| [Mumps2Py:] [83] lots of parsing bug fixes, added parsing support for structured system variables. |
[ Thread Index |
Date Index
| More lists.mumps2py.org/discuss Archives
]
Revision: 83
Author: pgallot
Date: 2008-02-04 23:33:14 +0000 (Mon, 04 Feb 2008)
Log Message:
-----------
lots of parsing bug fixes, added parsing support for structured system variables.
Modified Paths:
--------------
trunk/mumps2py/mumps2tok.py
Modified: trunk/mumps2py/mumps2tok.py
===================================================================
--- trunk/mumps2py/mumps2tok.py 2008-02-04 23:28:05 UTC (rev 82)
+++ trunk/mumps2py/mumps2tok.py 2008-02-04 23:33:14 UTC (rev 83)
@@ -19,8 +19,6 @@
import re, fileinput
from tokens import *
-GLOBALDB = 0 # debugging variable.
-
class ParseError(Exception):
""" Exception class for errors relating to parsing Mumps code."""
def __init__(self, the_module, line, dscr = "", pos = 0, lineno = -1):
@@ -45,15 +43,16 @@
return errormsg
MUMPS_RE_DICT = {
- "var":re.compile(r"((?P<globalvar>\^{1})|(?P<pbr>[.]{1}))?(?P<var>[A-Za-z%][A-Za-z0-9]*)(?P<indexed>[(]{1})?"),
+ "var":re.compile(r"((?P<globalvar>\^{1}))?(?P<var>[A-Za-z%][A-Za-z0-9]*)(?P<indexed>[(]{1})?"),
"indirection":re.compile(r"@"),
"op":re.compile(r"[-_/+']|([']?(([*]{1,2})|([]]{1,2})|\\|[[=><#&!/]))"),
"str":re.compile(r'"'),
"num":re.compile(r"([0-9]*[.])?[0-9]+([eE][+-][0-9]+)?"),
- "extr":re.compile(r"[$]{2}(?P<label>\w+)(\^(?P<routine>[%A-Za-z][A-Za-z0-9]*))?(?P<func>[(]{1})?"),
+ "extr":re.compile(r"[$]{2}"),
"intr":re.compile(r"[$]{1}(?P<intrinsic>[A-Za-z0-9]+)(?P<func>[(]{1})?"),
+ "structsysvar":re.compile(r"\^[$]{1}(?P<var>[A-Za-z0-9]+)(?P<indexed>[(]{1})?"),
"patmat":re.compile(r"[?]"),
- "patAtom":re.compile(r"(((?P<mnrp>\d+)?[.](?P<mxrp>\d+)?)|(?P<rpcount>\d+)){1}((?P<altlist>[(])|(?P<patcode>[ACELNPUacelnpu]+)|(?P<str>([\"]([\"]{2}|.)*?[\"])))"),
+ "patAtom":re.compile(r"(((?P<mnrp>\d+)?[.](?P<mxrp>\d+)?)|(?P<rpcount>\d+)){1}((?P<altlist>[(])|(?P<patcode>[ACELNPUacelnpu]+)|(?P<str>([\"])))"),
"nakedRef":re.compile(r"\^\("),
"subexpr":re.compile(r"[(]"),
"command":re.compile(r"(?P<indents>[ .]*)(?P<cmd>[A-Za-z]+)"),
@@ -68,7 +67,13 @@
""" parses a list of parameters """
actuallist = []
while line[pos] != ")":
+ pass_by_ref = False
+ if line[pos] == '.' and not MUMPS_RE_DICT['num'].match(line, pos):
+ pass_by_ref = True
+ pos = pos + 1
exprtok = parse_expr(mumps_module, line, pos, r"([:,)])")
+ if pass_by_ref:
+ exprtok.pass_by_ref = True
actuallist.append(exprtok)
pos = exprtok.end
if line[pos] == ":":
@@ -82,16 +87,12 @@
def consume_entry_ref(mumps_module, line, pos):
""" parses a Mumps Entry Ref"""
- #print "entry_ref:(%s)" % line[pos:]
- #global GLOBALDB
- #GLOBALDB=1
-
token = Token(ENTRYREF, pos)
if line[pos] == '@':
if MUMPS_RE_DICT["nakedRef"].match(line, pos + 1):
indirect = parse_expr(mumps_module, line, pos+1, r"([,+: ]|\s$)")
token.indirect = indirect
- pos = indirect.end + 1
+ pos = indirect.end
else:
indirect = parse_expr(mumps_module, line, pos, \
r"(\^|[ +),:]|\s*$)")
@@ -128,12 +129,9 @@
else:
token.routine = m_name.group()
pos = m_name.end()
- if not MUMPS_RE_DICT["cmdEnd"].match(line, pos) and \
- line[pos] == '(':
- pos, token.params = \
- consume_actuallist(mumps_module, line, pos + 1)
-
- #GLOBALDB=0
+ if not MUMPS_RE_DICT["cmdEnd"].match(line, pos) and line[pos] == '(':
+ pos, token.params = consume_actuallist(mumps_module, line, pos + 1)
+
token.end = pos
return (token, pos)
@@ -176,8 +174,7 @@
def parse_expr(mumps_module, line, startpos, terminator = r"([ ]|\s*$)"):
""" parses a Mumps expression by recursive descent"""
- if GLOBALDB > 0:
- print "parse_expr %d [%s]" % (startpos, line[startpos:])
+ #if __debug__: print "parse_expr %d [%s]" % (startpos, line[startpos:])
def consume_pattern(lineno, line, pos):
""" parses a mumps-style pattern-matchine pattern"""
@@ -194,8 +191,9 @@
if atom.group("mxrp"):
sub_token.maxrep = int(atom.group("mxrp"))
if atom.group("str"):
- sub_token.match_str = atom.group("str")
- pos = atom.end()
+ sub_token.match_str = parse_expr(\
+ mumps_module, line, atom.start("str"), r'\w|[ .,:)(]|\s*$')
+ pos = sub_token.match_str.end
elif atom.group("altlist"):
sub_token.alt_list = []
pos = atom.end()
@@ -249,11 +247,6 @@
token = Token(OP_TOKEN_DICT[op_type], mobj.start())
return (token, mobj.end())
- def consume_unknown(mobj):
- token = Token(UNKNOWNITEM, mobj.start())
- token.item = mobj.group()
- return (token, mobj.end())
-
def consume_naked_ref(mobj):
""" parses a Mumps naked reference"""
token = Token(OPNAKEDREF, mobj.start())
@@ -265,7 +258,7 @@
""" parses a Mumps indirection argument"""
token = Token(INDIRECTION, mobj.start())
expr = parse_expr(mumps_module, line, mobj.end(), \
- r"(\\|[-@:,+=><#&!*/ ')]|\s*$)")
+ r"(\^|\\|[-@:,+=><#&!*/ ')]|\s*$)")
token.expr = expr
pos = expr.end
if line[pos] == '@' and line[pos + 1] == '(':
@@ -279,30 +272,20 @@
token = VarToken(GLOBALVAR, mobj.start())
else:
token = VarToken(LOCALVAR, mobj.start())
- if mobj.group("pbr"):
- token.pass_by_reference = True
token.varname = mobj.group("var")
pos = mobj.end("var")
if mobj.group("indexed"):
- pos = pos + 1
- pos, indices = consume_actuallist(mumps_module, line, pos)
+ pos, indices = consume_actuallist(mumps_module, line, pos + 1)
token.indices = indices
return (token, pos)
def consume_extrinsic(mobj):
""" parses a Mumps extrinsic function or variable"""
- token = Token(USERFUNC, mobj.start())
pos = mobj.end()
- if mobj.group("label"):
- token.label = mobj.group("label")
- if mobj.group("routine"):
- token.routine = mobj.group("routine")
- if mobj.group("func"):
- pos, params = consume_actuallist(mumps_module, line, pos)
- token.params = params
- else: # vars...
- token.params = []
- return (token, pos)
+ token = Token(USERFUNC, pos)
+ (entry_ref, end_pos) = consume_entry_ref(mumps_module, line, pos)
+ token.entry_ref = entry_ref
+ return (token, end_pos)
def consume_intrinsic(mobj):
""" parses a Mumps intrinsic function or variable"""
@@ -326,6 +309,19 @@
token.name = name
return (token, pos)
+ def consume_structsysvar(mobj):
+ """ parses a Mumps Structured System variable"""
+ name = mobj.group("var").upper()
+ pos = mobj.end("var")
+ toktype = STRUCT_SYS_VARS.get(name, SSV_UNKNOWN)
+ token = Token(toktype, mobj.start())
+ if toktype == SSV_UNKNOWN:
+ token.name = name
+ if mobj.group("indexed"):
+ pos, indices = consume_actuallist(mumps_module, line, pos + 1)
+ token.indices = indices
+ return (token, pos)
+
def consume_pattmatch(mobj):
""" parses mumps-style pattern-matchine patterns"""
pos = mobj.end()
@@ -338,13 +334,13 @@
(MUMPS_RE_DICT["indirection"], consume_indirection, "indirection"),
(MUMPS_RE_DICT["var"], consume_var, "var"),
(MUMPS_RE_DICT["op"], consume_op, "op"),
+ (MUMPS_RE_DICT["patmat"], consume_pattmatch, "patmat"),
(MUMPS_RE_DICT["str"], consume_str, "str"),
(MUMPS_RE_DICT["num"], consume_num, "num"),
(MUMPS_RE_DICT["extr"], consume_extrinsic, "extr"),
(MUMPS_RE_DICT["intr"], consume_intrinsic, "intr"),
- (MUMPS_RE_DICT["patmat"], consume_pattmatch, "patmat"),
(MUMPS_RE_DICT["nakedRef"], consume_naked_ref, "nakedRef"),
- (re.compile(r"(\^\$JOB)"), consume_unknown, "unknown"),
+ (MUMPS_RE_DICT["structsysvar"], consume_structsysvar, "structsysvar"),
(MUMPS_RE_DICT["subexpr"], consume_sub_expr, "subexpr")]
pos = startpos
@@ -385,7 +381,7 @@
mterminated = MUMPS_RE_DICT["cmdEnd"]
if not mterminated.match(line, pos) and \
line[pos:pos+2] == '@(':
- indirection = parse_expr(mumps_module, line, pos, r"([ ,]|\s*$)")
+ indirection = parse_expr(mumps_module, line, pos, r"([ ,:]|\s*$)")
new_pos = indirection.end
if mterminated.match(line, new_pos):
@@ -548,10 +544,14 @@
pos = pos + 1
keyword_list = []
while line[pos] != ')':
- keyword = parse_expr(mumps_module, line, pos, r"([:)])")
+ keyword = parse_expr(mumps_module, line, pos, r"[,:=)]")
keyword_list.append(keyword)
pos = keyword.end
- if line[pos] == ':':
+ if line[pos] == "=":
+ kval = parse_expr(mumps_module, line, pos, r"[,:)]")
+ keyword.keyval = kval
+ pos = kval.end
+ if line[pos] in (':', ','):
pos = pos + 1
device["keyword_list"] = keyword_list
pos = pos + 1
@@ -622,6 +622,7 @@
if line[pos] == ',':
pos = pos + 1
lock_item["nrefs"] = nref_list
+ pos = pos + 1 # eat the closing parenthesis
else:
nref = parse_expr(mumps_module, line, pos, r"([ ,:]|\s*$)")
pos = nref.end
@@ -764,7 +765,6 @@
def parse_goto(mumps_module, line, pos):
"""parse the Mumps Goto command"""
- #NOTE: a goto seems very similar to a do, but with no arguments passed.
entry_ref_list = []
mterminated = MUMPS_RE_DICT["cmdEnd"]
while not mterminated.match(line, pos):
@@ -904,8 +904,7 @@
mumps_module.add_token(Token(toktype, startpos))
if toktype == UNKNOWNCMD:
mumps_module.last_token().name = cmd
- if re_match.group("indents"):
- # count the periods for the nesting level...
+ if re_match.group("indents"): # count the periods for the nesting level...
mumps_module.indent_token( \
len(re.findall("[.]", line[startpos:re_match.start("cmd")])))
if line[re_match.end("cmd")] == ':':
@@ -922,7 +921,7 @@
return mumps_module.last_token()
-def parseMumps(mumps_module):
+def parseMumps(mumps_module, open_fileinput= None):
""" Parse all the code of given Mumps module."""
mumps_module.empty_tokenlist()
@@ -932,8 +931,13 @@
(MUMPS_RE_DICT["emptyline"], parse_emptyline, "emptyline"),
(MUMPS_RE_DICT["cmd"], parse_command, "cmd")]
+ if open_fileinput:
+ input_file = open_fileinput
+ else:
+ input_file = fileinput.input(mumps_module.input_file)
+
try:
- for line in fileinput.input(mumps_module.input_file):
+ for line in input_file:
if fileinput.lineno() < mumps_module.start:
continue
@@ -970,7 +974,8 @@
raise ParseError(mumps_module, line, "no Pattern match", pos, \
fileinput.lineno())
finally:
- fileinput.close()
+ if not open_fileinput:
+ fileinput.close()
if __name__ == '__main__':
from mumps_module import parse_for_routines
@@ -983,7 +988,6 @@
parseMumps(the_module)
try:
- #mods = parse_for_routines("../testfiles/vista.m2p", "./out")
mods = parse_for_routines("../testfiles/fm22.m2p", "./out")
#parsemodule(mods,"ACKQUTL1")
for the_module in mods: