37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
|
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
|
-
-
-
-
-
-
+
+
-
-
+
-
-
+
+
-
+
-
+
-
-
+
+
-
+
-
+
-
+
-
-
+
+
-
-
+
+
-
+
+
+
+
+
+
+
-
+
-
+
|
s = re.sub(r"isNextNotCOD\(\s*\)", 'isNextNotCOD(dDA, s[m.end():], m.end())', s) # isNextNotCOD(s)
s = re.sub(r"isNextVerb\(\s*\)", 'isNextVerb(dDA, s[m.end():], m.end())', s) # isNextVerb(s)
s = re.sub(r"\bspell *[(]", '_oSpellChecker.isValid(', s)
s = re.sub(r"[\\](\d+)", 'lToken[\\1]', s)
return s
def changeReferenceToken (s, dPos):
for i in range(len(dPos), 0, -1):
s = s.replace("\\"+str(i), "\\"+str(dPos[i]))
return s
def genTokenLines (sTokenLine):
"tokenize a string and return a list of lines of tokens"
def genTokenRules (sTokenLine):
lToken = sTokenLine.split()
lTokenRules = None
lTokenLines = None
for i, sToken in enumerate(lToken):
if sToken.startswith("{") and sToken.endswith("}") and sToken in dDEF:
lToken[i] = dDEF[sToken]
if ( (sToken.startswith("[") and sToken.endswith("]")) or (sToken.startswith("([") and sToken.endswith("])")) ):
bSelectedGroup = sToken.startswith("(") and sToken.endswith(")")
if bSelectedGroup:
sToken = sToken[1:-1]
# multiple token
if not lTokenRules:
lTokenRules = [ [s] for s in sToken[1:-1].split("|") ]
if not lTokenLines:
lTokenLines = [ [s] for s in sToken[1:-1].split("|") ]
else:
lNewTemp = []
for aRule in lTokenRules:
for aRule in lTokenLines:
lElem = sToken[1:-1].split("|")
sElem1 = lElem.pop(0)
if bSelectedGroup:
sElem1 = "(" + sElem1 + ")"
for sElem in lElem:
if bSelectedGroup:
sElem = "(" + sElem + ")"
aNew = list(aRule)
aNew.append(sElem)
lNewTemp.append(aNew)
aRule.append(sElem1)
lTokenRules.extend(lNewTemp)
lTokenLines.extend(lNewTemp)
else:
# simple token
if not lTokenRules:
lTokenRules = [[sToken]]
if not lTokenLines:
lTokenLines = [[sToken]]
else:
for aRule in lTokenRules:
for aRule in lTokenLines:
aRule.append(sToken)
for aRule in lTokenRules:
for aRule in lTokenLines:
yield aRule
def createRule (iLine, sRuleName, sTokenLine, sActions, nPriority):
# print(iLine, "//", sRuleName, "//", sTokenLine, "//", sActions, "//", nPriority)
for lToken in genTokenRules(sTokenLine):
for lToken in genTokenLines(sTokenLine):
# Calculate positions
dPos = {}
nGroup = 0
dPos = {} # key: iGroup, value: iToken
iGroup = 0
for i, sToken in enumerate(lToken):
if sToken.startswith("(") and sToken.endswith(")"):
lToken[i] = sToken[1:-1]
nGroup += 1
dPos[nGroup] = i
iGroup += 1
dPos[iGroup] = i + 1 # we add 1, for we count tokens from 1 to n (not from 0)
# Parse actions
for nAction, sAction in enumerate(sActions.split(" <<- ")):
if sAction.strip():
sActionId = sRuleName + "_a" + str(nAction)
aAction = createAction(sActionId, sAction, nGroup, nPriority, dPos)
aAction = createAction(sActionId, sAction, nPriority, len(lToken), dPos)
if aAction:
dACTIONS[sActionId] = aAction
lResult = list(lToken)
lResult.extend(["##"+str(iLine), sActionId])
yield lResult
def changeReferenceToken (s, dPos):
for i in range(len(dPos), 0, -1):
s = s.replace("\\"+str(i), "\\"+str(dPos[i]))
return s
def createAction (sIdAction, sAction, nGroup, nPriority, dPos):
def createAction (sIdAction, sAction, nPriority, nToken, dPos):
m = re.search("([-~=])(\\d+|)(:\\d+|)>> ", sAction)
if not m:
print(" # Error. No action found at: ", sIdAction)
print(" ==", sAction, "==")
return None
# Condition
sCondition = sAction[:m.start()].strip()
if sCondition:
sCondition = prepareFunction(sCondition)
sCondition = changeReferenceToken(sCondition, dPos)
lFUNCTIONS.append(("g_c_"+sIdAction, sCondition))
sCondition = "g_c_"+sIdAction
else:
sCondition = ""
# Action
cAction = m.group(1)
sAction = sAction[m.end():].strip()
sAction = changeReferenceToken(sAction, dPos)
iStartAction = int(m.group(2)) if m.group(2) else 0
iEndAction = int(m.group(3)[1:]) if m.group(3) else iStartAction
if nGroup:
if dPos:
try:
iStartAction = dPos[iStartAction]
iEndAction = dPos[iEndAction]
except:
print("# Error. Wrong groups in: " + sIdAction)
if cAction == "-":
|
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
|
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
|
-
-
+
+
-
-
+
+
-
-
+
+
-
-
+
+
|
if mURL:
sURL = mURL.group(1).strip()
sMsg = sMsg[:mURL.start(0)].strip()
if sMsg[0:1] == "=":
sMsg = prepareFunction(sMsg[1:])
lFUNCTIONS.append(("g_m_"+sIdAction, sMsg))
for x in re.finditer("group[(](\\d+)[)]", sMsg):
if int(x.group(1)) > nGroup:
print("# Error in groups in message at line " + sIdAction + " ("+str(nGroup)+" groups only)")
if int(x.group(1)) > nToken:
print("# Error in token index in message at line " + sIdAction + " ("+str(nToken)+" tokens only)")
sMsg = "=g_m_"+sIdAction
else:
for x in re.finditer(r"\\(\d+)", sMsg):
if int(x.group(1)) > nGroup:
print("# Error in groups in message at line " + sIdAction + " ("+str(nGroup)+" groups only)")
if int(x.group(1)) > nToken:
print("# Error in token index in message at line " + sIdAction + " ("+str(nToken)+" tokens only)")
if re.search("[.]\\w+[(]", sMsg):
print("# Error in message at line " + sIdAction + ": This message looks like code. Line should begin with =")
if sAction[0:1] == "=" or cAction == "=":
if "define" in sAction and not re.search(r"define\(\\\d+ *, *\[.*\] *\)", sAction):
print("# Error in action at line " + sIdAction + ": second argument for define must be a list of strings")
sAction = prepareFunction(sAction)
for x in re.finditer("group[(](\\d+)[)]", sAction):
if int(x.group(1)) > nGroup:
print("# Error in groups in replacement at line " + sIdAction + " ("+str(nGroup)+" groups only)")
if int(x.group(1)) > nToken:
print("# Error in token index in replacement at line " + sIdAction + " ("+str(nToken)+" tokens only)")
else:
for x in re.finditer(r"\\(\d+)", sAction):
if int(x.group(1)) > nGroup:
print("# Error in groups in replacement at line " + sIdAction + " ("+str(nGroup)+" groups only)")
if int(x.group(1)) > nToken:
print("# Error in token index in replacement at line " + sIdAction + " ("+str(nToken)+" tokens only)")
if re.search("[.]\\w+[(]|sugg\\w+[(]", sAction):
print("# Error in action at line " + sIdAction + ": This action looks like code. Line should begin with =")
if cAction == "-":
## error detected --> suggestion
if not sAction:
print("# Error in action at line " + sIdAction + ": This action is empty.")
|
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
|
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
|
+
+
+
|
continue
sPyCallables += "def {} ({}):\n".format(sFuncName, sParams)
sPyCallables += " return " + sReturn + "\n"
#sJSCallables += " {}: function ({})".format(sFuncName, sParams) + " {\n"
#sJSCallables += " return " + jsconv.py2js(sReturn) + ";\n"
#sJSCallables += " },\n"
#sJSCallables += "}\n"
for sActionName, aAction in dACTIONS.items():
print(sActionName, aAction)
# Result
d = {
"graph_callables": sPyCallables,
"graph_gctests": sGCTests,
"rules_graph": oRuleGraph,
"rules_actions": dACTIONS
}
return d
|