better temp dir cleanup

README.md

@@ -17,9 +17,9 @@ take action on the parts that didn't.<br />
 
 ## Installation and Usage
 
-To use this script, you need Python 3 and a functioning and "recent" version of `binwalk` installed on your system.
+To use this script, you need Python 3 and a functioning and "recent" version of `binwalk` installed on your system. In practice, this means you're *probably* going to need to be on an x86/x86_64 Linux, but maybe you have better karma than I do.
 
-You'll probably also want to install optional `binwalk` dependencies such as `sasquatch`, `jefferson`, and others. You can learn how to do that in [binwalk's INSTALL.md](https://github.com/ReFirmLabs/binwalk/blob/master/INSTALL.md). `binhop` only "requires" `binwalk`, but it'll fail on binaries for which `binwalk` is dependent on optional modules.
+You'll probably also want to install optional `binwalk` dependencies such as `sasquatch`, `jefferson`, and others, depending on the binaries you want to submit. You can learn how to do that in [binwalk's INSTALL.md](https://github.com/ReFirmLabs/binwalk/blob/master/INSTALL.md). `binhop` only "requires" `binwalk`, but it'll fail on binaries for which `binwalk` is dependent on optional modules.
 
 When that's done, get `binhop` running with something like:
 
@@ -30,7 +30,7 @@ pip install -r requirements.txt
 ./binhop.py
 ```
 
-Once running, browse to [http://localhost:8080](http://localhost:8080) to upload a blob.
+Once running, browse to [http://localhost:8080](http://localhost:8080) and upload a blob.
 
 ## FAQ
 
@@ -42,7 +42,7 @@ Once running, browse to [http://localhost:8080](http://localhost:8080) to upload
 
 This repository is part of my coursework for CSC 842 - Security Tool Development at Dakota State University. Consequently, I may choose not to maintain this tool beyond the length of the course, but have selected a license that enables open contributions in any case.
 
-For aesthetics, the interface is browser-based. It'd be ideal to make it command-line only, but I ran out of time trying to summarize an arbitrarily large number of bytes and sections into a human-consumable CLI output. I'm open to ideas.
+For aesthetics, the interface is browser-based. It'd be ideal to make it command-line accessible, but I ran out of time trying to summarize an arbitrarily large number of bytes and sections into a human-consumable CLI output. I'm open to ideas.
 
 **Why did you select GPLv3? MIT is so much better.**
 

binhop.py

@@ -63,9 +63,9 @@ async def upload_file(request):
         sha1_hash.update(chunk)
         md5_hash.update(chunk)
 
+    with tempfile.TemporaryDirectory() as working_dir:
         try:
-        working_dir = tempfile.TemporaryDirectory(ignore_cleanup_errors=True)
+            scan = await scan_file(temp_file.name, working_dir)
-        scan = await scan_file(temp_file.name, working_dir.name)
             sigs = scan[0]
             extractor = sigs.extractor.output
         finally:
@@ -74,6 +74,7 @@ async def upload_file(request):
         carved, summary = [], []
         for sig in sigs.results:
             tmp_path = sig.file.path
+            summary.append("%s  0x%.8X  %s" % (sig.file.path, sig.offset, sig.description))
             if tmp_path in extractor:
                 if sig.offset in extractor[tmp_path].carved:
                     end_offset = sig.offset + os.path.getsize(extractor[tmp_path].carved[sig.offset])
@@ -86,17 +87,15 @@ async def upload_file(request):
                     extracted_files = [x for x in extractor[tmp_path].extracted[sig.offset].files if os.path.isfile(x)]
                     extracted_dirs = [x for x in extractor[tmp_path].extracted[sig.offset].files if os.path.isdir(x)]
                     summary.append(
-                    "Extracted %d files and %d directories from offset 0x%X to '%s' using '%s'"
+                        "Extracted %d files and %d directories from offset 0x%X to using '%s'"
                         % (
                             len(extracted_files),
                             len(extracted_dirs),
                             sig.offset,
-                        extractor[tmp_path].extracted[sig.offset].files[0],
                             sigs.extractor.output[tmp_path].extracted[sig.offset].command,
                         )
                     )
-    num_files, num_dirs, listing = await build_listing(working_dir.name)
+        num_files, num_dirs, listing = await build_listing(working_dir)
-    working_dir.cleanup()
 
     response_data = {
         "meta": {