Productionize manifest generator.

* Better output.
* Better error checking and reporting.
* Clearer names.
* Python style.

bin/python/ripple/util/Base58.py

@@ -0,0 +1,94 @@
+#!/usr/bin/env python
+
+from hashlib import sha256
+
+#
+# Human strings are base-58 with a
+# version prefix and a checksum suffix.
+#
+# Copied from ripple/protocol/RippleAddress.h
+#
+
+VER_NONE              = 1
+VER_NODE_PUBLIC       = 28
+VER_NODE_PRIVATE      = 32
+VER_ACCOUNT_ID        = 0
+VER_ACCOUNT_PUBLIC    = 35
+VER_ACCOUNT_PRIVATE   = 34
+VER_FAMILY_GENERATOR  = 41
+VER_FAMILY_SEED       = 33
+
+ALPHABET = 'rpshnaf39wBUDNEGHJKLM4PQRST7VWXYZ2bcdeCg65jkm8oFqi1tuvAxyz'
+
+VERSION_NAME = {
+    VER_NONE: 'VER_NONE',
+    VER_NODE_PUBLIC: 'VER_NODE_PUBLIC',
+    VER_NODE_PRIVATE: 'VER_NODE_PRIVATE',
+    VER_ACCOUNT_ID: 'VER_ACCOUNT_ID',
+    VER_ACCOUNT_PUBLIC: 'VER_ACCOUNT_PUBLIC',
+    VER_ACCOUNT_PRIVATE: 'VER_ACCOUNT_PRIVATE',
+    VER_FAMILY_GENERATOR: 'VER_FAMILY_GENERATOR',
+    VER_FAMILY_SEED: 'VER_FAMILY_SEED'
+}
+
+class Alphabet(object):
+    def __init__(self, radix, digit_to_char, char_to_digit):
+        self.radix = radix
+        self.digit_to_char = digit_to_char
+        self.char_to_digit = char_to_digit
+
+    def transcode_from(self, s, source_alphabet):
+        n, zero_count = source_alphabet._digits_to_number(s)
+        digits = []
+        while n > 0:
+            n, digit = divmod(n, self.radix)
+            digits.append(self.digit_to_char(digit))
+
+        s = ''.join(digits)
+        return self.digit_to_char(0) * zero_count + s[::-1]
+
+    def _digits_to_number(self, digits):
+        stripped = digits.lstrip(self.digit_to_char(0))
+        n = 0
+        for d in stripped:
+            n *= self.radix
+            n += self.char_to_digit(d)
+        return n, len(digits) - len(stripped)
+
+_INVERSE_INDEX = dict((c, i) for (i, c) in enumerate(ALPHABET))
+
+# In base 58 encoding, the digits come from the ALPHABET string.
+BASE58 = Alphabet(len(ALPHABET), ALPHABET.__getitem__, _INVERSE_INDEX.get)
+
+# In base 256 encoding, each digit is just a character between 0 and 255.
+BASE256 = Alphabet(256, chr, ord)
+
+def encode(b):
+    return BASE58.transcode_from(b, BASE256)
+
+def decode(b):
+    return BASE256.transcode_from(b, BASE58)
+
+def checksum(b):
+    """Returns a 4-byte checksum of a binary."""
+    return sha256(sha256(b).digest()).digest()[:4]
+
+def encode_version(ver, b):
+    """Encodes a version encoding and a binary as human string."""
+    b = chr(ver) + b
+    return encode(b + checksum(b))
+
+def decode_version(s):
+    """Decodes a human base-58 string into its version encoding and binary."""
+    b = decode(s)
+    body, check = b[:-4], b[-4:]
+    assert check == checksum(body), ('Bad checksum for', s)
+    return ord(body[0]), body[1:]
+
+def version_name(ver):
+    return VERSION_NAME.get(ver) or ('(unknown version %s)' % ver)
+
+def check_version(version, expected):
+    if version != expected:
+        raise ValueError('Expected version %s but got %s' % (
+            version_name(version), version_name(expected)))