Grammalecte  Diff

Differences From Artifact [53e75517b3]:

To Artifact [86b9b97d23]:


97
98
99
100
101
102
103

104
105
106
107
108
109
110
        self.dDef = dDef
        self.dDecl = dDecl
        self.dOptPriority = dOptPriority
        self.dAntiPatterns = {}
        self.dActions = {}
        self.dFuncName = {}
        self.dFunctions = {}


    def _genTokenLines (self, sTokenLine):
        "tokenize a string and return a list of lines of tokens"
        lTokenLines = []
        for sTokBlock in sTokenLine.split():
            # replace merger characters by spaces
            if "␣" in sTokBlock:







>







97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
        self.dDef = dDef
        self.dDecl = dDecl
        self.dOptPriority = dOptPriority
        self.dAntiPatterns = {}
        self.dActions = {}
        self.dFuncName = {}
        self.dFunctions = {}
        self.dURL = {}

    def _genTokenLines (self, sTokenLine):
        "tokenize a string and return a list of lines of tokens"
        lTokenLines = []
        for sTokBlock in sTokenLine.split():
            # replace merger characters by spaces
            if "␣" in sTokBlock:
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
            else:
                lToken.append(sToken)
        return lToken

    def createGraphAndActions (self, lRuleLine):
        "create a graph as a dictionary with <lRuleLine>"
        fStartTimer = time.time()
        print("{:>8,} rules in {:<30} ".format(len(lRuleLine), "<"+self.sGraphName+"|"+self.sGraphCode+">"), end="")
        lPreparedRule = []
        for i, sRuleName, sTokenLine, iActionBlock, lActions, nPriority in lRuleLine:
            for aRule in self.createRule(i, sRuleName, sTokenLine, iActionBlock, lActions, nPriority):
                lPreparedRule.append(aRule)
        # Debugging
        if False:
            print("\nRULES:")







|







179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
            else:
                lToken.append(sToken)
        return lToken

    def createGraphAndActions (self, lRuleLine):
        "create a graph as a dictionary with <lRuleLine>"
        fStartTimer = time.time()
        print("{:>8,} rules in {:<30} ".format(len(lRuleLine), f"<{self.sGraphName}|{self.sGraphCode}>"), end="")
        lPreparedRule = []
        for i, sRuleName, sTokenLine, iActionBlock, lActions, nPriority in lRuleLine:
            for aRule in self.createRule(i, sRuleName, sTokenLine, iActionBlock, lActions, nPriority):
                lPreparedRule.append(aRule)
        # Debugging
        if False:
            print("\nRULES:")
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252

253
254

255
256
257


258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
                        iGroup += 1
                        dPos[iGroup] = i + 1    # we add 1, for we count tokens from 1 to n (not from 0)

                # Parse actions
                for iAction, (iActionLine, sAction) in enumerate(lActions, 1):
                    sAction = sAction.strip()
                    if sAction:
                        sActionId = self.sGraphCode + "__" + sRuleName + "__b" + str(iActionBlock) + "_a" + str(iAction)
                        aAction = self.createAction(sActionId, sAction, nPriority, len(lToken), dPos, iActionLine)
                        if aAction:
                            sActionName = self.storeAction(sActionId, aAction)
                            lResult = list(lToken)
                            lResult.extend(["##"+str(iLine), sActionName])
                            #if iLine == 13341:
                            #    print("  ".join(lToken))
                            #    print(sActionId, aAction)
                            yield lResult
                        else:
                            print(" # Error on action at line:", iLine)
                            print(sTokenLine, "\n", lActions)

                    else:
                        print("No action found for ", iActionLine)


    def createAction (self, sActionId, sAction, nPriority, nToken, dPos, iActionLine):
        "create action rule as a list"


        # Option
        sOption = False
        m = re.match("/(\\w+)/", sAction)
        if m:
            sOption = m.group(1)
            sAction = sAction[m.end():].strip()
        if nPriority == -1:
            nPriority = self.dOptPriority.get(sOption, 4)

        # valid action?
        m = re.search(r"(?P<action>[-=~/!>])(?P<start>-?\d+\.?|)(?P<end>:\.?-?\d+|)(?P<casing>:|)>>", sAction)
        if not m:
            print("\n# Error. No action found at: ", sActionId)
            return None

        # Condition
        sCondition = sAction[:m.start()].strip()
        if sCondition:
            sCondition = changeReferenceToken(sCondition, dPos)
            sCondition = self.createFunction("cond", sCondition)
        else:







|










|

>


>



>
>












|
|







234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
                        iGroup += 1
                        dPos[iGroup] = i + 1    # we add 1, for we count tokens from 1 to n (not from 0)

                # Parse actions
                for iAction, (iActionLine, sAction) in enumerate(lActions, 1):
                    sAction = sAction.strip()
                    if sAction:
                        sActionId = f"{self.sGraphCode}__{sRuleName}__b{iActionBlock}_a{iAction}"
                        aAction = self.createAction(sActionId, sAction, nPriority, len(lToken), dPos, iActionLine)
                        if aAction:
                            sActionName = self.storeAction(sActionId, aAction)
                            lResult = list(lToken)
                            lResult.extend(["##"+str(iLine), sActionName])
                            #if iLine == 13341:
                            #    print("  ".join(lToken))
                            #    print(sActionId, aAction)
                            yield lResult
                        else:
                            print("# Error on action at line:", iLine)
                            print(sTokenLine, "\n", lActions)
                            exit()
                    else:
                        print("No action found for ", iActionLine)
                        exit()

    def createAction (self, sActionId, sAction, nPriority, nToken, dPos, iActionLine):
        "create action rule as a list"
        sLineId = "#" + str(iActionLine)

        # Option
        sOption = False
        m = re.match("/(\\w+)/", sAction)
        if m:
            sOption = m.group(1)
            sAction = sAction[m.end():].strip()
        if nPriority == -1:
            nPriority = self.dOptPriority.get(sOption, 4)

        # valid action?
        m = re.search(r"(?P<action>[-=~/!>])(?P<start>-?\d+\.?|)(?P<end>:\.?-?\d+|)(?P<casing>:|)>>", sAction)
        if not m:
            print("\n# Error. No action found at: ", sLineId, sActionId)
            exit()

        # Condition
        sCondition = sAction[:m.start()].strip()
        if sCondition:
            sCondition = changeReferenceToken(sCondition, dPos)
            sCondition = self.createFunction("cond", sCondition)
        else:
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
        cStartLimit = "<"
        cEndLimit = ">"
        if not m.group("start"):
            iStartAction = 1
            iEndAction = 0
        else:
            if cAction != "-" and (m.group("start").endswith(".") or m.group("end").startswith(":.")):
                print("\n# Error. Wrong selection on tokens.", sActionId)
                return None
            if m.group("start").endswith("."):
                cStartLimit = ">"
            iStartAction = int(m.group("start").rstrip("."))
            if not m.group("end"):
                iEndAction = iStartAction
            else:







|







294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
        cStartLimit = "<"
        cEndLimit = ">"
        if not m.group("start"):
            iStartAction = 1
            iEndAction = 0
        else:
            if cAction != "-" and (m.group("start").endswith(".") or m.group("end").startswith(":.")):
                print("\n# Error. Wrong selection on tokens at: ", sLineId ,sActionId)
                return None
            if m.group("start").endswith("."):
                cStartLimit = ">"
            iStartAction = int(m.group("start").rstrip("."))
            if not m.group("end"):
                iEndAction = iStartAction
            else:
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347

348
349
350
351
352
353
354
355
356
357
358
359

360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381

382
383
384
385
386
387
388
389
390
391

        if cAction == "-":
            ## error
            iMsg = sAction.find(" # ")
            if iMsg == -1:
                sMsg = "# Error. Error message not found."
                sURL = ""
                print("\n" + sMsg + " Action id: " + sActionId)
            else:
                sMsg = sAction[iMsg+3:].strip()
                sAction = sAction[:iMsg].strip()
                sURL = ""
                mURL = re.search("[|] *(https?://.*)", sMsg)
                if mURL:
                    sURL = mURL.group(1).strip()
                    sMsg = sMsg[:mURL.start(0)].strip()
                checkTokenNumbers(sMsg, sActionId, nToken)
                if sMsg[0:1] == "=":
                    sMsg = self.createFunction("msg", sMsg, True)
                else:
                    checkIfThereIsCode(sMsg, sActionId)

        # checking consistancy
        checkTokenNumbers(sAction, sActionId, nToken)

        sLineId = "#" + str(iActionLine)

        if cAction == ">":
            ## no action, break loop if condition is False
            return [sLineId, sOption, sCondition, cAction, ""]

        if not sAction and cAction != "!":
            print("\n# Error in action at line <" + sActionId + ">:  This action is empty.")


        if sAction[0:1] != "=" and cAction != "=":
            checkIfThereIsCode(sAction, sActionId)

        if cAction == "-":
            ## error detected --> suggestion
            if sAction[0:1] == "=":
                sAction = self.createFunction("sugg", sAction, True)
            elif sAction.startswith('"') and sAction.endswith('"'):
                sAction = sAction[1:-1]
            if not sMsg:
                print("\n# Error in action at line <" + sActionId + ">:  The message is empty.")

            return [sLineId, sOption, sCondition, cAction, sAction, iStartAction, iEndAction, cStartLimit, cEndLimit, bCaseSensitivity, nPriority, sMsg, sURL]
        if cAction == "~":
            ## text processor
            if sAction[0:1] == "=":
                sAction = self.createFunction("tp", sAction, True)
            elif sAction.startswith('"') and sAction.endswith('"'):
                sAction = sAction[1:-1]
            elif sAction not in "␣*_":
                nToken = sAction.count("|") + 1
                if iStartAction > 0 and iEndAction > 0:
                    if (iEndAction - iStartAction + 1) != nToken:
                        print("\n# Error in action at line <" + sActionId + ">: numbers of modified tokens modified.")
                elif iStartAction < 0 or iEndAction < 0 and iStartAction != iEndAction:
                    print("\n# Warning in action at line <" + sActionName + ">: rewriting with possible token position modified.")
            return [sLineId, sOption, sCondition, cAction, sAction, iStartAction, iEndAction, bCaseSensitivity]
        if cAction in "!/":
            ## tags
            return [sLineId, sOption, sCondition, cAction, sAction, iStartAction, iEndAction]
        if cAction == "=":
            ## disambiguator
            if "define(" in sAction and not re.search(r"define\(\\-?\d+ *, *\[.*\] *\)", sAction):
                print("\n# Error in action at line <" + sActionId + ">: second argument for <define> must be a list of strings")

            sAction = self.createFunction("da", sAction)
            return [sLineId, sOption, sCondition, cAction, sAction]
        print("\n# Unknown action.", sActionId)
        return None

    def storeAction (self, sActionId, aAction):
        "store <aAction> in <self.dActions> avoiding duplicates and return action name"
        nVar = 1
        while True:
            sActionName = sActionId + "_" + str(nVar)







|

















<
<





|
>











|
>











|

|







|
>


|







320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344


345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397

        if cAction == "-":
            ## error
            iMsg = sAction.find(" # ")
            if iMsg == -1:
                sMsg = "# Error. Error message not found."
                sURL = ""
                print("\n" + sMsg + " at: ", sLineId, sActionId)
            else:
                sMsg = sAction[iMsg+3:].strip()
                sAction = sAction[:iMsg].strip()
                sURL = ""
                mURL = re.search("[|] *(https?://.*)", sMsg)
                if mURL:
                    sURL = mURL.group(1).strip()
                    sMsg = sMsg[:mURL.start(0)].strip()
                checkTokenNumbers(sMsg, sActionId, nToken)
                if sMsg[0:1] == "=":
                    sMsg = self.createFunction("msg", sMsg, True)
                else:
                    checkIfThereIsCode(sMsg, sActionId)

        # checking consistancy
        checkTokenNumbers(sAction, sActionId, nToken)



        if cAction == ">":
            ## no action, break loop if condition is False
            return [sLineId, sOption, sCondition, cAction, ""]

        if not sAction and cAction != "!":
            print(f"\n# Error in action at line <{sLineId}/{sActionId}>:  This action is empty.")
            exit()

        if sAction[0:1] != "=" and cAction != "=":
            checkIfThereIsCode(sAction, sActionId)

        if cAction == "-":
            ## error detected --> suggestion
            if sAction[0:1] == "=":
                sAction = self.createFunction("sugg", sAction, True)
            elif sAction.startswith('"') and sAction.endswith('"'):
                sAction = sAction[1:-1]
            if not sMsg:
                print(f"\n# Error in action at line <{sLineId}/{sActionId}>:  The message is empty.")
                exit()
            return [sLineId, sOption, sCondition, cAction, sAction, iStartAction, iEndAction, cStartLimit, cEndLimit, bCaseSensitivity, nPriority, sMsg, sURL]
        if cAction == "~":
            ## text processor
            if sAction[0:1] == "=":
                sAction = self.createFunction("tp", sAction, True)
            elif sAction.startswith('"') and sAction.endswith('"'):
                sAction = sAction[1:-1]
            elif sAction not in "␣*_":
                nToken = sAction.count("|") + 1
                if iStartAction > 0 and iEndAction > 0:
                    if (iEndAction - iStartAction + 1) != nToken:
                        print(f"\n# Error in action at line <{sLineId}/{sActionId}>: numbers of modified tokens modified.")
                elif iStartAction < 0 or iEndAction < 0 and iStartAction != iEndAction:
                    print(f"\n# Warning in action at line <{sLineId}/{sActionId}>: rewriting with possible token position modified.")
            return [sLineId, sOption, sCondition, cAction, sAction, iStartAction, iEndAction, bCaseSensitivity]
        if cAction in "!/":
            ## tags
            return [sLineId, sOption, sCondition, cAction, sAction, iStartAction, iEndAction]
        if cAction == "=":
            ## disambiguator
            if "define(" in sAction and not re.search(r"define\(\\-?\d+ *, *\[.*\] *\)", sAction):
                print(f"\n# Error in action at line <{sLineId}/{sActionId}>: second argument for <define> must be a list of strings")
                exit()
            sAction = self.createFunction("da", sAction)
            return [sLineId, sOption, sCondition, cAction, sAction]
        print("\n# Unknown action at ", sLineId, sActionId)
        return None

    def storeAction (self, sActionId, aAction):
        "store <aAction> in <self.dActions> avoiding duplicates and return action name"
        nVar = 1
        while True:
            sActionName = sActionId + "_" + str(nVar)
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
                sParams = "lToken, nTokenOffset, nLastToken"
            elif sFuncName.startswith("_g_da_"): # disambiguator
                sParams = "lToken, nTokenOffset, nLastToken"
            else:
                print("# Unknown function type in [" + sFuncName + "]")
                continue
            # Python
            sPyCallables += "def {} ({}):\n".format(sFuncName, sParams)
            sPyCallables += "    return " + sReturn + "\n"
            # JavaScript
            sJSCallables += "    {}: function ({})".format(sFuncName, sParams) + " {\n"
            sJSCallables += "        return " + jsconv.py2js(sReturn) + ";\n"
            sJSCallables += "    },\n"
        return sPyCallables, sJSCallables


def processing (sGraphName, sGraphCode, sLang, lRuleLine, dDef, dDecl, dOptPriority):
    "to be run in a separate process"







|
|

|







438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
                sParams = "lToken, nTokenOffset, nLastToken"
            elif sFuncName.startswith("_g_da_"): # disambiguator
                sParams = "lToken, nTokenOffset, nLastToken"
            else:
                print("# Unknown function type in [" + sFuncName + "]")
                continue
            # Python
            sPyCallables += f"def {sFuncName} ({sParams}):\n"
            sPyCallables += f"    return {sReturn}\n"
            # JavaScript
            sJSCallables += f"    {sFuncName}: function ({sParams}) {{\n"
            sJSCallables += "        return " + jsconv.py2js(sReturn) + ";\n"
            sJSCallables += "    },\n"
        return sPyCallables, sJSCallables


def processing (sGraphName, sGraphCode, sLang, lRuleLine, dDef, dDecl, dOptPriority):
    "to be run in a separate process"
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
        elif sLine.startswith("@@@@GRAPH: "):
            # rules graph call
            m = re.match(r"@@@@GRAPH: *(\w+) *[|] *(\w+)", sLine.strip())
            if m:
                sGraphName = m.group(1)
                sGraphCode = m.group(2)
                if sGraphName in dAllGraph or sGraphCode in dGraphCode:
                    print("Error at line " + iLine + ". Graph name <" + sGraphName + "> or graph code <" + sGraphCode + "> already exists.")
                    exit()
                dAllGraph[sGraphName] = []
                dGraphCode[sGraphName] = sGraphCode
            else:
                print("Error. Graph name not found at line", iLine)
                exit()
        elif sLine.startswith("__") and sLine.endswith("__"):
            # new rule group
            m = re.match("__(\\w+)(!\\d|)__", sLine)
            if m:
                sRuleName = m.group(1)
                if sRuleName in aRuleName:
                    print("Error at line " + str(iLine) + ". Rule name <" + sRuleName + "> already exists.")
                    exit()
                aRuleName.add(sRuleName)
                iActionBlock = 1
                nPriority = int(m.group(2)[1:]) if m.group(2)  else -1
            else:
                print("Syntax error in rule group: ", sLine, " -- line:", iLine)
                exit()







|












|







483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
        elif sLine.startswith("@@@@GRAPH: "):
            # rules graph call
            m = re.match(r"@@@@GRAPH: *(\w+) *[|] *(\w+)", sLine.strip())
            if m:
                sGraphName = m.group(1)
                sGraphCode = m.group(2)
                if sGraphName in dAllGraph or sGraphCode in dGraphCode:
                    print(f"Error at line {iLine}. Graph name <{sGraphName}> or graph code <{sGraphCode}> already exists.")
                    exit()
                dAllGraph[sGraphName] = []
                dGraphCode[sGraphName] = sGraphCode
            else:
                print("Error. Graph name not found at line", iLine)
                exit()
        elif sLine.startswith("__") and sLine.endswith("__"):
            # new rule group
            m = re.match("__(\\w+)(!\\d|)__", sLine)
            if m:
                sRuleName = m.group(1)
                if sRuleName in aRuleName:
                    print(f"Error at line {iLine}. Rule name <{sRuleName}> already exists.")
                    exit()
                aRuleName.add(sRuleName)
                iActionBlock = 1
                nPriority = int(m.group(2)[1:]) if m.group(2)  else -1
            else:
                print("Syntax error in rule group: ", sLine, " -- line:", iLine)
                exit()
567
568
569
570
571
572
573















574
575
576
577
578
579
580
581
582

583
584
585
    sJSCallables = ""
    for xFuture in lResult:
        sGraphName, dGraph, dActions, sPy, sJS = xFuture.result()
        dAllGraph[sGraphName] = dGraph
        dAllActions.update(dActions)
        sPyCallables += sPy
        sJSCallables += sJS















    print("  Total: ", nRule, "rules, ", len(dAllActions), "actions")
    print("  Build time: {:.2f} s".format(time.time() - fStartTimer))

    return {
        # the graphs describe paths of tokens to actions which eventually execute callables
        "rules_graphs": str(dAllGraph),
        "rules_graphsJS": str(dAllGraph),
        "rules_actions": str(dAllActions),
        "rules_actionsJS": jsconv.pyActionsToString(dAllActions),

        "graph_callables": sPyCallables,
        "graph_callablesJS": sJSCallables
    }







>
>
>
>
>
>
>
>
>
>
>
>
>
>
>









>



573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
    sJSCallables = ""
    for xFuture in lResult:
        sGraphName, dGraph, dActions, sPy, sJS = xFuture.result()
        dAllGraph[sGraphName] = dGraph
        dAllActions.update(dActions)
        sPyCallables += sPy
        sJSCallables += sJS
    # create a dictionary of URL
    dTempURL = {}
    i = 1
    for sKey, lValue in dAllActions.items():
        if lValue[3] == "-":
            if lValue[-1]:
                if lValue[-1] not in dTempURL:
                    dTempURL[lValue[-1]] = i
                    i += 1
                lValue[-1] = i
            else:
                lValue[-1] = 0
    dURL = { v: k  for k, v in dTempURL.items() } # reversing key and values
    dURL[0] = ""
    # end
    print("  Total: ", nRule, "rules, ", len(dAllActions), "actions")
    print("  Build time: {:.2f} s".format(time.time() - fStartTimer))

    return {
        # the graphs describe paths of tokens to actions which eventually execute callables
        "rules_graphs": str(dAllGraph),
        "rules_graphsJS": str(dAllGraph),
        "rules_actions": str(dAllActions),
        "rules_actionsJS": jsconv.pyActionsToString(dAllActions),
        "rules_graph_URL": str(dURL),
        "graph_callables": sPyCallables,
        "graph_callablesJS": sJSCallables
    }