Allow qstr's with non-ident chars, construct good identifier for them.
Also, add qstr's for string appearing in unix REPL loop, gross effect
being less allocations for each command run.
diff --git a/py/makeqstrdata.py b/py/makeqstrdata.py
index c4b7be4..dbafd47 100644
--- a/py/makeqstrdata.py
+++ b/py/makeqstrdata.py
@@ -1,5 +1,6 @@
import argparse
import re
+from htmlentitydefs import codepoint2name
# this must match the equivalent function in qstr.c
def compute_hash(qstr):
@@ -10,7 +11,7 @@
def do_work(infiles):
# read the qstrs in from the input files
- qstrs = []
+ qstrs = {}
for infile in infiles:
with open(infile, 'rt') as f:
line_number = 0
@@ -23,28 +24,29 @@
continue
# verify line is of the correct form
- match = re.match(r'Q\(([0-9A-Za-z_]+)\)$', line)
+ match = re.match(r'Q\((.+)\)$', line)
if not match:
print('({}:{}) bad qstr format, got {}'.format(infile, line_number, line))
return False
# get the qstr value
qstr = match.group(1)
+ ident = re.sub(r'[^A-Za-z0-9_]', lambda s: "_" + codepoint2name[ord(s.group(0))] + "_", qstr)
# don't add duplicates
- if qstr in qstrs:
+ if ident in qstrs:
continue
# add the qstr to the list
- qstrs.append(qstr)
+ qstrs[ident] = qstr
# process the qstrs, printing out the generated C header file
print('// This file was automatically generated by makeqstrdata.py')
print('')
- for qstr in qstrs:
+ for ident, qstr in qstrs.items():
qhash = compute_hash(qstr)
qlen = len(qstr)
- print('Q({}, (const byte*)"\\x{:02x}\\x{:02x}\\x{:02x}\\x{:02x}" "{}")'.format(qstr, qhash & 0xff, (qhash >> 8) & 0xff, qlen & 0xff, (qlen >> 8) & 0xff, qstr))
+ print('Q({}, (const byte*)"\\x{:02x}\\x{:02x}\\x{:02x}\\x{:02x}" "{}")'.format(ident, qhash & 0xff, (qhash >> 8) & 0xff, qlen & 0xff, (qlen >> 8) & 0xff, qstr))
return True