add sha512, see #3

README.md

@@ -39,22 +39,23 @@ Run `chkbit -u PATH` to create/update the chkbit index.
 chkbit will
 
 - create a `.chkbit` index in every subdirectory of the path it was given.
-- update the index with md5 hashes for every file.
+- update the index with md5/sha512 hashes for every file.
 - report damage for files that failed the integrity check since the last run (check the exit status).
 
 Run `chkbit PATH` to verify only.
 
 ```
-usage: chkbit.py [-h] [-u] [-f] [-i] [-w N] [-q] [-v] [PATH [PATH ...]]
+usage: chkbit.py [-h] [-u] [--algo ALGO] [-f] [-i] [-w N] [-q] [-v] [PATH ...]
 
 Checks the data integrity of your files. See https://github.com/laktak/chkbit-py
 
 positional arguments:
   PATH                directories to check
 
-optional arguments:
+options:
   -h, --help          show this help message and exit
   -u, --update        update indices (without this chkbit will only verify files)
+  --algo ALGO         hash algorithm: md5, sha512
   -f, --force         force update of damaged items
   -i, --verify-index  verify files in the index only (will not report new files)
   -w N, --workers N   number of workers to use, default=5
@@ -112,12 +113,28 @@ The disadvantage is obviously that you get hidden `.chkbit` files in your conten
 
 chkbit operates on files.
 
-When run for the first time it records a md5 hash of the file contents as well as the file modification time.
+When run for the first time it records a hash of the file contents as well as the file modification time.
 
 When you run it again it first checks the modification time,
 
-- if the time changed (because you made an edit) it records a new md5 hash.
+- if the time changed (because you made an edit) it records a new hash.
-- otherwise it will compare the current md5 to the recorded value and report an error if they do not match.
+- otherwise it will compare the current hash to the recorded value and report an error if they do not match.
+
+### I wish to use a stronger hash algorithm
+
+chkbit now supports sha512. You can specify it with `--algo sha512`.
+
+Note that existing index files will use the hash that they were created with. If you wish to update all hashes you need to delete your existing indexes first.
+
+### How can I delete the index files?
+
+List them with
+
+```
+find . -name .chkbit
+```
+
+and add `-delete` to delete.
 
 ### Can I test if chkbit is working correctly?
 

chkbit/context.py

@@ -0,0 +1,13 @@
+import hashlib
+
+
+class Context:
+    def __init__(self, verify_index, update, force, hash_algo):
+
+        self.verify_index = verify_index
+        self.update = update
+        self.force = force
+        self.hash_algo = hash_algo
+
+        if hash_algo not in ["md5", "sha512"]:
+            raise Exception(f"{hash_algo} is unknown.")

chkbit/hashfile.py

@@ -1,18 +1,25 @@
 import hashlib
 
 
-BLOCKSIZE = 2 ** 10 * 128  # kb
+BLOCKSIZE = 2**10 * 128  # kb
 
 
-def hashfile(path):
+def hashfile(path, hash_algo=None):
-    md5 = hashlib.md5()
+
+    if not hash_algo or hash_algo == "md5":
+        h = hashlib.md5()
+    elif hash_algo == "sha512":
+        h = hashlib.sha512()
+    else:
+        raise Exception(f"{hash_algo} is unknown.")
+
     with open(path, "rb") as f:
         while True:
             buf = f.read(BLOCKSIZE)
             if len(buf) <= 0:
                 break
-            md5.update(buf)
+            h.update(buf)
-    return md5.hexdigest()
+    return h.hexdigest()
 
 
 def hashtext(text):

chkbit/index.py

@@ -58,12 +58,22 @@ class Index:
             self.log(stat, os.path.join(self.path, name))
 
     # calc new hashes for this index
-    def update(self):
+    def update(self, context):
         for name in self.files:
             if self.should_ignore(name):
                 self._log(Stat.SKIP, name)
                 continue
-            self.new[name] = self._calc_file(name)
+
+            a = context.hash_algo
+            # check previously used hash
+            if name in self.old:
+                old = self.old[name]
+                if "md5" in old:
+                    a = "md5"  # legacy structure
+                    self.old[name] = {"mod": old["mod"], "a": a, "h": old["md5"]}
+                elif "a" in old:
+                    a = old["a"]
+            self.new[name] = self._calc_file(name, a)
 
     # check/update the index (old vs new)
     def check_fix(self, force):
@@ -77,7 +87,7 @@ class Index:
             b = self.new[name]
             amod = a["mod"]
             bmod = b["mod"]
-            if a["md5"] == b["md5"]:
+            if a["h"] == b["h"]:
                 # ok, if the content stays the same the mod time does not matter
                 self._log(Stat.OK, name)
                 if amod != bmod:
@@ -101,11 +111,11 @@ class Index:
                 self._log(Stat.WARN_OLD, name)
                 self._setmod()
 
-    def _calc_file(self, name):
+    def _calc_file(self, name, a):
         path = os.path.join(self.path, name)
         info = os.stat(path)
         mtime = int(info.st_mtime * 1000)
-        return {"mod": mtime, "md5": hashfile(path)}
+        return {"mod": mtime, "a": a, "h": hashfile(path, a)}
 
     def save(self):
         if self.modified:
@@ -114,7 +124,7 @@ class Index:
             data["idx_hash"] = hashtext(text)
 
             with open(self.idx_file, "w", encoding="utf-8") as f:
-                json.dump(data, f)
+                json.dump(data, f, separators=(",", ":"))
             self.modified = False
             return True
         else:
@@ -129,7 +139,11 @@ class Index:
             if "data" in data:
                 # extract old format from js version
                 for item in json.loads(data["data"]):
-                    self.old[item["name"]] = {"mod": item["mod"], "md5": item["md5"]}
+                    self.old[item["name"]] = {
+                        "mod": item["mod"],
+                        "a": "md5",
+                        "h": item["md5"],
+                    }
             elif "idx" in data:
                 self.old = data["idx"]
                 text = json.dumps(self.old, separators=(",", ":"))

chkbit/indexthread.py

@@ -6,11 +6,11 @@ from chkbit import Index, Stat
 
 
 class IndexThread:
-    def __init__(self, idx, args, res_queue, todo_queue):
+    def __init__(self, idx, context, res_queue, todo_queue):
         self.idx = idx
-        self.verify_index_only = args.verify_index
+        self.verify_index_only = context.verify_index
-        self.update = args.update and not self.verify_index_only
+        self.update = context.update and not self.verify_index_only
-        self.force = args.force
+        self.context = context
         self.todo_queue = todo_queue
         self.res_queue = res_queue
         self.t = threading.Thread(target=self.run)
@@ -40,10 +40,10 @@ class IndexThread:
         if e.load() or not self.verify_index_only:
 
             # calc the new hashes
-            e.update()
+            e.update(self.context)
 
             # compare
-            e.check_fix(self.force)
+            e.check_fix(self.context.force)
 
             # save if update is set
             if self.update:

chkbit/main.py

@@ -4,7 +4,7 @@ import time
 import argparse
 import queue
 import threading
-from chkbit import IndexThread, Stat
+from chkbit import Context, IndexThread, Stat
 
 STATUS_CODES = """
 Status codes:
@@ -63,6 +63,13 @@ class Main:
             help="update indices (without this chkbit will only verify files)",
         )
 
+        parser.add_argument(
+            "--algo",
+            type=str,
+            default="md5",
+            help="hash algorithm: md5, sha512",
+        )
+
         parser.add_argument(
             "-f", "--force", action="store_true", help="force update of damaged items"
         )
@@ -90,6 +97,7 @@ class Main:
             action="store_true",
             help="quiet, don't show progress/information",
         )
+
         parser.add_argument(
             "-v", "--verbose", action="store_true", help="verbose output"
         )
@@ -120,9 +128,16 @@ class Main:
         for path in self.args.paths:
             todo_queue.put(path)
 
+        context = Context(
+            self.args.verify_index,
+            self.args.update,
+            self.args.force,
+            self.args.algo,
+        )
+
         # start indexing
         workers = [
-            IndexThread(idx, self.args, self.res_queue, todo_queue)
+            IndexThread(idx, context, self.res_queue, todo_queue)
             for idx in range(self.args.workers)
         ]