summaryrefslogtreecommitdiffstatshomepage
path: root/py
diff options
context:
space:
mode:
authorDamien George <damien.p.george@gmail.com>2016-09-02 14:32:47 +1000
committerDamien George <damien.p.george@gmail.com>2016-09-02 14:32:47 +1000
commitf127bef3e41f25eea6da73a52aab2fdc53be2464 (patch)
treee71c749ee10e9bfd1f52f8e8b1c3cca4e2ee6fd6 /py
parentf98bb2ddcbfed7033ae93ae8fe98b0540a9fb42f (diff)
downloadmicropython-f127bef3e41f25eea6da73a52aab2fdc53be2464.tar.gz
micropython-f127bef3e41f25eea6da73a52aab2fdc53be2464.zip
py/makeqstrdata.py: Compute the qstr hash from bytes, not characters.
Diffstat (limited to 'py')
-rw-r--r--py/makeqstrdata.py19
1 files changed, 10 insertions, 9 deletions
diff --git a/py/makeqstrdata.py b/py/makeqstrdata.py
index 8a3136b1f1..7249769f47 100644
--- a/py/makeqstrdata.py
+++ b/py/makeqstrdata.py
@@ -14,11 +14,13 @@ import sys
# - codepoint2name lives in a different module
import platform
if platform.python_version_tuple()[0] == '2':
- ord_bytes = ord
+ bytes_cons = lambda val, enc=None: bytearray(val)
from htmlentitydefs import codepoint2name
elif platform.python_version_tuple()[0] == '3':
- ord_bytes = lambda x:x
+ bytes_cons = bytes
from html.entities import codepoint2name
+# end compatibility code
+
codepoint2name[ord('-')] = 'hyphen';
# add some custom names to map characters that aren't in HTML
@@ -52,8 +54,8 @@ codepoint2name[ord('~')] = 'tilde'
# this must match the equivalent function in qstr.c
def compute_hash(qstr, bytes_hash):
hash = 5381
- for char in qstr:
- hash = (hash * 33) ^ ord(char)
+ for b in qstr:
+ hash = (hash * 33) ^ b
# Make sure that valid hash is never zero, zero means "hash not computed"
return (hash & ((1 << (8 * bytes_hash)) - 1)) or 1
@@ -115,16 +117,15 @@ def parse_input_headers(infiles):
return qcfgs, qstrs
def make_bytes(cfg_bytes_len, cfg_bytes_hash, qstr):
- qhash = compute_hash(qstr, cfg_bytes_hash)
+ qbytes = bytes_cons(qstr, 'utf8')
+ qlen = len(qbytes)
+ qhash = compute_hash(qbytes, cfg_bytes_hash)
if all(32 <= ord(c) <= 126 and c != '\\' and c != '"' for c in qstr):
# qstr is all printable ASCII so render it as-is (for easier debugging)
- qlen = len(qstr)
qdata = qstr
else:
# qstr contains non-printable codes so render entire thing as hex pairs
- qbytes = qstr.encode('utf8')
- qlen = len(qbytes)
- qdata = ''.join(('\\x%02x' % ord_bytes(b)) for b in qbytes)
+ qdata = ''.join(('\\x%02x' % b) for b in qbytes)
if qlen >= (1 << (8 * cfg_bytes_len)):
print('qstr is too long:', qstr)
assert False