Add BLAKE2 b2sum support

Changelog

@@ -1,6 +1,7 @@
 2025-xx-xx - v3.2.1.dev0:
 	* WARNING: This is a development snapshot, not a stable release.
 	* Support for BLAKE3 b3sum format (Thanks to Artur Ladka).
+	* Support for BLAKE2 b2sum format (Thanks to Artur Ladka).
 
 2025-11-02 - v3.2.0:
 	* Drop support for Python 3.6.

README.md

@@ -2,7 +2,7 @@
 
 cfv is a utility to test and create a wide range of checksum verification files.
 It currently supports testing and creating sfv, sfvmd5, csv, csv2, csv4, crc, md5, bsdmd5,
-sha1, sha224, sha256, sha384, sha512, b3 (BLAKE3) and torrent files.
+sha1, sha224, sha256, sha384, sha512, b2 (BLAKE2b), b3 (BLAKE3) and torrent files.
 Test-only support is available for par, par2.
 
 cfv was originally written by Matthew Mueller ([original project home](http://cfv.sourceforge.net/)).

cfv.1

@@ -1,6 +1,6 @@
 .TH cfv 1 "05 Feb 2005"
 .SH NAME
-cfv \- Verify file consistency with .sfv, .csv, .crc, .md5, md5sum, sha1sum, sha224sum, sha256sum, sha384sum, sha512sum, b3sum, .torrent, .par, or .par2 files
+cfv \- Verify file consistency with .sfv, .csv, .crc, .md5, md5sum, sha1sum, sha224sum, sha256sum, sha384sum, sha512sum, b2sum, b3sum, .torrent, .par, or .par2 files
 .SH SYNOPSIS
 .B cfv [\-p dir] [\-v|\-V|\-VV|\-q|\-Q] [\-\-progress VAL] [\-r|\-rr|\-R] [\-l|\-L] [\-n|\-N] [\-\-renameformat <s>] [\-s|\-S] [\-zz|\-z|\-Z|\-ZZ] [\-T|\-C] [\-m|\-M] [\-i|\-I] [\-u|\-uu|\-U] [\-\-encoding <e>] [\-\-unquote <b>] [\-\-fixpaths <s>] [\-\-strippaths <p>] [\-\-showpaths <i>] [\-\-list/\-\-list0 <l>] [\-\-announceurl <u>] [\-\-piece_size_pow2 <n>] [\-\-private_torrent] [\-\-length <n>] [\-t type] [\-f file] [files...]
 .SH DESCRIPTION
@@ -120,7 +120,7 @@ Specify the name of the checksum file to test or create.
 If file is \-, stdin (for \-T) or stdout (for \-C) will be used.
 .IP "\-t type"
 Specify the type of the file.
-Can be sfv, sfvmd5, csv, csv2, csv4, crc, md5, bsdmd5, sha1, sha224, sha256, sha384, sha512, b3, torrent, par, par2, auto, or help.
+Can be sfv, sfvmd5, csv, csv2, csv4, crc, md5, bsdmd5, sha1, sha224, sha256, sha384, sha512, b2, b3, torrent, par, par2, auto, or help.
 For details see supported checksum formats section.
 If the type is help, or an unknown type is given, a list of the types and their descriptions will be printed.
 The default is auto, which will detect the file type for you.
@@ -130,8 +130,9 @@ Prints a raw listing of files in the given set (ok, bad, unverified, notfound).
 .IP "\-\-list0 listset"
 Like \-\-list but files are separated by a null char.  Useful in combination with xargs \-0.
 .IP "\-\-length BITS"
-Set digest length in bits for hash types that support variable output (b3).
-The default for b3 is 256 bits. Must be a multiple of 8.
+Set digest length in bits for hash types that support variable output (b2, b3).
+The default is 512 bits for b2 and 256 bits for b3. Must be a multiple of 8.
+For b2 the maximum is 512 bits.
 .IP "\-\-announceurl URL"
 Tracker announce URL for .torrent file creation.
 .IP "\-\-piece_size_pow2 N"
@@ -185,6 +186,9 @@ is a GNU sha384sum format file
 .B sha512
 is a GNU sha512sum format file
 .br
+.B b2
+is a BLAKE2b b2sum format file (B2SUMS, .b2, .blk). Default digest size is 512 bits.
+.br
 .B b3
 is a BLAKE3 b3sum format file (B3SUMS, .b3, .bk3). Requires the Python blake3 module.
 .br

lib/cfv/common.py

@@ -785,78 +785,119 @@ def make_addfile(self, filename):
     pass
 
 
-# ---------- b3 (BLAKE3) ----------
+# ---------- HashSum base (shared by b2/b3) ----------
+
+# Match any even-length hex string (supports variable digest lengths when checking)
+_hashsum_rem = re.compile(r'((?:[0-9a-fA-F]{2})+) [ *]([^\r\n]+)[\r\n]*$')
 
-class BLAKE3_MixIn(object):
-    digest_size = 32
-    hash_name = 'blake3'
+
+class HashSumBase(TextChksumType):
+    """Shared base for coreutils-style hash checksum formats (b2sum, b3sum)."""
+    auto_chksumfile_order = 1
+    digest_size = None      # default digest size in bytes, set by subclass
+    max_digest_size = None  # max allowed digest size in bytes (None = unlimited)
+    hash_name = None        # e.g. 'blake2b', 'blake3'
+    _hash_func = None       # e.g. hash.getfileblake2b, hash.getfileblake3
+    _std_ext = None         # e.g. '.b2', '.b3'
 
     def do_test_file(self, filename, filecrc):
         # Derive digest size from the checksum in the file (supports variable lengths)
         digest_size = len(filecrc)
-
-        c = getfilehash(filename, '%s-%d' % (self.hash_name, digest_size), hash.getfileblake3, digest_size)[0]
+        c = getfilehash(filename, '%s-%d' % (self.hash_name, digest_size), self._hash_func, digest_size)[0]
         if c != filecrc:
             return c
 
-
-class BLAKE3(TextChksumType, BLAKE3_MixIn):
-    name = 'b3'
-    description = 'BLAKE3 b3sum'
-    descinfo = 'BLAKE3,name'
-    auto_chksumfile_order = 1
-    auto_filename_match = r'b3sum|\.(b3|bk3)$'
-
-    # Match any even-length hex string (supports variable digest lengths when checking)
-    _b3rem = re.compile(r'((?:[0-9a-fA-F]{2})+) [ *]([^\r\n]+)[\r\n]*$')
-
     def do_test_chksumfile_print_testingline(self, file):
         comment = parse_commentline(file.peekline(512).lstrip(), ';')
         TextChksumType.do_test_chksumfile_print_testingline(self, file, comment)
 
-    @staticmethod
-    def _is_b3_filename(filename):
-        if not filename:
-            return False
-        lname = filename.lower()
-        if lname.endswith('.gz'):
-            lname = lname[:-3]
-        return lname.endswith('.b3') or lname.endswith('.bk3') or 'b3sum' in os.path.basename(lname)
-
     @classmethod
     def auto_chksumfile_match(cls, file):
-        if not cls._is_b3_filename(file.name):
+        if not cls._is_valid_filename(file.name):
             return False
         line = file.peekline(4096)
         while line:
             stripped = line.lstrip()
             if not stripped or stripped.startswith(';'):
                 line = file.peeknextline(4096)
                 continue
-            return cls._b3rem.match(stripped) is not None
+            return _hashsum_rem.match(stripped) is not None
         return False
 
     def do_test_chksumline(self, line):
         stripped = line.lstrip()
         if not stripped or stripped.startswith(';'):
             return
-        x = self._b3rem.match(stripped)
+        x = _hashsum_rem.match(stripped)
         if not x:
             return -1
+        digest_size = len(x.group(1)) // 2
+        if self.max_digest_size and digest_size > self.max_digest_size:
+            return -1
         self.test_file(x.group(2), strutil.unhexlify(x.group(1)))
 
-    @staticmethod
-    def make_std_filename(filename):
-        return filename + '.b3'
+    @classmethod
+    def make_std_filename(cls, filename):
+        return filename + cls._std_ext
 
     def make_addfile(self, filename):
         digest_size = config.hash_length // 8 if config.hash_length else self.digest_size
+        if self.max_digest_size and digest_size > self.max_digest_size:
+            raise CFVValueError('--length for %s must be between 8 and %d' % (self.name, self.max_digest_size * 8))
         cache_key = '%s-%d' % (self.hash_name, digest_size)
-        digest = getfilehash(filename, cache_key, hash.getfileblake3, digest_size)[0]
+        digest = getfilehash(filename, cache_key, self._hash_func, digest_size)[0]
         hexdigest = strutil.hexlify(digest)
         return (hexdigest, -1), '%s  %s' % (hexdigest, filename) + os.linesep
 
 
+# ---------- b2 (BLAKE2) ----------
+
+class BLAKE2(HashSumBase):
+    name = 'b2'
+    description = 'BLAKE2b b2sum'
+    descinfo = 'BLAKE2b,name'
+    digest_size = 64
+    max_digest_size = 64
+    hash_name = 'blake2b'
+    _hash_func = staticmethod(hash.getfileblake2b)
+    _std_ext = '.b2'
+    auto_filename_match = r'b2sum|\.(b2|blk)$'
+
+    @staticmethod
+    def _is_valid_filename(filename):
+        if not filename:
+            return False
+        lname = filename.lower()
+        if lname.endswith('.gz'):
+            lname = lname[:-3]
+        return lname.endswith('.b2') or lname.endswith('.blk') or 'b2sum' in os.path.basename(lname)
+
+
+cftypes.register_cftype(BLAKE2)
+
+
+# ---------- b3 (BLAKE3) ----------
+
+class BLAKE3(HashSumBase):
+    name = 'b3'
+    description = 'BLAKE3 b3sum'
+    descinfo = 'BLAKE3,name'
+    digest_size = 32
+    hash_name = 'blake3'
+    _hash_func = staticmethod(hash.getfileblake3)
+    _std_ext = '.b3'
+    auto_filename_match = r'b3sum|\.(b3|bk3)$'
+
+    @staticmethod
+    def _is_valid_filename(filename):
+        if not filename:
+            return False
+        lname = filename.lower()
+        if lname.endswith('.gz'):
+            lname = lname[:-3]
+        return lname.endswith('.b3') or lname.endswith('.bk3') or 'b3sum' in os.path.basename(lname)
+
+
 cftypes.register_cftype(BLAKE3)
 
 
@@ -1797,6 +1838,10 @@ def make(cftype, ifilename, testfiles):
         view.ev_make_cf_typenotsupported(filename, cftype)
         stats.cferror += 1
         return
+    max_digest_size = getattr(cftype, 'max_digest_size', None)
+    if max_digest_size and config.hash_length and config.hash_length // 8 > max_digest_size:
+        view.perror('cfv: --length for %s must be between 8 and %d' % (cftype.name, max_digest_size * 8))
+        sys.exit(1)
     if os.path.exists(filename):
         view.ev_make_cf_alreadyexists(filename)
         stats.cferror += 1
@@ -2005,7 +2050,7 @@ def show_unverified_files(filelist):
 # md5sum/sha1sum files have no standard extension, so just search for
 # files with md5/sha1 in the name anywhere, and let the test func see
 # if it really is one.
-atrem = re.compile(r'md5|sha1|sha224|sha256|sha384|sha512|b3sum|\.(b3|bk3|csv|sfv|par|p[0-9][0-9]|par2|torrent|crc)(\.gz)?$', re.IGNORECASE)
+atrem = re.compile(r'md5|sha1|sha224|sha256|sha384|sha512|b2sum|b3sum|\.(b2|blk|b3|bk3|csv|sfv|par|p[0-9][0-9]|par2|torrent|crc)(\.gz)?$', re.IGNORECASE)
 
 
 def autotest(typename):
@@ -2071,8 +2116,8 @@ def printusage(err=0):
     phelp(' --progress=VAL  show progress meter (yes, no, or auto(default))')
     phelp(' --help/-h show help')
     phelp(' --version show cfv and module versions')
-    phelp('creation options (b3):')
-    phelp(' --length=BITS        digest length in bits (default: 256 for b3)')
+    phelp('creation options (b2, b3):')
+    phelp(' --length=BITS        digest length in bits (default: 512 for b2, 256 for b3)')
     phelp('creation options (torrent):')
     phelp(' --announceurl=URL    tracker announce url')
     phelp(' --piece_size_pow2=N  power of two to set the piece size to (default 18)')

lib/cfv/hash.py

@@ -126,3 +126,9 @@ def getfileblake3(filename, callback, digest_size):
         callback,
         finalize=lambda h: h.digest(digest_size),
     )
+
+
+def getfileblake2b(filename, callback, digest_size):
+    def hasher(data=b''):
+        return hashlib.blake2b(data, digest_size=digest_size)
+    return _getfilechecksum(filename, hasher, callback)