From 6c8dcd46a59deb550f96c8910f1d6199c6db14e2 Mon Sep 17 00:00:00 2001
From: RichardG867 <richardg867@gmail.com>
Date: Wed, 22 Nov 2023 21:21:56 -0300
Subject: [PATCH] PEExtractor: Add Inno Setup (used by NEC) extraction

---
 Dockerfile              |  2 +-
 biostools/extractors.py | 33 +++++++++++++++++++++++++++++++++
 2 files changed, 34 insertions(+), 1 deletion(-)

diff --git a/Dockerfile b/Dockerfile
index 2beff13..561977b 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -47,7 +47,7 @@ FROM debian:bullseye
 # Install runtime dependencies.
 RUN sed -i -e 's/main/main contrib non-free/' /etc/apt/sources.list && \
 	apt-get update && \
-	DEBIAN_FRONTEND=noninteractive apt-get install -y python3 python3-pip p7zip-full p7zip-rar unshield qemu-system-x86 && \
+	DEBIAN_FRONTEND=noninteractive apt-get install -y python3 python3-pip p7zip-full p7zip-rar innoextract unshield qemu-system-x86 && \
 	apt-get clean && \
 	rm -rf /var/lib/apt/lists
 
diff --git a/biostools/extractors.py b/biostools/extractors.py
index a9105be..4b518aa 100644
--- a/biostools/extractors.py
+++ b/biostools/extractors.py
@@ -2300,6 +2300,13 @@ class PEExtractor(ArchiveExtractor):
 			if ret:
 				return ret
 
+		# Cover Inno Setup installers.
+		if file_header[48:52] == b'Inno':
+			# Determine if this executable can be extracted with innoextract.
+			ret = self._extract_inno(file_path, file_header, dest_dir)
+			if ret:
+				return ret
+
 		# Read up to 16 MB as a safety net.
 		file_header += util.read_complement(file_path, file_header)
 
@@ -2455,6 +2462,32 @@ class PEExtractor(ArchiveExtractor):
 		# Return destination directory path.
 		return dest_dir
 
+	def _extract_inno(self, file_path, file_header, dest_dir):
+		# Create destination directory and stop if it couldn't be created.
+		if not util.try_makedirs(dest_dir):
+			return True
+
+		# Run innoextract command.
+		try:
+			subprocess.run(['innoextract', '-e', os.path.abspath(file_path)], stdout=self._devnull, stderr=subprocess.STDOUT, cwd=dest_dir)
+		except:
+			pass
+
+		# Assume failure if nothing was extracted.
+		files_extracted = os.listdir(dest_dir)
+		if len(files_extracted) < 1:
+			self.debug_print('Extraction produced no files:', file_path)
+			return False
+
+		# Remove file.
+		try:
+			os.remove(file_path)
+		except:
+			pass
+
+		# Return destination directory path.
+		return dest_dir
+
 
 class TarExtractor(ArchiveExtractor):
 	"""Extract tar archives."""