Make get_mozilla_ciphers script a little more regexy and readable

1 files changed, 62 insertions, 61 deletions

diff --git a/src/common/get_mozilla_ciphers.py b/src/common/get_mozilla_ciphers.py
index 59f8b5989..378ce01e8 100644
--- a/src/common/get_mozilla_ciphers.py
+++ b/src/common/get_mozilla_ciphers.py
@@ -1,54 +1,64 @@
 #!/usr/bin/python
+# coding=utf-8
 # Copyright 2011, The Tor Project, Inc
-# original version by Arturo Filastò
+# original version by Arturo Filastò
 
 # This script parses Firefox and OpenSSL sources, and uses this information
 # to generate a ciphers.inc file.
 
-# Read the cpp file to understand what Ciphers map to what name
+import re
+
+#####
+# Read the cpp file to understand what Ciphers map to what name :
+# Make "ciphers" a map from name used in the javascript to a cipher macro name
 fileA = open('security/manager/ssl/src/nsNSSComponent.cpp','r')
 
-start = None
-lines = fileA.readlines()
-for i, line in enumerate(lines):
-    if line.strip().startswith('static CipherPref CipherPrefs'):
+# The input format is a file containing exactly one section of the form:
+# static CipherPref CipherPrefs[] = {
+#  {"name", MACRO_NAME}, // comment
+#  ...
+#  {NULL, 0}
+# }
+
+inCipherSection = False
+cipherLines = []
+for line in fileA:
+    if line.startswith('static CipherPref CipherPrefs'):
         # Get the starting boundary of the Cipher Preferences
-        start = i
-
-    if start and line.strip().startswith('{NULL, 0}'):
-        # Get the ending boundary of the Cipher Prefs
-        end = i
-        break
+        inCipherSection = True
+    elif inCipherSection:
+        line = line.strip()
+        if line.startswith('{NULL, 0}'):
+            # At the ending boundary of the Cipher Prefs
+            break
+        else:
+            cipherLines.append(line)
 fileA.close()
 
 # Parse the lines and put them into a dict
 ciphers = {}
-for x in lines[start:end]:
-    line = x.strip()
-    if line.startswith('{'):
-        for i, y in enumerate(line):
-            if y == '}':
-                parsed = line[1:i]
-        key, value = parsed.split(',')
-        ciphers[key.replace("\"","")] = value.strip()
-
-# Read the JS file to understand what ciphers are enabled
+for line in cipherLines:
+    m = re.search(r'^{\s*\"([^\"]+)\",\s*(\S*)\s*}', line)
+    if m:
+        key,value = m.groups()
+        ciphers[key] = value
+
+#####
+# Read the JS file to understand what ciphers are enabled.  The format is
+#  pref("name", true/false);
+# Build a map enabled_ciphers from javascript name to "true" or "false",
+# and an (unordered!) list of the macro names for those ciphers that are
+# enabled.
 fileB = open('netwerk/base/public/security-prefs.js', 'r')
 
 enabled_ciphers = {}
-for x in fileB.readlines():
-    start = None
-    line = x.strip()
-    for i, y in enumerate(line):
-        if y == "(":
-            start = i
-        if start and y == ")":
-            end = i
-    if start:
-        inner = line[start+1:end]
-        key, value = inner.replace("\"","").split(",")
-        if key.startswith('security.ssl3'):
-            enabled_ciphers[key.strip()] = value.strip()
+for line in fileB:
+    m = re.match(r'pref\(\"([^\"]+)\"\s*,\s*(\S*)\s*\)', line)
+    if not m:
+        continue
+    key, val = m.groups()
+    if key.startswith("security.ssl3"):
+        enabled_ciphers[key] = val
 fileB.close()
 
 used_ciphers = []
@@ -56,27 +66,21 @@ for k, v in enabled_ciphers.items():
     if v == "true":
         used_ciphers.append(ciphers[k])
 
-
 oSSLinclude = ('/usr/include/openssl/ssl3.h', '/usr/include/openssl/ssl.h',
                '/usr/include/openssl/ssl2.h', '/usr/include/openssl/ssl23.h',
                '/usr/include/openssl/tls1.h')
 
+#####
+# This reads the hex code for the ciphers that are used by firefox.
+# sslProtoD is set to a map from macro name to macro value in sslproto.h;
+# cipher_codes is set to an (unordered!) list of these hex values.
 sslProto = open('security/nss/lib/ssl/sslproto.h', 'r')
 sslProtoD = {}
 
-# This reads the HEX code for the ciphers that are used
-# by firefox.
-for x in sslProto.readlines():
-    line = x.strip()
-    if line.lower().startswith("#define"):
-        # If I ever see somebody putting mixed tab
-        # and spaces I will cut their fingers :)
-        line = line.replace('\t', ' ')
-        key = line.split(' ')[1]
-        value = line.split(' ')[-1]
-        key = key.strip()
-        value = value.strip()
-        print "%s %s\n\n" % (key, value)
+for line in sslProto:
+    m = re.match('#define\s+(\S+)\s+(\S+)', line)
+    if m:
+        key, value = m.groups()
         sslProtoD[key] = value
 sslProto.close()
 
@@ -84,23 +88,23 @@ cipher_codes = []
 for x in used_ciphers:
     cipher_codes.append(sslProtoD[x].lower())
 
+####
+# Now read through all the openssl include files, and try to find the openssl
+# macro names for those files.
 cipher_hex = {}
 for fl in oSSLinclude:
     fp = open(fl, 'r')
-    for x in fp.readlines():
-        line = x.strip()
-        if line.lower().startswith("#define"):
-            line = line.replace('\t', ' ')
-            value = line.split(' ')[1]
-            key = line.split(' ')[-1]
-            key = key.strip()
-            value = value.strip()
+    for line in fp.readlines():
+        m = re.match('#define\s+(\S+)\s+(\S+)', line)
+        if m:
+            value,key = m.groups()
             if key.startswith('0x'):
                 key = key.replace('0x0300','0x').lower()
                 #print "%s %s" % (key, value)
                 cipher_hex[key] = value
     fp.close()
 
+# Now generate the output.
 for x in cipher_codes:
     try:
         res = """#ifdef %s
@@ -109,9 +113,6 @@ for x in cipher_codes:
        XCIPHER(%s, %s)
     #endif""" % (cipher_hex[x], x, cipher_hex[x], x, cipher_hex[x])
         print res
-    except:
+    except KeyError:
         print "Not found %s" % x
 
-
-#print enabled_ciphers
-#print ciphers