| 1 |
2 |
sinclairrf |
################################################################################
|
| 2 |
|
|
#
|
| 3 |
|
|
# Copyright 2012, Sinclair R.F., Inc.
|
| 4 |
|
|
#
|
| 5 |
|
|
# Collection of utilities for the assembler.
|
| 6 |
|
|
#
|
| 7 |
|
|
################################################################################
|
| 8 |
|
|
|
| 9 |
|
|
import os
|
| 10 |
|
|
import re
|
| 11 |
|
|
|
| 12 |
|
|
class AsmException(Exception):
|
| 13 |
|
|
"""
|
| 14 |
|
|
Exception class for the assembler.\n
|
| 15 |
|
|
This allows the top-level module to capture error messages other than internal
|
| 16 |
|
|
errors and program bugs so that users see a single-line error relevant to
|
| 17 |
|
|
their code rather than the usual Python mess.
|
| 18 |
|
|
"""
|
| 19 |
|
|
def __init__(self,message):
|
| 20 |
|
|
self.msg = message;
|
| 21 |
|
|
def __str__(self):
|
| 22 |
|
|
return self.msg;
|
| 23 |
|
|
|
| 24 |
|
|
class FileBodyIterator:
|
| 25 |
|
|
"""
|
| 26 |
|
|
Iterator for files that returns bodies of lines of the file.\n
|
| 27 |
|
|
The directive must be the first non-white spaces on a line.\n
|
| 28 |
|
|
The iterator outputs a list whos first element is the line number for the
|
| 29 |
|
|
first line of the block and whose subsequent elements are the lines with the
|
| 30 |
|
|
content of the block.\n
|
| 31 |
|
|
The iterator handles the ".include" directive.
|
| 32 |
|
|
"""
|
| 33 |
|
|
|
| 34 |
|
|
def __init__(self, fps, ad):
|
| 35 |
|
|
"""
|
| 36 |
|
|
Initialize the iterator.\n
|
| 37 |
|
|
fps list of file pointers from the argument line
|
| 38 |
|
|
ad asmDef_9x8 object (required to identify the directives)
|
| 39 |
|
|
"""
|
| 40 |
|
|
# Do sanity check on arguments.
|
| 41 |
|
|
if ad.IsDirective(".include"):
|
| 42 |
|
|
raise Exception('Program Bug: The ".include" directive is defined by FileBodyIterator');
|
| 43 |
|
|
# Initialize the raw processing states
|
| 44 |
|
|
self.ixConstants = 0;
|
| 45 |
|
|
self.fpPending = list(fps);
|
| 46 |
|
|
self.ad = ad;
|
| 47 |
|
|
self.current = list();
|
| 48 |
|
|
self.pending = list();
|
| 49 |
|
|
# Initialize the include search paths
|
| 50 |
|
|
self.searchPaths = list();
|
| 51 |
|
|
self.searchPaths.append('.');
|
| 52 |
|
|
# Prepare the file parsing
|
| 53 |
|
|
self.included = list();
|
| 54 |
|
|
for fp in self.fpPending:
|
| 55 |
|
|
if fp.name in self.included:
|
| 56 |
|
|
raise AsmException('Input file %s listed more than once' % fp.name);
|
| 57 |
|
|
self.included.append(fp.name);
|
| 58 |
|
|
self.fpStack = list();
|
| 59 |
|
|
self.fpStack.append(dict(fp=self.fpPending.pop(0), line=0));
|
| 60 |
|
|
self.pendingInclude = None;
|
| 61 |
|
|
|
| 62 |
|
|
def __iter__(self):
|
| 63 |
|
|
"""
|
| 64 |
|
|
Required function for an iterable object.
|
| 65 |
|
|
"""
|
| 66 |
|
|
return self;
|
| 67 |
|
|
|
| 68 |
|
|
def next(self):
|
| 69 |
|
|
"""
|
| 70 |
|
|
Return the next directive body from the iterator.\n
|
| 71 |
|
|
The body is a list with the following content:
|
| 72 |
|
|
the name of the file
|
| 73 |
|
|
the line number for the first line of the body
|
| 74 |
|
|
the body consisting of lines from the source file\n
|
| 75 |
|
|
The body contains comment lines preceding the directive, the line with the
|
| 76 |
|
|
directive, and optional lines following the directive up to the optional
|
| 77 |
|
|
comments preceding the next directive.
|
| 78 |
|
|
"""
|
| 79 |
|
|
# Discard the body emitted by the previous call.
|
| 80 |
|
|
self.current = self.pending;
|
| 81 |
|
|
self.pending = list();
|
| 82 |
|
|
# If the current body is an include directive, then process it immediately.
|
| 83 |
|
|
if self.current and re.match(r'\s*\.include\b',self.current[-1]):
|
| 84 |
|
|
return self.current;
|
| 85 |
|
|
# Loop until all of the files have been processed
|
| 86 |
|
|
while self.fpStack or self.fpPending or self.pendingInclude:
|
| 87 |
|
|
# Indicate when a new file is started.
|
| 88 |
|
|
if 'started' not in self.fpStack[-1]:
|
| 89 |
|
|
if not self.current:
|
| 90 |
|
|
self.fpStack[-1]['started'] = True;
|
| 91 |
|
|
self.current.append(self.fpStack[-1]['fp'].name);
|
| 92 |
|
|
self.current.append(0);
|
| 93 |
|
|
return self.current;
|
| 94 |
|
|
# Ensure the bodies in closed files are all emitted before continuing to
|
| 95 |
|
|
# the next/enclosing file.
|
| 96 |
|
|
if 'closed' in self.fpStack[-1]:
|
| 97 |
|
|
# Provide end-of-file indication if there is not a pending body fragment.
|
| 98 |
|
|
if not self.current:
|
| 99 |
|
|
self.current.append(self.fpStack[-1]['fp'].name);
|
| 100 |
|
|
self.current.append(-1);
|
| 101 |
|
|
self.fpStack.pop();
|
| 102 |
|
|
return self.current;
|
| 103 |
|
|
# Handle a queued ".include" directive.
|
| 104 |
|
|
if self.pendingInclude:
|
| 105 |
|
|
# Don't open the include file until all previous content has been emitted.
|
| 106 |
|
|
if self.current:
|
| 107 |
|
|
return self.current;
|
| 108 |
|
|
self.included.append(self.pendingInclude);
|
| 109 |
|
|
fp_pending = None;
|
| 110 |
|
|
for path in self.searchPaths:
|
| 111 |
|
|
fullInclude = os.path.join(path,self.pendingInclude);
|
| 112 |
|
|
if os.path.exists(fullInclude):
|
| 113 |
|
|
fp_pending = open('%s/%s' % (path,self.pendingInclude),'r');
|
| 114 |
|
|
break;
|
| 115 |
|
|
else:
|
| 116 |
|
|
raise AsmException('%s not found' % self.pendingInclude);
|
| 117 |
|
|
self.fpStack.append(dict(fp=fp_pending, line=0));
|
| 118 |
|
|
self.pendingInclude = None;
|
| 119 |
|
|
# Provide start-of-file indication.
|
| 120 |
|
|
self.fpStack[-1]['started'] = True;
|
| 121 |
|
|
self.current.append(fp_pending.name);
|
| 122 |
|
|
self.current.append(0);
|
| 123 |
|
|
return self.current;
|
| 124 |
|
|
# Get the next file to process if fpStack is empty.
|
| 125 |
|
|
if not self.fpStack:
|
| 126 |
|
|
self.fpStack.append(dict(fp=self.fpPending.pop(0), line=0));
|
| 127 |
|
|
# Process/continue processing the top file.
|
| 128 |
|
|
fp = self.fpStack[-1];
|
| 129 |
|
|
for line in fp['fp']:
|
| 130 |
|
|
fp['line'] += 1;
|
| 131 |
|
|
# Handle single-line directives.
|
| 132 |
|
|
if re.match(r'\s*\.(IFDEF|IFNDEF|ELSE|ENDIF|include)\b',line):
|
| 133 |
|
|
if not self.pending:
|
| 134 |
|
|
self.pending.append(fp['fp'].name);
|
| 135 |
|
|
self.pending.append(fp['line']);
|
| 136 |
|
|
self.pending.append(line);
|
| 137 |
|
|
if not self.current:
|
| 138 |
|
|
self.current = self.pending;
|
| 139 |
|
|
self.pending = list();
|
| 140 |
|
|
return self.current;
|
| 141 |
|
|
# Append empty and comment lines to the pending block.
|
| 142 |
|
|
if re.match(r'\s*(;|$)', line):
|
| 143 |
|
|
if not self.pending:
|
| 144 |
|
|
self.pending.append(fp['fp'].name);
|
| 145 |
|
|
self.pending.append(fp['line']);
|
| 146 |
|
|
self.pending.append(line);
|
| 147 |
|
|
continue;
|
| 148 |
|
|
# See if the line starts with a directive.
|
| 149 |
|
|
tokens = re.findall(r'\s*(\S+)',line);
|
| 150 |
|
|
if self.ad.IsDirective(tokens[0]):
|
| 151 |
|
|
if not self.pending:
|
| 152 |
|
|
self.pending.append(fp['fp'].name);
|
| 153 |
|
|
self.pending.append(fp['line']);
|
| 154 |
|
|
self.pending.append(line);
|
| 155 |
|
|
if self.current:
|
| 156 |
|
|
return self.current;
|
| 157 |
|
|
self.current = self.pending;
|
| 158 |
|
|
self.pending = list();
|
| 159 |
|
|
continue;
|
| 160 |
|
|
# Otherwise, this line belongs to the body of the preceding directive.
|
| 161 |
5 |
sinclairrf |
if not self.pending:
|
| 162 |
|
|
self.pending.append(fp['fp'].name);
|
| 163 |
|
|
self.pending.append(fp['line']);
|
| 164 |
2 |
sinclairrf |
if not self.current:
|
| 165 |
|
|
self.current += self.pending[0:2];
|
| 166 |
|
|
self.current += self.pending[2:];
|
| 167 |
|
|
self.current.append(line);
|
| 168 |
|
|
self.pending = list();
|
| 169 |
|
|
# Past the last line of the current file -- close it.
|
| 170 |
|
|
self.fpStack[-1]['fp'].close();
|
| 171 |
|
|
self.fpStack[-1]['closed'] = True;
|
| 172 |
|
|
# Prepare to emit pending bodies if any.
|
| 173 |
|
|
if not self.current:
|
| 174 |
|
|
self.current = self.pending;
|
| 175 |
|
|
self.pending = list();
|
| 176 |
|
|
raise StopIteration;
|
| 177 |
|
|
|
| 178 |
|
|
def AddSearchPath(self,path):
|
| 179 |
|
|
"""
|
| 180 |
|
|
Use by the top level assembler to add search paths for opening included
|
| 181 |
|
|
files.
|
| 182 |
|
|
"""
|
| 183 |
|
|
self.searchPaths.append(path);
|
| 184 |
|
|
|
| 185 |
|
|
def Include(self,filename):
|
| 186 |
|
|
self.pendingInclude = filename;
|
| 187 |
|
|
|
| 188 |
|
|
################################################################################
|
| 189 |
|
|
#
|
| 190 |
|
|
# Parse strings into the desired types.
|
| 191 |
|
|
#
|
| 192 |
|
|
################################################################################
|
| 193 |
|
|
|
| 194 |
|
|
def ParseNumber(inString):
|
| 195 |
|
|
"""
|
| 196 |
|
|
Test for recognized integer values and return the value if recognized,
|
| 197 |
|
|
otherwise return None.
|
| 198 |
|
|
"""
|
| 199 |
|
|
# look for single-digit 0
|
| 200 |
|
|
if inString == '0':
|
| 201 |
|
|
return 0;
|
| 202 |
5 |
sinclairrf |
# look for a binary value
|
| 203 |
|
|
a = re.match(r'0b[01_]+$',inString);
|
| 204 |
|
|
if a:
|
| 205 |
|
|
b = re.sub(r'_','',a.group(0)[2:]);
|
| 206 |
|
|
return int(b,2);
|
| 207 |
|
|
# look for an octal value
|
| 208 |
|
|
a = re.match(r'0[0-7_]+$',inString);
|
| 209 |
|
|
if a:
|
| 210 |
|
|
return int(a.group(0)[1:],8);
|
| 211 |
2 |
sinclairrf |
# look for decimal value
|
| 212 |
5 |
sinclairrf |
a = re.match(r'[+\-]?[1-9_]\d*$',inString);
|
| 213 |
2 |
sinclairrf |
if a:
|
| 214 |
|
|
return int(a.group(0),10);
|
| 215 |
|
|
# look for a hex value
|
| 216 |
5 |
sinclairrf |
a = re.match(r'0x[0-9A-Fa-f_]+$',inString);
|
| 217 |
2 |
sinclairrf |
if a:
|
| 218 |
5 |
sinclairrf |
return int(a.group(0)[2:],16);
|
| 219 |
2 |
sinclairrf |
# Everything else is an error
|
| 220 |
|
|
return None;
|
| 221 |
|
|
|
| 222 |
|
|
def ParseChar(inchar):
|
| 223 |
|
|
"""
|
| 224 |
|
|
Parse single characters including escaped characters. Return the character
|
| 225 |
|
|
value and the number of characters in the input string matched.
|
| 226 |
|
|
"""
|
| 227 |
|
|
if re.match(r'\\[0-7]{3}',inchar):
|
| 228 |
|
|
return (int(inchar[1:4],8),4,);
|
| 229 |
|
|
elif re.match(r'\\[0-7]{2}',inchar):
|
| 230 |
|
|
return (int(inchar[1:3],8),3,);
|
| 231 |
|
|
elif re.match(r'\\[0-7]{1}',inchar):
|
| 232 |
|
|
return (int(inchar[1],8),2,);
|
| 233 |
|
|
elif re.match(r'\\[xX][0-9A-Fa-f]{2}',inchar):
|
| 234 |
|
|
return (int(inchar[2:4],16),4,);
|
| 235 |
|
|
elif re.match(r'\\[xX][0-9A-Fa-f]{1}',inchar):
|
| 236 |
|
|
return (int(inchar[2],16),3,);
|
| 237 |
|
|
elif re.match(r'\\.',inchar):
|
| 238 |
|
|
if inchar[1] == 'a': # bell ==> control-G
|
| 239 |
|
|
return (7,2,);
|
| 240 |
|
|
elif inchar[1] == 'b': # backspace ==> control-H
|
| 241 |
|
|
return (8,2,);
|
| 242 |
|
|
elif inchar[1] == 'f': # form feed ==> control-L
|
| 243 |
|
|
return (12,2,);
|
| 244 |
|
|
elif inchar[1] == 'n': # line feed ==> control-J
|
| 245 |
|
|
return (10,2,);
|
| 246 |
|
|
elif inchar[1] == 'r': # carriage return ==> control-M
|
| 247 |
|
|
return (13,2,);
|
| 248 |
|
|
elif inchar[1] == 't': # horizontal tab ==> control-I
|
| 249 |
|
|
return (9,2,);
|
| 250 |
|
|
else: # unrecognized escaped character ==> return that character
|
| 251 |
|
|
return (ord(inchar[1]),2,);
|
| 252 |
|
|
else:
|
| 253 |
|
|
return (ord(inchar[0]),1,);
|
| 254 |
|
|
|
| 255 |
|
|
def ParseString(inString):
|
| 256 |
|
|
"""
|
| 257 |
|
|
Parse strings recognized by the assembler.\n
|
| 258 |
|
|
A string consists of the following:
|
| 259 |
|
|
an optional count/termination character -- one of CNc
|
| 260 |
|
|
a starting double-quote character
|
| 261 |
|
|
the body of the string including escape sequences
|
| 262 |
|
|
a terminating double-quote character
|
| 263 |
|
|
Errors are indicated by returning the location (an integer) within the string
|
| 264 |
|
|
where the error occurs.
|
| 265 |
|
|
"""
|
| 266 |
|
|
# Detect optional count/termination character.
|
| 267 |
|
|
ix = 1 if inString[0] in 'CNc' else 0;
|
| 268 |
|
|
# Ensure the required start double quote is preset.
|
| 269 |
|
|
if inString[ix] != '"' or inString[-1] != '"':
|
| 270 |
|
|
raise Exception('Program Bug -- missing one or more double quotes around string');
|
| 271 |
|
|
ix = ix + 1;
|
| 272 |
|
|
# Convert the characters and escape sequences in the string to a list of their
|
| 273 |
|
|
# integer values.
|
| 274 |
|
|
outString = list();
|
| 275 |
|
|
while ix < len(inString)-1:
|
| 276 |
|
|
(thisChar,thisLen,) = ParseChar(inString[ix:-1]);
|
| 277 |
|
|
outString.append(thisChar);
|
| 278 |
|
|
ix += thisLen;
|
| 279 |
|
|
# Insert the optional character count or append the optional nul terminating
|
| 280 |
|
|
# character.
|
| 281 |
|
|
if inString[0] == 'C':
|
| 282 |
|
|
outString.insert(0,len(outString));
|
| 283 |
|
|
elif inString[0] == 'N':
|
| 284 |
|
|
outString.append(0);
|
| 285 |
|
|
elif inString[0] == 'c':
|
| 286 |
|
|
outString.insert(0,len(outString)-1);
|
| 287 |
|
|
# That's all.
|
| 288 |
|
|
return outString;
|
| 289 |
|
|
|
| 290 |
|
|
def ParseToken(ad,fl_loc,col,raw,allowed):
|
| 291 |
|
|
"""
|
| 292 |
|
|
Examine the raw tokens and convert them into dictionary objects consisting of
|
| 293 |
|
|
the following:
|
| 294 |
|
|
type the type of token
|
| 295 |
|
|
value the value of the token
|
| 296 |
|
|
this can be the name of a symbol, a numeric value, a string body, ...
|
| 297 |
|
|
loc start location of the token
|
| 298 |
|
|
this is is required by subsequent stages of the assembler for
|
| 299 |
|
|
error messages
|
| 300 |
|
|
argument optional entry required for macros arguments
|
| 301 |
|
|
range optional entry required when a range is provided for a parameter\n
|
| 302 |
|
|
The token type is compared against the allowed tokens.\n
|
| 303 |
|
|
Detect syntax errors and display error messages consisting of the error and
|
| 304 |
|
|
the location within the file where the error occurs.
|
| 305 |
|
|
"""
|
| 306 |
|
|
flc_loc = fl_loc + ':' + str(col+1);
|
| 307 |
|
|
# look for instructions
|
| 308 |
|
|
# Note: Do this before anything else because instructions can be a
|
| 309 |
|
|
# strange mix of symbols.
|
| 310 |
|
|
if ad.IsInstruction(raw):
|
| 311 |
|
|
if 'instruction' not in allowed:
|
| 312 |
|
|
raise AsmException('instruction "%s" not allowed at %s' % (raw,flc_loc));
|
| 313 |
|
|
return dict(type='instruction', value=raw, loc=flc_loc);
|
| 314 |
|
|
# look for computation
|
| 315 |
|
|
a = re.match(r'\${\S+}$',raw);
|
| 316 |
|
|
if a:
|
| 317 |
|
|
if 'singlevalue' not in allowed:
|
| 318 |
|
|
raise AsmException('Computed value not allowed at %s' % flc_loc);
|
| 319 |
|
|
try:
|
| 320 |
|
|
tParseNumber = eval(raw[2:-1],ad.SymbolDict());
|
| 321 |
|
|
except:
|
| 322 |
|
|
raise AsmException('Malformed computed value at %s: "%s"' % (flc_loc,raw,));
|
| 323 |
|
|
if type(tParseNumber) != int:
|
| 324 |
|
|
raise AsmException('Malformed single-byte value at %s' % flc_loc);
|
| 325 |
|
|
return dict(type='value', value=tParseNumber, loc=flc_loc);
|
| 326 |
|
|
# look for a repeated single-byte numeric value (N*M where M is the repeat count)
|
| 327 |
5 |
sinclairrf |
matchString=r'(0|0b[01_]+|0[0-7]+|[+\-]?[1-9]\d*|0x[0-9A-Fa-f]{1,2})\*([1-9]\d*|C_\w+|\$\{\S+\})$';
|
| 328 |
2 |
sinclairrf |
a = re.match(matchString,raw);
|
| 329 |
|
|
if a:
|
| 330 |
|
|
if 'multivalue' not in allowed:
|
| 331 |
|
|
raise AsmException('Multi-byte value not allowed at %s' % flc_loc);
|
| 332 |
|
|
b = re.findall(matchString,a.group(0));
|
| 333 |
|
|
if not b:
|
| 334 |
|
|
raise Exception('Program Bug -- findall failed after match worked');
|
| 335 |
|
|
b = b[0];
|
| 336 |
|
|
try:
|
| 337 |
|
|
tParseNumber = ParseNumber(b[0]);
|
| 338 |
|
|
except:
|
| 339 |
|
|
raise AsmException('Malformed multi-byte value at %s' % (fl_loc + ':' + str(col+1)));
|
| 340 |
|
|
tValue = list();
|
| 341 |
|
|
fl_loc2 = fl_loc+':'+str(col+1+len(b[0])+1);
|
| 342 |
|
|
if re.match(r'[1-9]',b[1]):
|
| 343 |
|
|
repeatCount = int(b[1]);
|
| 344 |
|
|
elif re.match(r'C_',b[1]):
|
| 345 |
|
|
if not ad.IsConstant(b[1]):
|
| 346 |
|
|
raise AsmException('Unrecognized symbol "%s" at %s' % (b[1],fl_loc2,));
|
| 347 |
|
|
ix = ad.symbols['list'].index(b[1]);
|
| 348 |
|
|
if len(ad.symbols['body'][ix]) != 1:
|
| 349 |
|
|
raise asmDef.AsmException('constant can only be one byte at %s' % fl_loc2);
|
| 350 |
|
|
repeatCount = ad.symbols['body'][ix][0];
|
| 351 |
|
|
elif re.match(r'\$',b[1]):
|
| 352 |
|
|
repeatCount = eval(b[1][2:-1],ad.SymbolDict());
|
| 353 |
|
|
else:
|
| 354 |
|
|
raise Exception('Program Bug -- unrecognized repeat count');
|
| 355 |
|
|
if repeatCount <= 0:
|
| 356 |
|
|
raise AsmException('Repeat count must be positive at %s' % fl_loc2);
|
| 357 |
|
|
for ix in range(repeatCount):
|
| 358 |
|
|
tValue.append(tParseNumber);
|
| 359 |
|
|
return dict(type='value', value=tValue, loc=flc_loc);
|
| 360 |
|
|
# look for a single-byte numeric value
|
| 361 |
5 |
sinclairrf |
a = re.match(r'(0|0b[01_]+|0[0-7]+|[+\-]?[1-9]\d*|0x[0-9A-Fa-f]+)$',raw);
|
| 362 |
2 |
sinclairrf |
if a:
|
| 363 |
|
|
if 'singlevalue' not in allowed:
|
| 364 |
|
|
raise AsmException('Value not allowed at %s' % flc_loc);
|
| 365 |
|
|
try:
|
| 366 |
|
|
tParseNumber = ParseNumber(raw);
|
| 367 |
|
|
except:
|
| 368 |
|
|
raise AsmException('Malformed single-byte value at %s' % flc_loc);
|
| 369 |
|
|
return dict(type='value', value=tParseNumber, loc=flc_loc);
|
| 370 |
|
|
# capture double-quoted strings
|
| 371 |
|
|
if re.match(r'[CNc]?"',raw):
|
| 372 |
|
|
if 'string' not in allowed:
|
| 373 |
|
|
raise AsmException('String not allowed at %s' % flc_loc);
|
| 374 |
|
|
parsedString = ParseString(raw);
|
| 375 |
|
|
if type(parsedString) == int:
|
| 376 |
|
|
raise AsmException('Malformed string at %s' % (fl_loc + ':' + str(col+parsedString)));
|
| 377 |
|
|
return dict(type='value', value=parsedString, loc=flc_loc);
|
| 378 |
|
|
# capture single-quoted character
|
| 379 |
|
|
if raw[0] == "'":
|
| 380 |
|
|
if 'singlevalue' not in allowed:
|
| 381 |
|
|
raise AsmException('Character not allowed at %s' % flc_loc);
|
| 382 |
|
|
(thisChar,thisLen,) = ParseChar(raw[1:-1]);
|
| 383 |
|
|
if len(raw) != thisLen+2:
|
| 384 |
|
|
raise AsmException('Malformed \'.\' in %s' % flc_loc);
|
| 385 |
|
|
return dict(type='value', value=thisChar, loc=flc_loc);
|
| 386 |
|
|
# look for directives
|
| 387 |
|
|
if ad.IsDirective(raw):
|
| 388 |
|
|
if 'directive' not in allowed:
|
| 389 |
|
|
raise AsmException('Directive not allowed at %s' % flc_loc);
|
| 390 |
|
|
return dict(type='directive', value=raw, loc=flc_loc);
|
| 391 |
|
|
# look for macros
|
| 392 |
6 |
sinclairrf |
# Note: Macro arguments can contain a single layer of macros.
|
| 393 |
|
|
a = re.match(r'\.[A-Za-z]',raw);
|
| 394 |
2 |
sinclairrf |
if a:
|
| 395 |
|
|
b = re.match(r'\.[^(]+',raw);
|
| 396 |
|
|
if not ad.IsMacro(b.group(0)):
|
| 397 |
6 |
sinclairrf |
raise AsmException('Unrecognized directive or macro at %s:%d' % (fl_loc,col+1,));
|
| 398 |
2 |
sinclairrf |
if ('macro' not in allowed) and not ('singlemacro' in allowed and ad.IsSingleMacro(b.group(0))):
|
| 399 |
|
|
raise AsmException('Macro "%s" not allowed at %s:%d' % (b.group(0),fl_loc,col+1,));
|
| 400 |
6 |
sinclairrf |
macroArgs = list();
|
| 401 |
|
|
if len(b.group(0)) == len(raw):
|
| 402 |
|
|
pass;
|
| 403 |
|
|
elif (raw[len(b.group(0))] != '(') or (raw[-1] != ')'):
|
| 404 |
|
|
raise AsmException('Malformed macro invokaction "%" at %s:%d' % (raw,fl_loc,col+1,));
|
| 405 |
|
|
else:
|
| 406 |
|
|
tcol = len(b.group(0))+1;
|
| 407 |
|
|
while tcol < len(raw):
|
| 408 |
|
|
c = re.match(r'[^,(]*(\([^)]*\))?',raw[tcol:-1]);
|
| 409 |
|
|
macroArgs.append(c.group(0));
|
| 410 |
|
|
tcol += len(c.group(0))+1;
|
| 411 |
2 |
sinclairrf |
nArgs = ad.MacroNumberArgs(b.group(0))
|
| 412 |
|
|
if len(macroArgs) not in nArgs:
|
| 413 |
|
|
raise AsmException('Wrong number of arguments to macro "%s" at %s:%d' % (b.group(0),fl_loc,col+1));
|
| 414 |
|
|
while len(macroArgs) < nArgs[-1]:
|
| 415 |
|
|
macroArgs.append(ad.MacroDefault(b.group(0),len(macroArgs)));
|
| 416 |
|
|
outArgs = list();
|
| 417 |
6 |
sinclairrf |
tcol = col + len(b.group(0)) + 1;
|
| 418 |
2 |
sinclairrf |
for ixArg in range(len(macroArgs)):
|
| 419 |
6 |
sinclairrf |
outArgs.append(ParseToken(ad,fl_loc,tcol,macroArgs[ixArg],ad.MacroArgTypes(b.group(0),ixArg)));
|
| 420 |
|
|
tcol += len(macroArgs[ixArg]) + 1;
|
| 421 |
2 |
sinclairrf |
return dict(type='macro', value=b.group(0), loc=fl_loc + ':' + str(col+1), argument=outArgs);
|
| 422 |
|
|
# look for a label definition
|
| 423 |
|
|
a = re.match(r':[A-Za-z]\w*$',raw);
|
| 424 |
|
|
if a:
|
| 425 |
|
|
if 'label' not in allowed:
|
| 426 |
|
|
raise AsmException('Label not allowed at %s' % flc_loc);
|
| 427 |
|
|
return dict(type='label', value=raw[1:], loc=flc_loc);
|
| 428 |
|
|
# look for parameters with range specification
|
| 429 |
|
|
a = re.match('[LG]_\w+[[]\d+\+?:\d+]$',raw);
|
| 430 |
|
|
if a:
|
| 431 |
|
|
if 'symbol' not in allowed:
|
| 432 |
|
|
raise AsmException('Symbol not allowed at %s' % flc_loc);
|
| 433 |
|
|
a = re.findall('([LG]_\w+)([[].*)',raw)[0];
|
| 434 |
|
|
return dict(type='symbol', value=a[0], loc=flc_loc, range=a[1]);
|
| 435 |
|
|
# look for symbols
|
| 436 |
|
|
# Note: This should be the last check performed as every other kind of
|
| 437 |
|
|
# token should be recognizable
|
| 438 |
|
|
a = re.match(r'[A-Za-z]\w*$',raw);
|
| 439 |
|
|
if a:
|
| 440 |
|
|
if 'symbol' not in allowed:
|
| 441 |
|
|
raise AsmException('Symbol not allowed at %s' % flc_loc);
|
| 442 |
|
|
return dict(type='symbol', value=a.group(0), loc=flc_loc);
|
| 443 |
|
|
# anything else is an error
|
| 444 |
|
|
raise AsmException('Malformed entry at %s: "%s"' % (flc_loc,raw,));
|
| 445 |
|
|
|
| 446 |
|
|
################################################################################
|
| 447 |
|
|
#
|
| 448 |
|
|
# Extract the tokens from a block of code.
|
| 449 |
|
|
#
|
| 450 |
|
|
# These blocks of code should be generated by FileBodyIterator.
|
| 451 |
|
|
#
|
| 452 |
|
|
################################################################################
|
| 453 |
|
|
|
| 454 |
|
|
def RawTokens(ad,filename,startLineNumber,lines):
|
| 455 |
|
|
"""
|
| 456 |
|
|
Extract the list of tokens from the provided list of lines.
|
| 457 |
|
|
Convert the directive body into a list of individual tokens.\n
|
| 458 |
|
|
Tokens are directive names, symbol names, values, strings, labels, etc.\n
|
| 459 |
|
|
The return is a list of the tokens in the sequence they are encountered. Each
|
| 460 |
|
|
of these tokens is a dictionary object constructed by ParseToken.
|
| 461 |
|
|
"""
|
| 462 |
|
|
allowed = [
|
| 463 |
|
|
'instruction',
|
| 464 |
|
|
'label',
|
| 465 |
|
|
'macro',
|
| 466 |
|
|
'multivalue',
|
| 467 |
|
|
'singlevalue',
|
| 468 |
|
|
'string',
|
| 469 |
|
|
'symbol'
|
| 470 |
|
|
];
|
| 471 |
|
|
ifstack = list();
|
| 472 |
|
|
tokens = list();
|
| 473 |
|
|
lineNumber = startLineNumber - 1;
|
| 474 |
|
|
for line in lines:
|
| 475 |
|
|
lineNumber = lineNumber + 1;
|
| 476 |
|
|
fl_loc = '%s:%d' % (filename,lineNumber);
|
| 477 |
|
|
col = 0;
|
| 478 |
|
|
spaceFound = True;
|
| 479 |
|
|
while col < len(line):
|
| 480 |
|
|
flc_loc = fl_loc + ':' + str(col+1);
|
| 481 |
|
|
# Identify and then ignore white-space characters.
|
| 482 |
|
|
if re.match(r'\s',line[col:]):
|
| 483 |
|
|
spaceFound = True;
|
| 484 |
|
|
col = col + 1;
|
| 485 |
|
|
continue;
|
| 486 |
|
|
# Ensure tokens start on new lines or are separated by spaces.
|
| 487 |
|
|
if not spaceFound:
|
| 488 |
|
|
raise AsmException('Missing space in %s:%d' % (fl_loc,col+1));
|
| 489 |
|
|
spaceFound = False;
|
| 490 |
|
|
# Ignore comments.
|
| 491 |
|
|
if line[col] == ';':
|
| 492 |
|
|
break;
|
| 493 |
|
|
# Catch N"" string.
|
| 494 |
|
|
if re.match(r'N""',line[col:]):
|
| 495 |
|
|
a = re.match(r'N""',line[col:]);
|
| 496 |
|
|
# Catch strings.
|
| 497 |
|
|
elif re.match(r'[CNc]?"',line[col:]):
|
| 498 |
|
|
a = re.match(r'[CNc]?"([^\\"]|\\.)+"',line[col:]);
|
| 499 |
|
|
if not a:
|
| 500 |
|
|
raise AsmException('Malformed string at %s' % flc_loc);
|
| 501 |
|
|
# Catch single-quoted characters
|
| 502 |
|
|
elif re.match(r'\'',line[col:]):
|
| 503 |
|
|
a = re.match(r'\'(.|\\.|\\[xX][0-9A-Fa-f]{1,2}|\\[0-7]{1,3})\'',line[col:]);
|
| 504 |
|
|
if not a:
|
| 505 |
|
|
raise AsmException('Malformed \'.\' at %s' % flc_loc);
|
| 506 |
|
|
else:
|
| 507 |
|
|
# Everything else is a white-space delimited token.
|
| 508 |
|
|
a = re.match(r'\S+',line[col:]);
|
| 509 |
|
|
# Get the candidate token.
|
| 510 |
|
|
candToken = a.group(0);
|
| 511 |
|
|
# Catch conditional code inclusion constructs before parsing the token
|
| 512 |
|
|
if candToken == '.else':
|
| 513 |
|
|
if not ifstack:
|
| 514 |
|
|
raise AsmException('Unmatched ".else" at %s' % flc_loc);
|
| 515 |
|
|
ifstack[-1] = not ifstack[-1];
|
| 516 |
|
|
col += 5;
|
| 517 |
|
|
continue;
|
| 518 |
|
|
if candToken == '.endif':
|
| 519 |
|
|
if not ifstack:
|
| 520 |
|
|
raise AsmException('Unmatched ".endif" at %s' % flc_loc);
|
| 521 |
|
|
ifstack.pop();
|
| 522 |
|
|
col += 6;
|
| 523 |
|
|
continue;
|
| 524 |
|
|
elif re.match(r'\.ifdef\(',candToken):
|
| 525 |
|
|
a = re.findall(r'\.ifdef\((\w+)\)$',candToken);
|
| 526 |
|
|
if not a:
|
| 527 |
|
|
raise AsmException('Malformed ".ifdef" at %s' % flc_loc);
|
| 528 |
|
|
ifstack.append(ad.IsSymbol(a[0]));
|
| 529 |
|
|
col += 8+len(a[0]);
|
| 530 |
|
|
continue;
|
| 531 |
|
|
elif re.match(r'\.ifndef\(',candToken):
|
| 532 |
|
|
a = re.findall(r'\.ifndef\((\w+)\)$',candToken);
|
| 533 |
|
|
if not a:
|
| 534 |
|
|
raise AsmException('Malformed ".ifndef" at %s' % flc_loc);
|
| 535 |
|
|
ifstack.append(not ad.IsSymbol(a[0]));
|
| 536 |
|
|
col += 9+len(a[0]);
|
| 537 |
|
|
continue;
|
| 538 |
|
|
if ifstack and not ifstack[-1]:
|
| 539 |
|
|
col += len(candToken);
|
| 540 |
|
|
continue;
|
| 541 |
|
|
# Determine which kinds of tokens are allowed at this location in the
|
| 542 |
|
|
# directive body.
|
| 543 |
|
|
if not tokens:
|
| 544 |
|
|
selAllowed = 'directive';
|
| 545 |
|
|
else:
|
| 546 |
|
|
selAllowed = allowed;
|
| 547 |
|
|
# Append the parsed token to the list of tokens.
|
| 548 |
|
|
tokens.append(ParseToken(ad,fl_loc,col,candToken,selAllowed));
|
| 549 |
|
|
col += len(candToken);
|
| 550 |
|
|
if ifstack:
|
| 551 |
|
|
raise AsmException('%d unmatched conditionals at line %d' % (len(ifstack),lineNumber,));
|
| 552 |
|
|
return tokens;
|
