Using sha256 instead of sha1 for file hashes now

CHANGELOG.adoc

@@ -11,6 +11,11 @@ For a detailed view of what has changed, refer to the {uri-repo}/commits/master[
 
 == v1.2.0 - 2021-09-??
 
+=== Backwards Compatibility Changes
+
+* Collected files are now stored as their SHA-256 hash value instead of SHA-1 ({uri-issue}389[#389])
+* The log field `shasum` now holds the SHA-256 hash value of files instead of SHA-1 ({uri-issue}389[#389])
+
 === Security
 
 * Backported security fixes from rdesktop to our Python C extension doing RLE processing.

README.md

@@ -242,7 +242,7 @@ pyrdp_output/
 │   ├── WinDev2108Eval.crt
 │   └── WinDev2108Eval.pem
 ├── files
-│   ├── 3dc9575a72ea896a3a910af8f4e43c92939a4421
+│   ├── e91c6a5eb3ca15df5a5cb4cf4ebb6f33b2d379a3a12d7d6de8c412d4323feb4c
 ├── filesystems
 │   ├── Kimberly835337
 │   │   └── device1
@@ -265,7 +265,7 @@ pyrdp_output/
 ```
 
 * `certs/` contains the certificates generated stored using the `CN` of the certificate as the file name
-* `files/` contains all files captured and are deduplicated by saving them using the SHA1 hash of the content as the filename
+* `files/` contains all files captured and are deduplicated by saving them using the SHA-256 hash of the content as the filename
 * `filesystems/` contains a recreation of the filesystem of the targets classified by session IDs.
    To save space on similar sessions, files are symbolic links to the actual files under `files/`.
 * `logs/` contains all the various logs with most in both JSON and plaintext formats:

pyrdp/mitm/FileMapping.py

@@ -40,19 +40,23 @@ def write(self, data: bytes):
         self.file.write(data)
         self.written = True
 
-    def getSha1Hash(self):
+    def getShaHash(self):
         with open(self.dataPath, "rb") as f:
-            sha1 = hashlib.sha1()
+            # Note: In early 2022 we switched to sha256 for file hashes. If you
+            #       want to use sha1, uncomment the next line and comment the
+            #       other one below.
+            #hash = hashlib.sha1()
+            hash = hashlib.sha256()
 
             while True:
                 buffer = f.read(65536)
 
                 if len(buffer) == 0:
                     break
 
-                sha1.update(buffer)
+                hash.update(buffer)
 
-        return sha1.hexdigest()
+        return hash.hexdigest()
 
     def finalize(self):
         if self.file.closed:
@@ -61,7 +65,7 @@ def finalize(self):
         self.log.debug("Closing file %(path)s", {"path": self.dataPath})
         self.file.close()
 
-        fileHash = self.getSha1Hash()
+        fileHash = self.getShaHash()
 
         # Go up one directory because files are saved to outDir / tmp while we're downloading them
         hashPath = (self.dataPath.parents[1] / fileHash)
@@ -82,7 +86,7 @@ def finalize(self):
             # Make the symlink relative so you can move the output folder around and it will still work.
             self.filesystemPath.symlink_to(Path(os.path.relpath(hashPath, self.filesystemPath.parent)))
 
-            self.log.info("SHA1 '%(path)s' = '%(shasum)s'", {
+            self.log.info("SHA-256 '%(path)s' = '%(shasum)s'", {
                 "path": str(self.filesystemPath.relative_to(self.filesystemDir)), "shasum": fileHash
             })
 

test/test_FileMapping.py

@@ -22,7 +22,7 @@ def setUp(self):
     def createMapping(self, mkdir: MagicMock, mkstemp: MagicMock, mock_open_object):
         mkstemp.return_value = (1, str(self.outDir / "tmp" / "tmp_test"))
         mapping = FileMapping.generate("/test", self.outDir, Path("filesystems"), self.log)
-        mapping.getSha1Hash = Mock(return_value = self.hash)
+        mapping.getShaHash = Mock(return_value = self.hash)
         mapping.file.closed = False
         return mapping, mkdir, mkstemp, mock_open_object