Grammalecte  Check-in [57b72370f1]

                        lToken.append(sToken+sSuffix)
                    break
        else:
            lToken.append(sToken)
    return lToken


def createRule (iLine, sRuleName, sTokenLine, iActionBlock, lActions, nPriority, dOptPriority, dDef, dDecl):
    "generator: create rule as list"
    # print(iLine, "//", sRuleName, "//", sTokenLine, "//", lActions, "//", nPriority)
    if sTokenLine.startswith("!!") and sTokenLine.endswith("¡¡"):
        # antipattern
        sTokenLine = sTokenLine[2:-2].strip()
        if sRuleName not in dANTIPATTERNS:
            dANTIPATTERNS[sRuleName]= []
        for lToken in genTokenLines(sTokenLine, dDef, dDecl):
            dANTIPATTERNS[sRuleName].append(lToken)

            for i, sToken in enumerate(lToken):
                if sToken.startswith("(") and sToken.endswith(")"):
                    lToken[i] = sToken[1:-1]
                    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):
                sAction = sAction.strip()
                if sAction:
                    sActionId = sRuleName + "__b" + str(iActionBlock) + "_a" + str(iAction)
                    aAction = createAction(sActionId, sAction, nPriority, dOptPriority, len(lToken), dPos, iActionLine)
                    if aAction:
                        sActionName = 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 changeReferenceToken (sText, dPos):
    "change group reference in <sText> with values in <dPos>"
    if "\\" not in sText:
        return sText
    for i in range(len(dPos), 0, -1):

def checkIfThereIsCode (sText, sActionId):
    "check if there is code in <sText> (debugging)"
    if re.search(r"[.]\w+[(]|sugg\w+[(]|\(\\[0-9]|\[[0-9]", sText):
        print("# Warning at line " + sActionId + ":  This message looks like code. Line should probably begin with =")
        print(sText)


def createAction (sActionId, sAction, nPriority, dOptPriority, 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()

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


def make (lRule, sLang, dDef, dDecl, dOptPriority):
    "compile rules, returns a dictionary of values"
    # for clarity purpose, don’t create any file here

    # removing comments, zeroing empty lines, creating definitions, storing tests, merging rule lines
    print("  parsing rules...")
    lTokenLine = []
    lActions = []
    bActionBlock = False
    nPriority = -1
    dAllGraph = {}
    sGraphName = ""
    iActionBlock = 0
    aRuleName = set()

    for iLine, sLine in lRule:
        sLine = sLine.rstrip()
        if "\t" in sLine:
            # tabulation not allowed
            print("Error. Tabulation at line: ", iLine)
            exit()
        elif sLine.startswith("@@@@GRAPH: "):
            # rules graph call
            m = re.match(r"@@@@GRAPH: *(\w+)", sLine.strip())
            if m:
                sGraphName = m.group(1)
                if sGraphName in dAllGraph:
                    print("Error at line " + iLine + ". Graph name <" + sGraphName + "> already exists.")
                    exit()
                dAllGraph[sGraphName] = []
            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()
        elif re.match("    \\S", sLine):
            # tokens line
            lTokenLine.append([iLine, sLine.strip()])
        elif sLine.startswith("        ||"):
            # tokens line continuation
            iPrevLine, sPrevLine = lTokenLine[-1]
            lTokenLine[-1] = [iPrevLine, sPrevLine + " " + sLine.strip()[2:]]
        elif sLine.startswith("        <<- "):
            # actions
            lActions.append([iLine, sLine[12:].strip()])
            if not re.search(r"[-=~/!>](?:-?\d\.?(?::\.?-?\d+|)|):?>>", sLine):
                bActionBlock = True
        elif sLine.startswith("        # "):
            # action message
            iPrevLine, sPrevLine = lActions[-1]
            lActions[-1] = [iPrevLine, sPrevLine + sLine]
        elif sLine.startswith("        ") and bActionBlock:
            # action line continuation
            iPrevLine, sPrevLine = lActions[-1]
            lActions[-1] = [iPrevLine, sPrevLine + " " + sLine.strip()]
            if re.search(r"[-=~/!>](?:-?\d\.?(?::\.?-?\d+|)|):?>>", sLine):
                bActionBlock = False
        elif re.match("[  ]*$", sLine):
            # empty line to end merging
            if not lTokenLine:
                continue
            if bActionBlock or not lActions:
                print("Error. No action found at line:", iLine)
                print(bActionBlock, lActions)
                exit()
            if not sGraphName:
                print("Error. All rules must belong to a named graph. Line: ", iLine)
                exit()
            for j, sTokenLine in lTokenLine:
                dAllGraph[sGraphName].append((j, sRuleName, sTokenLine, iActionBlock, list(lActions), nPriority))
            lTokenLine.clear()
            lActions.clear()
            iActionBlock += 1






        else:


            print("Unknown line at:", iLine)
            print(sLine)

    # processing rules
    print("  preparing rules...")
    nRule = 0
    for sGraphName, lRuleLine in dAllGraph.items():
        print("{:>8,} rules in {:<24} ".format(len(lRuleLine), "<"+sGraphName+">"), end="")
        lPreparedRule = []
        for i, sRuleGroup, sTokenLine, iActionBlock, lActions, nPriority in lRuleLine:
            for aRule in createRule(i, sRuleGroup, sTokenLine, iActionBlock, lActions, nPriority, dOptPriority, dDef, dDecl):
                lPreparedRule.append(aRule)
        nRule += len(lRuleLine)
        # Graph creation
        oDARG = darg.DARG(lPreparedRule, sLang)
        dAllGraph[sGraphName] = oDARG.createGraph()
        # Debugging
        if False:

    # Debugging
    if False:
        print("\nActions:")
        for sActionName, aAction in dACTIONS.items():
            print(sActionName, aAction)
        print("\nFunctions:")
        print(sPyCallables)

    print("Nombre d’actions: ", len(dACTIONS))

    # Result
    return {
        "graph_callables": sPyCallables,
        "graph_callablesJS": sJSCallables,
        "rules_graphs": str(dAllGraph),
        "rules_graphsJS": str(dAllGraph),
        "rules_actions": str(dACTIONS),
        "rules_actionsJS": jsconv.pyActionsToString(dACTIONS)
    }

    return [ int(sCode)  if sCode.isdigit() or (sCode[0:1] == "-" and sCode[1:].isdigit())  else sCode \
             for sCode in sGroupsPositioningCode.split(",") ]


def pyActionsToString (dActions):
    "returns dictionary as string (nbsp -> nnbsp, True -> true, etc.)"
    for _, aValue in dActions.items():
        if aValue[3] == "-":
            aValue[4] = aValue[4].replace(" ", " ") # nbsp --> nnbsp
            aValue[11] = aValue[11].replace("« ", "« ").replace(" »", " »").replace(" :", " :").replace(" :", " :")
    return str(dActions).replace("True", "true").replace("False", "false")

        for (let [sLineId, nextNodeKey] of Object.entries(oNode)) {
            let bCondMemo = null;
            for (let sRuleId of oGraph[nextNodeKey]) {
                try {
                    if (bDebug) {
                        console.log("   >TRY: " + sRuleId + " " + sLineId);
                    }
                    let [_, sOption, sFuncCond, cActionType, sWhat, ...eAct] = gc_rules_graph.dRule[sRuleId];
                    // Suggestion    [ option, condition, "-", replacement/suggestion/action, iTokenStart, iTokenEnd, cStartLimit, cEndLimit, bCaseSvty, nPriority, sMessage, sURL ]
                    // TextProcessor [ option, condition, "~", replacement/suggestion/action, iTokenStart, iTokenEnd, bCaseSvty ]
                    // Disambiguator [ option, condition, "=", replacement/suggestion/action ]
                    // Tag           [ option, condition, "/", replacement/suggestion/action, iTokenStart, iTokenEnd ]
                    // Immunity      [ option, condition, "!", "",                            iTokenStart, iTokenEnd ]
                    // Test          [ option, condition, ">", "" ]
                    if (!sOption || dOptions.gl_get(sOption, false)) {

    for sOption, lRuleGroup in chain(_getRules(True), _getRules(False)):
        if sOption != "@@@@":
            for _, _, sLineId, sRuleId, _, _ in lRuleGroup:
                if not sFilter or zFilter.search(sRuleId):
                    yield ("RegEx", sOption, sLineId, sRuleId)
    # tokens rules
    for sRuleName, lActions in _rules_graph.dRule.items():
        sLineId, sOption, _, cActionType, *_ = lActions
        if cActionType == "-":
            yield("Tokens", sOption, sLineId, sRuleName)


def displayRules (sFilter=None):
    "display the name of rules, with the filter <sFilter>"
    echo("List of rules. Filter: << " + str(sFilter) + " >>")
    for sOption, sLineId, sRuleId, sType in listRules(sFilter):
        echo("{:<8} {:<10} {:<10} {}".format(sOption, sLineId, sRuleId, sType))

        bChange = False
        for sLineId, nextNodeKey in dNode.items():
            bCondMemo = None
            for sRuleId in dGraph[nextNodeKey]:
                try:
                    if bDebug:
                        echo("   >TRY: " + sRuleId + " " + sLineId)
                    _, sOption, sFuncCond, cActionType, sWhat, *eAct = _rules_graph.dRule[sRuleId]
                    # Suggestion    [ option, condition, "-", replacement/suggestion/action, iTokenStart, iTokenEnd, cStartLimit, cEndLimit, bCaseSvty, nPriority, sMessage, sURL ]
                    # TextProcessor [ option, condition, "~", replacement/suggestion/action, iTokenStart, iTokenEnd, bCaseSvty ]
                    # Disambiguator [ option, condition, "=", replacement/suggestion/action ]
                    # Tag           [ option, condition, "/", replacement/suggestion/action, iTokenStart, iTokenEnd ]
                    # Immunity      [ option, condition, "!", "",                            iTokenStart, iTokenEnd ]
                    # Test          [ option, condition, ">", "" ]
                    if not sOption or dOptions.get(sOption, False):

TEST: __ocr__ {{J }}arrive demain


# Note: l’option “mapos” cherche les apostrophes manquantes après les lettres l, d, n, m, t, s, j, c, ç
__ocr_lettres_isolées2!2__
    [á|â|ä|b|c|ç|d|e|é|è|ê|ë|f|g|h|i|í|ì|î|ï|j|k|l|m|n|o|ó|ò|ô|ö|p|q|r|s|t|u|ú|ù|û|ü|v|w|x|z]
        <<- not before("\\d[   ]+$") and not (\1.isupper() and value(>1, "|.|<end>|"))
        ->> _           # Lettre isolée : erreur de numérisation ?

TEST: __ocr__ des verres luisent sur {{i}} le bureau blanc.
TEST: __ocr__ la voix, {{e}} est celle de…
TEST: __ocr__ ressemble {{h}} une fenêtre de serre.
TEST: __ocr__ Ça a duré 3 h.
TEST: __ocr__ c’est alors que je suis fort.
TEST: __ocr__ X

TEST: __ocr__ elles seront là tôt ou {{lard}}.


# tandis que / taudis
__ocr_tandis__
    taudis [que|qu’]
        <<- /ocr/ -1>> tandis                                                                       # Erreur de numérisation ?

TEST: __ocr__ mais {{taudis}} qu’elle œuvrait à leur salut, les nuages s’amoncelaient.


# l’est / Test
__ocr_l_est__
    Test

__conf_a_à_faim_peur_honte_soif_chaud_froid__
    à [faim|peur|honte|soif|sommeil]
        <<- /conf/ -1>> a                       # Avoir “\2”. Confusion : “à” est une préposition. Pour le verbe avoir, écrivez “a”.

    <start> elle à [chaud|froid]
        <<- /conf/ -3>> a                       # Avoir “\4”. Confusion : “à” est une préposition. Pour le verbe avoir, écrivez “a”.

TEST: Elle {{à}} chaud.
TEST: Elle {{à}} froid.
TEST: cet homme {{à}} faim
TEST: Votre réaction à froid ?

TEST: c’en est fini d’eux, ils sont comme morts
TEST: c’en est assez, ça suffit


# en butte à / but / bute
__conf_en_butte_à_au__
    en [>but|>bute]  [à|au|aux]
        <<- /conf/ -2>> butte                                           # Confusion. Écrivez « en butte \3 ».

TEST: Et moi toujours en {{but}} à de nouveaux dangers


# cane / canne (from LanguageTool)
__conf_canne_cane__
    >cane [à|a] [sucre|pêche|selfie]

__conf_gène_gêne__
    sans >gène
    sans-gènes
    sans-gène
        <<- /conf/ ->> sans-gêne                    # Confusion. Les gènes sont des éléments des chromosomes. Pour le synonyme d’embarras, écrivez “gêne”.

    [ces|des|mes|tes|ses|nos|vos|leurs] gênes
        <<- /conf/ -2>> gènes                       # Confusion probable. La gêne est un embarras. Pour parles des éléments des chromosomes, écrivez “gènes”.

TEST: Quel {{sans gène}}, celui-là !
TEST: Il croit que ses {{gênes}} décident de sa santé…


# gent [nf] / gent(e)(s) [adj]
__conf_gent__

TEST: il en a toujours {{était}} ainsi                                              ->> été
TEST: celle-ci avait {{tenue}} compte de notre passé                                ->> tenu
TEST: Ils avaient barre sur lui.


__ppas_nous_vous_avoir__
    [nous|vous]  ?[ne|n’]¿  ?[lui|leur]¿  >avoir  *WORD
        <<- /ppas/ morph(\1, ":Os")
            and not value(\-1, "|barre|confiance|charge|cours|envie|peine|prise|crainte|cure|affaire|hâte|force|recours|")
            and value(<1, "|<start>|,|comme|comment|et|lorsque|mais|où|ou|quand|qui|pourquoi|puisque|quoique|si|sinon|")
            and not \-1.isupper() and morph(\-1, ":(?:[123][sp]|Q.*:[fp])", ":(?:G|W|Q.*:m:[si])")
        --1>> =suggVerbPpas(\-1, ":m:s")
        # Ce verbe devrait être un participe passé au masculin singulier.|http://fr.wikipedia.org/wiki/Accord_du_participe_pass%C3%A9_en_fran%C3%A7ais

TEST: Nous avons {{donne}} tout notre potentiel.
TEST: mais nous avons {{était}} surpris par cette annonce.
TEST: Nous lui avons {{donnée}} un cadeau.
TEST: Vous lui avez {{donnés}} un cadeau.
TEST: nous avions quelque peu {{tempérés}} leurs ardeurs
TEST: D’un côté, le modèle occidental, […], nous a libérés de […]


__ppas_det_nom_avoir__
    [un|une|des|le|la|l’|les|ce|cet|cette|ces|mon|ton|son|ma|ta|sa|mes|tes|ses|notre|votre|nos|vos|leur|leurs|certains|certaines|quelques|plusieurs]  *WORD  ?[ne|n’]¿  ?[lui|leur]¿  >avoir  *WORD
        <<- /ppas/ not value(\-1, "|barre|confiance|charge|cours|envie|peine|prise|crainte|cure|affaire|hâte|force|recours|")
            and value(<1, "|<start>|,|comme|comment|et|lorsque|mais|où|ou|quand|qui|pourquoi|puisque|quoique|si|sinon|")
            and morph(\2, ":[NA]", ":G") and not \-1.isupper() and morph(\-1, ":(?:[123][sp]|Y|Q.*:[fp])", ":(?:G|W|Q.*:m:[si])")
            and not (\-2 == "avions" and morph(\-1, ":3[sp]"))
        --1>> =suggVerbPpas(\-1, ":m:s")
        # Ce verbe devrait être un participe passé au masculin singulier.|http://fr.wikipedia.org/wiki/Accord_du_participe_pass%C3%A9_en_fran%C3%A7ais

TEST: Les femmes lui avait {{conseillées}} de se taire.
TEST: le mur avait {{était}} détruit
TEST: Lorsque les femmes ont {{apprit}} la nouvelle…
TEST: Les élèves lui ont {{données}}.
TEST: Les élèves lui ont {{donnés}} une réponse.
TEST: Les élèves leur ont {{donnée}} ça.

    [<start>|,|comment|pourquoi|combien|que|qu’|quoique|quoiqu’|où|puis|quand|qui]  [ai|avais|eus|eussé|eusse|aurai|aurais|suis|étais|fus|fussé|fusse|serai|serais]  je
        <<- /inte/ space_after(\2, 1, 1)
        -2:3>> \2-je                                                                                # Forme interrogative ? Mettez un trait d’union.

    *WORD  ~.[is]$  je  [<end>|,]
    *WORD  ~.[is]$  je  @:¬:1s
        <<- /inte/ space_after(\2, 1, 1) and morph(\2, ":V.*:1s", ":[GNW]") and not value(\1, "|je|j’|tu|")
        -2:3>> \2-je                                                                                # Forme interrogative ? Mettez un trait d’union.

TEST: quel animal {{dessine je}}
TEST: {{mangé je}} {{ça}} avec dégoût ?
TEST: {{viendrais je}} à la fête ?
TEST: {{ai je}} enfin trouvé la réponse à mes questions ?
TEST: quel amour {{connaîtrai je}} si je juge sans cesse ?