43 insertions(+), 1 deletion(-)

diff --git a/src/fuzz_introspector/commands.py b/src/fuzz_introspector/commands.py
index 9278b0033..9b82c017c 100644
--- a/src/fuzz_introspector/commands.py
+++ b/src/fuzz_introspector/commands.py
@@ -46,6 +46,9 @@ def correlate_binaries_to_logs(binaries_dir: str) -> int:
 
 def end_to_end(args) -> int:
     """Runs both frontend and backend."""
+    if not args.language:
+        args.language = utils.detect_language(args.target_dir)
+
     oss_fuzz.analyse_folder(args.language, args.target_dir,
                             'LLVMFuzzerTestOneInput')
 
diff --git a/src/fuzz_introspector/frontends/oss_fuzz.py b/src/fuzz_introspector/frontends/oss_fuzz.py
index 1994c0854..f81c2e9f9 100644
--- a/src/fuzz_introspector/frontends/oss_fuzz.py
+++ b/src/fuzz_introspector/frontends/oss_fuzz.py
@@ -223,7 +223,13 @@ def process_rust_project(target_dir, out):
             f.write(f'Call tree\n{calltree}')
 
 
-def analyse_folder(language, directory, entrypoint, out='', module_only=False):
+def analyse_folder(language: str = '',
+                   directory: str = '',
+                   entrypoint: str = '',
+                   out='',
+                   module_only=False):
+    """Runs a full frontend analysis on a given directory"""
+
     if language == 'c':
         process_c_project(directory, entrypoint, out, module_only)
     if language.lower() in ['cpp', 'c++']:
diff --git a/src/fuzz_introspector/utils.py b/src/fuzz_introspector/utils.py
index 9e6295e28..8bd37bf91 100644
--- a/src/fuzz_introspector/utils.py
+++ b/src/fuzz_introspector/utils.py
@@ -21,6 +21,7 @@
 import re
 import shutil
 import yaml
+import pathlib
 
 from bs4 import BeautifulSoup
 
@@ -564,3 +565,35 @@ def locate_rust_fuzz_item(funcname: str, item_list: List[str]) -> str:
             break
 
     return ''
+
+
+def detect_language(directory) -> str:
+    """Given a folder finds the likely programming language of the project"""
+    language_extensions = {
+        'c': ['.c', '.h'],
+        'cpp': ['.cpp', '.cc', '.c++', '.h', '.hpp'],
+        'jvm': ['.java'],
+        'rust': ['.rs']
+    }
+    paths_to_avoid = [
+        '/src/aflplusplus', '/src/honggfuzz', '/src/libfuzzer', '/src/fuzztest'
+    ]
+
+    language_counts: Dict[str, int] = {}
+
+    for dirpath, _, filenames in os.walk(directory):
+        if any([x for x in paths_to_avoid if dirpath.startswith(x)]):
+            continue
+        for filename in filenames:
+            for language, extensions in language_extensions.items():
+                if pathlib.Path(filename).suffix in extensions:
+                    curr_count = language_counts.get(language, 0)
+                    language_counts[language] = curr_count + 1
+
+    max_lang = ''
+    max_count = -1
+    for language, count in language_counts.items():
+        if count >= max_count:
+            max_count = count
+            max_lang = language
+    return max_lang