Grammalecte  Changes On Branch cbc9afbecda7b823

"""
Grammalecte: compile rules
"""

import re
import traceback
import json

import compile_rules_js_convert as jsconv
import compile_rules_graph as crg


dDEF = {}
lFUNCTIONS = []

aRULESET = set()     # set of rule-ids to check if there is several rules with the same id
nRULEWITHOUTNAME = 0

dJSREGEXES = {}

sWORDLIMITLEFT  = r"(?<![\w.,–-])"   # r"(?<![-.,—])\b"  seems slower
sWORDLIMITRIGHT = r"(?![\w–-])"      # r"\b(?!-—)"       seems slower


def prepareFunction (s):
    "convert simple rule syntax to a string of Python code"
    s = s.replace("__also__", "bCondMemo")
    s = s.replace("__else__", "not bCondMemo")
    s = re.sub(r"isStart *\(\)", 'before("^ *$|, *$")', s)
    s = re.sub(r"isRealStart *\(\)", 'before("^ *$")', s)
    s = re.sub(r"isStart0 *\(\)", 'before0("^ *$|, *$")', s)
    s = re.sub(r"isRealStart0 *\(\)", 'before0("^ *$")', s)
    s = re.sub(r"isEnd *\(\)", 'after("^ *$|^,")', s)
    s = re.sub(r"isRealEnd *\(\)", 'after("^ *$")', s)
    s = re.sub(r"isEnd0 *\(\)", 'after0("^ *$|^,")', s)
    s = re.sub(r"isRealEnd0 *\(\)", 'after0("^ *$")', s)
    s = re.sub(r"(select|exclude)[(][\\](\d+)", '\\1(dTokenPos, m.start(\\2), m.group(\\2)', s)
    s = re.sub(r"define[(][\\](\d+)", 'define(dTokenPos, m.start(\\1)', s)
    s = re.sub(r"(morph|morphex|displayInfo)[(][\\](\d+)", '\\1((m.start(\\2), m.group(\\2))', s)
    s = re.sub(r"(morph|morphex|displayInfo)[(]", '\\1(dTokenPos, ', s)
    s = re.sub(r"(sugg\w+|switch\w+)\(@", '\\1(m.group(i[4])', s)
    s = re.sub(r"word\(\s*1\b", 'nextword1(s, m.end()', s)                                  # word(1)
    s = re.sub(r"word\(\s*-1\b", 'prevword1(s, m.start()', s)                               # word(-1)
    s = re.sub(r"word\(\s*(\d)", 'nextword(s, m.end(), \\1', s)                             # word(n)
    s = re.sub(r"word\(\s*-(\d)", 'prevword(s, m.start(), \\1', s)                          # word(-n)
    s = re.sub(r"before\(\s*", 'look(s[:m.start()], ', s)                                   # before(s)
    s = re.sub(r"after\(\s*", 'look(s[m.end():], ', s)                                      # after(s)
    s = re.sub(r"textarea\(\s*", 'look(s, ', s)                                             # textarea(s)
    s = re.sub(r"before_chk1\(\s*", 'look_chk1(dTokenPos, s[:m.start()], 0, ', s)           # before_chk1(s)
    s = re.sub(r"after_chk1\(\s*", 'look_chk1(dTokenPos, s[m.end():], m.end(), ', s)        # after_chk1(s)
    s = re.sub(r"textarea_chk1\(\s*", 'look_chk1(dTokenPos, s, 0, ', s)                     # textarea_chk1(s)
    s = re.sub(r"/0", 'sx[m.start():m.end()]', s)                                           # /0
    s = re.sub(r"before0\(\s*", 'look(sx[:m.start()], ', s)                                 # before0(s)
    s = re.sub(r"after0\(\s*", 'look(sx[m.end():], ', s)                                    # after0(s)
    s = re.sub(r"textarea0\(\s*", 'look(sx, ', s)                                           # textarea0(s)
    s = re.sub(r"before0_chk1\(\s*", 'look_chk1(dTokenPos, sx[:m.start()], 0, ', s)         # before0_chk1(s)
    s = re.sub(r"after0_chk1\(\s*", 'look_chk1(dTokenPos, sx[m.end():], m.end(), ', s)      # after0_chk1(s)
    s = re.sub(r"textarea0_chk1\(\s*", 'look_chk1(dTokenPos, sx, 0, ', s)                   # textarea0_chk1(s)
    s = re.sub(r"isEndOfNG\(\s*\)", 'isEndOfNG(dTokenPos, s[m.end():], m.end())', s)        # isEndOfNG(s)
    s = re.sub(r"isNextNotCOD\(\s*\)", 'isNextNotCOD(dTokenPos, s[m.end():], m.end())', s)  # isNextNotCOD(s)
    s = re.sub(r"isNextVerb\(\s*\)", 'isNextVerb(dTokenPos, s[m.end():], m.end())', s)      # isNextVerb(s)
    s = re.sub(r"\bspell *[(]", '_oSpellChecker.isValid(', s)
    s = re.sub(r"[\\](\d+)", 'm.group(\\1)', s)
    return s


def uppercase (s, sLang):
    "(flag i is not enough): converts regex to uppercase regex: 'foo' becomes '[Ff][Oo][Oo]', but 'Bar' becomes 'B[Aa][Rr]'."

            nState = 4
        elif nState == 4:
            nState = 0
    return sUp


def countGroupInRegex (sRegex):
    "returns the number of groups in <sRegex>"
    try:
        return re.compile(sRegex).groups
    except:
        traceback.print_exc()
        print(sRegex)
    return 0


def createRule (s, nIdLine, sLang, bParagraph, dOptPriority):
    "returns rule as list [option name, regex, bCaseInsensitive, identifier, list of actions]"
    global dJSREGEXES
    global nRULEWITHOUTNAME


    sLineId = str(nIdLine) + ("p" if bParagraph else "s")
    sRuleId = sLineId

    #### GRAPH CALL
    if s.startswith("@@@@"):
        if bParagraph:
            print("Error. Graph call can be made only after the first pass (sentence by sentence)")
            exit()
        return ["@@@@", s[4:], sLineId]

    #### OPTIONS
    sOption = False         # False or [a-z0-9]+ name
    nPriority = 4           # Default is 4, value must be between 0 and 9
    tGroups = None          # code for groups positioning (only useful for JavaScript)
    cCaseMode = 'i'         # i: case insensitive,  s: case sensitive,  u: uppercasing allowed
    cWordLimitLeft = '['    # [: word limit, <: no specific limit
    cWordLimitRight = ']'   # ]: word limit, >: no specific limit
    m = re.match("^__(?P<borders_and_case>[[<]\\w[]>])(?P<option>/[a-zA-Z0-9]+|)(?P<ruleid>\\(\\w+\\)|)(?P<priority>![0-9]|)__ *", s)

    #### REGEX TRIGGER
    i = s.find(" <<-")
    if i == -1:
        print("# Error: no condition at line " + sLineId)
        return None
    sRegex = s[:i].strip()
    s = s[i+4:]

    # JS groups positioning codes
    m = re.search("@@\\S+", sRegex)
    if m:
        tGroups = jsconv.groupsPositioningCodeToList(sRegex[m.start()+2:])
        sRegex = sRegex[:m.start()].strip()
    # JS regex
    m = re.search("<js>.+</js>i?", sRegex)

        sRegex = sRegex.replace("(?i)", "")
        sRegex = uppercase(sRegex, sLang)
    else:
        print("# Unknown case mode [" + cCaseMode + "] at line " + sLineId)

    ## check regex
    try:
        re.compile(sRegex)
    except:
        print("# Regex error at line ", nIdLine)
        print(sRegex)
        traceback.print_exc()
        return None
    ## groups in non grouping parenthesis
    for x in re.finditer(r"\(\?:[^)]*\([[\w -]", sRegex):
        print("# Warning: groups inside non grouping parenthesis in regex at line " + sLineId)

    #### PARSE ACTIONS
    lActions = []
    nAction = 1
    for sAction in s.split(" <<- "):
        t = createAction(sRuleId + "_" + str(nAction), sAction, nGroup)
        nAction += 1
        if t:
            lActions.append(t)
    if not lActions:
        return None

    return [sOption, sRegex, bCaseInsensitive, sLineId, sRuleId, nPriority, lActions, tGroups]


def checkReferenceNumbers (sText, sActionId, nToken):
    "check if token references in <sText> greater than <nToken> (debugging)"
    for x in re.finditer(r"\\(\d+)", sText):
        if int(x.group(1)) > nToken:
            print("# Error in token index at line " + sActionId + " ("+str(nToken)+" tokens only)")
            print(sText)


def checkIfThereIsCode (sText, sActionId):
    "check if there is code in <sText> (debugging)"
    if re.search("[.]\\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 (sIdAction, sAction, nGroup):
    "returns an action to perform as a tuple (condition, action type, action[, iGroup [, message, URL ]])"


    m = re.search(r"([-~=>])(\d*|)>>", sAction)
    if not m:
        print("# No action at line " + sIdAction)
        return None

    #### CONDITION
    sCondition = sAction[:m.start()].strip()
    if sCondition:
        sCondition = prepareFunction(sCondition)
        lFUNCTIONS.append(("_c_"+sIdAction, sCondition))
        checkReferenceNumbers(sCondition, sIdAction, nGroup)


        if ".match" in sCondition:
            print("# Error. JS compatibility. Don't use .match() in condition, use .search()")
        sCondition = "_c_"+sIdAction
    else:
        sCondition = None

    #### iGroup / positioning
    iGroup = int(m.group(2)) if m.group(2) else 0
    if iGroup > nGroup:
        print("# Selected group > group number in regex at line " + sIdAction)

    #### ACTION
    sAction = sAction[m.end():].strip()
    cAction = m.group(1)
    if cAction == "-":
        ## error
        iMsg = sAction.find(" # ")
        if iMsg == -1:
            sMsg = "# Error. Error message not found."
            sURL = ""
            print(sMsg + " Action id: " + sIdAction)
        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()
            checkReferenceNumbers(sMsg, sIdAction, nGroup)
            if sMsg[0:1] == "=":
                sMsg = prepareFunction(sMsg[1:])
                lFUNCTIONS.append(("_m_"+sIdAction, sMsg))



                sMsg = "=_m_"+sIdAction
            else:


                checkIfThereIsCode(sMsg, sIdAction)



    checkReferenceNumbers(sAction, sIdAction, nGroup)
    if sAction[0:1] == "=" or cAction == "=":


        sAction = prepareFunction(sAction)
        sAction = sAction.replace("m.group(i[4])", "m.group("+str(iGroup)+")")



    else:
        checkIfThereIsCode(sAction, sIdAction)

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

    if not sAction:
        print("# Error in action at line " + sIdAction + ":  This action is empty.")
        return None

    if cAction == "-":
        ## error detected --> suggestion


        if sAction[0:1] == "=":
            lFUNCTIONS.append(("_s_"+sIdAction, sAction[1:]))
            sAction = "=_s_"+sIdAction
        elif sAction.startswith('"') and sAction.endswith('"'):
            sAction = sAction[1:-1]
        if not sMsg:
            print("# Error in action at line " + sIdAction + ":  the message is empty.")
        return [sCondition, cAction, sAction, iGroup, sMsg, sURL]
    elif cAction == "~":
        ## text processor


        if sAction[0:1] == "=":
            lFUNCTIONS.append(("_p_"+sIdAction, sAction[1:]))
            sAction = "=_p_"+sIdAction
        elif sAction.startswith('"') and sAction.endswith('"'):
            sAction = sAction[1:-1]
        return [sCondition, cAction, sAction, iGroup]
    elif cAction == "=":
        ## disambiguator
        if sAction[0:1] == "=":
            sAction = sAction[1:]
        if "define" in sAction and not re.search(r"define\(dTokenPos, *m\.start.*, \[.*\] *\)", sAction):
            print("# Error in action at line " + sIdAction + ": second argument for define must be a list of strings")
            print(sAction)
        lFUNCTIONS.append(("_d_"+sIdAction, sAction))
        sAction = "_d_"+sIdAction
        return [sCondition, cAction, sAction]



    else:
        print("# Unknown action at line " + sIdAction)
        return None


def _calcRulesStats (lRules):
    "count rules and actions"
    d = {'=':0, '~': 0, '-': 0, '>': 0}
    for aRule in lRules:
        if aRule[0] != "@@@@":
            for aAction in aRule[6]:
                d[aAction[1]] = d[aAction[1]] + 1
    return (d, len(lRules))


def displayStats (lParagraphRules, lSentenceRules):
    "display rules numbers"
    print("  {:>18} {:>18} {:>18} {:>18}".format("DISAMBIGUATOR", "TEXT PROCESSOR", "GRAMMAR CHECKING", "REGEX"))
    d, nRule = _calcRulesStats(lParagraphRules)
    print("§ {:>10} actions {:>10} actions {:>10} actions  in {:>8} rules".format(d['='], d['~'], d['-'], nRule))
    d, nRule = _calcRulesStats(lSentenceRules)
    print("s {:>10} actions {:>10} actions {:>10} actions  in {:>8} rules".format(d['='], d['~'], d['-'], nRule))

            m = re.match("OPTGROUP/([a-z0-9]+):(.+)$", sLine)
            lStructOpt.append( (m.group(1), list(map(str.split, m.group(2).split(",")))) )
        elif sLine.startswith("OPTSOFTWARE:"):
            lOpt = [ [s, {}]  for s in sLine[12:].strip().split() ]  # don’t use tuples (s, {}), because unknown to JS
        elif sLine.startswith("OPT/"):
            m = re.match("OPT/([a-z0-9]+):(.+)$", sLine)
            for i, sOpt in enumerate(m.group(2).split()):
                lOpt[i][1][m.group(1)] = eval(sOpt)
        elif sLine.startswith("OPTPRIORITY/"):
            m = re.match("OPTPRIORITY/([a-z0-9]+): *([0-9])$", sLine)
            dOptPriority[m.group(1)] = int(m.group(2))
        elif sLine.startswith("OPTLANG/"):
            m = re.match("OPTLANG/([a-z][a-z](?:_[A-Z][A-Z]|)):(.+)$", sLine)
            sLang = m.group(1)[:2]
            dOptLabel[sLang] = { "__optiontitle__": m.group(2).strip() }

    print("  options defined for: " + ", ".join([ t[0] for t in lOpt ]))
    dOptions = { "lStructOpt": lStructOpt, "dOptLabel": dOptLabel, "sDefaultUILang": sDefaultUILang }
    dOptions.update({ "dOpt"+k: v  for k, v in lOpt })
    return dOptions, dOptPriority


def printBookmark (nLevel, sComment, nLine):
    "print bookmark within the rules file"
    print("  {:>6}:  {}".format(nLine, "  " * nLevel + sComment))


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

    print("> read rules file...")
    try:
        lRules = open(spLang + "/rules.grx", 'r', encoding="utf-8").readlines()
    except:
        print("Error. Rules file in project [" + sLang + "] not found.")
        exit()

    # removing comments, zeroing empty lines, creating definitions, storing tests, merging rule lines
    print("  parsing rules...")


    lRuleLine = []
    lTest = []
    lOpt = []
    bGraph = False
    lGraphRule = []

    for i, sLine in enumerate(lRules, 1):
        if sLine.startswith('#END'):
            # arbitrary end
            printBookmark(0, "BREAK BY #END", i)
            break
        elif sLine.startswith("#"):
            # comment
            pass
        elif sLine.startswith("DEF:"):

            # definition



            m = re.match("DEF: +([a-zA-Z_][a-zA-Z_0-9]*) +(.+)$", sLine.strip())
            if m:
                dDEF["{"+m.group(1)+"}"] = m.group(2)
            else:
                print("Error in definition: ", end="")
                print(sLine.strip())
        elif sLine.startswith("TEST:"):
            # test
            lTest.append("{:<8}".format(i) + "  " + sLine[5:].strip())
        elif sLine.startswith("TODO:"):
            # todo
            pass
        elif sLine.startswith(("OPTGROUP/", "OPTSOFTWARE:", "OPT/", "OPTLANG/", "OPTDEFAULTUILANG:", "OPTLABEL/", "OPTPRIORITY/")):
            # options
            lOpt.append(sLine)


        elif sLine.startswith("!!"):
            # bookmark
            m = re.match("!!+", sLine)
            nExMk = len(m.group(0))
            if sLine[nExMk:].strip():
                printBookmark(nExMk-2, sLine[nExMk:-3].strip(), i)
        # Graph rules
        elif sLine.startswith("@@@@GRAPH:"):
            # rules graph call
            m = re.match(r"@@@@GRAPH: *(\w+)", sLine.strip())
            if m:
                printBookmark(1, "____ GRAPH: " + m.group(1) + " ____", i)
                lRuleLine.append([i, "@@@@"+m.group(1)])
                bGraph = True
            lGraphRule.append([i, sLine])
            bGraph = True
        elif sLine.startswith("@@@@END_GRAPH"):
            #lGraphRule.append([i, sLine])
            bGraph = False
        elif re.match("@@@@ *$", sLine):
            pass
        elif bGraph:
            lGraphRule.append([i, sLine])
        # Regex rules
        elif re.match("[  \t]*$", sLine):
            # empty line
            pass
        elif sLine.startswith(("    ", "\t")):
            # rule (continuation)
            lRuleLine[-1][1] += " " + sLine.strip()
        else:
            # new rule
            lRuleLine.append([i, sLine.strip()])

    # generating options files
    print("  parsing options...")
    try:
        dOptions, dOptPriority = prepareOptions(lOpt)
    except:

                        lSentenceRulesJS.append(jsconv.pyRuleToJS(aRule, dJSREGEXES, sWORDLIMITLEFT))

    # creating file with all functions callable by rules
    print("  creating callables...")
    sPyCallables = "# generated code, do not edit\n"
    sJSCallables = "// generated code, do not edit\nconst oEvalFunc = {\n"
    for sFuncName, sReturn in lFUNCTIONS:

        if sFuncName.startswith("_c_"): # condition
            sParams = "s, sx, m, dTokenPos, sCountry, bCondMemo"
        elif sFuncName.startswith("_m_"): # message
            sParams = "s, m"
        elif sFuncName.startswith("_s_"): # suggestion
            sParams = "s, m"
        elif sFuncName.startswith("_p_"): # preprocessor
            sParams = "s, m"
        elif sFuncName.startswith("_d_"): # disambiguator
            sParams = "s, m, dTokenPos"
        else:
            print("# Unknown function type in [" + sFuncName + "]")
            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"

    displayStats(lParagraphRules, lSentenceRules)

    print("Unnamed rules: " + str(nRULEWITHOUTNAME))

    dVars = {   "callables": sPyCallables,
                "callablesJS": sJSCallables,
                "gctests": sGCTests,
                "gctestsJS": sGCTestsJS,
                "paragraph_rules": mergeRulesByOption(lParagraphRules),
                "sentence_rules": mergeRulesByOption(lSentenceRules),
                "paragraph_rules_JS": jsconv.writeRulesToJSArray(mergeRulesByOption(lParagraphRulesJS)),
                "sentence_rules_JS": jsconv.writeRulesToJSArray(mergeRulesByOption(lSentenceRulesJS)) }
    dVars.update(dOptions)

    # compile graph rules
    dVars2 = crg.make(lGraphRule, dDEF, sLang, bJavaScript)
    dVars.update(dVars2)

    return dVars

"""
Grammalecte: compile rules
Create a Direct Acyclic Rule Graphs (DARGs)
"""

import re
import traceback
import json

import darg


dACTIONS = {}
dFUNCTIONS = {}


def prepareFunction (s, bTokenValue=False):
    "convert simple rule syntax to a string of Python code"
    s = s.replace("__also__", "bCondMemo")
    s = s.replace("__else__", "not bCondMemo")
    s = re.sub(r"(morph|analyse|value|displayInfo)[(]\\(\d+)", 'g_\\1(lToken[\\2+nTokenOffset]', s)
    s = re.sub(r"(select|exclude|define)[(][\\](\d+)", 'g_\\1(lToken[\\2+nTokenOffset]', s)
    s = re.sub(r"(tag_before|tag_after)[(][\\](\d+)", 'g_\\1(lToken[\\2+nTokenOffset], dTags', s)
    s = re.sub(r"(switchGender|has(?:Mas|Fem)Form)[(]\\(\d+)", '\\1(lToken[\\2+nTokenOffset]["sValue"]', s)
    s = re.sub(r"(morph|analyse|value)\(>1", 'g_\\1(lToken[nLastToken+1]', s)                       # next token
    s = re.sub(r"(morph|analyse|value)\(<1", 'g_\\1(lToken[nTokenOffset]', s)                       # previous token
    s = re.sub(r"[\\](\d+)\.is(upper|lower|title)\(\)", 'lToken[\\1+nTokenOffset]["sValue"].is\\2()', s)
    s = re.sub(r"\bspell *[(]", '_oSpellChecker.isValid(', s)
    s = re.sub(r"\bbefore\(\s*", 'look(sSentence[:lToken[1+nTokenOffset]["nStart"]], ', s)          # before(s)
    s = re.sub(r"\bafter\(\s*", 'look(sSentence[lToken[nLastToken]["nEnd"]:], ', s)                 # after(s)
    s = re.sub(r"\bbefore0\(\s*", 'look(sSentence0[:lToken[1+nTokenOffset]["nStart"]], ', s)        # before0(s)
    s = re.sub(r"\bafter0\(\s*", 'look(sSentence[lToken[nLastToken]["nEnd"]:], ', s)                # after0(s)
    if bTokenValue:
        # token values are used as parameter
        s = re.sub(r"[\\](\d+)", 'lToken[\\1+nTokenOffset]["sValue"]', s)
    else:
        # tokens used as parameter
        s = re.sub(r"[\\](\d+)", 'lToken[\\1+nTokenOffset]', s)
    return s


def genTokenLines (sTokenLine, dDef):
    "tokenize a string and return a list of lines of tokens"
    lToken = sTokenLine.split()
    lTokenLines = None
    for sToken in lToken:
        # optional token?
        bNullPossible = sToken.startswith("?") and sToken.endswith("¿")
        if bNullPossible:
            sToken = sToken[1:-1]
        # token with definition?
        if sToken.startswith("({") and sToken.endswith("})") and sToken[1:-1] in dDef:
            sToken = "(" + dDef[sToken[1:-1]] + ")"
        elif sToken.startswith("{") and sToken.endswith("}") and sToken in dDef:
            sToken = dDef[sToken]
        if ( (sToken.startswith("[") and sToken.endswith("]")) or (sToken.startswith("([") and sToken.endswith("])")) ):
            # multiple token
            bSelectedGroup = sToken.startswith("(") and sToken.endswith(")")
            if bSelectedGroup:
                sToken = sToken[1:-1]
            lNewToken = sToken[1:-1].split("|")
            if not lTokenLines:
                lTokenLines = [ [s]  for s  in lNewToken ]
                if bNullPossible:
                    lTokenLines.extend([ []  for i  in range(len(lNewToken)+1) ])
            else:
                lNewTemp = []
                if bNullPossible:
                    for aRule in lTokenLines:
                        for sElem in lNewToken:
                            aNewRule = list(aRule)
                            aNewRule.append(sElem)
                            lNewTemp.append(aNewRule)
                else:
                    sElem1 = lNewToken.pop(0)
                    for aRule in lTokenLines:
                        for sElem in lNewToken:
                            aNewRule = list(aRule)
                            aNewRule.append("(" + sElem + ")"  if bSelectedGroup  else sElem)
                            lNewTemp.append(aNewRule)
                        aRule.append("(" + sElem1 + ")"  if bSelectedGroup  else sElem1)
                lTokenLines.extend(lNewTemp)
        else:
            # simple token
            if not lTokenLines:
                lTokenLines = [[sToken], []]  if bNullPossible  else [[sToken]]
            else:
                if bNullPossible:
                    lNewTemp = []
                    for aRule in lTokenLines:
                        lNew = list(aRule)
                        lNew.append(sToken)
                        lNewTemp.append(lNew)
                    lTokenLines.extend(lNewTemp)
                else:
                    for aRule in lTokenLines:
                        aRule.append(sToken)
    for aRule in lTokenLines:
        yield aRule


def createRule (iLine, sRuleName, sTokenLine, iActionBlock, sActions, nPriority, dDef):
    "generator: create rule as list"
    # print(iLine, "//", sRuleName, "//", sTokenLine, "//", sActions, "//", nPriority)
    for lToken in genTokenLines(sTokenLine, dDef):
        # Calculate positions
        dPos = {}   # key: iGroup, value: iToken
        iGroup = 0
        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, sAction in enumerate(sActions.split(" <<- ")):
            sAction = sAction.strip()
            if sAction:
                sActionId = sRuleName + "__b" + str(iActionBlock) + "_a" + str(iAction) + "_" + str(len(lToken))
                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 (sText, dPos):
    "change group reference in <sText> with values in <dPos>"
    for i in range(len(dPos), 0, -1):
        sText = sText.replace("\\"+str(i), "\\"+str(dPos[i]))
    return sText


def checkTokenNumbers (sText, sActionId, nToken):
    "check if token references in <sText> greater than <nToken> (debugging)"
    for x in re.finditer(r"\\(\d+)", sText):
        if int(x.group(1)) > nToken:
            print("# Error in token index at line " + sActionId + " ("+str(nToken)+" tokens only)")
            print(sText)


def checkIfThereIsCode (sText, sActionId):
    "check if there is code in <sText> (debugging)"
    if re.search("[.]\\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, nToken, dPos):
    "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()
    # valid action?
    m = re.search("(?P<action>[-~=/])(?P<start>\\d+|)(?P<end>:\\d+|)>> ", sAction)
    if not m:
        print(" # Error. No action found at: ", sActionId)
        print("   ==", sAction, "==")
        return None
    # Condition
    sCondition = sAction[:m.start()].strip()
    if sCondition:
        sCondition = prepareFunction(sCondition)
        sCondition = changeReferenceToken(sCondition, dPos)
        dFUNCTIONS["_g_c_"+sActionId] = sCondition
        sCondition = "_g_c_"+sActionId
    else:
        sCondition = ""
    # Action
    cAction = m.group("action")
    sAction = sAction[m.end():].strip()
    sAction = changeReferenceToken(sAction, dPos)
    if not m.group("start"):
        iStartAction = 1
        iEndAction = 0
    else:
        iStartAction = int(m.group("start"))
        iEndAction = int(m.group("end")[1:])  if m.group("end")  else iStartAction
    if dPos and m.group("start"):
        try:
            iStartAction = dPos[iStartAction]
            if iEndAction:
                iEndAction = dPos[iEndAction]
        except:
            print("# Error. Wrong groups in: " + sActionId)
            print("  iStartAction:", iStartAction, "iEndAction:", iEndAction)
            print(" ", dPos)

    if cAction == "-":
        ## error
        iMsg = sAction.find(" # ")
        if iMsg == -1:
            sMsg = "# Error. Error message not found."
            sURL = ""
            print(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 = prepareFunction(sMsg[1:], True)
                dFUNCTIONS["g_m_"+sActionId] = sMsg
                sMsg = "=g_m_"+sActionId
            else:
                checkIfThereIsCode(sMsg, sActionId)

    # checking consistancy
    checkTokenNumbers(sAction, sActionId, nToken)

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

    if not sAction:
        print("# Error in action at line " + sActionId + ":  This action is empty.")

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

    if cAction == "-":
        ## error detected --> suggestion
        if sAction[0:1] == "=":
            sAction = prepareFunction(sAction, True)
            dFUNCTIONS["_g_s_"+sActionId] = sAction[1:]
            sAction = "=_g_s_"+sActionId
        elif sAction.startswith('"') and sAction.endswith('"'):
            sAction = sAction[1:-1]
        if not sMsg:
            print("# Error in action at line " + sActionId + ":  The message is empty.")
        return [sOption, sCondition, cAction, sAction, iStartAction, iEndAction, nPriority, sMsg, sURL]
    elif cAction == "~":
        ## text processor
        if sAction[0:1] == "=":
            sAction = prepareFunction(sAction, True)
            dFUNCTIONS["_g_p_"+sActionId] = sAction[1:]
            sAction = "=_g_p_"+sActionId
        elif sAction.startswith('"') and sAction.endswith('"'):
            sAction = sAction[1:-1]
        return [sOption, sCondition, cAction, sAction, iStartAction, iEndAction]
    elif cAction == "/":
        ## tags
        return [sOption, sCondition, cAction, sAction, iStartAction, iEndAction]
    elif cAction == "=":
        ## disambiguator
        if sAction[0:1] == "=":
            sAction = sAction[1:]
        if "define" in sAction and not re.search(r"define\(\\\d+ *, *\[.*\] *\)", sAction):
            print("# Error in action at line " + sActionId + ": second argument for <define> must be a list of strings")
        sAction = prepareFunction(sAction)
        dFUNCTIONS["_g_d_"+sActionId] = sAction
        sAction = "_g_d_"+sActionId
        return [sOption, sCondition, cAction, sAction]
    else:
        print("# Unknown action at line " + sActionId)
        return None


def make (lRule, dDef, sLang, bJavaScript):
    "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 = []
    sActions = ""
    nPriority = 4
    dAllGraph = {}
    sGraphName = ""
    iActionBlock = 0

    for i, sLine in lRule:
        sLine = sLine.rstrip()
        if "\t" in sLine:
            # tabulation not allowed
            print("Error. Tabulation at line: ", i)
            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. Group name " + sGraphName + " already exists.")
                    exit()
                dAllGraph[sGraphName] = []
            else:
                print("Error. Graph name not found at line", i)
                exit()
        elif sLine.startswith("__") and sLine.endswith("__"):
            # new rule group
            m = re.match("__(\\w+)(!\\d|)__", sLine)
            if m:
                sRuleName = m.group(1)
                iActionBlock = 1
                nPriority = int(m.group(2)[1:]) if m.group(2)  else 4
            else:
                print("Error at rule group: ", sLine, " -- line:", i)
                break
        elif re.search("^    +<<- ", sLine) or sLine.startswith("        ") \
                or re.search("^    +#", sLine) or re.search(r"^    [-~=>/](?:\d(?::\d+|)|)>> ", sLine) :
            # actions
            sActions += " " + sLine.strip()
        elif re.match("[  ]*$", sLine):
            # empty line to end merging
            if not lTokenLine:
                continue
            if not sActions:
                print("Error. No action found at line:", i)
                exit()
            if not sGraphName:
                print("Error. All rules must belong to a named graph. Line: ", i)
                exit()
            for j, sTokenLine in lTokenLine:
                dAllGraph[sGraphName].append((j, sRuleName, sTokenLine, iActionBlock, sActions, nPriority))
            lTokenLine.clear()
            sActions = ""
            iActionBlock += 1
        elif sLine.startswith(("    ")):
            # tokens
            lTokenLine.append([i, sLine.strip()])
        else:
            print("Unknown line:")
            print(sLine)

    # processing rules
    print("  preparing rules...")
    for sGraphName, lRuleLine in dAllGraph.items():
        lPreparedRule = []
        for i, sRuleGroup, sTokenLine, iActionBlock, sActions, nPriority in lRuleLine:
            for lRule in createRule(i, sRuleGroup, sTokenLine, iActionBlock, sActions, nPriority, dDef):
                lPreparedRule.append(lRule)
        # Graph creation
        oDARG = darg.DARG(lPreparedRule, sLang)
        dAllGraph[sGraphName] = oDARG.createGraph()
        # Debugging
        #print("\nGRAPH:", sGraphName)
        #for e in lPreparedRule:
        #    if e[-2] == "##4239":
        #        print(e)
        #    print(e)
        #for k, v in dAllGraph[sGraphName].items():
        #    print(k, "\t", v)

    # creating file with all functions callable by rules
    print("  creating callables...")
    sPyCallables = "# generated code, do not edit\n"
    #sJSCallables = "// generated code, do not edit\nconst oEvalFunc = {\n"
    for sFuncName, sReturn in dFUNCTIONS.items():
        if sFuncName.startswith("_g_c_"): # condition
            sParams = "lToken, nTokenOffset, nLastToken, sCountry, bCondMemo, dTags, sSentence, sSentence0"
        elif sFuncName.startswith("g_m_"): # message
            sParams = "lToken, nTokenOffset"
        elif sFuncName.startswith("_g_s_"): # suggestion
            sParams = "lToken, nTokenOffset"
        elif sFuncName.startswith("_g_p_"): # preprocessor
            sParams = "lToken, nTokenOffset"
        elif sFuncName.startswith("_g_d_"): # disambiguator
            sParams = "lToken, nTokenOffset"
        else:
            print("# Unknown function type in [" + sFuncName + "]")
            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"

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

    # Result
    return {
        "graph_callables": sPyCallables,
        "rules_graphs": dAllGraph,
        "rules_actions": dACTIONS
    }

"""
Convert Python code and regexes to JavaScript code
"""

import copy
import re
import json


def py2js (sCode):

        sRegex = sRegex + "i"
    if not lNegLookBeforeRegex:
        lNegLookBeforeRegex = None
    return (sRegex, lNegLookBeforeRegex)


def pyRuleToJS (lRule, dJSREGEXES, sWORDLIMITLEFT):
    "modify Python rules -> JS rules"
    lRuleJS = copy.deepcopy(lRule)
    # graph rules
    if lRuleJS[0] == "@@@@":
        return lRuleJS
    del lRule[-1] # tGroups positioning codes are useless for Python
    # error messages
    for aAction in lRuleJS[6]:
        if aAction[1] == "-":
            aAction[2] = aAction[2].replace(" ", " ") # nbsp --> nnbsp
            aAction[4] = aAction[4].replace("« ", "« ").replace(" »", " »").replace(" :", " :").replace(" :", " :")
    # js regexes
    lRuleJS[1], lNegLookBehindRegex = regex2js(dJSREGEXES.get(lRuleJS[3], lRuleJS[1]), sWORDLIMITLEFT)
    lRuleJS.append(lNegLookBehindRegex)
    return lRuleJS


def writeRulesToJSArray (lRules):
    "create rules as a string of arrays (to be bundled in a JSON string)"
    sArray = "[\n"
    for sOption, aRuleGroup in lRules:
        if sOption != "@@@@":
            sArray += '  ["' + sOption + '", [\n'  if sOption  else  "  [false, [\n"
            for sRegex, bCaseInsensitive, sLineId, sRuleId, nPriority, lActions, aGroups, aNegLookBehindRegex in aRuleGroup:
                sArray += '    [' + sRegex + ", "
                sArray += "true, " if bCaseInsensitive  else "false, "
                sArray += '"' + sLineId + '", '
                sArray += '"' + sRuleId + '", '
                sArray += str(nPriority) + ", "
                sArray += json.dumps(lActions, ensure_ascii=False) + ", "
                sArray += json.dumps(aGroups, ensure_ascii=False) + ", "
                sArray += json.dumps(aNegLookBehindRegex, ensure_ascii=False) + "],\n"
            sArray += "  ]],\n"
        else:
            sArray += '  ["' + sOption + '", [\n'
            for sGraphName, sLineId in aRuleGroup:
                sArray += '    ["' + sGraphName + '", "' + sLineId + '"],\n"'
            sArray += "  ]],\n"
    sArray += "]"
    return sArray


def groupsPositioningCodeToList (sGroupsPositioningCode):
    "convert <sGroupsPositioningCode> to a list of codes (numbers or strings)"
    if not sGroupsPositioningCode:
        return None
    return [ int(sCode)  if sCode.isdigit() or (sCode[0:1] == "-" and sCode[1:].isdigit())  else sCode \
             for sCode in sGroupsPositioningCode.split(",") ]

#!python3

"""
RULE GRAPH BUILDER
"""

# by Olivier R.
# License: MPL 2

import re
import traceback


from graphspell.progressbar import ProgressBar


class DARG:
    """DIRECT ACYCLIC RULE GRAPH"""
    # This code is inspired from Steve Hanov’s DAWG, 2011. (http://stevehanov.ca/blog/index.php?id=115)

    def __init__ (self, lRule, sLangCode):
        print("===== Direct Acyclic Rule Graph - Minimal Acyclic Finite State Automaton =====")

        # Preparing DARG
        print(" > Preparing list of tokens")
        self.sLangCode = sLangCode
        self.nRule = len(lRule)
        self.aPreviousRule = []
        Node.resetNextId()
        self.oRoot = Node()
        self.lUncheckedNodes = []  # list of nodes that have not been checked for duplication.
        self.lMinimizedNodes = {}  # list of unique nodes that have been checked for duplication.
        self.nNode = 0
        self.nArc = 0

        # build
        lRule.sort()
        oProgBar = ProgressBar(0, len(lRule))
        for aRule in lRule:
            self.insert(aRule)
            oProgBar.increment(1)
        oProgBar.done()
        self.finish()
        self.countNodes()
        self.countArcs()
        self.displayInfo()

    # BUILD DARG
    def insert (self, aRule):
        "insert a new rule (tokens must be inserted in order)"
        if aRule < self.aPreviousRule:
            exit("# Error: tokens must be inserted in order.")

        # find common prefix between word and previous word
        nCommonPrefix = 0
        for i in range(min(len(aRule), len(self.aPreviousRule))):
            if aRule[i] != self.aPreviousRule[i]:
                break
            nCommonPrefix += 1

        # Check the lUncheckedNodes for redundant nodes, proceeding from last
        # one down to the common prefix size. Then truncate the list at that point.
        self._minimize(nCommonPrefix)

        # add the suffix, starting from the correct node mid-way through the graph
        if len(self.lUncheckedNodes) == 0:
            oNode = self.oRoot
        else:
            oNode = self.lUncheckedNodes[-1][2]

        iToken = nCommonPrefix
        for sToken in aRule[nCommonPrefix:]:
            oNextNode = Node()
            oNode.dArcs[sToken] = oNextNode
            self.lUncheckedNodes.append((oNode, sToken, oNextNode))
            if iToken == (len(aRule) - 2):
                oNode.bFinal = True
            iToken += 1
            oNode = oNextNode
        oNode.bFinal = True
        self.aPreviousRule = aRule

    def finish (self):
        "minimize unchecked nodes"
        self._minimize(0)

    def _minimize (self, downTo):
        # proceed from the leaf up to a certain point
        for i in range( len(self.lUncheckedNodes)-1, downTo-1, -1 ):
            oNode, sToken, oChildNode = self.lUncheckedNodes[i]
            if oChildNode in self.lMinimizedNodes:
                # replace the child with the previously encountered one
                oNode.dArcs[sToken] = self.lMinimizedNodes[oChildNode]
            else:
                # add the state to the minimized nodes.
                self.lMinimizedNodes[oChildNode] = oChildNode
            self.lUncheckedNodes.pop()

    def countNodes (self):
        "count nodes within the whole graph"
        self.nNode = len(self.lMinimizedNodes)

    def countArcs (self):
        "count arcs within the whole graph"
        self.nArc = 0
        for oNode in self.lMinimizedNodes:
            self.nArc += len(oNode.dArcs)

    def displayInfo (self):
        "display informations about the rule graph"
        print(" * {:<12} {:>16,}".format("Rules:", self.nRule))
        print(" * {:<12} {:>16,}".format("Nodes:", self.nNode))
        print(" * {:<12} {:>16,}".format("Arcs:", self.nArc))

    def createGraph (self):
        "create the graph as a dictionary"
        dGraph = { 0: self.oRoot.getNodeAsDict() }
        for oNode in self.lMinimizedNodes:
            sHashId = oNode.__hash__()
            if sHashId not in dGraph:
                dGraph[sHashId] = oNode.getNodeAsDict()
            else:
                print("Error. Double node… same id: ", sHashId)
                print(str(oNode.getNodeAsDict()))
        dGraph = self._rewriteKeysOfDARG(dGraph)
        self._checkRegexes(dGraph)
        return dGraph

    def _rewriteKeysOfDARG (self, dGraph):
        "keys of DARG are long numbers (hashes): this function replace these hashes with smaller numbers (to reduce storing size)"
        # create translation dictionary
        dKeyTrans = {}
        for i, nKey in enumerate(dGraph):
            dKeyTrans[nKey] = i
        # replace keys
        dNewGraph = {}
        for nKey, dVal in dGraph.items():
            dNewGraph[dKeyTrans[nKey]] = dVal
        for nKey, dVal in dGraph.items():
            for sArc, val in dVal.items():
                if type(val) is int:
                    dVal[sArc] = dKeyTrans[val]
                else:
                    for sArc, nKey in val.items():
                        val[sArc] = dKeyTrans[nKey]
        return dNewGraph

    def _checkRegexes (self, dGraph):
        "check validity of regexes"
        aRegex = set()
        for nKey, dVal in dGraph.items():
            if "<re_value>" in dVal:
                for sRegex in dVal["<re_value>"]:
                    if sRegex not in aRegex:
                        self._checkRegex(sRegex)
                        aRegex.add(sRegex)
            if "<re_morph>" in dVal:
                for sRegex in dVal["<re_morph>"]:
                    if sRegex not in aRegex:
                        self._checkRegex(sRegex)
                        aRegex.add(sRegex)
        aRegex.clear()

    def _checkRegex (self, sRegex):
        #print(sRegex)
        if "¬" in sRegex:
            sPattern, sNegPattern = sRegex.split("¬")
            try:
                if not sNegPattern:
                    print("# Warning! Empty negpattern:", sRegex)
                re.compile(sPattern)
                re.compile(sNegPattern)
            except:
                print("# Error. Wrong regex:", sRegex)
                traceback.print_exc()
                exit()
        else:
            try:
                if not sRegex:
                    print("# Warning! Empty pattern:", sRegex)
                re.compile(sRegex)
            except:
                print("# Error. Wrong regex:", sRegex)
                traceback.print_exc()
                exit()


class Node:
    """Node of the rule graph"""

    NextId = 0

    def __init__ (self):
        self.i = Node.NextId
        Node.NextId += 1
        self.bFinal = False
        self.dArcs = {}          # key: arc value; value: a node

    @classmethod
    def resetNextId (cls):
        "reset to 0 the node counter"
        cls.NextId = 0

    def __str__ (self):
        # Caution! this function is used for hashing and comparison!
        cFinal = "1"  if self.bFinal  else "0"
        l = [cFinal]
        for (key, oNode) in self.dArcs.items():
            l.append(str(key))
            l.append(str(oNode.i))
        return "_".join(l)

    def __hash__ (self):
        # Used as a key in a python dictionary.
        return self.__str__().__hash__()

    def __eq__ (self, other):
        # Used as a key in a python dictionary.
        # Nodes are equivalent if they have identical arcs, and each identical arc leads to identical states.
        return self.__str__() == other.__str__()

    def getNodeAsDict (self):
        "returns the node as a dictionary structure"
        dNode = {}
        dReValue = {}
        dReMorph = {}
        dRule = {}
        dLemma = {}
        dMeta = {}
        for sArc, oNode in self.dArcs.items():
            if sArc.startswith("@") and len(sArc) > 1:
                dReMorph[sArc[1:]] = oNode.__hash__()
            elif sArc.startswith("~") and len(sArc) > 1:
                dReValue[sArc[1:]] = oNode.__hash__()
            elif sArc.startswith(">") and len(sArc) > 1:
                dLemma[sArc[1:]] = oNode.__hash__()
            elif sArc.startswith("*") and len(sArc) > 1:
                dMeta[sArc[1:]] = oNode.__hash__()
            elif sArc.startswith("##"):
                dRule[sArc[1:]] = oNode.__hash__()
            else:
                dNode[sArc] = oNode.__hash__()
        if dReValue:
            dNode["<re_value>"] = dReValue
        if dReMorph:
            dNode["<re_morph>"] = dReMorph
        if dLemma:
            dNode["<lemmas>"] = dLemma
        if dMeta:
            dNode["<meta>"] = dMeta
        if dRule:
            dNode["<rules>"] = dRule
        #if self.bFinal:
        #    dNode["<final>"] = 1
        return dNode

// data
let _sAppContext = "";                                  // what software is running
let _dOptions = null;
let _aIgnoredRules = new Set();
let _oSpellChecker = null;



var gc_engine = {

    //// Informations

    lang: "${lang}",

                var spellchecker = require("resource://grammalecte/graphspell/spellchecker.js");
                _oSpellChecker = new spellchecker.SpellChecker("${lang}", "", "${dic_main_filename_js}", "${dic_extended_filename_js}", "${dic_community_filename_js}", "${dic_personal_filename_js}");
            } else {
                _oSpellChecker = new SpellChecker("${lang}", sPath, "${dic_main_filename_js}", "${dic_extended_filename_js}", "${dic_community_filename_js}", "${dic_personal_filename_js}");
            }
            _sAppContext = sContext;
            _dOptions = gc_options.getOptions(sContext).gl_shallowCopy();     // duplication necessary, to be able to reset to default
            _oSpellChecker.activateStorage();
        }
        catch (e) {
            helpers.logerror(e);
        }
    },

    getSpellChecker: function () {

function displayInfo (dDA, aWord) {
    // for debugging: info of word
    if (!aWord) {
        helpers.echo("> nothing to find");
        return true;
    }
    let lMorph = _oSpellChecker.getMorph(aWord[1]);
    if (lMorph.length === 0) {
        helpers.echo("> not in dictionary");
        return true;
    }
    if (dDA.has(aWord[0])) {
        helpers.echo("DA: " + dDA.get(aWord[0]));
    }
    helpers.echo("FSA: " + lMorph);
    return true;
}








function morph (dDA, aWord, sPattern, bStrict=true, bNoWord=false) {
    // analyse a tuple (position, word), return true if sPattern in morphologies (disambiguation on)
    if (!aWord) {
        //helpers.echo("morph: noword, returns " + bNoWord);
        return bNoWord;
    }
    //helpers.echo("aWord: "+aWord.toString());



    let lMorph = dDA.has(aWord[0]) ? dDA.get(aWord[0]) : _oSpellChecker.getMorph(aWord[1]);
    //helpers.echo("lMorph: "+lMorph.toString());
    if (lMorph.length === 0) {
        return false;
    }
    //helpers.echo("***");
    if (bStrict) {
        return lMorph.every(s  =>  (s.search(sPattern) !== -1));
    }
    return lMorph.some(s  =>  (s.search(sPattern) !== -1));
}

function morphex (dDA, aWord, sPattern, sNegPattern, bNoWord=false) {
    // analyse a tuple (position, word), returns true if not sNegPattern in word morphologies and sPattern in word morphologies (disambiguation on)
    if (!aWord) {
        //helpers.echo("morph: noword, returns " + bNoWord);
        return bNoWord;
    }
    //helpers.echo("aWord: "+aWord.toString());



    let lMorph = dDA.has(aWord[0]) ? dDA.get(aWord[0]) : _oSpellChecker.getMorph(aWord[1]);
    //helpers.echo("lMorph: "+lMorph.toString());
    if (lMorph.length === 0) {
        return false;
    }
    //helpers.echo("***");
    // check negative condition
    if (lMorph.some(s  =>  (s.search(sNegPattern) !== -1))) {
        return false;
    }
    // search sPattern
    return lMorph.some(s  =>  (s.search(sPattern) !== -1));
}

function analyse (sWord, sPattern, bStrict=true) {
    // analyse a word, return true if sPattern in morphologies (disambiguation off)
    let lMorph = _oSpellChecker.getMorph(sWord);
    if (lMorph.length === 0) {
        return false;
    }
    if (bStrict) {
        return lMorph.every(s  =>  (s.search(sPattern) !== -1));
    }
    return lMorph.some(s  =>  (s.search(sPattern) !== -1));
}

function analysex (sWord, sPattern, sNegPattern) {
    // analyse a word, returns True if not sNegPattern in word morphologies and sPattern in word morphologies (disambiguation off)
    let lMorph = _oSpellChecker.getMorph(sWord);
    if (lMorph.length === 0) {
        return false;
    }
    // check negative condition
    if (lMorph.some(s  =>  (s.search(sNegPattern) !== -1))) {
        return false;
    }
    // search sPattern
    return lMorph.some(s  =>  (s.search(sPattern) !== -1));











}


//// functions to get text outside pattern scope

// warning: check compile_rules.py to understand how it works

function select (dDA, nPos, sWord, sPattern, lDefault=null) {
    if (!sWord) {
        return true;
    }
    if (dDA.has(nPos)) {
        return true;
    }
    let lMorph = _oSpellChecker.getMorph(sWord);


    if (lMorph.length === 0  ||  lMorph.length === 1) {
        return true;
    }
    let lSelect = lMorph.filter( sMorph => sMorph.search(sPattern) !== -1 );
    if (lSelect.length > 0) {
        if (lSelect.length != lMorph.length) {
            dDA.set(nPos, lSelect);
        }
    } else if (lDefault) {
        dDA.set(nPos, lDefaul);
    }
    return true;
}

function exclude (dDA, nPos, sWord, sPattern, lDefault=null) {
    if (!sWord) {
        return true;
    }
    if (dDA.has(nPos)) {
        return true;
    }
    let lMorph = _oSpellChecker.getMorph(sWord);


    if (lMorph.length === 0  ||  lMorph.length === 1) {
        return true;
    }
    let lSelect = lMorph.filter( sMorph => sMorph.search(sPattern) === -1 );
    if (lSelect.length > 0) {
        if (lSelect.length != lMorph.length) {
            dDA.set(nPos, lSelect);
        }
    } else if (lDefault) {
        dDA.set(nPos, lDefault);
    }
    return true;
}

"""
Grammar checker
"""

from .grammar_checker import *

"""
Grammalecte, grammar checker

"""

import importlib
import json

from . import text


class GrammarChecker:
    "GrammarChecker: Wrapper for the grammar checker engine"

    def __init__ (self, sLangCode, sContext="Python"):
        self.sLangCode = sLangCode
        # Grammar checker engine
        self.gce = importlib.import_module("."+sLangCode, "grammalecte")
        self.gce.load(sContext)
        # Spell checker
        self.oSpellChecker = self.gce.getSpellChecker()
        # Lexicographer
        self.oLexicographer = None
        # Text formatter
        self.oTextFormatter = None

    def getGCEngine (self):
        "return the grammar checker object"
        return self.gce

    def getSpellChecker (self):
        "return the spell checker object"
        return self.oSpellChecker

    def getTextFormatter (self):
        "load and return the text formatter"
        if self.oTextFormatter is None:
            tf = importlib.import_module("."+self.sLangCode+".textformatter", "grammalecte")
            self.oTextFormatter = tf.TextFormatter()
        return self.oTextFormatter

    def getLexicographer (self):
        "load and return the lexicographer"
        if self.oLexicographer is None:
            lxg = importlib.import_module("."+self.sLangCode+".lexicographe", "grammalecte")
            self.oLexicographer = lxg.Lexicographe(self.oSpellChecker)
        return self.oLexicographer

    def displayGCOptions (self):
        "display the grammar checker options"
        self.gce.displayOptions()

    def getParagraphErrors (self, sText, dOptions=None, bContext=False, bSpellSugg=False, bDebug=False):
        "returns a tuple: (grammar errors, spelling errors)"
        aGrammErrs = self.gce.parse(sText, "FR", bDebug=bDebug, dOptions=dOptions, bContext=bContext)
        aSpellErrs = self.oSpellChecker.parseParagraph(sText, bSpellSugg)
        return aGrammErrs, aSpellErrs

    def generateText (self, sText, bEmptyIfNoErrors=False, bSpellSugg=False, nWidth=100, bDebug=False):
        "[todo]"
        pass

    def generateTextAsJSON (self, sText, bContext=False, bEmptyIfNoErrors=False, bSpellSugg=False, bReturnText=False, bDebug=False):
        "[todo]"
        pass

    def generateParagraph (self, sText, dOptions=None, bEmptyIfNoErrors=False, bSpellSugg=False, nWidth=100, bDebug=False):
        "parse text and return a readable text with underline errors"
        aGrammErrs, aSpellErrs = self.getParagraphErrors(sText, dOptions, False, bSpellSugg, bDebug)
        if bEmptyIfNoErrors and not aGrammErrs and not aSpellErrs:
            return ""
        return text.generateParagraph(sText, aGrammErrs, aSpellErrs, nWidth)

    def generateParagraphAsJSON (self, iIndex, sText, dOptions=None, bContext=False, bEmptyIfNoErrors=False, bSpellSugg=False, bReturnText=False, lLineSet=None, bDebug=False):
        "parse text and return errors as a JSON string"
        aGrammErrs, aSpellErrs = self.getParagraphErrors(sText, dOptions, bContext, bSpellSugg, bDebug)
        aGrammErrs = list(aGrammErrs)
        if bEmptyIfNoErrors and not aGrammErrs and not aSpellErrs:
            return ""
        if lLineSet:
            aGrammErrs, aSpellErrs = text.convertToXY(aGrammErrs, aSpellErrs, lLineSet)
            return json.dumps({ "lGrammarErrors": aGrammErrs, "lSpellingErrors": aSpellErrs }, ensure_ascii=False)
        if bReturnText:
            return json.dumps({ "iParagraph": iIndex, "sText": sText, "lGrammarErrors": aGrammErrs, "lSpellingErrors": aSpellErrs }, ensure_ascii=False)
        return json.dumps({ "iParagraph": iIndex, "lGrammarErrors": aGrammErrs, "lSpellingErrors": aSpellErrs }, ensure_ascii=False)

"""
Grammalecte
Grammar checker engine
"""

import re
import sys
import os
import traceback
#import unicodedata
from itertools import chain

from ..graphspell.spellchecker import SpellChecker
from ..graphspell.echo import echo
from . import gc_options

from ..graphspell.tokenizer import Tokenizer
from .gc_rules_graph import dAllGraph, dRule

try:
    # LibreOffice / OpenOffice
    from com.sun.star.linguistic2 import SingleProofreadingError
    from com.sun.star.text.TextMarkupType import PROOFREADING
    from com.sun.star.beans import PropertyValue
    #import lightproof_handler_${implname} as opt
    _bWriterError = True
except ImportError:
    _bWriterError = False


__all__ = [ "lang", "locales", "pkg", "name", "version", "author", \
            "load", "parse", "getSpellChecker", \
            "setOption", "setOptions", "getOptions", "getDefaultOptions", "getOptionsLabels", "resetOptions", "displayOptions", \
            "ignoreRule", "resetIgnoreRules", "reactivateRule", "listRules", "displayRules" ]

author = "${author}"

_rules = None                               # module gc_rules

# data
_sAppContext = ""                           # what software is running
_dOptions = None

_oSpellChecker = None









_oTokenizer = None


















































_aIgnoredRules = set()


















































































































































































#### Initialization












def load (sContext="Python"):
    "initialization of the grammar checker"
    global _oSpellChecker
    global _sAppContext
    global _dOptions
    global _oTokenizer
    try:
        _oSpellChecker = SpellChecker("${lang}", "${dic_main_filename_py}", "${dic_extended_filename_py}", "${dic_community_filename_py}", "${dic_personal_filename_py}")
        _sAppContext = sContext
        _dOptions = dict(gc_options.getOptions(sContext))   # duplication necessary, to be able to reset to default
        _oTokenizer = _oSpellChecker.getTokenizer()
        _oSpellChecker.activateStorage()
    except:
        traceback.print_exc()








































def _getRules (bParagraph):
    try:
        if not bParagraph:
            return _rules.lSentenceRules
        return _rules.lParagraphRules
    except:
        _loadRules()
    if not bParagraph:
        return _rules.lSentenceRules
    return _rules.lParagraphRules


def _loadRules ():
    from . import gc_rules
    global _rules
    _rules = gc_rules
    # compile rules regex
    for sOption, lRuleGroup in chain(_rules.lParagraphRules, _rules.lSentenceRules):
        if sOption != "@@@@":
            for aRule in lRuleGroup:
                try:
                    aRule[0] = re.compile(aRule[0])
                except:
                    echo("Bad regular expression in # " + str(aRule[2]))
                    aRule[0] = "(?i)<Grammalecte>"


#### Parsing

_zEndOfSentence = re.compile(r'([.?!:;…][ .?!… »”")]*|.$)')
_zBeginOfParagraph = re.compile(r"^\W*")
_zEndOfParagraph = re.compile(r"\W*$")

def _getSentenceBoundaries (sText):
    iStart = _zBeginOfParagraph.match(sText).end()
    for m in _zEndOfSentence.finditer(sText):
        yield (iStart, m.end())
        iStart = m.end()


def parse (sText, sCountry="${country_default}", bDebug=False, dOptions=None, bContext=False):
    "analyses the paragraph sText and returns list of errors"
    #sText = unicodedata.normalize("NFC", sText)
    aErrors = None
    sRealText = sText
    dPriority = {}  # Key = position; value = priority
    dOpt = _dOptions  if not dOptions  else dOptions
    bShowRuleId = option('idrule')

    # parse paragraph
    try:
        sNew, aErrors = _proofread(None, sText, sRealText, 0, True, dPriority, sCountry, dOpt, bShowRuleId, bDebug, bContext)
        if sNew:
            sText = sNew
    except:
        raise

    # cleanup
    if " " in sText:
        sText = sText.replace(" ", ' ') # nbsp
    if " " in sText:
        sText = sText.replace(" ", ' ') # nnbsp
    if "'" in sText:
        sText = sText.replace("'", "’")
    if "‑" in sText:
        sText = sText.replace("‑", "-") # nobreakdash

    # parse sentences
    for iStart, iEnd in _getSentenceBoundaries(sText):
        if 4 < (iEnd - iStart) < 2000:
            try:
                oSentence = TokenSentence(sText[iStart:iEnd], sRealText[iStart:iEnd], iStart)
                _, errs = _proofread(oSentence, sText[iStart:iEnd], sRealText[iStart:iEnd], iStart, False, dPriority, sCountry, dOpt, bShowRuleId, bDebug, bContext)
                aErrors.update(errs)
            except:
                raise
    return aErrors.values() # this is a view (iterable)


def _proofread (oSentence, s, sx, nOffset, bParagraph, dPriority, sCountry, dOptions, bShowRuleId, bDebug, bContext):
    dErrs = {}
    bParagraphChange = False
    bSentenceChange = False
    dTokenPos = oSentence.dTokenPos if oSentence else {}
    for sOption, lRuleGroup in _getRules(bParagraph):
        if sOption == "@@@@":
            # graph rules
            if not bParagraph and bSentenceChange:
                oSentence.update(s, bDebug)
                bSentenceChange = False
            for sGraphName, sLineId in lRuleGroup:
                if bDebug:
                    print("\n>>>> GRAPH:", sGraphName, sLineId)
                bParagraphChange, s = oSentence.parse(dAllGraph[sGraphName], dPriority, sCountry, dOptions, bShowRuleId, bDebug, bContext)
                dErrs.update(oSentence.dError)
                dTokenPos = oSentence.dTokenPos
        elif not sOption or dOptions.get(sOption, False):
            # regex rules
            for zRegex, bUppercase, sLineId, sRuleId, nPriority, lActions in lRuleGroup:
                if sRuleId not in _aIgnoredRules:
                    for m in zRegex.finditer(s):
                        bCondMemo = None
                        for sFuncCond, cActionType, sWhat, *eAct in lActions:
                            # action in lActions: [ condition, action type, replacement/suggestion/action[, iGroup[, message, URL]] ]
                            try:
                                bCondMemo = not sFuncCond or globals()[sFuncCond](s, sx, m, dTokenPos, sCountry, bCondMemo)
                                if bCondMemo:
                                    if bDebug:
                                        print("RULE:", sLineId)
                                    if cActionType == "-":
                                        # grammar error
                                        nErrorStart = nOffset + m.start(eAct[0])
                                        if nErrorStart not in dErrs or nPriority > dPriority.get(nErrorStart, -1):
                                            dErrs[nErrorStart] = _createError(s, sx, sWhat, nOffset, m, eAct[0], sLineId, sRuleId, bUppercase, eAct[1], eAct[2], bShowRuleId, sOption, bContext)
                                            dPriority[nErrorStart] = nPriority
                                    elif cActionType == "~":
                                        # text processor
                                        s = _rewrite(s, sWhat, eAct[0], m, bUppercase)
                                        bParagraphChange = True
                                        bSentenceChange = True
                                        if bDebug:
                                            echo("~ " + s + "  -- " + m.group(eAct[0]) + "  # " + sLineId)
                                    elif cActionType == "=":
                                        # disambiguation
                                        if not bParagraph:
                                            globals()[sWhat](s, m, dTokenPos)
                                            if bDebug:
                                                echo("= " + m.group(0) + "  # " + sLineId)
                                    elif cActionType == ">":
                                        # we do nothing, this test is just a condition to apply all following actions
                                        pass
                                    else:
                                        echo("# error: unknown action at " + sLineId)
                                elif cActionType == ">":
                                    break
                            except Exception as e:
                                raise Exception(str(e), "# " + sLineId + " # " + sRuleId)
    if bParagraphChange:
        return (s, dErrs)
    return (False, dErrs)


def _createError (s, sx, sRepl, nOffset, m, iGroup, sLineId, sRuleId, bUppercase, sMsg, sURL, bShowRuleId, sOption, bContext):
    nStart = nOffset + m.start(iGroup)
    nEnd = nOffset + m.end(iGroup)
    # suggestions
    if sRepl[0:1] == "=":
        sSugg = globals()[sRepl[1:]](s, m)
        lSugg = sSugg.split("|")  if sSugg  else []
    elif sRepl == "_":
        lSugg = []
    else:
        lSugg = m.expand(sRepl).split("|")
    if bUppercase and lSugg and m.group(iGroup)[0:1].isupper():
        lSugg = list(map(str.capitalize, lSugg))
    # Message
    sMessage = globals()[sMsg[1:]](s, m)  if sMsg[0:1] == "="  else  m.expand(sMsg)
    if bShowRuleId:
        sMessage += "  # " + sLineId + " # " + sRuleId
    #
    if _bWriterError:
        xErr = SingleProofreadingError()    # uno.createUnoStruct( "com.sun.star.linguistic2.SingleProofreadingError" )
        xErr.nErrorStart = nStart
        xErr.nErrorLength = nEnd - nStart
        xErr.nErrorType = PROOFREADING
        xErr.aRuleIdentifier = sRuleId
        xErr.aShortComment = sMessage   # sMessage.split("|")[0]     # in context menu
        xErr.aFullComment = sMessage   # sMessage.split("|")[-1]    # in dialog
        if bShowRuleId:
            xErr.aShortComment += "  " + sLineId + " # " + sRuleId
        xErr.aSuggestions = tuple(lSugg)
        if sURL:
            xProperty = PropertyValue()
            xProperty.Name = "FullCommentURL"
            xProperty.Value = sURL
            xErr.aProperties = (xProperty,)
        else:
            xErr.aProperties = ()
        return xErr
    else:
        dErr = {}
        dErr["nStart"] = nStart
        dErr["nEnd"] = nEnd
        dErr["sLineId"] = sLineId
        dErr["sRuleId"] = sRuleId
        dErr["sType"] = sOption  if sOption  else "notype"
        dErr["sMessage"] = sMessage
        dErr["aSuggestions"] = lSugg
        dErr["URL"] = sURL  if sURL  else ""
        if bContext:
            dErr['sUnderlined'] = self.sSentence0[nStart:nEnd]
            dErr['sBefore'] = self.sSentence0[max(0,nStart-80):nStart]
            dErr['sAfter'] = self.sSentence0[nEnd:nEnd+80]
        return dErr


def _rewrite (sSentence, sRepl, iGroup, m, bUppercase):
    "text processor: write <sRepl> in <sSentence> at <iGroup> position"
    nLen = m.end(iGroup) - m.start(iGroup)
    if sRepl == "*":
        sNew = " " * nLen
    elif sRepl == "_":
        sNew = sRepl + " " * (nLen-1)
    elif sRepl[0:1] == "=":
        sNew = globals()[sRepl[1:]](sSentence, m)
        sNew = sNew + " " * (nLen-len(sNew))
        if bUppercase and m.group(iGroup)[0:1].isupper():
            sNew = sNew.capitalize()
    else:
        sNew = m.expand(sRepl)
        sNew = sNew + " " * (nLen-len(sNew))
    return sSentence[0:m.start(iGroup)] + sNew + sSentence[m.end(iGroup):]


def ignoreRule (sRuleId):
    "disable rule <sRuleId>"
    _aIgnoredRules.add(sRuleId)


def resetIgnoreRules ():
    "clear all ignored rules"
    _aIgnoredRules.clear()


def reactivateRule (sRuleId):
    "(re)activate rule <sRuleId>"
    _aIgnoredRules.discard(sRuleId)


def listRules (sFilter=None):
    "generator: returns typle (sOption, sLineId, sRuleId)"
    if sFilter:
        try:
            zFilter = re.compile(sFilter)
        except:
            echo("# Error. List rules: wrong regex.")
            sFilter = None
    for sOption, lRuleGroup in chain(_getRules(True), _getRules(False)):
        if sOption != "@@@@":
            for _, _, sLineId, sRuleId, _, _ in lRuleGroup:
                if not sFilter or zFilter.search(sRuleId):
                    yield (sOption, sLineId, sRuleId)


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


def setOption (sOpt, bVal):
    "set option <sOpt> with <bVal> if it exists"
    if sOpt in _dOptions:
        _dOptions[sOpt] = bVal


def setOptions (dOpt):
    "update the dictionary of options with <dOpt>"
    for sKey, bVal in dOpt.items():
        if sKey in _dOptions:
            _dOptions[sKey] = bVal


def getOptions ():
    "return the dictionary of current options"
    return _dOptions


def getDefaultOptions ():
    "return the dictionary of default options"
    return dict(gc_options.getOptions(_sAppContext))


def getOptionsLabels (sLang):
    "return options labels"
    return gc_options.getUI(sLang)


def displayOptions (sLang):
    "display the list of grammar checking options"
    echo("List of options")
    echo("\n".join( [ k+":\t"+str(v)+"\t"+gc_options.getUI(sLang).get(k, ("?", ""))[0]  for k, v  in sorted(_dOptions.items()) ] ))
    echo("")


def resetOptions ():
    "set options to default values"
    global _dOptions
    _dOptions = dict(gc_options.getOptions(_sAppContext))


def getSpellChecker ():
    "return the spellchecker object"
    return _oSpellChecker


def _getPath ():
    return os.path.join(os.path.dirname(sys.modules[__name__].__file__), __name__ + ".py")



#### common functions









def option (sOpt):
    "return True if option <sOpt> is active"
    return _dOptions.get(sOpt, False)


def displayInfo (dTokenPos, tWord):
    "for debugging: retrieve info of word"
    if not tWord:
        echo("> nothing to find")
        return True
    lMorph = _oSpellChecker.getMorph(tWord[1])
    if not lMorph:
        echo("> not in dictionary")
        return True
    print("TOKENS:", dTokenPos)
    if tWord[0] in dTokenPos and "lMorph" in dTokenPos[tWord[0]]:
        echo("DA: " + str(dTokenPos[tWord[0]]["lMorph"]))
    echo("FSA: " + str(lMorph))
    return True









def morph (dTokenPos, tWord, sPattern, bStrict=True, bNoWord=False):
    "analyse a tuple (position, word), return True if sPattern in morphologies (disambiguation on)"
    if not tWord:
        return bNoWord


    lMorph = dTokenPos[tWord[0]]["lMorph"]  if tWord[0] in dTokenPos and "lMorph" in dTokenPos[tWord[0]]  else _oSpellChecker.getMorph(tWord[1])
    if not lMorph:
        return False
    zPattern = re.compile(sPattern)
    if bStrict:
        return all(zPattern.search(s)  for s in lMorph)
    return any(zPattern.search(s)  for s in lMorph)


def morphex (dTokenPos, tWord, sPattern, sNegPattern, bNoWord=False):
    "analyse a tuple (position, word), returns True if not sNegPattern in word morphologies and sPattern in word morphologies (disambiguation on)"
    if not tWord:
        return bNoWord
    lMorph = dTokenPos[tWord[0]]["lMorph"]  if tWord[0] in dTokenPos and "lMorph" in dTokenPos[tWord[0]]  else _oSpellChecker.getMorph(tWord[1])
    if not lMorph:
        return False

    # check negative condition
    zNegPattern = re.compile(sNegPattern)
    if any(zNegPattern.search(s)  for s in lMorph):
        return False
    # search sPattern
    zPattern = re.compile(sPattern)
    return any(zPattern.search(s)  for s in lMorph)


def analyse (sWord, sPattern, bStrict=True):
    "analyse a word, return True if sPattern in morphologies (disambiguation off)"
    lMorph = _oSpellChecker.getMorph(sWord)

    if not lMorph:
        return False
    zPattern = re.compile(sPattern)
    if bStrict:
        return all(zPattern.search(s)  for s in lMorph)
    return any(zPattern.search(s)  for s in lMorph)


def analysex (sWord, sPattern, sNegPattern):
    "analyse a word, returns True if not sNegPattern in word morphologies and sPattern in word morphologies (disambiguation off)"
    lMorph = _oSpellChecker.getMorph(sWord)
    if not lMorph:
        return False
    # check negative condition
    zNegPattern = re.compile(sNegPattern)
    if any(zNegPattern.search(s)  for s in lMorph):
        return False
    # search sPattern
    zPattern = re.compile(sPattern)
    return any(zPattern.search(s)  for s in lMorph)











## functions to get text outside pattern scope

# warning: check compile_rules.py to understand how it works

_zNextWord = re.compile(r" +(\w[\w-]*)")
_zPrevWord = re.compile(r"(\w[\w-]*) +$")

def nextword (s, iStart, n):
    "get the nth word of the input string or empty string"
    m = re.match("(?: +[\\w%-]+){" + str(n-1) + "} +([\\w%-]+)", s[iStart:])
    if not m:
        return None
    return (iStart+m.start(1), m.group(1))

    if sNegPattern and re.search(sNegPattern, s):
        return False
    if re.search(sPattern, s):
        return True
    return False


def look_chk1 (dTokenPos, s, nOffset, sPattern, sPatternGroup1, sNegPatternGroup1=None):
    "returns True if s has pattern sPattern and m.group(1) has pattern sPatternGroup1"
    m = re.search(sPattern, s)
    if not m:
        return False
    try:
        sWord = m.group(1)
        nPos = m.start(1) + nOffset
    except:
        return False
    if sNegPatternGroup1:
        return morphex(dTokenPos, (nPos, sWord), sPatternGroup1, sNegPatternGroup1)
    return morph(dTokenPos, (nPos, sWord), sPatternGroup1, False)


#### Disambiguator

def select (dTokenPos, nPos, sWord, sPattern, lDefault=None):
    "Disambiguation: select morphologies of <sWord> matching <sPattern>"
    if not sWord:
        return True
    if nPos not in dTokenPos:
        print("Error. There should be a token at this position: ", nPos)
        return True
    lMorph = _oSpellChecker.getMorph(sWord)
    if not lMorph or len(lMorph) == 1:
        return True
    lSelect = [ sMorph  for sMorph in lMorph  if re.search(sPattern, sMorph) ]
    if lSelect:
        if len(lSelect) != len(lMorph):
            dTokenPos[nPos]["lMorph"] = lSelect
    elif lDefault:
        dTokenPos[nPos]["lMorph"] = lDefault
    return True


def exclude (dTokenPos, nPos, sWord, sPattern, lDefault=None):
    "Disambiguation: exclude morphologies of <sWord> matching <sPattern>"
    if not sWord:
        return True
    if nPos not in dTokenPos:
        print("Error. There should be a token at this position: ", nPos)
        return True
    lMorph = _oSpellChecker.getMorph(sWord)
    if not lMorph or len(lMorph) == 1:
        return True
    lSelect = [ sMorph  for sMorph in lMorph  if not re.search(sPattern, sMorph) ]
    if lSelect:
        if len(lSelect) != len(lMorph):
            dTokenPos[nPos]["lMorph"] = lSelect
    elif lDefault:
        dTokenPos[nPos]["lMorph"] = lDefault
    return True


def define (dTokenPos, nPos, lMorph):
    "Disambiguation: set morphologies of token at <nPos> with <lMorph>"
    if nPos not in dTokenPos:
        print("Error. There should be a token at this position: ", nPos)
        return True
    dTokenPos[nPos]["lMorph"] = lMorph
    return True




#### TOKEN SENTENCE CHECKER

class TokenSentence:
    "Text parser"

    def __init__ (self, sSentence, sSentence0, nOffset):
        self.sSentence = sSentence
        self.sSentence0 = sSentence0
        self.nOffsetWithinParagraph = nOffset
        self.lToken = list(_oTokenizer.genTokens(sSentence, True))
        self.dTokenPos = { dToken["nStart"]: dToken  for dToken in self.lToken  if dToken["sType"] != "INFO" }
        self.dTags = {}
        self.dError = {}

    def __str__ (self):
        s = "sentence: " + self.sSentence0 + "\n"
        s += "now:      " + self.sSentence  + "\n"
        for dToken in self.lToken:
            s += f'{dToken["nStart"]}\t{dToken["nEnd"]}\t{dToken["sValue"]}'
            if "lMorph" in dToken:
                s += "\t" + str(dToken["lMorph"])
            s += "\n"
        for nPos, dToken in self.dTokenPos.items():
            s += f"{nPos}\t{dToken}\n"
        return s

    def update (self, sSentence, bDebug=False):
        "update <sSentence> and retokenize"
        self.sSentence = sSentence
        lNewToken = list(_oTokenizer.genTokens(sSentence, True))
        for dToken in lNewToken:
            if "lMorph" in self.dTokenPos.get(dToken["nStart"], {}):
                dToken["lMorph"] = self.dTokenPos[dToken["nStart"]]["lMorph"]
        self.lToken = lNewToken
        self.dTokenPos = { dToken["nStart"]: dToken  for dToken in self.lToken  if dToken["sType"] != "INFO" }
        if bDebug:
            print("UPDATE:")
            print(self)

    def _getNextMatchingNodes (self, dToken, dGraph, dNode, bDebug=False):
        "generator: return nodes where <dToken> “values” match <dNode> arcs"
        # token value
        if dToken["sValue"] in dNode:
            if bDebug:
                print("  MATCH:", dToken["sValue"])
            yield dGraph[dNode[dToken["sValue"]]]
        if dToken["sValue"][0:2].istitle(): # we test only 2 first chars, to make valid words such as "Laissez-les", "Passe-partout".
            sValue = dToken["sValue"].lower()
            if sValue in dNode:
                if bDebug:
                    print("  MATCH:", sValue)
                yield dGraph[dNode[sValue]]
        elif dToken["sValue"].isupper():
            sValue = dToken["sValue"].lower()
            if sValue in dNode:
                if bDebug:
                    print("  MATCH:", sValue)
                yield dGraph[dNode[sValue]]
            sValue = dToken["sValue"].capitalize()
            if sValue in dNode:
                if bDebug:
                    print("  MATCH:", sValue)
                yield dGraph[dNode[sValue]]
        # regex value arcs
        if "<re_value>" in dNode:
            for sRegex in dNode["<re_value>"]:
                if "¬" not in sRegex:
                    # no anti-pattern
                    if re.search(sRegex, dToken["sValue"]):
                        if bDebug:
                            print("  MATCH: ~" + sRegex)
                        yield dGraph[dNode["<re_value>"][sRegex]]
                else:
                    # there is an anti-pattern
                    sPattern, sNegPattern = sRegex.split("¬", 1)
                    if sNegPattern and re.search(sNegPattern, dToken["sValue"]):
                        continue
                    if not sPattern or re.search(sPattern, dToken["sValue"]):
                        if bDebug:
                            print("  MATCH: ~" + sRegex)
                        yield dGraph[dNode["<re_value>"][sRegex]]
        # analysable tokens
        if dToken["sType"][0:4] == "WORD":
            # token lemmas
            if "<lemmas>" in dNode:
                for sLemma in _oSpellChecker.getLemma(dToken["sValue"]):
                    if sLemma in dNode["<lemmas>"]:
                        if bDebug:
                            print("  MATCH: >" + sLemma)
                        yield dGraph[dNode["<lemmas>"][sLemma]]
            # regex morph arcs
            if "<re_morph>" in dNode:
                for sRegex in dNode["<re_morph>"]:
                    if "¬" not in sRegex:
                        # no anti-pattern
                        if any(re.search(sRegex, sMorph)  for sMorph in _oSpellChecker.getMorph(dToken["sValue"])):
                            if bDebug:
                                print("  MATCH: @" + sRegex)
                            yield dGraph[dNode["<re_morph>"][sRegex]]
                    else:
                        # there is an anti-pattern
                        sPattern, sNegPattern = sRegex.split("¬", 1)
                        if sNegPattern == "*":
                            # all morphologies must match with <sPattern>
                            if sPattern and all(re.search(sPattern, sMorph)  for sMorph in _oSpellChecker.getMorph(dToken["sValue"])):
                                if bDebug:
                                    print("  MATCH: @" + sRegex)
                                yield dGraph[dNode["<re_morph>"][sRegex]]
                        else:
                            if sNegPattern and any(re.search(sNegPattern, sMorph)  for sMorph in _oSpellChecker.getMorph(dToken["sValue"])):
                                continue
                            if not sPattern or any(re.search(sPattern, sMorph)  for sMorph in _oSpellChecker.getMorph(dToken["sValue"])):
                                if bDebug:
                                    print("  MATCH: @" + sRegex)
                                yield dGraph[dNode["<re_morph>"][sRegex]]
        # meta arc (for token type)
        if "<meta>" in dNode:
            for sMeta in dNode["<meta>"]:
                # not regex here, we just search if <dNode["sType"]> exists within <sMeta>
                if sMeta == "*":
                    if bDebug:
                        print("  MATCH: *" + sMeta)
                    yield dGraph[dNode["<meta>"]["*"]]
                elif "¬" in sMeta:
                    if dToken["sType"] not in sMeta:
                        if bDebug:
                            print("  MATCH: *" + sMeta)
                        yield dGraph[dNode["<meta>"][sMeta]]
                elif dToken["sType"] in sMeta:
                    if bDebug:
                        print("  MATCH: *" + sMeta)
                    yield dGraph[dNode["<meta>"][sMeta]]

    def parse (self, dGraph, dPriority, sCountry="${country_default}", dOptions=None, bShowRuleId=False, bDebug=False, bContext=False):
        "parse tokens from the text and execute actions encountered"
        self.dError = {}
        dPriority = {}  # Key = position; value = priority
        dOpt = _dOptions  if not dOptions  else dOptions
        lPointer = []
        bTagAndRewrite = False
        for dToken in self.lToken:
            if bDebug:
                print("TOKEN:", dToken["sValue"])
            # check arcs for each existing pointer
            lNextPointer = []
            for dPointer in lPointer:
                for dNode in self._getNextMatchingNodes(dToken, dGraph, dPointer["dNode"], bDebug):
                    lNextPointer.append({"iToken": dPointer["iToken"], "dNode": dNode})
            lPointer = lNextPointer
            # check arcs of first nodes
            for dNode in self._getNextMatchingNodes(dToken, dGraph, dGraph[0], bDebug):
                lPointer.append({"iToken": dToken["i"], "dNode": dNode})
            # check if there is rules to check for each pointer
            for dPointer in lPointer:
                #if bDebug:
                #    print("+", dPointer)
                if "<rules>" in dPointer["dNode"]:
                    bChange, dErr = self._executeActions(dGraph, dPointer["dNode"]["<rules>"], dPointer["iToken"]-1, dToken["i"], dPriority, dOpt, sCountry, bShowRuleId, bDebug, bContext)
                    self.dError.update(dErr)
                    if bChange:
                        bTagAndRewrite = True
        if bTagAndRewrite:
            self.rewrite(bDebug)
        if bDebug:
            print(self)
        return (bTagAndRewrite, self.sSentence)

    def _executeActions (self, dGraph, dNode, nTokenOffset, nLastToken, dPriority, dOptions, sCountry, bShowRuleId, bDebug, bContext):
        "execute actions found in the DARG"
        dError = {}
        bChange = False
        for sLineId, nextNodeKey in dNode.items():
            bCondMemo = None
            for sRuleId in dGraph[nextNodeKey]:
                try:
                    if bDebug:
                        print("  TRY:", sRuleId)
                    sOption, sFuncCond, cActionType, sWhat, *eAct = dRule[sRuleId]
                    # Suggestion    [ option, condition, "-", replacement/suggestion/action, iTokenStart, iTokenEnd, nPriority, message, URL ]
                    # TextProcessor [ option, condition, "~", replacement/suggestion/action, iTokenStart, iTokenEnd ]
                    # Disambiguator [ option, condition, "=", replacement/suggestion/action ]
                    # Sentence Tag  [ option, condition, "/", replacement/suggestion/action, iTokenStart, iTokenEnd ]
                    # Test          [ option, condition, ">", "" ]
                    if not sOption or dOptions.get(sOption, False):
                        bCondMemo = not sFuncCond or globals()[sFuncCond](self.lToken, nTokenOffset, nLastToken, sCountry, bCondMemo, self.dTags, self.sSentence, self.sSentence0)
                        if bCondMemo:
                            if cActionType == "-":
                                # grammar error
                                nTokenErrorStart = nTokenOffset + eAct[0]
                                if "bImmune" not in self.lToken[nTokenErrorStart]:
                                    nTokenErrorEnd = (nTokenOffset + eAct[1])  if eAct[1]  else nLastToken
                                    nErrorStart = self.nOffsetWithinParagraph + self.lToken[nTokenErrorStart]["nStart"]
                                    nErrorEnd = self.nOffsetWithinParagraph + self.lToken[nTokenErrorEnd]["nEnd"]
                                    if nErrorStart not in dError or eAct[2] > dPriority.get(nErrorStart, -1):
                                        dError[nErrorStart] = self._createError(sWhat, nTokenOffset, nTokenErrorStart, nErrorStart, nErrorEnd, sLineId, sRuleId, True, eAct[3], eAct[4], bShowRuleId, "notype", bContext)
                                        dPriority[nErrorStart] = eAct[2]
                                        if bDebug:
                                            print("  NEW_ERROR:", dError[nErrorStart], "\n  ", dRule[sRuleId])
                            elif cActionType == "~":
                                # text processor
                                if bDebug:
                                    print("  TAG_PREPARE:\n  ", dRule[sRuleId])
                                nEndToken = (nTokenOffset + eAct[1])  if eAct[1]  else nLastToken
                                self._tagAndPrepareTokenForRewriting(sWhat, nTokenOffset + eAct[0], nEndToken, nTokenOffset, True, bDebug)
                                bChange = True
                            elif cActionType == "=":
                                # disambiguation
                                if bDebug:
                                    print("  DISAMBIGUATOR:\n  ", dRule[sRuleId])
                                globals()[sWhat](self.lToken, nTokenOffset)
                            elif cActionType == ">":
                                # we do nothing, this test is just a condition to apply all following actions
                                if bDebug:
                                    print("  COND_OK")
                                pass
                            elif cActionType == "/":
                                # sentence tags
                                if bDebug:
                                    print("  SENTENCE_TAG:\n  ", dRule[sRuleId])
                                nTokenTag = nTokenOffset + eAct[0]
                                if sWhat not in self.dTags:
                                    self.dTags[sWhat] = (nTokenTag, nTokenTag)
                                elif nTokenTag > self.dTags[sWhat][1]:
                                    self.dTags[sWhat] = (self.dTags[sWhat][0], nTokenTag)
                            else:
                                print("# error: unknown action at " + sLineId)
                        elif cActionType == ">":
                            if bDebug:
                                print("  COND_BREAK")
                            break
                except Exception as e:
                    raise Exception(str(e), sLineId, sRuleId, self.sSentence)
        return bChange, dError

    def _createError (self, sSugg, nTokenOffset, iFirstToken, nStart, nEnd, sLineId, sRuleId, bUppercase, sMsg, sURL, bShowRuleId, sOption, bContext):
        # suggestions
        if sSugg[0:1] == "=":
            sSugg = globals()[sSugg[1:]](self.lToken, nTokenOffset)
            lSugg = sSugg.split("|")  if sSugg  else []
        elif sSugg == "_":
            lSugg = []
        else:
            lSugg = self._expand(sSugg, nTokenOffset).split("|")
        if bUppercase and lSugg and self.lToken[iFirstToken]["sValue"][0:1].isupper():
            lSugg = list(map(str.capitalize, lSugg))
        # Message
        sMessage = globals()[sMsg[1:]](self.lToken, nTokenOffset)  if sMsg[0:1] == "="  else self._expand(sMsg, nTokenOffset)
        if bShowRuleId:
            sMessage += "  " + sLineId + " # " + sRuleId
        #
        if _bWriterError:
            xErr = SingleProofreadingError()    # uno.createUnoStruct( "com.sun.star.linguistic2.SingleProofreadingError" )
            xErr.nErrorStart = nStart
            xErr.nErrorLength = nEnd - nStart
            xErr.nErrorType = PROOFREADING
            xErr.aRuleIdentifier = sRuleId
            xErr.aShortComment = sMessage   # sMessage.split("|")[0]     # in context menu
            xErr.aFullComment = sMessage   # sMessage.split("|")[-1]    # in dialog
            if bShowRuleId:
                xErr.aShortComment += "  " + sLineId + " # " + sRuleId
            xErr.aSuggestions = tuple(lSugg)
            if sURL:
                xProperty = PropertyValue()
                xProperty.Name = "FullCommentURL"
                xProperty.Value = sURL
                xErr.aProperties = (xProperty,)
            else:
                xErr.aProperties = ()
            return xErr
        else:
            dErr = {}
            dErr["nStart"] = nStart
            dErr["nEnd"] = nEnd
            dErr["sLineId"] = sLineId
            dErr["sRuleId"] = sRuleId
            dErr["sType"] = sOption  if sOption  else "notype"
            dErr["sMessage"] = sMessage
            dErr["aSuggestions"] = lSugg
            dErr["URL"] = sURL  if sURL  else ""
            if bContext:
                dErr['sUnderlined'] = self.sSentence0[nStart:nEnd]
                dErr['sBefore'] = self.sSentence0[max(0,nStart-80):nStart]
                dErr['sAfter'] = self.sSentence0[nEnd:nEnd+80]
            return dErr

    def _expand (self, sText, nTokenOffset):
        #print("*", sText)
        for m in re.finditer(r"\\([0-9]+)", sText):
            sText = sText.replace(m.group(0), self.lToken[int(m.group(1))+nTokenOffset]["sValue"])
        #print(">", sText)
        return sText

    def _tagAndPrepareTokenForRewriting (self, sWhat, nTokenRewriteStart, nTokenRewriteEnd, nTokenOffset, bUppercase=True, bDebug=False):
        "text processor: rewrite tokens between <nTokenRewriteStart> and <nTokenRewriteEnd> position"
        if bDebug:
            print("   START:", nTokenRewriteStart, "END:", nTokenRewriteEnd)
        if sWhat == "*":
            # purge text
            if nTokenRewriteEnd - nTokenRewriteStart == 0:
                self.lToken[nTokenRewriteStart]["bToRemove"] = True
            else:
                for i in range(nTokenRewriteStart, nTokenRewriteEnd+1):
                    self.lToken[i]["bToRemove"] = True
        elif sWhat == "␣":
            # merge tokens
            self.lToken[nTokenRewriteStart]["nMergeUntil"] = nTokenRewriteEnd
        elif sWhat == "!":
            # immunity
            if nTokenRewriteEnd - nTokenRewriteStart == 0:
                self.lToken[nTokenRewriteStart]["bImmune"] = True
            else:
                for i in range(nTokenRewriteStart, nTokenRewriteEnd+1):
                    self.lToken[i]["bImmune"] = True
        elif sWhat == "_":
            # neutralized token
            if nTokenRewriteEnd - nTokenRewriteStart == 0:
                self.lToken[nTokenRewriteStart]["sNewValue"] = "_"
            else:
                for i in range(nTokenRewriteStart, nTokenRewriteEnd+1):
                    self.lToken[i]["sNewValue"] = "_"
        else:
            if sWhat.startswith("="):
                sWhat = globals()[sWhat[1:]](self.lToken, nTokenOffset)
            else:
                sWhat = self._expand(sWhat, nTokenOffset)
            bUppercase = bUppercase and self.lToken[nTokenRewriteStart]["sValue"][0:1].isupper()
            if nTokenRewriteEnd - nTokenRewriteStart == 0:
                # one token
                sWhat = sWhat + " " * (len(self.lToken[nTokenRewriteStart]["sValue"])-len(sWhat))
                if bUppercase:
                    sWhat = sWhat[0:1].upper() + sWhat[1:]
                self.lToken[nTokenRewriteStart]["sNewValue"] = sWhat
            else:
                # several tokens
                lTokenValue = sWhat.split("|")
                if len(lTokenValue) != (nTokenRewriteEnd - nTokenRewriteStart + 1):
                    print("Error. Text processor: number of replacements != number of tokens.")
                    return
                for i, sValue in zip(range(nTokenRewriteStart, nTokenRewriteEnd+1), lTokenValue):
                    if not sValue or sValue == "*":
                        self.lToken[i]["bToRemove"] = True
                    else:
                        if bUppercase:
                            sValue = sValue[0:1].upper() + sValue[1:]
                        self.lToken[i]["sNewValue"] = sValue

    def rewrite (self, bDebug=False):
        "rewrite the sentence, modify tokens, purge the token list"
        if bDebug:
            print("REWRITE")
        lNewToken = []
        nMergeUntil = 0
        dTokenMerger = None
        for dToken in self.lToken:
            bKeepToken = True
            if dToken["sType"] != "INFO":
                if "bImmune" in dToken:
                    nErrorStart = self.nOffsetWithinParagraph + dToken["nStart"]
                    if nErrorStart in self.dError:
                        if bDebug:
                            print("immunity -> error removed:", self.dError[nErrorStart])
                        del self.dError[nErrorStart]
                if nMergeUntil and dToken["i"] <= nMergeUntil:
                    dTokenMerger["sValue"] += " " * (dToken["nStart"] - dTokenMerger["nEnd"]) + dToken["sValue"]
                    dTokenMerger["nEnd"] = dToken["nEnd"]
                    if bDebug:
                        print("  MERGED TOKEN:", dTokenMerger["sValue"])
                    bKeepToken = False
                if "nMergeUntil" in dToken:
                    if dToken["i"] > nMergeUntil: # this token is not already merged with a previous token
                        dTokenMerger = dToken
                    if dToken["nMergeUntil"] > nMergeUntil:
                        nMergeUntil = dToken["nMergeUntil"]
                    del dToken["nMergeUntil"]
                elif "bToRemove" in dToken:
                    if bDebug:
                        print("  REMOVED:", dToken["sValue"])
                    self.sSentence = self.sSentence[:dToken["nStart"]] + " " * (dToken["nEnd"] - dToken["nStart"]) + self.sSentence[dToken["nEnd"]:]
                    bKeepToken = False
            #
            if bKeepToken:
                lNewToken.append(dToken)
                if "sNewValue" in dToken:
                    # rewrite token and sentence
                    if bDebug:
                        print(dToken["sValue"], "->", dToken["sNewValue"])
                    dToken["sRealValue"] = dToken["sValue"]
                    dToken["sValue"] = dToken["sNewValue"]
                    nDiffLen = len(dToken["sRealValue"]) - len(dToken["sNewValue"])
                    sNewRepl = (dToken["sNewValue"] + " " * nDiffLen)  if nDiffLen >= 0  else dToken["sNewValue"][:len(dToken["sRealValue"])]
                    self.sSentence = self.sSentence[:dToken["nStart"]] + sNewRepl + self.sSentence[dToken["nEnd"]:]
                    del dToken["sNewValue"]
            else:
                try:
                    del self.dTokenPos[dToken["nStart"]]
                except:
                    print(self)
                    print(dToken)
                    exit()
        if bDebug:
            print("  REWRITED:", self.sSentence)
        self.lToken.clear()
        self.lToken = lNewToken



#### Analyse tokens

def g_value (dToken, sValues, nLeft=None, nRight=None):
    "test if <dToken['sValue']> is in sValues (each value should be separated with |)"
    sValue = "|"+dToken["sValue"]+"|"  if nLeft is None  else "|"+dToken["sValue"][slice(nLeft, nRight)]+"|"
    if sValue in sValues:
        return True
    if dToken["sValue"][0:2].istitle(): # we test only 2 first chars, to make valid words such as "Laissez-les", "Passe-partout".
        if sValue.lower() in sValues:
            return True
    elif dToken["sValue"].isupper():
        #if sValue.lower() in sValues:
        #    return True
        sValue = "|"+sValue[1:].capitalize()
        if sValue.capitalize() in sValues:
            return True
    return False


def g_morph (dToken, sPattern, sNegPattern="", nLeft=None, nRight=None, bMemorizeMorph=True):
    "analyse a token, return True if <sNegPattern> not in morphologies and <sPattern> in morphologies"
    if "lMorph" in dToken:
        lMorph = dToken["lMorph"]
    else:
        if nLeft is not None:
            lMorph = _oSpellChecker.getMorph(dToken["sValue"][slice(nLeft, nRight)])
            if bMemorizeMorph:
                dToken["lMorph"] = lMorph
        else:
            lMorph = _oSpellChecker.getMorph(dToken["sValue"])
        if not lMorph:
            return False
    # check negative condition
    if sNegPattern:
        if sNegPattern == "*":
            # all morph must match sPattern
            zPattern = re.compile(sPattern)
            return all(zPattern.search(sMorph)  for sMorph in lMorph)
        else:
            zNegPattern = re.compile(sNegPattern)
            if any(zNegPattern.search(sMorph)  for sMorph in lMorph):
                return False
    # search sPattern
    zPattern = re.compile(sPattern)
    return any(zPattern.search(sMorph)  for sMorph in lMorph)


def g_analyse (dToken, sPattern, sNegPattern="", nLeft=None, nRight=None, bMemorizeMorph=True):
    "analyse a token, return True if <sNegPattern> not in morphologies and <sPattern> in morphologies (disambiguation off)"
    if nLeft is not None:
        lMorph = _oSpellChecker.getMorph(dToken["sValue"][slice(nLeft, nRight)])
        if bMemorizeMorph:
            dToken["lMorph"] = lMorph
    else:
        lMorph = _oSpellChecker.getMorph(dToken["sValue"])
    if not lMorph:
        return False
    # check negative condition
    if sNegPattern:
        if sNegPattern == "*":
            zPattern = re.compile(sPattern)
            return all(zPattern.search(sMorph)  for sMorph in lMorph)
        else:
            zNegPattern = re.compile(sNegPattern)
            if any(zNegPattern.search(sMorph)  for sMorph in lMorph):
                return False
    # search sPattern
    zPattern = re.compile(sPattern)
    return any(zPattern.search(sMorph)  for sMorph in lMorph)


def g_tag_before (dToken, dTags, sTag):
    if sTag not in dTags:
        return False
    if dToken["i"] > dTags[sTag][0]:
        return True
    return False


def g_tag_after (dToken, dTags, sTag):
    if sTag not in dTags:
        return False
    if dToken["i"] < dTags[sTag][1]:
        return True
    return False


#### Disambiguator

def g_select (dToken, sPattern, lDefault=None):
    "select morphologies for <dToken> according to <sPattern>, always return True"
    lMorph = dToken["lMorph"]  if "lMorph" in dToken  else _oSpellChecker.getMorph(dToken["sValue"])
    if not lMorph or len(lMorph) == 1:

        if lDefault:

            dToken["lMorph"] = lDefault
            #print("DA:", dToken["sValue"], dToken["lMorph"])

        return True
    lSelect = [ sMorph  for sMorph in lMorph  if re.search(sPattern, sMorph) ]
    if lSelect:
        if len(lSelect) != len(lMorph):
            dToken["lMorph"] = lSelect

    elif lDefault:
        dToken["lMorph"] = lDefault

    #print("DA:", dToken["sValue"], dToken["lMorph"])
    return True


def g_exclude (dToken, sPattern, lDefault=None):
    "select morphologies for <dToken> according to <sPattern>, always return True"
    lMorph = dToken["lMorph"]  if "lMorph" in dToken  else _oSpellChecker.getMorph(dToken["sValue"])
    if not lMorph or len(lMorph) == 1:
        if lDefault:
            dToken["lMorph"] = lDefault
            #print("DA:", dToken["sValue"], dToken["lMorph"])
        return True






    lSelect = [ sMorph  for sMorph in lMorph  if not re.search(sPattern, sMorph) ]
    if lSelect:
        if len(lSelect) != len(lMorph):
            dToken["lMorph"] = lSelect

    elif lDefault:
        dToken["lMorph"] = lDefault

    #print("DA:", dToken["sValue"], dToken["lMorph"])
    return True


def g_define (dToken, lMorph):
    "set morphologies of <dToken>, always return True"
    dToken["lMorph"] = lMorph

    #print("DA:", dToken["sValue"], lMorph)
    return True



#### GRAMMAR CHECKER PLUGINS

${plugins}


#### CALLABLES FOR REGEX RULES (generated code)

${callables}


#### CALLABLES FOR GRAPH RULES (generated code)

${graph_callables}

"""
Grammar checker default options
"""

# generated code, do not edit

def getUI (sLang):
    "returns dictionary of UI labels"
    if sLang in _dOptLabel:
        return _dOptLabel[sLang]
    return _dOptLabel["fr"]


def getOptions (sContext="Python"):
    "returns dictionary of options"
    if sContext in dOpt:
        return dOpt[sContext]
    return dOpt["Python"]


lStructOpt = ${lStructOpt}

"""
Grammar checker regex rules
"""

# generated code, do not edit

lParagraphRules = ${paragraph_rules}

lSentenceRules = ${sentence_rules}

"""
Grammar checker graph rules
"""

# generated code, do not edit

dAllGraph = ${rules_graphs}

dRule = ${rules_actions}

#!python3

"""
Text tools
"""

import textwrap
from itertools import chain


def getParagraph (sText):
    "generator: returns paragraphs of text"

        return ""
    lGrammErrs = sorted(aGrammErrs, key=lambda d: d["nStart"])
    lSpellErrs = sorted(aSpellErrs, key=lambda d: d['nStart'])
    sText = ""
    nOffset = 0
    for sLine in wrap(sParagraph, nWidth): # textwrap.wrap(sParagraph, nWidth, drop_whitespace=False)
        sText += sLine + "\n"
        nLineLen = len(sLine)
        sErrLine = ""
        nLenErrLine = 0
        nGrammErr = 0
        nSpellErr = 0
        for dErr in lGrammErrs:
            nStart = dErr["nStart"] - nOffset
            if nStart < nLineLen:
                nGrammErr += 1
                if nStart >= nLenErrLine:
                    sErrLine += " " * (nStart - nLenErrLine) + "^" * (dErr["nEnd"] - dErr["nStart"])
                    nLenErrLine = len(sErrLine)
            else:
                break
        for dErr in lSpellErrs:
            nStart = dErr['nStart'] - nOffset
            if nStart < nLineLen:
                nSpellErr += 1
                nEnd = dErr['nEnd'] - nOffset
                if nEnd > len(sErrLine):
                    sErrLine += " " * (nEnd - len(sErrLine))
                sErrLine = sErrLine[:nStart] + "°" * (nEnd - nStart) + sErrLine[nEnd:]
            else:
                break
        if sErrLine:
            sText += sErrLine + "\n"
        if nGrammErr:
            sText += getReadableErrors(lGrammErrs[:nGrammErr], nWidth)
            del lGrammErrs[0:nGrammErr]
        if nSpellErr:
            sText += getReadableErrors(lSpellErrs[:nSpellErr], nWidth, True)
            del lSpellErrs[0:nSpellErr]
        nOffset += nLineLen
    return sText


def getReadableErrors (lErrs, nWidth, bSpell=False):
    "Returns lErrs errors as readable errors"
    sErrors = ""
    for dErr in lErrs:

    return sErrors


def getReadableError (dErr, bSpell=False):
    "Returns an error dErr as a readable error"
    try:
        if bSpell:
            sText = u"* {nStart}:{nEnd}  # {sValue}:".format(**dErr)
        else:
            sText = u"* {nStart}:{nEnd}  # {sLineId} / {sRuleId}:\n".format(**dErr)
            sText += "  " + dErr.get("sMessage", "# error : message not found")
        if dErr.get("aSuggestions", None):
            sText += "\n  > Suggestions : " + " | ".join(dErr.get("aSuggestions", "# error : suggestions not found"))
        if dErr.get("URL", None):
            sText += "\n  > URL: " + dErr["URL"]
        return sText
    except KeyError:
        return u"* Non-compliant error: {}".format(dErr)


def createParagraphWithLines (lLine):
    "Returns a text as merged lines and a set of data about lines (line_number_y, start_x, end_x)"
    sText = ""

# NOTES SUR LA LANGUE FRANÇAISE

## CE QUI ENTOURE UN VERBE

    PRONOMS (avant)
        COD         COI
        le / l’
        la / l’
        les
        en
        me / m’     me / m’
        te / t’     te / t’
        se / s’     lui
        nous        nous
        vous        nous
        se / s’     leur
                    y

    ADVERBE DE NÉGATION (avant)
        ne / n’

    SOMME
        [le|la|l’|les|en|me|m’|te|t’|se|s’|nous|vous|lui|leur|y]

    COMBINAISONS VALIDES
        ?[ne|n’]¿   [me|te|se]      [le|la|l’|les]
        ?[ne|n’]¿   [m’|t’|s’]      [le|la|l’|les|en|y]
        ?[ne|n’]¿   [le|la]         [lui|leur]
        ?[ne|n’]¿   [l’|les]        [lui|leur|en|y]
        ?[ne|n’]¿   [lui|leur]      en
        ?[ne|n’]¿   [nous|vous]     [le|la|l’|les|en|y]
        ne          [le|la|l’|les|me|m’|te|t’|se|s’|nous|vous|lui|leur]
        n’          [en|y]

    RÉSUMÉ & SIMPLIFICATION
        ?[ne|n’]¿   [le|la|l’|les|en|me|m’|te|t’|se|s’|nous|vous|lui|leur|y]
        ?[ne|n’]¿   [me|m’|te|t’|se|s’|nous|vous]   [le|la|l’|les|en|y]
        ?[ne|n’]¿   [le|la|l’|les]                  [lui|leur|en|y]
        ?[ne|n’]¿   [lui|leur]                      en

    ADVERBE DE NÉGATION (après)
        guère
        jamais
        pas
        plus
        point
        que / qu’
        rien

    PRONOMS À L’IMPÉRATIF
        APRÈS
            -moi
            -toi
            -lui
            -leur
            -nous
            -vous
            -le
            -la
            -les
            -en
            -y

        AVANT
            Uniquement les combinaisons avec l’adverbe de négation [ne|n’]

[args]
lang = fr
lang_name = French
locales = fr_FR fr_BE fr_CA fr_CH fr_LU fr_BF fr_BJ fr_CD fr_CI fr_CM fr_MA fr_ML fr_MU fr_NE fr_RE fr_SN fr_TG
country_default = FR
name = Grammalecte
implname = grammalecte
# always use 3 numbers for version: x.y.z
version = 0.7
author = Olivier R.
provider = Dicollecte
link = http://grammalecte.net
description = Correcteur grammatical pour le français.
extras = README_fr.txt
logo = logo.png

        return this._lVtyp[this._dVerb[sVerb][0]];
    },

    getSimil: function (sWord, sMorph, bSubst=false) {
        if (!sMorph.includes(":V")) {
            return new Set();
        }
        let sInfi = sMorph.slice(1, sMorph.indexOf("/"));
        let aSugg = new Set();
        let tTags = this._getTags(sInfi);
        if (tTags) {
            if (!bSubst) {
                // we suggest conjugated forms
                if (sMorph.includes(":V1")) {
                    aSugg.add(sInfi);

    if (s2 == "vous") {
        return "vous";
    }
    if (s2 == "eux") {
        return "ils";
    }
    if (s2 == "elle" || s2 == "elles") {

        if (cregex.mbNprMasNotFem(_oSpellChecker.getMorph(s1))) {
            return "ils";
        }
        // si épicène, indéterminable, mais OSEF, le féminin l’emporte
        return "elles";
    }
    return s1 + " et " + s2;
}

function apposition (sWord1, sWord2) {
    // returns true if nom + nom (no agreement required)

    return cregex.mbNomNotAdj(_oSpellChecker.getMorph(sWord2)) && cregex.mbPpasNomNotAdj(_oSpellChecker.getMorph(sWord1));
}

function isAmbiguousNAV (sWord) {
    // words which are nom|adj and verb are ambiguous (except être and avoir)
    let lMorph = _oSpellChecker.getMorph(sWord);
    if (lMorph.length === 0) {
        return false;
    }
    if (!cregex.mbNomAdj(lMorph) || sWord == "est") {
        return false;
    }
    if (cregex.mbVconj(lMorph) && !cregex.mbMG(lMorph)) {
        return true;
    }
    return false;
}

function isAmbiguousAndWrong (sWord1, sWord2, sReqMorphNA, sReqMorphConj) {
    //// use it if sWord1 won’t be a verb; word2 is assumed to be true via isAmbiguousNAV

    let a2 = _oSpellChecker.getMorph(sWord2);
    if (a2.length === 0) {
        return false;
    }
    if (cregex.checkConjVerb(a2, sReqMorphConj)) {
        // verb word2 is ok
        return false;
    }
    let a1 = _oSpellChecker.getMorph(sWord1);
    if (a1.length === 0) {
        return false;
    }
    if (cregex.checkAgreement(a1, a2) && (cregex.mbAdj(a2) || cregex.mbAdj(a1))) {
        return false;
    }
    return true;
}

function isVeryAmbiguousAndWrong (sWord1, sWord2, sReqMorphNA, sReqMorphConj, bLastHopeCond) {
    //// use it if sWord1 can be also a verb; word2 is assumed to be true via isAmbiguousNAV

    let a2 = _oSpellChecker.getMorph(sWord2);
    if (a2.length === 0) {
        return false;
    }
    if (cregex.checkConjVerb(a2, sReqMorphConj)) {
        // verb word2 is ok
        return false;
    }
    let a1 = _oSpellChecker.getMorph(sWord1);
    if (a1.length === 0) {
        return false;
    }
    if (cregex.checkAgreement(a1, a2) && (cregex.mbAdj(a2) || cregex.mbAdjNb(a1))) {
        return false;
    }
    // now, we know there no agreement, and conjugation is also wrong
    if (cregex.isNomAdj(a1)) {
        return true;
    }
    //if cregex.isNomAdjVerb(a1): # considered true
    if (bLastHopeCond) {
        return true;
    }
    return false;
}

function checkAgreement (sWord1, sWord2) {

    let a2 = _oSpellChecker.getMorph(sWord2);
    if (a2.length === 0) {
        return true;
    }
    let a1 = _oSpellChecker.getMorph(sWord1);
    if (a1.length === 0) {
        return true;
    }
    return cregex.checkAgreement(a1, a2);
}

function mbUnit (s) {
    if (/[µ\/⁰¹²³⁴⁵⁶⁷⁸⁹Ωℓ·]/.test(s)) {

    var phonet = require("resource://grammalecte/fr/phonet.js");
}


//// verbs

function suggVerb (sFlex, sWho, funcSugg2=null) {

    let aSugg = new Set();
    for (let sStem of _oSpellChecker.getLemma(sFlex)) {
        let tTags = conj._getTags(sStem);
        if (tTags) {
            // we get the tense
            let aTense = new Set();
            for (let sMorph of _oSpellChecker.getMorph(sFlex)) {
                let m;
                let zVerb = new RegExp (">"+sStem+"/.*?(:(?:Y|I[pqsf]|S[pq]|K))", "g");
                while ((m = zVerb.exec(sMorph)) !== null) {
                    // stem must be used in regex to prevent confusion between different verbs (e.g. sauras has 2 stems: savoir and saurer)
                    if (m) {
                        if (m[1] === ":Y") {
                            aTense.add(":Ip");
                            aTense.add(":Iq");
                            aTense.add(":Is");

        return Array.from(aSugg).join("|");
    }
    return "";
}

function suggVerbPpas (sFlex, sWhat=null) {
    let aSugg = new Set();
    for (let sStem of _oSpellChecker.getLemma(sFlex)) {
        let tTags = conj._getTags(sStem);
        if (tTags) {
            if (!sWhat) {
                aSugg.add(conj._getConjWithTags(sStem, tTags, ":PQ", ":Q1"));
                aSugg.add(conj._getConjWithTags(sStem, tTags, ":PQ", ":Q2"));
                aSugg.add(conj._getConjWithTags(sStem, tTags, ":PQ", ":Q3"));
                aSugg.add(conj._getConjWithTags(sStem, tTags, ":PQ", ":Q4"));

        return Array.from(aSugg).join("|");
    }
    return "";
}

function suggVerbTense (sFlex, sTense, sWho) {
    let aSugg = new Set();
    for (let sStem of _oSpellChecker.getLemma(sFlex)) {
        if (conj.hasConj(sStem, sTense, sWho)) {
            aSugg.add(conj.getConj(sStem, sTense, sWho));
        }
    }
    if (aSugg.size > 0) {
        return Array.from(aSugg).join("|");
    }
    return "";
}

function suggVerbImpe (sFlex) {
    let aSugg = new Set();
    for (let sStem of _oSpellChecker.getLemma(sFlex)) {
        let tTags = conj._getTags(sStem);
        if (tTags) {
            if (conj._hasConjWithTags(tTags, ":E", ":2s")) {
                aSugg.add(conj._getConjWithTags(sStem, tTags, ":E", ":2s"));
            }
            if (conj._hasConjWithTags(tTags, ":E", ":1p")) {
                aSugg.add(conj._getConjWithTags(sStem, tTags, ":E", ":1p"));
            }
            if (conj._hasConjWithTags(tTags, ":E", ":2p")) {
                aSugg.add(conj._getConjWithTags(sStem, tTags, ":E", ":2p"));
            }
        }
    }
    if (aSugg.size > 0) {
        return Array.from(aSugg).join("|");
    }
    return "";
}

function suggVerbInfi (sFlex) {
    return _oSpellChecker.getLemma(sFlex).filter(sStem => conj.isVerb(sStem)).join("|");
}


const _dQuiEst = new Map ([
    ["je", ":1s"], ["j’", ":1s"], ["j’en", ":1s"], ["j’y", ":1s"],
    ["tu", ":2s"], ["il", ":3s"], ["on", ":3s"], ["elle", ":3s"],
    ["nous", ":1p"], ["vous", ":2p"], ["ils", ":3p"], ["elles", ":3p"]

    if (!sWho) {
        if (sSuj[0].gl_isLowerCase()) { // pas un pronom, ni un nom propre
            return "";
        }
        sWho = ":3s";
    }
    let aSugg = new Set();
    for (let sStem of _oSpellChecker.getLemma(sFlex)) {
        let tTags = conj._getTags(sStem);
        if (tTags) {
            for (let sTense of lMode) {
                if (conj._hasConjWithTags(tTags, sTense, sWho)) {
                    aSugg.add(conj._getConjWithTags(sStem, tTags, sTense, sWho));
                }
            }
        }
    }
    if (aSugg.size > 0) {
        return Array.from(aSugg).join("|");
    }
    return "";
}

//// Nouns and adjectives

function suggPlur (sFlex, sWordToAgree=null) {
    // returns plural forms assuming sFlex is singular
    if (sWordToAgree) {
        let lMorph = _oSpellChecker.getMorph(sWordToAgree);
        if (lMorph.length === 0) {
            return "";
        }
        let sGender = cregex.getGender(lMorph);
        if (sGender == ":m") {
            return suggMasPlur(sFlex);
        } else if (sGender == ":f") {
            return suggFemPlur(sFlex);
        }
    }
    let aSugg = new Set();

        return Array.from(aSugg).join("|");
    }
    return "";
}

function suggMasSing (sFlex, bSuggSimil=false) {
    // returns masculine singular forms

    let aSugg = new Set();
    for (let sMorph of _oSpellChecker.getMorph(sFlex)) {
        if (!sMorph.includes(":V")) {
            // not a verb
            if (sMorph.includes(":m") || sMorph.includes(":e")) {
                aSugg.add(suggSing(sFlex));
            } else {
                let sStem = cregex.getLemmaOfMorph(sMorph);
                if (mfsp.isFemForm(sStem)) {

        return Array.from(aSugg).join("|");
    }
    return "";
}

function suggMasPlur (sFlex, bSuggSimil=false) {
    // returns masculine plural forms

    let aSugg = new Set();
    for (let sMorph of _oSpellChecker.getMorph(sFlex)) {
        if (!sMorph.includes(":V")) {
            // not a verb
            if (sMorph.includes(":m") || sMorph.includes(":e")) {
                aSugg.add(suggPlur(sFlex));
            } else {
                let sStem = cregex.getLemmaOfMorph(sMorph);
                if (mfsp.isFemForm(sStem)) {

    }
    return "";
}


function suggFemSing (sFlex, bSuggSimil=false) {
    // returns feminine singular forms

    let aSugg = new Set();
    for (let sMorph of _oSpellChecker.getMorph(sFlex)) {
        if (!sMorph.includes(":V")) {
            // not a verb
            if (sMorph.includes(":f") || sMorph.includes(":e")) {
                aSugg.add(suggSing(sFlex));
            } else {
                let sStem = cregex.getLemmaOfMorph(sMorph);
                if (mfsp.isFemForm(sStem)) {

        return Array.from(aSugg).join("|");
    }
    return "";
}

function suggFemPlur (sFlex, bSuggSimil=false) {
    // returns feminine plural forms

    let aSugg = new Set();
    for (let sMorph of _oSpellChecker.getMorph(sFlex)) {
        if (!sMorph.includes(":V")) {
            // not a verb
            if (sMorph.includes(":f") || sMorph.includes(":e")) {
                aSugg.add(suggPlur(sFlex));
            } else {
                let sStem = cregex.getLemmaOfMorph(sMorph);
                if (mfsp.isFemForm(sStem)) {

    if (aSugg.size > 0) {
        return Array.from(aSugg).join("|");
    }
    return "";
}

function hasFemForm (sFlex) {
    for (let sStem of _oSpellChecker.getLemma(sFlex)) {
        if (mfsp.isFemForm(sStem) || conj.hasConj(sStem, ":PQ", ":Q3")) {
            return true;
        }
    }
    if (phonet.hasSimil(sFlex, ":f")) {
        return true;
    }
    return false;
}

function hasMasForm (sFlex) {
    for (let sStem of _oSpellChecker.getLemma(sFlex)) {
        if (mfsp.isFemForm(sStem) || conj.hasConj(sStem, ":PQ", ":Q1")) {
            // what has a feminine form also has a masculine form
            return true;
        }
    }
    if (phonet.hasSimil(sFlex, ":m")) {
        return true;
    }
    return false;
}

function switchGender (sFlex, bPlur=null) {

    let aSugg = new Set();
    if (bPlur === null) {
        for (let sMorph of _oSpellChecker.getMorph(sFlex)) {
            if (sMorph.includes(":f")) {
                if (sMorph.includes(":s")) {
                    aSugg.add(suggMasSing(sFlex));
                } else if (sMorph.includes(":p")) {
                    aSugg.add(suggMasPlur(sFlex));
                }
            } else if (sMorph.includes(":m")) {
                if (sMorph.includes(":s")) {
                    aSugg.add(suggFemSing(sFlex));
                } else if (sMorph.includes(":p")) {
                    aSugg.add(suggFemPlur(sFlex));
                } else {
                    aSugg.add(suggFemSing(sFlex));
                    aSugg.add(suggFemPlur(sFlex));
                }
            }
        }
    } else if (bPlur) {
        for (let sMorph of _oSpellChecker.getMorph(sFlex)) {
            if (sMorph.includes(":f")) {
                aSugg.add(suggMasPlur(sFlex));
            } else if (sMorph.includes(":m")) {
                aSugg.add(suggFemPlur(sFlex));
            }
        }
    } else {
        for (let sMorph of _oSpellChecker.getMorph(sFlex)) {
            if (sMorph.includes(":f")) {
                aSugg.add(suggMasSing(sFlex));
            } else if (sMorph.includes(":m")) {
                aSugg.add(suggFemSing(sFlex));
            }
        }
    }
    if (aSugg.size > 0) {
        return Array.from(aSugg).join("|");
    }
    return "";
}

function switchPlural (sFlex) {
    let aSugg = new Set();
    for (let sMorph of _oSpellChecker.getMorph(sFlex)) { 
        if (sMorph.includes(":s")) {
            aSugg.add(suggPlur(sFlex));
        } else if (sMorph.includes(":p")) {
            aSugg.add(suggSing(sFlex));
        }
    }
    if (aSugg.size > 0) {
        return Array.from(aSugg).join("|");
    }
    return "";
}

function hasSimil (sWord, sPattern=null) {
    return phonet.hasSimil(sWord, sPattern);
}

function suggSimil (sWord, sPattern=null, bSubst=false) {
    // return list of words phonetically similar to sWord and whom POS is matching sPattern
    let aSugg = phonet.selectSimil(sWord, sPattern);
    for (let sMorph of _oSpellChecker.getMorph(sWord)) {
        for (let e of conj.getSimil(sWord, sMorph, bSubst)) {
            aSugg.add(e);
        }
    }
    if (aSugg.size > 0) {
        return Array.from(aSugg).join("|");
    }

    if (sWord[0] == "h" || sWord[0] == "H") {
        return "ce|cet";
    }
    return "ce";
}

function suggLesLa (sWord) {

    if (_oSpellChecker.getMorph(sWord).some(s  =>  s.includes(":p"))) {
        return "les|la";
    }
    return "la";
}

function formatNumber (s) {
    let nLen = s.length;

    [':O2', [" 2ᵉ pers.,", "Pronom : 2ᵉ personne"]],
    [':O3', [" 3ᵉ pers.,", "Pronom : 3ᵉ personne"]],
    [':C', [" conjonction,", "Conjonction"]],
    [':Ĉ', [" conjonction (él.),", "Conjonction (élément)"]],
    [':Cc', [" conjonction de coordination,", "Conjonction de coordination"]],
    [':Cs', [" conjonction de subordination,", "Conjonction de subordination"]],
    [':Ĉs', [" conjonction de subordination (él.),", "Conjonction de subordination (élément)"]],

    [':Ñ', [" locution nominale (él.),", "Locution nominale (élément)"]],
    [':Â', [" locution adjectivale (él.),", "Locution adjectivale (élément)"]],
    [':Ṽ', [" locution verbale (él.),", "Locution verbale (élément)"]],
    [':Ŵ', [" locution adverbiale (él.),", "Locution adverbiale (élément)"]],
    [':Ŕ', [" locution prépositive (él.),", "Locution prépositive (élément)"]],
    [':Ĵ', [" locution interjective (él.),", "Locution interjective (élément)"]],

    ['en', " pronom adverbial"],
    ["m'en", " (me) pronom personnel objet + (en) pronom adverbial"],
    ["t'en", " (te) pronom personnel objet + (en) pronom adverbial"],
    ["s'en", " (se) pronom personnel objet + (en) pronom adverbial"]
]);

const _dChar = new Map([
    ['.', "point"],
    ['·', "point médian"],
    ['…', "points de suspension"],
    [':', "deux-points"],
    [';', "point-virgule"],
    [',', "virgule"],
    ['?', "point d’interrogation"],

    ['–', "tiret demi-cadratin"],
    ['«', "guillemet ouvrant (chevrons)"],
    ['»', "guillemet fermant (chevrons)"],
    ['“', "guillemet ouvrant double"],
    ['”', "guillemet fermant double"],
    ['‘', "guillemet ouvrant"],
    ['’', "guillemet fermant"],
    ['"', "guillemets droits (déconseillé en typographie)"],
    ['/', "signe de la division"],
    ['+', "signe de l’addition"],
    ['*', "signe de la multiplication"],
    ['=', "signe de l’égalité"],
    ['<', "inférieur à"],
    ['>', "supérieur à"],
    ['⩽', "inférieur ou égal à"],
    ['⩾', "supérieur ou égal à"],
    ['%', "signe de pourcentage"],
    ['‰', "signe pour mille"],
]);


class Lexicographe {

    constructor (oSpellChecker, oTokenizer, oLocGraph) {
        this.oSpellChecker = oSpellChecker;

    getInfoForToken (oToken) {
        // Token: .sType, .sValue, .nStart, .nEnd
        // return a object {sType, sValue, aLabel}
        let m = null;
        try {
            switch (oToken.sType) {
                case 'SEPARATOR':
                case 'SIGN':
                    return {
                        sType: oToken.sType,
                        sValue: oToken.sValue,
                        aLabel: [_dChar.gl_get(oToken.sValue, "caractère indéterminé")]
                    };
                    break;
                case 'NUM':
                    return {
                        sType: oToken.sType,
                        sValue: oToken.sValue,
                        aLabel: ["nombre"]
                    };
                    break;
                case 'LINK':
                    return {
                        sType: oToken.sType,
                        sValue: oToken.sValue.slice(0, 40) + "…",
                        aLabel: ["hyperlien"]
                    };
                    break;
                case 'WORD_ELIDED':
                    let sTemp = oToken.sValue.replace("’", "").replace("'", "").replace("`", "").toLowerCase();
                    return {
                        sType: oToken.sType,
                        sValue: oToken.sValue,
                        aLabel: [_dElidedPrefix.gl_get(sTemp, "préfixe élidé inconnu")]
                    };
                    break;
                case 'WORD_ORDINAL':
                    return {
                        sType: oToken.sType,
                        sValue: oToken.sValue,
                        aLabel: ["nombre ordinal"]
                    };
                    break;
                case 'FOLDERUNIX':
                    return {
                        sType: oToken.sType,
                        sValue: oToken.sValue.slice(0, 40) + "…",
                        aLabel: ["dossier UNIX (et dérivés)"]
                    };
                    break;
                case 'FOLDERWIN':
                    return {
                        sType: oToken.sType,
                        sValue: oToken.sValue.slice(0, 40) + "…",
                        aLabel: ["dossier Windows"]
                    };
                    break;
                case 'WORD_ACRONYM':
                    return {
                        sType: oToken.sType,
                        sValue: oToken.sValue,
                        aLabel: ["Sigle ou acronyme"]
                    };
                    break;
                case 'WORD':

        let aTokenList = this.getListOfTokens(sText.replace("'", "’").trim(), false);
        let iKey = 0;
        let aElem = [];
        do {
            let oToken = aTokenList[iKey];
            let sMorphLoc = '';
            let aTokenTempList = [oToken];
            if (oToken.sType == "WORD" || oToken.sType == "WORD_ELIDED"){
                let iKeyTree = iKey + 1;
                let oLocNode = this.oLocGraph[oToken.sValue.toLowerCase()];
                while (oLocNode) {
                    let oTokenNext = aTokenList[iKeyTree];
                    iKeyTree++;
                    if (oTokenNext) {
                        oLocNode = oLocNode[oTokenNext.sValue.toLowerCase()];

${regex_gctestsJS}

"""
Grammalecte - Conjugueur
"""

# License: GPL 3

import re
import traceback

from .conj_data import lVtyp as _lVtyp
from .conj_data import lTags as _lTags

_dGroup = { "0": "auxiliaire", "1": "1ᵉʳ groupe", "2": "2ᵉ groupe", "3": "3ᵉ groupe" }

_dTenseIdx = { ":PQ": 0, ":Ip": 1, ":Iq": 2, ":Is": 3, ":If": 4, ":K": 5, ":Sp": 6, ":Sq": 7, ":E": 8 }



def isVerb (sVerb):
    "return True if it’s a existing verb"
    return sVerb in _dVerb


def getConj (sVerb, sTense, sWho):
    "returns conjugation (can be an empty string)"
    if sVerb not in _dVerb:
        return None

    "returns raw informations about sVerb"
    if sVerb not in _dVerb:
        return None
    return _lVtyp[_dVerb[sVerb][0]]


def getSimil (sWord, sMorph, bSubst=False):
    "returns a set of verbal forms similar to <sWord>, according to <sMorph>"
    if ":V" not in sMorph:
        return set()
    sInfi = sMorph[1:sMorph.find("/")]
    aSugg = set()
    tTags = _getTags(sInfi)
    if tTags:
        if not bSubst:
            # we suggest conjugated forms
            if ":V1" in sMorph:
                aSugg.add(sInfi)

            # if there is only one past participle (epi inv), unreliable.
            if len(aSugg) == 1:
                aSugg.clear()
    return aSugg


def getConjSimilInfiV1 (sInfi):
    "returns verbal forms phonetically similar to infinitive form (for verb in group 1)"
    if sInfi not in _dVerb:
        return set()
    aSugg = set()
    tTags = _getTags(sInfi)
    if tTags:
        aSugg.add(_getConjWithTags(sInfi, tTags, ":Iq", ":2s"))
        aSugg.add(_getConjWithTags(sInfi, tTags, ":Iq", ":3s"))

    "returns sWord modified by sSfx"
    if not sSfx:
        return ""
    if sSfx == "0":
        return sWord
    try:
        return sWord[:-(ord(sSfx[0])-48)] + sSfx[1:]  if sSfx[0] != '0'  else  sWord + sSfx[1:]  # 48 is the ASCII code for "0"
    except (IndexError, TypeError):
        return "## erreur, code : " + str(sSfx) + " ##"



class Verb ():
    "Verb and its conjugation"

    def __init__ (self, sVerb, sVerbPattern=""):
        # conjugate a unknown verb with rules from sVerbPattern
        if not isinstance(sVerb, str):
            raise TypeError("sVerb should be a string")
        if not sVerb:
            raise ValueError("Empty string.")

        if sVerbPattern == "":
            sVerbPattern = sVerb

        self.sVerb = sVerb
        self.sVerbAux = ""
        self._sRawInfo = getVtyp(sVerbPattern)
        self.sInfo = self._readableInfo()
        self.bProWithEn = (self._sRawInfo[5] == "e")
        self._tTags = _getTags(sVerbPattern)
        if not self._tTags:
            raise ValueError("Unknown verb.")
        self._tTagsAux = _getTags(self.sVerbAux)
        self.cGroup = self._sRawInfo[0]
        self.dConj = {
            ":Y": {
                "label": "Infinitif",
                ":": sVerb,
            },
            ":P": {
                "label": "Participe présent",

                sInfo = "# erreur - code : " + self._sRawInfo
            return sGroup + " · " + sInfo
        except:
            traceback.print_exc()
            return "# erreur"

    def infinitif (self, bPro, bNeg, bTpsCo, bInt, bFem):
        "returns string (conjugaison à l’infinitif)"
        try:
            if bTpsCo:
                sInfi = self.sVerbAux  if not bPro  else  "être"
            else:
                sInfi = self.sVerb
            if bPro:
                if self.bProWithEn:

                sInfi += " … ?"
            return sInfi
        except:
            traceback.print_exc()
            return "# erreur"

    def participePasse (self, sWho):
        "returns past participle according to <sWho>"
        try:
            return self.dConj[":Q"][sWho]
        except:
            traceback.print_exc()
            return "# erreur"

    def participePresent (self, bPro, bNeg, bTpsCo, bInt, bFem):
        "returns string (conjugaison du participe présent)"
        try:
            if not self.dConj[":P"][":"]:
                return ""
            if bTpsCo:
                sPartPre = _getConjWithTags(self.sVerbAux, self._tTagsAux, ":PQ", ":P")  if not bPro  else  getConj("être", ":PQ", ":P")
            else:
                sPartPre = self.dConj[":P"][":"]

                sPartPre += " … ?"
            return sPartPre
        except:
            traceback.print_exc()
            return "# erreur"

    def conjugue (self, sTemps, sWho, bPro, bNeg, bTpsCo, bInt, bFem):
        "returns string (conjugue le verbe au temps <sTemps> pour <sWho>) "
        try:
            if not self.dConj[sTemps][sWho]:
                return ""
            if not bTpsCo and bInt and sWho == ":1s" and self.dConj[sTemps].get(":1ś", False):
                sWho = ":1ś"
            if bTpsCo:
                sConj = _getConjWithTags(self.sVerbAux, self._tTagsAux, sTemps, sWho)  if not bPro  else  getConj("être", sTemps, sWho)

                else:
                    sConj = _dProObjEl[sWho] + "en " + sConj
            if bNeg:
                sConj = "n’" + sConj  if bEli and not bPro  else  "ne " + sConj
            if bInt:
                if sWho == ":3s" and not _zNeedTeuph.search(sConj):
                    sConj += "-t"
                sConj += "-" + self._getPronomSujet(sWho, bFem)
            else:
                if sWho == ":1s" and bEli and not bNeg and not bPro:
                    sConj = "j’" + sConj
                else:
                    sConj = self._getPronomSujet(sWho, bFem) + " " + sConj
            if bNeg:
                sConj += " pas"
            if bTpsCo:
                sConj += " " + self._seekPpas(bPro, bFem, sWho.endswith("p") or self._sRawInfo[5] == "r")
            if bInt:
                sConj += " … ?"
            return sConj
        except:
            traceback.print_exc()
            return "# erreur"

    def _getPronomSujet (self, sWho, bFem):
        try:
            if sWho == ":3s":
                if self._sRawInfo[5] == "r":
                    return "on"
                elif bFem:
                    return "elle"
            elif sWho == ":3p" and bFem:
                return "elles"
            return _dProSuj[sWho]
        except:
            traceback.print_exc()
            return "# erreur"

    def imperatif (self, sWho, bPro, bNeg, bTpsCo, bFem):
        "returns string (conjugaison à l’impératif)"
        try:
            if not self.dConj[":E"][sWho]:
                return ""
            if bTpsCo:
                sImpe = _getConjWithTags(self.sVerbAux, self._tTagsAux, ":E", sWho)  if not bPro  else  getConj(u"être", ":E", sWho)
            else:
                sImpe = self.dConj[":E"][sWho]

"""
Conjugation generator
beta stage, unfinished, the root for a new way to generate flexions…
"""

import re


def conjugate (sVerb, sVerbTag="i_____a", bVarPpas=True):
    "conjugate <sVerb> and returns a list of tuples (conjugation form, tags)"
    lConj = []
    cGroup = getVerbGroupChar(sVerb)
    for nCut, sAdd, sFlexTags, sPattern in getConjRules(sVerb, bVarPpas):
        if not sPattern or re.search(sPattern, sVerb):
            sFlexion = sVerb[0:-nCut] + sAdd  if nCut  else sVerb + sAdd
            lConj.append((sFlexion, ":V" + cGroup + "_" + sVerbTag + sFlexTags))
    return lConj


def getVerbGroupChar (sVerb):
    "returns the group number of <sVerb> guessing on its ending"
    sVerb = sVerb.lower()
    if sVerb.endswith("er"):
        return "1"
    if sVerb.endswith("ir"):
        return "2"
    if sVerb == "être" or sVerb == "avoir":
        return "0"
    if sVerb.endswith("re"):
        return "3"
    return "4"


def getConjRules (sVerb, bVarPpas=True, nGroup=2):
    "returns a list of lists to conjugate a verb, guessing on its ending"
    if sVerb.endswith("er"):
        # premier groupe, conjugaison en fonction de la terminaison du lemme
        # 5 lettres
        if sVerb[-5:] in oConj["V1"]:
            lConj = list(oConj["V1"][sVerb[-5:]])
        # 4 lettres
        elif sVerb[-4:] in oConj["V1"]:

        [2,     "isses",        ":Sp:Sq:2s/*",      False],
        [2,     "isse",         ":Sp:3s/*",         False],
        [2,     "ît",           ":Sq:3s/*",         False],
        [2,     "is",           ":E:2s/*",          False],
        [2,     "issons",       ":E:1p/*",          False],
        [2,     "issez",        ":E:2p/*",          False]
    ],

    # premier groupe (bien plus irrégulier que prétendu)
    "V1": {
        # a
        # verbes en -er, -ger, -yer, -cer
        "er": [
            [2,      "er",        ":Y/*",               False],
            [2,      "ant",       ":P/*",               False],

"""
Grammalecte - Compiled regular expressions
"""

import re

#### Lemme
Lemma = re.compile(r"^>(\w[\w-]*)")

#### Analyses
Gender = re.compile(":[mfe]")
Number = re.compile(":[spi]")

#### Nom et adjectif
NA = re.compile(":[NA]")

NPf = re.compile(":(?:M[12P]|T):f")
NPe = re.compile(":(?:M[12P]|T):e")


#### FONCTIONS

def getLemmaOfMorph (s):
    "return lemma in morphology <s>"
    return Lemma.search(s).group(1)

def checkAgreement (l1, l2):
    "returns True if agreement in gender and number is possible between morphologies <l1> and <l2>"
    # check number agreement
    if not mbInv(l1) and not mbInv(l2):
        if mbSg(l1) and not mbSg(l2):
            return False
        if mbPl(l1) and not mbPl(l2):
            return False
    # check gender agreement
    if mbEpi(l1) or mbEpi(l2):
        return True
    if mbMas(l1) and not mbMas(l2):
        return False
    if mbFem(l1) and not mbFem(l2):
        return False
    return True

def checkConjVerb (lMorph, sReqConj):
    "returns True if <sReqConj> in <lMorph>"
    return any(sReqConj in s  for s in lMorph)

def getGender (lMorph):
    "returns gender of word (':m', ':f', ':e' or empty string)."
    sGender = ""
    for sMorph in lMorph:
        m = Gender.search(sMorph)
        if m:
            if not sGender:
                sGender = m.group(0)
            elif sGender != m.group(0):
                return ":e"
    return sGender

def getNumber (lMorph):
    "returns number of word (':s', ':p', ':i' or empty string)."
    sNumber = ""
    for sMorph in lMorph:
        m = Number.search(sMorph)
        if m:
            if not sNumber:
                sNumber = m.group(0)
            elif sNumber != m.group(0):
                return ":i"
    return sNumber

# NOTE :  isWhat (lMorph)    returns True   if lMorph contains nothing else than What
#         mbWhat (lMorph)    returns True   if lMorph contains What at least once

## isXXX = it’s certain

def isNom (lMorph):
    "returns True if all morphologies are “nom”"
    return all(":N" in s  for s in lMorph)

def isNomNotAdj (lMorph):
    "returns True if all morphologies are “nom”, but not “adjectif”"
    return all(NnotA.search(s)  for s in lMorph)

def isAdj (lMorph):
    "returns True if all morphologies are “adjectif”"
    return all(":A" in s  for s in lMorph)

def isNomAdj (lMorph):
    "returns True if all morphologies are “nom” or “adjectif”"
    return all(NA.search(s)  for s in lMorph)

def isNomVconj (lMorph):
    "returns True if all morphologies are “nom” or “verbe conjugué”"
    return all(NVconj.search(s)  for s in lMorph)

def isInv (lMorph):
    "returns True if all morphologies are “invariable”"
    return all(":i" in s  for s in lMorph)

def isSg (lMorph):
    "returns True if all morphologies are “singulier”"
    return all(":s" in s  for s in lMorph)

def isPl (lMorph):
    "returns True if all morphologies are “pluriel”"
    return all(":p" in s  for s in lMorph)

def isEpi (lMorph):
    "returns True if all morphologies are “épicène”"
    return all(":e" in s  for s in lMorph)

def isMas (lMorph):
    "returns True if all morphologies are “masculin”"
    return all(":m" in s  for s in lMorph)

def isFem (lMorph):
    "returns True if all morphologies are “féminin”"
    return all(":f" in s  for s in lMorph)


## mbXXX = MAYBE XXX

def mbNom (lMorph):
    "returns True if one morphology is “nom”"
    return any(":N" in s  for s in lMorph)

def mbAdj (lMorph):
    "returns True if one morphology is “adjectif”"
    return any(":A" in s  for s in lMorph)

def mbAdjNb (lMorph):
    "returns True if one morphology is “adjectif” or “nombre”"
    return any(AD.search(s)  for s in lMorph)

def mbNomAdj (lMorph):
    "returns True if one morphology is “nom” or “adjectif”"
    return any(NA.search(s)  for s in lMorph)

def mbNomNotAdj (lMorph):
    "returns True if one morphology is “nom”, but not “adjectif”"
    bResult = False
    for s in lMorph:
        if ":A" in s:
            return False
        if ":N" in s:
            bResult = True
    return bResult

def mbPpasNomNotAdj (lMorph):
    "returns True if one morphology is “nom” or “participe passé”, but not “adjectif”"
    return any(PNnotA.search(s)  for s in lMorph)

def mbVconj (lMorph):
    "returns True if one morphology is “nom” or “verbe conjugué”"
    return any(Vconj.search(s)  for s in lMorph)

def mbVconj123 (lMorph):
    "returns True if one morphology is “nom” or “verbe conjugué” (but not “avoir” or “être”)"
    return any(Vconj123.search(s)  for s in lMorph)

def mbMG (lMorph):
    "returns True if one morphology is “mot grammatical”"
    return any(":G" in s  for s in lMorph)

def mbInv (lMorph):
    "returns True if one morphology is “invariable”"
    return any(":i" in s  for s in lMorph)

def mbSg (lMorph):
    "returns True if one morphology is “singulier”"
    return any(":s" in s  for s in lMorph)

def mbPl (lMorph):
    "returns True if one morphology is “pluriel”"
    return any(":p" in s  for s in lMorph)

def mbEpi (lMorph):
    "returns True if one morphology is “épicène”"
    return any(":e" in s  for s in lMorph)

def mbMas (lMorph):
    "returns True if one morphology is “masculin”"
    return any(":m" in s  for s in lMorph)

def mbFem (lMorph):
    "returns True if one morphology is “féminin”"
    return any(":f" in s  for s in lMorph)

def mbNpr (lMorph):
    "returns True if one morphology is “nom propre” or “titre de civilité”"
    return any(NP.search(s)  for s in lMorph)

def mbNprMasNotFem (lMorph):
    "returns True if one morphology is “nom propre masculin” but not “féminin”"
    if any(NPf.search(s)  for s in lMorph):
        return False
    return any(NPm.search(s)  for s in lMorph)

#### GRAMMAR CHECKING ENGINE PLUGIN: Parsing functions for French language

from . import cregex as cr


def rewriteSubject (s1, s2):
    "rewrite complex subject: <s1> a prn/patr/npr (M[12P]) followed by “et” and <s2>"
    if s2 == "lui":
        return "ils"
    if s2 == "moi":
        return "nous"
    if s2 == "toi":
        return "vous"
    if s2 == "nous":
        return "nous"
    if s2 == "vous":
        return "vous"
    if s2 == "eux":
        return "ils"
    if s2 == "elle" or s2 == "elles":

        if cr.mbNprMasNotFem(_oSpellChecker.getMorph(s1)):
            return "ils"
        # si épicène, indéterminable, mais OSEF, le féminin l’emporte
        return "elles"
    return s1 + " et " + s2


def apposition (sWord1, sWord2):
    "returns True if nom + nom (no agreement required)"

    return cr.mbNomNotAdj(_oSpellChecker.getMorph(sWord2)) and cr.mbPpasNomNotAdj(_oSpellChecker.getMorph(sWord1))


def isAmbiguousNAV (sWord):
    "words which are nom|adj and verb are ambiguous (except être and avoir)"
    lMorph = _oSpellChecker.getMorph(sWord)

    if not cr.mbNomAdj(lMorph) or sWord == "est":
        return False
    if cr.mbVconj(lMorph) and not cr.mbMG(lMorph):
        return True
    return False


def isAmbiguousAndWrong (sWord1, sWord2, sReqMorphNA, sReqMorphConj):
    "use it if <sWord1> won’t be a verb; <sWord2> is assumed to be True via isAmbiguousNAV"

    a2 = _oSpellChecker.getMorph(sWord2)
    if not a2:
        return False
    if cr.checkConjVerb(a2, sReqMorphConj):
        # verb word2 is ok
        return False
    a1 = _oSpellChecker.getMorph(sWord1)
    if not a1:
        return False
    if cr.checkAgreement(a1, a2) and (cr.mbAdj(a2) or cr.mbAdj(a1)):
        return False
    return True


def isVeryAmbiguousAndWrong (sWord1, sWord2, sReqMorphNA, sReqMorphConj, bLastHopeCond):
    "use it if <sWord1> can be also a verb; <sWord2> is assumed to be True via isAmbiguousNAV"

    a2 = _oSpellChecker.getMorph(sWord2)
    if not a2:
        return False
    if cr.checkConjVerb(a2, sReqMorphConj):
        # verb word2 is ok
        return False
    a1 = _oSpellChecker.getMorph(sWord1)
    if not a1:
        return False
    if cr.checkAgreement(a1, a2) and (cr.mbAdj(a2) or cr.mbAdjNb(a1)):
        return False
    # now, we know there no agreement, and conjugation is also wrong
    if cr.isNomAdj(a1):
        return True
    #if cr.isNomAdjVerb(a1): # considered True
    if bLastHopeCond:
        return True
    return False


def checkAgreement (sWord1, sWord2):
    "check agreement between <sWord1> and <sWord1>"
    a2 = _oSpellChecker.getMorph(sWord2)
    if not a2:
        return True
    a1 = _oSpellChecker.getMorph(sWord1)
    if not a1:
        return True
    return cr.checkAgreement(a1, a2)


_zUnitSpecial = re.compile("[µ/⁰¹²³⁴⁵⁶⁷⁸⁹Ωℓ·]")
_zUnitNumbers = re.compile("[0-9]")

def mbUnit (s):
    "returns True it can be a measurement unit"
    if _zUnitSpecial.search(s):
        return True
    if 1 < len(s) < 16 and s[0:1].islower() and (not s[1:].islower() or _zUnitNumbers.search(s)):
        return True
    return False


#### Syntagmes

_zEndOfNG1 = re.compile(" *$| +(?:, +|)(?:n(?:’|e |o(?:u?s|tre) )|l(?:’|e(?:urs?|s|) |a )|j(?:’|e )|m(?:’|es? |a |on )|t(?:’|es? |a |u )|s(?:’|es? |a )|c(?:’|e(?:t|tte|s|) )|ç(?:a |’)|ils? |vo(?:u?s|tre) )")
_zEndOfNG2 = re.compile(r" +(\w[\w-]+)")
_zEndOfNG3 = re.compile(r" *, +(\w[\w-]+)")

def isEndOfNG (dDA, s, iOffset):
    "returns True if next word doesn’t belong to a noun group"
    if _zEndOfNG1.match(s):
        return True
    m = _zEndOfNG2.match(s)
    if m and morphex(dDA, (iOffset+m.start(1), m.group(1)), ":[VR]", ":[NAQP]"):
        return True
    m = _zEndOfNG3.match(s)
    if m and not morph(dDA, (iOffset+m.start(1), m.group(1)), ":[NA]", False):
        return True
    return False


_zNextIsNotCOD1 = re.compile(" *,")
_zNextIsNotCOD2 = re.compile(" +(?:[mtsnj](e +|’)|[nv]ous |tu |ils? |elles? )")
_zNextIsNotCOD3 = re.compile(r" +([a-zéèî][\w-]+)")

def isNextNotCOD (dDA, s, iOffset):
    "returns True if next word is not a COD"
    if _zNextIsNotCOD1.match(s) or _zNextIsNotCOD2.match(s):
        return True
    m = _zNextIsNotCOD3.match(s)
    if m and morphex(dDA, (iOffset+m.start(1), m.group(1)), ":[123][sp]", ":[DM]"):
        return True
    return False


_zNextIsVerb1 = re.compile(" +[nmts](?:e |’)")
_zNextIsVerb2 = re.compile(r" +(\w[\w-]+)")

def isNextVerb (dDA, s, iOffset):
    "returns True if next word is a verb"
    if _zNextIsVerb1.match(s):
        return True
    m = _zNextIsVerb2.match(s)
    if m and morph(dDA, (iOffset+m.start(1), m.group(1)), ":[123][sp]", False):
        return True
    return False


#### Exceptions

aREGULARPLURAL = frozenset(["abricot", "amarante", "aubergine", "acajou", "anthracite", "brique", "caca", "café", \
                            "carotte", "cerise", "chataigne", "corail", "citron", "crème", "grave", "groseille", \
                            "jonquille", "marron", "olive", "pervenche", "prune", "sable"])
aSHOULDBEVERB = frozenset(["aller", "manger"])

#### GRAMMAR CHECKING ENGINE PLUGIN: Suggestion mechanisms

from . import conj
from . import mfsp
from . import phonet


## Verbs

def suggVerb (sFlex, sWho, funcSugg2=None):
    "change <sFlex> conjugation according to <sWho>"
    aSugg = set()
    for sStem in _oSpellChecker.getLemma(sFlex):
        tTags = conj._getTags(sStem)
        if tTags:
            # we get the tense
            aTense = set()
            for sMorph in _oSpellChecker.getMorph(sFlex):
                for m in re.finditer(">"+sStem+"/.*?(:(?:Y|I[pqsf]|S[pq]|K|P))", sMorph):
                    # stem must be used in regex to prevent confusion between different verbs (e.g. sauras has 2 stems: savoir and saurer)
                    if m:
                        if m.group(1) == ":Y":
                            aTense.add(":Ip")
                            aTense.add(":Iq")
                            aTense.add(":Is")
                        elif m.group(1) == ":P":

        if aSugg2:
            aSugg.add(aSugg2)
    if aSugg:
        return "|".join(aSugg)
    return ""


def suggVerbPpas (sFlex, sPattern=None):
    "suggest past participles for <sFlex>"
    aSugg = set()
    for sStem in _oSpellChecker.getLemma(sFlex):
        tTags = conj._getTags(sStem)
        if tTags:
            if not sPattern:
                aSugg.add(conj._getConjWithTags(sStem, tTags, ":PQ", ":Q1"))
                aSugg.add(conj._getConjWithTags(sStem, tTags, ":PQ", ":Q2"))
                aSugg.add(conj._getConjWithTags(sStem, tTags, ":PQ", ":Q3"))
                aSugg.add(conj._getConjWithTags(sStem, tTags, ":PQ", ":Q4"))
                aSugg.discard("")
            elif sPattern == ":m:s":
                aSugg.add(conj._getConjWithTags(sStem, tTags, ":PQ", ":Q1"))
            elif sPattern == ":m:p":
                if conj._hasConjWithTags(tTags, ":PQ", ":Q2"):
                    aSugg.add(conj._getConjWithTags(sStem, tTags, ":PQ", ":Q2"))
                else:
                    aSugg.add(conj._getConjWithTags(sStem, tTags, ":PQ", ":Q1"))
            elif sPattern == ":f:s":
                if conj._hasConjWithTags(tTags, ":PQ", ":Q3"):
                    aSugg.add(conj._getConjWithTags(sStem, tTags, ":PQ", ":Q3"))
                else:
                    aSugg.add(conj._getConjWithTags(sStem, tTags, ":PQ", ":Q1"))
            elif sPattern == ":f:p":
                if conj._hasConjWithTags(tTags, ":PQ", ":Q4"):
                    aSugg.add(conj._getConjWithTags(sStem, tTags, ":PQ", ":Q4"))
                else:
                    aSugg.add(conj._getConjWithTags(sStem, tTags, ":PQ", ":Q1"))
            elif sPattern == ":s":
                aSugg.add(conj._getConjWithTags(sStem, tTags, ":PQ", ":Q1"))
                aSugg.add(conj._getConjWithTags(sStem, tTags, ":PQ", ":Q3"))
                aSugg.discard("")
            elif sPattern == ":p":
                aSugg.add(conj._getConjWithTags(sStem, tTags, ":PQ", ":Q2"))
                aSugg.add(conj._getConjWithTags(sStem, tTags, ":PQ", ":Q4"))
                aSugg.discard("")
            else:
                aSugg.add(conj._getConjWithTags(sStem, tTags, ":PQ", ":Q1"))
    if aSugg:
        return "|".join(aSugg)
    return ""


def suggVerbTense (sFlex, sTense, sWho):
    "change <sFlex> to a verb according to <sTense> and <sWho>"
    aSugg = set()
    for sStem in _oSpellChecker.getLemma(sFlex):
        if conj.hasConj(sStem, sTense, sWho):
            aSugg.add(conj.getConj(sStem, sTense, sWho))
    if aSugg:
        return "|".join(aSugg)
    return ""


def suggVerbImpe (sFlex):
    "change <sFlex> to a verb at imperative form"
    aSugg = set()
    for sStem in _oSpellChecker.getLemma(sFlex):
        tTags = conj._getTags(sStem)
        if tTags:
            if conj._hasConjWithTags(tTags, ":E", ":2s"):
                aSugg.add(conj._getConjWithTags(sStem, tTags, ":E", ":2s"))
            if conj._hasConjWithTags(tTags, ":E", ":1p"):
                aSugg.add(conj._getConjWithTags(sStem, tTags, ":E", ":1p"))
            if conj._hasConjWithTags(tTags, ":E", ":2p"):
                aSugg.add(conj._getConjWithTags(sStem, tTags, ":E", ":2p"))
    if aSugg:
        return "|".join(aSugg)
    return ""


def suggVerbInfi (sFlex):
    "returns infinitive forms of <sFlex>"
    return "|".join([ sStem  for sStem in _oSpellChecker.getLemma(sFlex)  if conj.isVerb(sStem) ])


_dQuiEst = { "je": ":1s", "j’": ":1s", "j’en": ":1s", "j’y": ":1s", \
             "tu": ":2s", "il": ":3s", "on": ":3s", "elle": ":3s", "nous": ":1p", "vous": ":2p", "ils": ":3p", "elles": ":3p" }
_lIndicatif = [":Ip", ":Iq", ":Is", ":If"]
_lSubjonctif = [":Sp", ":Sq"]

def suggVerbMode (sFlex, cMode, sSuj):
    "returns other conjugations of <sFlex> acconding to <cMode> and <sSuj>"
    if cMode == ":I":
        lMode = _lIndicatif
    elif cMode == ":S":
        lMode = _lSubjonctif
    elif cMode.startswith((":I", ":S")):
        lMode = [cMode]
    else:
        return ""
    sWho = _dQuiEst.get(sSuj.lower(), None)
    if not sWho:
        if sSuj[0:1].islower(): # pas un pronom, ni un nom propre
            return ""
        sWho = ":3s"
    aSugg = set()
    for sStem in _oSpellChecker.getLemma(sFlex):
        tTags = conj._getTags(sStem)
        if tTags:
            for sTense in lMode:
                if conj._hasConjWithTags(tTags, sTense, sWho):
                    aSugg.add(conj._getConjWithTags(sStem, tTags, sTense, sWho))
    if aSugg:
        return "|".join(aSugg)
    return ""


## Nouns and adjectives

def suggPlur (sFlex, sWordToAgree=None):
    "returns plural forms assuming sFlex is singular"
    if sWordToAgree:
        lMorph = _oSpellChecker.getMorph(sFlex)
        if not lMorph:
            return ""
        sGender = cr.getGender(lMorph)
        if sGender == ":m":
            return suggMasPlur(sFlex)
        elif sGender == ":f":
            return suggFemPlur(sFlex)
    aSugg = set()
    if "-" not in sFlex:
        if sFlex.endswith("l"):

    if aSugg:
        return "|".join(aSugg)
    return ""


def suggMasSing (sFlex, bSuggSimil=False):
    "returns masculine singular forms"

    aSugg = set()
    for sMorph in _oSpellChecker.getMorph(sFlex):
        if not ":V" in sMorph:
            # not a verb
            if ":m" in sMorph or ":e" in sMorph:
                aSugg.add(suggSing(sFlex))
            else:
                sStem = cr.getLemmaOfMorph(sMorph)
                if mfsp.isFemForm(sStem):

    if aSugg:
        return "|".join(aSugg)
    return ""


def suggMasPlur (sFlex, bSuggSimil=False):
    "returns masculine plural forms"

    aSugg = set()
    for sMorph in _oSpellChecker.getMorph(sFlex):
        if not ":V" in sMorph:
            # not a verb
            if ":m" in sMorph or ":e" in sMorph:
                aSugg.add(suggPlur(sFlex))
            else:
                sStem = cr.getLemmaOfMorph(sMorph)
                if mfsp.isFemForm(sStem):

    if aSugg:
        return "|".join(aSugg)
    return ""


def suggFemSing (sFlex, bSuggSimil=False):
    "returns feminine singular forms"

    aSugg = set()
    for sMorph in _oSpellChecker.getMorph(sFlex):
        if not ":V" in sMorph:
            # not a verb
            if ":f" in sMorph or ":e" in sMorph:
                aSugg.add(suggSing(sFlex))
            else:
                sStem = cr.getLemmaOfMorph(sMorph)
                if mfsp.isFemForm(sStem):

    if aSugg:
        return "|".join(aSugg)
    return ""


def suggFemPlur (sFlex, bSuggSimil=False):
    "returns feminine plural forms"

    aSugg = set()
    for sMorph in _oSpellChecker.getMorph(sFlex):
        if not ":V" in sMorph:
            # not a verb
            if ":f" in sMorph or ":e" in sMorph:
                aSugg.add(suggPlur(sFlex))
            else:
                sStem = cr.getLemmaOfMorph(sMorph)
                if mfsp.isFemForm(sStem):

            aSugg.add(e)
    if aSugg:
        return "|".join(aSugg)
    return ""


def hasFemForm (sFlex):
    "return True if there is a feminine form of <sFlex>"
    for sStem in _oSpellChecker.getLemma(sFlex):
        if mfsp.isFemForm(sStem) or conj.hasConj(sStem, ":PQ", ":Q3"):
            return True
    if phonet.hasSimil(sFlex, ":f"):
        return True
    return False


def hasMasForm (sFlex):
    "return True if there is a masculine form of <sFlex>"
    for sStem in _oSpellChecker.getLemma(sFlex):
        if mfsp.isFemForm(sStem) or conj.hasConj(sStem, ":PQ", ":Q1"):
            # what has a feminine form also has a masculine form
            return True
    if phonet.hasSimil(sFlex, ":m"):
        return True
    return False


def switchGender (sFlex, bPlur=None):
    "return feminine or masculine form(s) of <sFlex>"
    aSugg = set()
    if bPlur == None:
        for sMorph in _oSpellChecker.getMorph(sFlex):
            if ":f" in sMorph:
                if ":s" in sMorph:
                    aSugg.add(suggMasSing(sFlex))
                elif ":p" in sMorph:
                    aSugg.add(suggMasPlur(sFlex))
            elif ":m" in sMorph:
                if ":s" in sMorph:
                    aSugg.add(suggFemSing(sFlex))
                elif ":p" in sMorph:
                    aSugg.add(suggFemPlur(sFlex))
                else:
                    aSugg.add(suggFemSing(sFlex))
                    aSugg.add(suggFemPlur(sFlex))
    elif bPlur:
        for sMorph in _oSpellChecker.getMorph(sFlex):
            if ":f" in sMorph:
                aSugg.add(suggMasPlur(sFlex))
            elif ":m" in sMorph:
                aSugg.add(suggFemPlur(sFlex))
    else:
        for sMorph in _oSpellChecker.getMorph(sFlex):
            if ":f" in sMorph:
                aSugg.add(suggMasSing(sFlex))
            elif ":m" in sMorph:
                aSugg.add(suggFemSing(sFlex))
    if aSugg:
        return "|".join(aSugg)
    return ""


def switchPlural (sFlex):
    "return plural or singular form(s) of <sFlex>"
    aSugg = set()
    for sMorph in _oSpellChecker.getMorph(sFlex):
        if ":s" in sMorph:
            aSugg.add(suggPlur(sFlex))
        elif ":p" in sMorph:
            aSugg.add(suggSing(sFlex))
    if aSugg:
        return "|".join(aSugg)
    return ""


def hasSimil (sWord, sPattern=None):
    "return True if there is words phonetically similar to <sWord> (according to <sPattern> if required)"
    return phonet.hasSimil(sWord, sPattern)


def suggSimil (sWord, sPattern=None, bSubst=False):
    "return list of words phonetically similar to sWord and whom POS is matching sPattern"

    aSugg = phonet.selectSimil(sWord, sPattern)
    for sMorph in _oSpellChecker.getMorph(sWord):
        aSugg.update(conj.getSimil(sWord, sMorph, bSubst))
        break
    if aSugg:
        return "|".join(aSugg)
    return ""


def suggCeOrCet (sWord):
    "suggest “ce” or “cet” or both according to the first letter of <sWord>"
    if re.match("(?i)[aeéèêiouyâîï]", sWord):
        return "cet"
    if sWord[0:1] == "h" or sWord[0:1] == "H":
        return "ce|cet"
    return "ce"


def suggLesLa (sWord):
    "suggest “les” or “la” according to <sWord>"
    if any( ":p" in sMorph  for sMorph in _oSpellChecker.getMorph(sWord) ):
        return "les|la"
    return "la"


_zBinary = re.compile("^[01]+$")

def formatNumber (s):
    "add spaces or hyphens to big numbers"
    nLen = len(s)
    if nLen < 4:
        return s
    sRes = ""
    # nombre ordinaire
    nEnd = nLen
    while nEnd > 0:

    elif nLen == 9 and s.startswith("0"):
        sRes += "|" + s[0:3] + " " + s[3:5] + " " + s[5:7] + " " + s[7:9]                   # fixe belge 1
        sRes += "|" + s[0:2] + " " + s[2:5] + " " + s[5:7] + " " + s[7:9]                   # fixe belge 2
    return sRes


def formatNF (s):
    "typography: format NF reference (norme française)"
    try:
        m = re.match("NF[  -]?(C|E|P|Q|S|X|Z|EN(?:[  -]ISO|))[  -]?([0-9]+(?:[/‑-][0-9]+|))", s)
        if not m:
            return ""
        return "NF " + m.group(1).upper().replace(" ", " ").replace("-", " ") + " " + m.group(2).replace("/", "‑").replace("-", "‑")
    except:
        traceback.print_exc()
        return "# erreur #"


def undoLigature (c):
    "typography: split ligature character <c> in several chars"
    if c == "fi":
        return "fi"
    elif c == "fl":
        return "fl"
    elif c == "ff":
        return "ff"
    elif c == "ffi":

_xNormalizedCharsForInclusiveWriting = str.maketrans({
    '(': '_',  ')': '_',
    '.': '_',  '·': '_',
    '–': '_',  '—': '_',
    '/': '_'
})


def normalizeInclusiveWriting (sToken):
    "typography: replace word separators used in inclusive writing by underscore (_)"
    return sToken.translate(_xNormalizedCharsForInclusiveWriting)

"""
Grammalecte - Lexicographe
"""

# License: MPL 2


import re
import traceback


_dTAGS = {
    ':N': (" nom,", "Nom"),
    ':A': (" adjectif,", "Adjectif"),
    ':M1': (" prénom,", "Prénom"),
    ':M2': (" patronyme,", "Patronyme, matronyme, nom de famille…"),
    ':MP': (" nom propre,", "Nom propre"),
    ':W': (" adverbe,", "Adverbe"),
    ':J': (" interjection,", "Interjection"),

    ':O2': (" 2ᵉ pers.,", "Pronom : 2ᵉ personne"),
    ':O3': (" 3ᵉ pers.,", "Pronom : 3ᵉ personne"),
    ':C': (" conjonction,", "Conjonction"),
    ':Ĉ': (" conjonction (él.),", "Conjonction (élément)"),
    ':Cc': (" conjonction de coordination,", "Conjonction de coordination"),
    ':Cs': (" conjonction de subordination,", "Conjonction de subordination"),
    ':Ĉs': (" conjonction de subordination (él.),", "Conjonction de subordination (élément)"),

    ':Ñ': (" locution nominale (él.),", "Locution nominale (élément)"),
    ':Â': (" locution adjectivale (él.),", "Locution adjectivale (élément)"),
    ':Ṽ': (" locution verbale (él.),", "Locution verbale (élément)"),
    ':Ŵ': (" locution adverbiale (él.),", "Locution adverbiale (élément)"),
    ':Ŕ': (" locution prépositive (él.),", "Locution prépositive (élément)"),
    ':Ĵ': (" locution interjective (él.),", "Locution interjective (élément)"),

    'il': " pronom personnel sujet, 3ᵉ pers. masc. sing.",
    'on': " pronom personnel sujet, 3ᵉ pers. sing. ou plur.",
    'elle': " pronom personnel sujet, 3ᵉ pers. fém. sing.",
    'nous': " pronom personnel sujet/objet, 1ʳᵉ pers. plur.",
    'vous': " pronom personnel sujet/objet, 2ᵉ pers. plur.",
    'ils': " pronom personnel sujet, 3ᵉ pers. masc. plur.",
    'elles': " pronom personnel sujet, 3ᵉ pers. masc. plur.",

    "là": " particule démonstrative",
    "ci": " particule démonstrative",

    'le': " COD, masc. sing.",
    'la': " COD, fém. sing.",
    'les': " COD, plur.",

    'moi': " COI (à moi), sing.",
    'toi': " COI (à toi), sing.",
    'lui': " COI (à lui ou à elle), sing.",
    'nous2': " COI (à nous), plur.",
    'vous2': " COI (à vous), plur.",
    'leur': " COI (à eux ou à elles), plur.",

    'y': " pronom adverbial",
    "m'y": " (me) pronom personnel objet + (y) pronom adverbial",
    "t'y": " (te) pronom personnel objet + (y) pronom adverbial",
    "s'y": " (se) pronom personnel objet + (y) pronom adverbial",

    'en': " pronom adverbial",
    "m'en": " (me) pronom personnel objet + (en) pronom adverbial",
    "t'en": " (te) pronom personnel objet + (en) pronom adverbial",
    "s'en": " (se) pronom personnel objet + (en) pronom adverbial",
}


class Lexicographe:
    "Lexicographer - word analyzer"

    def __init__ (self, oSpellChecker):
        self.oSpellChecker = oSpellChecker
        self._zElidedPrefix = re.compile("(?i)^([dljmtsncç]|quoiqu|lorsqu|jusqu|puisqu|qu)['’](.+)")
        self._zCompoundWord = re.compile("(?i)(\\w+)-((?:les?|la)-(?:moi|toi|lui|[nv]ous|leur)|t-(?:il|elle|on)|y|en|[mts][’'](?:y|en)|les?|l[aà]|[mt]oi|leur|lui|je|tu|ils?|elles?|on|[nv]ous)$")
        self._zTag = re.compile("[:;/][\\w*][^:;/]*")

    def analyzeWord (self, sWord):
        "returns a tuple (a list of morphologies, a set of verb at infinitive form)"
        try:
            if not sWord:
                return (None, None)
            if sWord.count("-") > 4:
                return (["élément complexe indéterminé"], None)
            if sWord.isdigit():
                return (["nombre"], None)

                aMorph.append( "{} : {}".format(sWord, self.formatTags(lMorph[0])) )
            else:
                aMorph.append( "{} :  inconnu du dictionnaire".format(sWord) )
            # suffixe d’un mot composé
            if m2:
                aMorph.append( "-{} : {}".format(m2.group(2), self._formatSuffix(m2.group(2).lower())) )
            # Verbes
            aVerb = set([ s[1:s.find("/")]  for s in lMorph  if ":V" in s ])
            return (aMorph, aVerb)
        except:
            traceback.print_exc()
            return (["#erreur"], None)

    def formatTags (self, sTags):
        "returns string: readable tags"
        sRes = ""
        sTags = re.sub("(?<=V[1-3])[itpqnmr_eaxz]+", "", sTags)
        sTags = re.sub("(?<=V0[ea])[itpqnmr_eaxz]+", "", sTags)
        for m in self._zTag.finditer(sTags):
            sRes += _dTAGS.get(m.group(0), " [{}]".format(m.group(0)))[0]
        if sRes.startswith(" verbe") and not sRes.endswith("infinitif"):
            sRes += " [{}]".format(sTags[1:sTags.find(" ")])

"""
Masculins, féminins, singuliers et pluriels
"""

from .mfsp_data import lTagMiscPlur as _lTagMiscPlur
from .mfsp_data import lTagMasForm as _lTagMasForm
from .mfsp_data import dMiscPlur as _dMiscPlur
from .mfsp_data import dMasForm as _dMasForm

"""
Grammalecte - Suggestion phonétique
"""

# License: GPL 3

import re

from .phonet_data import dWord as _dWord
from .phonet_data import lSet as _lSet
from .phonet_data import dMorph as _dMorph

#! python3

"""
Grammar checker tests for French language
"""

import unittest
import os
import re
import time

                        sExceptedSuggs = sExceptedSuggs[1:-1]
                else:
                    sErrorText = sLine.strip()
                    sExceptedSuggs = ""
                sExpectedErrors = self._getExpectedErrors(sErrorText)
                sTextToCheck = sErrorText.replace("}}", "").replace("{{", "")
                sFoundErrors, sListErr, sFoundSuggs = self._getFoundErrors(sTextToCheck, sOption)
                # tests
                if sExpectedErrors != sFoundErrors:
                    print("\n# Line num: " + sLineNum + \
                          "\n> to check: " + _fuckBackslashUTF8(sTextToCheck) + \
                          "\n  expected: " + sExpectedErrors + \
                          "\n  found:    " + sFoundErrors + \
                          "\n  errors:   \n" + sListErr)
                if sExceptedSuggs:
                    if sExceptedSuggs != sFoundSuggs:
                        print("\n# Line num: " + sLineNum + \
                              "\n> to check: " + _fuckBackslashUTF8(sTextToCheck) + \
                              "\n  errors:   \n" + sListErr)
        # untested rules
        i = 0
        for sOpt, sLineId, sRuleId in gce.listRules():
            if sLineId not in self._aRuleTested and not re.search("^[0-9]+[sp]$|^[pd]_", sRuleId):
                echo(sRuleId, end= ", ")
                i += 1
        if i:

#!python3

"""
Text formatter
"""

import re


dReplTable = {
    # surnumerary_spaces
    "start_of_paragraph":          [("^[  ]+", "")],

    "erase_non_breaking_hyphens":  [("­", "")],
    ## typographic signs
    "ts_apostrophe":          [ ("(?i)\\b([ldnjmtscç])['´‘′`](?=\\w)", "\\1’"),
                                ("(?i)(qu|jusqu|lorsqu|puisqu|quoiqu|quelqu|presqu|entr|aujourd|prud)['´‘′`]", "\\1’") ],
    "ts_ellipsis":            [ ("\\.\\.\\.", "…"),
                                ("(?<=…)[.][.]", "…"),
                                ("…[.](?![.])", "…") ],
    "ts_n_dash_middle":       [ (" [-—] ", " – "),
                                (" [-—],", " –,") ],
    "ts_m_dash_middle":       [ (" [-–] ", " — "),
                                (" [-–],", " —,") ],
    "ts_n_dash_start":        [ ("^[-—][  ]", "– "),
                                ("^– ", "– "),
                                ("^[-–—](?=[\\w.…])", "– ") ],
    "ts_m_dash_start":        [ ("^[-–][  ]", "— "),

# ERREURS COURANTES
# http://fr.wikipedia.org/wiki/Wikip%C3%A9dia:Fautes_d%27orthographe/Courantes


!!
!!
!! Options                                                                                        !!
!!
!!

OPTGROUP/basic: typo apos, esp tab, nbsp unit, tu maj, num virg, nf chim, ocr mapos, liga
OPTGROUP/gramm: conf sgpl gn
OPTGROUP/verbs: infi conj ppas, imp inte vmode
OPTGROUP/style: bs pleo, redon1 redon2, neg

OPTLABEL/debug:     Debug
OPTLABEL/idrule:    Display control rule identifier [!]|Display control rule identifier in the context menu message.



!!
!!
!! Définitions pour les regex                                                                     !!
!!
!!

# REGEX
DEF: avoir          [aeo]\w*
DEF: etre           [êeésf]\w+
DEF: avoir_etre     [aeêésfo]\w*
DEF: aller          (?:ai?ll|v[ao]|ir[aio])\w*
DEF: ppas           \w[\w-]+[éiust]e?s?
DEF: infi           \w[\w-]+(?:er|ir|re)
DEF: w_1            \w[\w-]*
DEF: w_2            \w[\w-]+
DEF: w_3            \w[\w-]+\w
DEF: w_4            \w[\w-][\w-]+\w
DEF: w1             \w+
DEF: w2             \w\w+
DEF: w3             \w\w\w+
DEF: w4             \w\w\w\w+

# GRAPH
DEF: mois           [>janvier|>février|>mars|>avril|>mai|>juin|>juillet|>août|>aout|>septembre|>octobre|>novembre|>décembre|>vendémiaire|>brumaire|>frimaire|>nivôse|>pluviôse|>ventôse|>germinal|>floréal|>prairial|>messidor|>thermidor|>fructidor]
DEF: mi_mois        [>mi-janvier|>mi-février|>mi-mars|>mi-avril|>mi-mai|>mi-juin|>mi-juillet|>mi-août|>mi-aout|>mi-septembre|>mi-octobre|>mi-novembre|>mi-décembre|mi-vendémiaire|mi-brumaire|mi-frimaire|mi-nivôse|mi-pluviôse|mi-ventôse|mi-germinal|mi-floréal|mi-prairial|mi-messidor|mi-thermidor|mi-fructidor]
DEF: pronom_obj     [moi|toi|soi|lui|elle|nous|vous|eux|elles|moi-même|toi-même|soi-même|lui-même|elle-même|nous-mêmes|vous-même|vous-mêmes|eux-mêmes|elles-mêmes]


!!
!!
!!
!!
!!
!!
!!
!!
!!
!!
!!
!!
!!
!!
!!
!!
!!
!!
!!
!!
!! PASSE 0: PARAGRAPHE PAR PARAGRAPHE                                                             !!
!!
!!
!!
!!
!!
!!
!!
!!
!!
!!
!!
!!
!!
!!
!!
!!
!!
!!
!!
!!



!!
!!
!!! Espaces & tabulations                                                                         !!
!!
!!

# Espaces surnuméraires
# Note : les tabulations ne sont pas soulignées dans LibreOffice. Mais l’erreur est bien présente.
__<s>/tab(tab_début_ligne)__           ^[      ]+ <<- ->> ""                   # Espace(s) en début de ligne à supprimer : utilisez les retraits de paragraphe.
__<s>/tab(tab_fin_ligne)__             [       ]+$ <<- ->> ""                  # Espace(s) en fin de ligne à supprimer.

# À partir d’ici, toute règle est susceptible de supprimer des caractères et les remplacer par des espaces ou des chaînes de longueur égale.
#
#


!!!
!!!
!!! Processeur: efface les ponctuations gênantes (URL, sigles, abréviations, IP, heures, etc.)    !!
!!!
!!!

# e-mail
__<i>(p_email)__
    \w[\w.-]*@\w[\w.-]*\w[.]\w+ <<- ~>> *

# URL
__<i>(p_URL)__
    https?://[\w./?&!%=+*"'@$#-]+ <<- ~>> *
__<i](p_URL2)__
    ((?:{w_1}[.])*)({w_2})([.](?:com|net|org|info|fr|ca|be|ch|i[ot]|co[.]uk|tk|es|jp|zh|ru|us|nl|xyz)) @@0,**,$
    <<- ~1>> *
    <<- ~2>> =\2.capitalize()

    <<- ~3>> *

# Numéro de chapitre
__<i>(p_chapitre)__
    ^\d+[.][\d.-]* <<- ~>> *

# Numéro suivi de plusieurs espaces, considéré comme une numérotation de chapitre

# crochets
__[s](p_points_suspension_entre_crochets)__
    \[…\] <<- ~>> *
__[s](p_mot_entre_crochets)__
    \[({w_1})\] @@1
    <<- \1.isdigit() ~>> *
    <<- __else__ and morph(\1, ":G", False) ~>> =" " + \1 + " "
    <<- __else__ and \1.isalpha() ~>> " _"
__[s](points_suspension_entre_parenthèses)__
    \(…\)
    <<- ->> […]                 # Pour indiquer une troncature de texte, on utilise usuellement des crochets.
    <<- ~>> *

# Divers
__[i](p_FranceTV)__

TEST: Dossier C:\Program Files (x86)\LibreOffice
TEST: [1] Dossier à revoir.



!!!
!!!
!!! Processeur: balises HTML et LaTeX                                                             !!
!!!
!!!

# HTML
__<i>/html(p_html_amp_xxx)__            &amp;[a-zA-Z]+; <<- ~>> _
__<i>/html(p_html_lt)__                 &lt; <<- ~>> "   <"
__<i>/html(p_html_gt)__                 &gt; <<- ~>> >

TEST: __html__  des <b>{{homme}}</b>
TEST: __html__  des [b]{{femme}}[/b]



!!
!!
!!!! Écritures épicènes dystypographiques                                                         !!
!!
!!

# (attention aux modifs: brainfuck d’intensité non négligeable)
# La désambiguïsation est faite lors de la deuxième passe

__[i](typo_écriture_épicène_tous_toutes)__

TEST: je suis {{déconsidéré.e}} par ma hiérarchie.
TEST: il faut en parler à l’{{auteur(e)}} et à son agent.


!!
!!
!!!! Majuscules manquantes                                                                        !!
!!
!!

# Majuscules après un point
__[s]/maj(majuscule_après_point)__
    ({w_2})[.] ([a-zéàèîçô]\w*)  @@0,$
    <<- not re.search("(?i)^(?:etc|[A-Z]|chap|cf|fig|hab|litt|circ|coll|r[eé]f|étym|suppl|bibl|bibliogr|cit|op|vol|déc|nov|oct|janv|juil|avr|sept)$", \1)

    <<- after("\w\w[.] +\w+") -1>> =\1.capitalize()                                                 # Majuscule en début de phrase, sauf éventuellement lors d’une énumération.

TEST: {{je}} suis disponible quand tu veux. Mais pas aujourd’hui.


!!
!!
!!!! Virgules                                                                                     !!
!!
!!

# virgules manquantes
__[i>/virg(virgule_manquante_avant_etc)__   {w_1}( etc[.])  @@$ <<- -1>> , etc.                     # Avant « etc. », il faut mettre une virgule.
__[i>/virg(virgule_manquante_avant_car)__
    ({w_1})( car)(?= (?:j[e’]|tu|ils?|nous|vous|elles?|on|les?|l[a’]|ces?|des?|cette|[mts](?:on|a|es))\b)  @@0,$
    <<- not morph(\1, ":[DR]", False) -2>> , car
    # Si « car » est la conjonction de coordination, une virgule est peut-être souhaitable.|http://bdl.oqlf.gouv.qc.ca/bdl/gabarit_bdl.asp?id=3447
__[i>/virg(virgule_manquante_avant_mais)__
    ({w_1})( mais)(?= (?:j[e’]|tu|ils?|nous|vous|elles?|on)\b)  @@0,$
    <<- not morph(\1, ">(?:[mtscl]es|[nv]os|quels)/", False) -2>> , mais
    # Si « mais » est la conjonction de coordination, une virgule est souhaitable si elle introduit une nouvelle proposition.|http://bdl.oqlf.gouv.qc.ca/bdl/gabarit_bdl.asp?id=3445
__[i>/virg(virgule_manquante_avant_donc)__
    ({w_1})( donc)(?= (?:j[e’]|tu|ils?|elles?|on)\b)  @@0,$
    <<- not morph(\1, ":V", False) -2>> , donc
    # Si « mais » est la conjonction de coordination, une virgule est souhaitable si elle introduit une nouvelle proposition.|http://bdl.oqlf.gouv.qc.ca/bdl/gabarit_bdl.asp?id=3448

TEST: Un chien, un chat{{ etc.}}                                            ->> , etc.

#__<i>/typo__  ({w_2}) *$   @@0
#    <<- before("[a-zéàùèê][.] ") and not before(r"^(?:\d|\w[.])")
#    -1>> \1.|\1…|\1 !|\1 ?|\1 :|\1 ;|\1,                                               # Il manque une ponctuation finale.


!!
!!
!!!! Espaces manquants                                                                            !!
!!
!!

__<s>/typo(typo_espace_manquant_après1)__
    ({w1})[,:]({w_1})  @@0,$  <<- not \1.isdigit() -2>> " \2"                                       # Il manque un espace.
__<s>/typo(typo_espace_manquant_après2)__
    [?!;%‰‱˚»}]({w_1})  @@$  <<- -1>> " \1"                                                       # Il manque un espace.

TEST: espace.{{Manquant}}                                                   ->> " Manquant"
TEST: qui ont refuse.{{de}} se sauver                                       ->> " de"
TEST: pic.twitter.com/PICNAME


!!
!!
!!!! Points                                                                                       !!
!!
!!

# Points superflus
__[i>/typo(typo_et_cetera)__             etc(?:[.]{3,5}|…) <<- ->> etc.                             # Un seul point après « etc. »

TEST: {{etc…}}

TEST: Ligne de points{{...........................}}
TEST: Aujourd’hui, M. Camus va nous lire son texte.



!!
!!
!!!! Tirets                                                                                       !!
!!
!!

# Tirets, énumérations et dialogues
__<s>/typo(typo_tiret_début_ligne)__    ^[-_][  ] <<- ->> "— |– "           # Dialogues et énumérations : un tiret cadratin ou demi-cadratin, suivi d’un espace insécable, est requis.
__<s>/typo(typo_tiret_dans_dialogue)__  ^ *«[  ](-[  ]) @@$ <<- -1>> "— |– "
    # Dialogues : un tiret cadratin ou demi-cadratin, suivi d’un espace insécable, est requis. (Attention : à moins qu’il s’agisse d’un dialogue inclus dans un autre dialogue, cette manière d’écrire est erronée.)

TEST: {{—}}Viens là.                                                        ->> "— |– "
TEST: Cette phrase contient{{ - }}c’est certain – des tirets d’incise.      ->> " – | — "
TEST: « {{- }}Viens ! On va en finir avec cette affaire, sale fils de pute. ->> "— |– "


!!
!!
!!!! Ponctuations redondantes                                                                     !!
!!
!!

# virgules et points
__<s>/typo(typo_virgules_points)__
    ,[.,]{2,}
    <<- ->> =\0.replace(",", ".").replace("...", "…")                                               # Erreur de numérisation ? Virgules au lieu de points ?

TEST: Ah !{{,.,}} et en quoi consistait le festin ?


__<s>/typo(typo_ponctuation_superflue1)__    , ?([.,;:!?]) @@$ <<- ->> =",|" + \1       # Une de ces ponctuations est superflue.

TEST: Succession : a{{;}} b.                                                ->> " ;"
TEST: C’est assez {{: }}parlez !                                            ->> ": "
TEST: {{» }}C’est ce qu’on croit savoir.                                    ->> "» |« "


!!
!!
!!!! Signes typographiques                                                                        !!
!!
!!

__<s>/typo(typo_signe_multiplication)__
    (\d+) ?[x*] ?(\d+)  @@0,$ <<- not \0.startswith("0x") ->> \1 × \2                               # Signe de multiplication typographique.

TEST: {{800 x 600}}                                                         ->> 800 × 600

TEST: __liga__ un {{fl}}ottement                                           ->> fl
TEST: __liga__ une belle {{fi}}gure                                        ->> fi


!!
!!
!!!! Apostrophes                                                                                  !!
!!
!!
## Apostrophe typographique
__[i>/apos(apostrophe_typographique)__
    ([ldsncjmç]|jusqu|lorsqu|aujourd|presqu|quelqu|puisqu|qu|prud|entr)['´‘′`ʼ](?=[\w"«]) @@0 <<- ->> \1’    # Apostrophe typographique.
__[i>/apos(apostrophe_typographique_après_t)__
    t['´‘′`ʼ](?!ils?|elles?|on)(?=[\w"«]) <<- ->> t’                                                         # Apostrophe typographique.

TEST: {{C'}}est l’apostrophe typographique qui manque.                    ->> C’
TEST: C’est {{l'}}apostrophe typographique qui manque.                    ->> l’
TEST: {{Aujourd'}}hui, je suis fatigué
TEST: Je pense {{qu`}}elle se trompe
TEST: Que {{t'}}arrive-t-il ?


__[i>/typo(typo_apostrophe_incorrecte)__
    (([ldsncjmtç]|lorsqu|puisqu|presqu|quoiqu|quelqu|jusqu|qu|aujourd|entr)(?:[’'´‘′`ʼ][’'´‘′`ʼ ]|”|“|"| [’'´‘′`ʼ] ?))[aeéiouhœæyîèêôû]  @@0,0
    <<- not (\2.__len__() == 1  and  \1.endswith("′ ")) -1>> \2’                                    # Apostrophe incorrecte.

TEST: {{C’ }}est rageant.                                                 ->> C’
TEST: {{Lorsqu’ }}il vient, on s’ennuie.
TEST: {{l’'}}électricien est parti
TEST: Je pense {{qu ` }}elle ne viendra pas.

TEST: Il {{s }}était dit qu’il ne le ferait pas.                                ->> s’
TEST: Je {{m }}étais imaginé que je pourrais pas retrouver cet objet.           ->> m’
TEST: Je {{n }}ai pas retrouvé l’ambiance de mes années de lycée.               ->> n’


!!
!!
!!!! Guillemets typographiques                                                                    !!
!!
!!

__<s>/typo(typo_guillemets_typographiques_doubles_ouvrants)__
    "(?=\w)
    <<- not before(r"[a-zA-Zéïîùàâäôö]$") ->> « |“                                                  # Guillemets typographiques ouvrants.
__<s>/typo(typo_guillemets_typographiques_doubles_fermants)__

TEST: Le {{'}}grand’ homme.                                                 ->> “|‘|‹
TEST: Le ‘grand{{'}} homme.                                                 ->> ”|’|›
TEST: J’en ai '''marre''' (syntaxe wiki).


!!
!!
!!!! Élisions                                                                                     !!
!!
!!

__[i]/typo(typo_élision_déterminants)__
    (l[ea] |de )([aâeéêiîoôu]\w+)  @@0,$
    <<- not re.search("(?i)^(?:onz[ei]|énième|iourte|ouistiti|ouate|one-?step|ouf|Ouagadougou|I(?:I|V|X|er|ᵉʳ|ʳᵉ|è?re))", \2) and not \2.isupper() and not morph(\2, ":G", False)
    -1>> =\1[0]+"’"                                         # Élision de l’article devant un mot commençant par une voyelle.|http://fr.wikipedia.org/wiki/Élision

    -1>> cet                                                # Euphonie : « ce » devient « cet » devant un mot commençant par une voyelle.

TEST: {{ce}} animal est dangereux


!!
!!
!!!! Divers                                                                                       !!
!!
!!

## NF (memo: don’t use flag i)
__[s]/nf(nf_norme_française)__
    [nN][fF][  -]?(?:c|C|e|E|p|P|q|Q|s|S|x|X|z|Z|[eE][nN](?:[  -][iI][sS][oO]|))[  -]?[0-9]+(?:[/ ‑-][0-9]+|)
    <<- not re.search("^NF (?:C|E|P|Q|S|X|Z|EN(?: ISO|)) [0-9]+(?:‑[0-9]+|)", \0)

    <<- ->> =\0.replace("2", "₂").replace("3", "₃").replace("4", "₄")                               # Typographie des composés chimiques. [!]

TEST: __chim__ les molécules {{CaCO3}} et {{H2O}}…


!!
!!
!!!! Cohérence des guillemets                                                                     !!
!!
!!

__<s>/typo(typo_cohérence_guillemets_chevrons_ouvrants)__
    («)[^»“]+?(”)  @@0,$
    <<- not before(r"\w$") -1>> “                                                                   # Guillemet ouvrant différent du guillemet fermant (”).
    <<- not after(r"^\w") -2>> " »"                                                                 # Guillemet fermant différent du guillemet ouvrant («).

TEST: « Je suis là.
TEST: — Oui, nous t’avons vu arriver. »
TEST: « J’en suis “malade”. »


!!
!!
!!!! Espaces insécables avant unités de mesure                                                    !!
!!
!!

__[s]/unit(unit_nbsp_avant_unités1)__
    ((\d+(?:,\d+[⁰¹²³⁴⁵⁶⁷⁸⁹]?|[⁰¹²³⁴⁵⁶⁷⁸⁹]|)) ?)(?:[kcmµn]?(?:[slgJKΩ]|m[²³]?|Wh?|Hz|dB)|[%‰€$£¥Åℓhj]|min|°C|℃)(?![’'])  @@0,0
    <<- -1>> "\2 "
    # Avec une unité de mesure, mettez un espace insécable.

TEST: 1998 s’annonce grandiose.
TEST: il y en a 3,43 milliards.
TEST: je veux 200 euros.


!!
!!
!!!! Grands nombres                                                                               !!
!!
!!

__[s]/num(num_grand_nombre_soudé)__
    \d\d\d\d+
    <<- not before("NF[  -]?(C|E|P|Q|X|Z|EN(?:[  -]ISO|)) *$") >>>
    <<- \0.__len__() > 4 ->> =formatNumber(\0)                                                      # Formatage des grands nombres.

    <<- ~>> =\0.replace(" ", "_")

TEST: Il a perdu {{20 000}} euros à la Bourse en un seul mois.


!!
!!
!!!! Dates                                                                                        !!
!!
!!
__[i]/date(date_nombres)__
    (?<!\d[ /.-])(\d\d?)[ /.-](\d\d?)[ /.-](\d\d\d+)(?![ /.-]\d)  @@0,w,$
    <<- not checkDate(\1, \2, \3) and not before(r"(?i)\bversions? +$") ->> _                       # Cette date est invalide.
    <<- ~>> =\0.replace(".", "-").replace(" ", "-").replace("\/", "-")

TEST: le {{29 02 2011}}
TEST: le {{40-02-2011}}
TEST: le {{32.03.2018}}
TEST: le {{81/01/2012}}
TEST: 12-12-2012


!!
!!
!!!! Redondances                                                                                  !!
!!
!!
__[i]/redon1(redondances_paragraphe)__
    ({w_4})[  ,.;!?:].*[  ](\1)  @@0,$
    <<- not morph(\1, ":(?:G|V0)|>(?:t(?:antôt|emps|rès)|loin|souvent|parfois|quelquefois|côte|petit|même)/", False) and not \1[0].isupper()
    -2>> _                                                      # Dans ce paragraphe, répétition de « \1 » (à gauche).
    <<- __also__ -1>> _                                         # Dans ce paragraphe, répétition de « \1 » (à droite).

TEST: __redon1__ Tu es son {{avenir}}. Et lui aussi est ton {{avenir}}.
TEST: __redon1__ Car parfois il y en a. Mais parfois il n’y en a pas.




!!!
!!!
!!! Processeur: Dernier nettoyage avant coupure du paragraphe en phrases                          !!
!!!
!!!

# Trait d’union conditionnel (u00AD)
__<i>(p_trait_union_conditionnel1)__    \w+‑\w+‑\w+ <<- ~>> =\0.replace("‑", "")
__<i>(p_trait_union_conditionnel2)__    \w+‑\w+ <<- ~>> =\0.replace("‑", "")

# empêcher la scission en fin de dialogue
__<s>(p_fin_dialogue1)__    ([?!…][?!…  ]*)[ "'”» ]*,  @@0 <<- ~1>> *
__<s>(p_fin_dialogue2)__    ([?!…][?!…  ]*)[ "'”» ]*[a-zéèêîô]  @@0 <<- ~1>> ,

TEST: « Je suis donc perdu ? », dit Paul.
TEST: “C’est bon !”, croit savoir Marie.
TEST: “Parce que… ?” finit par demander Paul.
TEST: « Dans quel pays sommes-nous ? » demanda un manifestant.


!!
!!
!!
!!
!!

!!
!!
!!
!!
!!
!!
!!
!! PASSE 1: PHRASE PAR PHRASE                                                                     !!
!!
!!
!!
!!
!!
!!
!!

!!
!!

[++]



!!!! Doublons (casse identique)                                                                   !!

__[s](doublon)__
    ({w1}) {1,3}\1  @@0
    <<- not re.search("(?i)^([nv]ous|faire|en|la|lui|donnant|œuvre|h[éoa]|hou|olé|joli|Bora|couvent|dément|sapiens|très|vroum|[0-9]+)$", \1)
        and not (re.search("^(?:est|une?)$", \1) and before("[’']$"))
        and not (\1 == "mieux" and before("(?i)qui +$"))
    ->> \1   # Doublon.

TEST: Il y a un {{doublon doublon}}.


!!!! Nombres: typographie                                                                         !!

#(\d\d\d\d)-(\d\d\d\d)   <<- ->> \1–\2                              # Ne pas séparer deux dates par un trait d’union, mais par un tiret demi-cadratin.

__[s]/num(num_lettre_O_zéro1)__  [\dO]+[O][\dO]+ <<- not option("ocr") ->> =\0.replace("O", "0")    # S’il s’agit d’un nombre, utilisez le chiffre « 0 » plutôt que la lettre « O ».
__[s]/num(num_lettre_O_zéro2)__  [1-9]O <<- not option("ocr") ->> =\0.replace("O", "0")             # S’il s’agit d’un nombre, utilisez le chiffre « 0 » plutôt que la lettre « O ».

TEST: année {{2O11}}                                                      ->> 2011
TEST: {{3O}} (chiffre avec un O).                                         ->> 30


# Nombres ordinaux
__[s]/typo(typo_ordinaux_premier)__
    1 ?(?:ier|i?ère)s?
    <<- ->> =\0.replace(" ", "").replace("è", "").replace("i", "").replace("e", "ᵉ").replace("r", "ʳ").replace("s", "ˢ")
    # Nombre ordinal. Premier : 1ᵉʳ. Première : 1ʳᵉ. Premiers : 1ᵉʳˢ. Premières : 1ʳᵉˢ.|http://bdl.oqlf.gouv.qc.ca/bdl/gabarit_bdl.asp?id=4271
__[s]/typo(typo_ordinaux_deuxième)__
    2 ?nde?s?
    <<- ->> =\0.replace(" ", "").replace("n", "").replace("d", "ᵈ").replace("e", "ᵉ").replace("s", "ˢ")
    # Nombre ordinal. Second : 2ᵈ. Seconde : 2ᵈᵉ. Seconds : 2ᵈˢ. Secondes : 2ᵈᵉˢ.|http://bdl.oqlf.gouv.qc.ca/bdl/gabarit_bdl.asp?id=4271
__[s]/typo(typo_ordinaux_nième)__
    ([0-9]+) ?(?:è|i?[èe]me)s?  @@0
    <<- \0.endswith("s") ->> \1ᵉˢ                                                                   # Nombre ordinal pluriel. Exemples : 2ᵉˢ, 3ᵉˢ, 4ᵉˢ…|http://bdl.oqlf.gouv.qc.ca/bdl/gabarit_bdl.asp?id=4271
    <<- __else__ ->> \1ᵉ                                                                            # Nombre ordinal singulier. Exemples : 2ᵉ, 3ᵉ, 4ᵉ…|http://bdl.oqlf.gouv.qc.ca/bdl/gabarit_bdl.asp?id=4271
__[s]/typo(typo_ordinaux_romain_premier)__
    I ?(?:ier|i?ère)s?
    <<- ->> =\0.replace(" ", "").replace("è", "").replace("i", "").replace("e", "ᵉ").replace("r", "ʳ").replace("s", "ˢ")
    # Nombre ordinal romain. Premier : 1ᵉʳ. Première : Iʳᵉ. Premiers : Iᵉʳˢ. Premières : Iʳᵉˢ.|http://bdl.oqlf.gouv.qc.ca/bdl/gabarit_bdl.asp?id=4271
__[s]/typo(typo_ordinaux_romain_deuxième)__
    II ?nde?s?
    <<- ->> =\0.replace(" ", "").replace("n", "").replace("d", "ᵈ").replace("e", "ᵉ").replace("s", "ˢ")
    # Nombre ordinal romain. Second : IIᵈ. Seconde : IIᵈᵉ. Seconds : IIᵈˢ. Secondes : IIᵈᵉˢ.|http://bdl.oqlf.gouv.qc.ca/bdl/gabarit_bdl.asp?id=4271
__[s]/typo(typo_ordinaux_romains_nième)__
    ([IVXLCM]{1,3}) ?(?:è|i?[èe]me)s?  @@0
    <<- not morph(\0, ":G", False) >>>

TEST: le {{IInd}} siècle                        ->> IIᵈ
TEST: le {{VIième}} siècle                      ->> VIᵉ
TEST: Le {{XXIème}} siècle.                     ->> XXIᵉ
TEST: le {{XXè}} siècle.                        ->> XXᵉ



!!!! Écritures épicènes invariables                                                               !!

__[i](d_typo_écriture_épicène_pluriel)__
    ({w_1}[éuitsrn])_(?:[nt]|)e_s  @@0
    <<- morphex(\1, ":[NAQ]", ":G") =>> define(\1, [":N:A:Q:e:p"])

__[i](d_typo_écriture_épicène_singulier)__
    ({w_2}[éuitsrn])_e  @@0
    <<- morph(\1, ":[NAQ]", False) =>> define(\1, [":N:A:Q:e:s"])


!!!! Dates                                                                                        !!

__[i]/date(date_jour_mois_année)__
    (\d\d?) (janvier|février|ma(?:rs|i)|a(?:vril|o[ûu]t)|jui(?:n|llet)|septembre|octobre|novembre|décembre) (\d\d\d+)  @@0,w,$
    <<- not checkDateWithString(\1, \2, \3) ->> _                                                   # Cette date est invalide.

TEST: {{29 février 2011}}


__[i]/date(date_journée_jour_mois_année1)__
    (lundi|m(?:ardi|ercredi)|jeudi|vendredi|samedi|dimanche),? (?:le |)(\d\d?)-(\d\d?)-(\d\d+)  @@0,w,w,$
    <<- not after(r"^ +av(?:ant|) +J(?:C|ésus-Christ)") and not checkDay(\1, \2, \3, \4)
    -1>> =getDay(\2, \3, \4)                                                                        # Le jour de la date suivante est incorrect.

TEST: {{mercredi}}, le 10-06-2014                                         ->> mardi


__[i]/date(date_journée_jour_mois_année2)__
    (lundi|m(?:ardi|ercredi)|jeudi|vendredi|samedi|dimanche),? (?:le |)(\d\d?) (janvier|février|ma(?:rs|i)|a(?:vril|o[ûu]t)|jui(?:n|llet)|septembre|octobre|novembre|décembre) (\d\d+)  @@0,w,w,$
    <<- not after(r"^ +av(?:ant|) +J(?:C|ésus-Christ)") and not checkDayWithString(\1, \2, \3, \4)
    -1>> =getDayWithString(\2, \3, \4)                                                              # Le jour de la date suivante est incorrect.

TEST: {{mercredi}}, le 10 juin 2014                                       ->> mardi
TEST: {{lundi}}, 18 août 1912                                             ->> dimanche
TEST: lundi, 18 août 1912 avant Jésus-Christ (date imaginaire)


__[i]/date(date_mois_31)__ 31 (avril|juin|septembre|novembre)  @@3
    <<- ->> 30 \1                                                                                   # Cette date est invalide. Il n’y a que 30 jours en \1.

TEST: le {{31 avril}}


__[i]/date(date_février)__ 3[01] février
    <<- ->> 28 février|29 février                                                                   # Cette date est invalide. Il n’y a que 28 ou 29 jours en février.

TEST: le {{30 février}}



!!!
!!!
!!! Processeur: épuration des signes inutiles et quelques simplifications                         !!
!!!
!!!

# fin de phrase
__<s>(p_fin_de_phrase)__        [.?!:;…][ .?!… »”")]*$  <<- ~>> *

# début de phrase

__[i](p_références_aux_notes)__ [a-zéèâàôîù][a-zéèâàôîù-]+(\d+) @@$ <<- not morph(\0, ":", False) ~1>> *

# faux positifs avec adverbes de négation
__[i](p_pas_mal)__              pas mal <<- not morph(word(-1), ":D", False) ~>> *
__[i](p_pas_assez)__            pas assez ({w_2}) @@$ <<- morph(\1, ":A", False) and not morph(word(-1), ":D", False) ~>> *

# faux positifs avec «à chez»
__[i](p_de_chez_à_chez_pronom)__    de chez \w+ (?:à|jusqu à) chez (?:moi|toi|lui|elles?|eux|nous|vous) <<- ~>> *
__[i](p_de_chez)__                  (jusqu à|de) chez @@0 <<- ~1>> *

# faux positifs
__[i](p_en_tout_et_pour_tout)__ en tout et pour tout <<- ~>> *
__[i](p_au_sortir_de)__         au (sortir) de?s?  @@3 <<- ~1>> *
__[i](p_au_revoir)__            au revoir <<- ~>> au_revoir

TEST: New York {{étaient}} {{devenue}} la plaque tournante de tous les trafics.


!!
!!
!!!! Traits d’union                                                                               !!
!!
!!

__<i]/tu(tu_t_euphonique_incorrect)__
    ([-–—− ]t(?:[’' ][-–—−]?|[-–—−][’' ]?))(ils?|elles?|on|tu)  @@0,$
    <<- re.search("(?i)^(?:ils|elles|tu)$", \2) -1>> -      # Le “t” euphonique n’est pas nécessaire avec “\2”.|http://bdl.oqlf.gouv.qc.ca/bdl/gabarit_bdl.asp?T1=t+euphonique&id=2513
    <<- __else__ and \1 != "-t-" and \1 != "-T-" -1>> -t-   # Pour le “t” euphonique, il faut deux traits d’union. Pas d’apostrophe. Pas d’espace.
    <<- \1 != "-t-" ~1>> -t-
__<i]/tu(tu_t_euphonique_superflu)__
    [td]([- ]t[-’' ])(?:il|elle|on)  @@1
    <<- -1>> -                                              # Le “t” euphonique est superflu quand le verbe se termine par “t” ou “d”.|http://bdl.oqlf.gouv.qc.ca/bdl/gabarit_bdl.asp?T1=t+euphonique&id=2513
    <<- \1 != "-t-" ~1>> -t-
__<i]/tu(tu_t_euphonique_manquant)__
    [aec](-(il|elle|on))  @@1,2  <<- -1>> -t-\2             # Il faut un “t” euphonique.|http://bdl.oqlf.gouv.qc.ca/bdl/gabarit_bdl.asp?T1=t+euphonique&id=2513

TEST: va{{ t’}}il y parvenir ?                          ->> -t-
TEST: A{{ t’}}elle soif ?                               ->> -t-
TEST: A{{ t-}}elle faim ?                               ->> -t-
TEST: a{{ t'}}elle                                      ->> -t-

TEST: Cet {{in folio}} est incroyable.
TEST: c’est {{in}} stage de réinsertion.
TEST: Dans ce drive in douze hommes mangent.


__[i]/tu(tu_préfixe_mi)__
    mi ({w2}) @@$  <<- morph(\1, ":[NAQ]", False) ->> mi-\1                                         # S’il s’agit d’un seul mot, il manque un trait d’union.

TEST: J’ai été engagé pour un {{mi temps}}.


__[i]/tu(tu_préfixe_quasi)__
    (?:l(?:es?|a|eurs?)|ce(?:tte|t|s|)|des?|m(?:a|on|es)|[ts](?:es|a)) (quasi ({w1}))  @@$,$
    <<- morphex(\2, ":N", ":[AGW]") -1>> quasi-\2                                                   # Il manque un trait d’union : « quasi » s’accroche au mot qui le suit s’il s’agit d’un nom.
    <<- ~1>> \2

TEST: leurs {{quasi indifférences}}


__[i]/tu(tu_préfixe_semi)__

__[i]/tu(tu_là_adverbe)__       là (bas|haut|desso?us|dedans|devant|derrière) @@3 <<- ->> là-\1     # Il manque un trait d’union.
__[i]/tu(tu_n_est_ce_pas)__             n’est(?: ce[ -]pas|-ce-pas) <<- ->> n’est-ce pas            # Un trait d’union.
__[i]/tu(tu_nord_sud_ouest)__           (nord|sud) ouest @@0 <<- ->> \1-ouest                       # Il manque un trait d’union.
__[i]/tu(tu_nord_sud_est)__             (nord|sud) est @@0 <<- isEnd() ->> \1-est                   # Il manque un trait d’union.
__[i]/tu(tu_outre_mer)__                outre mer <<- ->> outre-mer                                 # Il manque un trait d’union.
__[i]/tu(tu_ouï_dire)__                 ou[iï] dire <<- morph(word(-1), ":G") ->> ouï-dire          # Il manque un trait d’union.
__[i]/tu(tu_par_préposition)__
    par (desso?us|devant|delà|derrière|dehors|dedans|devers)  @@$
    <<- ->> par-\1                                                                                  # Il manque un trait d’union.
__[i]/tu(tu_par_ci_par_là)__            par ci,? par là <<- ->> par-ci par-là|par-ci, par-là        # Trait(s) d’union manquant(s).
__[i]/tu(tu_prêt_à_porter)__
    (prêts?) à porter  @@0
    <<- before(r"(?i)\b(?:les?|du|des|un|ces?|[mts]on) +") ->> \1-à-porter                          # Il manque les traits d’union.
__[i]/tu(tu_plate_forme)__  plates? formes? <<- ->> plate-forme|plates-formes|plateforme|plateformes    # Il manque un trait d’union. Vous pouvez aussi souder les deux mots.
__[i]/tu(tu_quelques_uns_unes)__       quelques (une?s) @@$ <<- ->> quelques-\1                     # Il manque un trait d’union.
__[i]/tu(tu_plus_moins_values)__       (plus|moins) (values?)  @@0,$  <<- ->> \1-\2                 # Il manque un trait d’union.
__[i]/tu(tu_rez_de_chaussée)__         rez de chaussées? <<- ->> rez-de-chaussée                    # Il manque un trait d’union.
__[i]/tu(tu_science_fiction)__         science fiction <<- ->> science-fiction                      # Il manque un trait d’union.
__[i]/tu(tu_stock_option)__            stock options? <<- ->> =\0.replace(" ", "-")                 # Il manque un trait d’union.

TEST: on a fait {{à la va vite}}…
TEST: marre de faire des {{allers retours}}
TEST: {{le}} {{bouche à oreille}}
TEST: à ce {{moment là}}                                                        ->> moment-là
TEST: une {{plus value}}                                                        ->> plus-value
TEST: Il est {{en-dessous}} de tout.                                            ->> en dessous
TEST: Ils sont {{en-deçà}} de tout                                              ->> en deçà
TEST: {{Là bas}}.                                                               ->> Là-bas
TEST: {{Au dessus}}                                                             ->> Au-dessus
TEST: {{ci dessus}}                                                             ->> ci-dessus
TEST: {{par dessus}}                                                            ->> par-dessus
TEST: {{au delà}}                                                               ->> au-delà
TEST: {{ci devant}}.                                                            ->> ci-devant
TEST: Il faisait froid cet {{hiver la}}.                                        ->> hiver-là
TEST: {{Ceux la}} vont à la pêche.                                              ->> Ceux-là

TEST: des {{stock options}}
TEST: Un autre chantier important, celui si sensible de la préservation des données personnelles des élèves


# est-ce … ?
__[i]/tu(tu_est_ce)__
    (?<![cCdDlL][’'])(est ce) ({w_2})  @@0,$
    <<- morphex(\2, ":", ":N.*:[me]:[si]|>qui/") and morph(word(-1), ":Cs", False, True)
    -1>> est-ce                                                                                     # S’il s’agit d’une interrogation, il manque un trait d’union.

TEST: {{est ce}} que c’est grave ?                                              ->> est-ce
TEST: qu’{{est ce}} que c’est ?                                                 ->> est-ce
TEST: elles reviendront, {{n’est ce pas}} ?
TEST: nous en sommes à l’étape où nous voulons définir ce qu’est ce projet

__[u]/tu(tu_Nouveau_Brunswick)__        Nouveau Brunswick <<- ->> Nouveau-Brunswick                     # Il manque un trait d’union.
__[u]/tu(tu_Nouvelle_Qqch)__            Nouvelle (Zélande|Calédonie|Orléans) @@$ <<- ->> Nouvelle-\1    # Il manque un trait d’union.
__[u]/tu(tu_Pas_de_Calais)__            Pas de Calais <<- ->> Pas-de-Calais                             # Il manque les traits d’union.
__[u]/tu(tu_Rhône_Alpes)__              Rhône Alpes <<- ->> Rhône-Alpes                                 # Il manque les traits d’union.
__[u]/tu(tu_Saône_et_Loire)__           Saône et Loire <<- ->> Saône-et-Loire                           # Il manque un trait d’union.
__[u]/tu(tu_Jésus_Christ)__             Jésus Christ <<- ->> Jésus-Christ                               # Il manque un trait d’union.
__[u]/tu(tu_Jean_prénom)__
    Jean (Baptiste|Claude|François|Jacques|Louis|Luc|Marc|Marie|Michel|Paul|Philippe|Pierre)  @@$
    <<- ->> Jean-\1                                                                                 # Il manque un trait d’union.
__[u]/tu(tu_Marie_Prénom)__
    Marie (Ange|Agnès|Anne|Antoinette|Cécile|Chantal|Charlotte|Christine|Claire|Claude|Dominique|France|Françoise|Hélène|Jeanne|José|Josèphe|Line|Louise|Madeleine|Noëlle|Odile|Paule|Pierre|Rose|Thérèse)  @@$
    <<- ->> Marie-\1                                                                                # Il manque un trait d’union.
__[s]/tu(tu_St_Ste_Bidule)__
    ((Ste?) )[A-ZÉÈÎ]\w+ @@0,0
    <<- -1>> \2-
    # Il manque un trait d’union s’il s’agit d’une église, d’une cité, d’une communauté… S’il s’agit d’une personne, écrivez « saint(e) » sans majuscule, sans trait d’union.

TEST: Émigrer au {{Royaume Uni}}

TEST: {{Parce-qu}}’ils y croient plus que tout.
TEST: Oui, ça{{-aussi}}.
TEST: Peu {{d’entre-nous}} savent ce dont il s’agit.


__[i]/tu(tu_y_attaché)__
    (y[’-])({avoir_etre})(?:-(?:t-|)(?:ils?|elles?|je|tu|on|nous|vous)|) @@0,2
    <<- morph(\2, ":V0|>en/", False) -1>> "y "                                                      # Ici, ni apostrophe, ni trait d’union.

TEST: {{Y’}}a trop de malheureux sur Terre.
TEST: {{Y’}}en a marre, de ces conneries.
TEST: {{y-}}a-t-il des beignets ?                     ->> "y "


__[i]/tu(tu_lorsque)__
    lors que?
    <<- not before(r"(?i)\bd[eè]s +$") ->> =\0.replace(" ", "")                                     # Attachez les deux mots.|https://fr.wiktionary.org/wiki/lorsque

TEST: Elle y arriva {{lors qu}}’elle trouva l’astuce permettant l’ouverture de la porte.
TEST: Dès lors qu’on sait comment s’y prendre, aucune raison de faillir.


!!!! Virgules                                                                                     !!

# Dialogues
__[u]/virg(virgule_dialogue_après_nom_propre)__
    ([A-ZÉÈ][\w-]+) (\w+-(?:moi|toi|l(?:ui|a|e(?:ur|s|))|nous|vous|je|tu|ils|elles))  @@0,$
    <<- morphex(\1, ":M", ":G") and not morph(\2, ":N", False) and isStart()
    -1>> \1,                                                                        # Dialogue ? Ajoutez une virgule pour mettre en incise la personne à qui s’adresse la réplique.

TEST: donne{{-le-moi}} Camille.
TEST: donne-moi Alice.
TEST: Tape-toi Patrick.


__[u]/virg(virgule_après_verbe_COD)__
    l(?:es?|a) ({w_2}(?:[ei]r|re)) ([A-ZÉÂÔÈ][\w-]+)  @@w,$
    <<- morph(\1, ":Y", False) and morph(\2, ":M", False) and not morph(word(-1), ">à/", False, False)
    -1>> \1,                                                                                        # Une virgule est probablement souhaitable.

TEST: Tu vas les {{donner}} Rachel.
TEST: Il va la {{tuer}} Paul.
TEST: Cependant les promesses n’engagent que ceux qui les croient, comme aimait à le dire Jacques Chirac.



!!!! Apostrophe manquante (2)                                                                     !!

__<s>/typo(typo_apostrophe_manquante_audace2)__
    ^ *([LDSNCJMTÇ] )[aeéiouhAEÉIOUHyîèêôûYÎÈÊÔÛ]  @@*
    <<- option("mapos") -1>> =\1[:-1]+"’"                                                           # Il manque peut-être une apostrophe.

TEST: __mapos__ {{L }}opinion des gens, elle s’en moquait.



!!!! A / À: accentuation la préposition en début de phrase                                        !!

__<s]/typo(typo_À_début_phrase1)__
    ^ *(A) (?!t[’-](?:ils?|elles?|on))({w_2})  @@*,$
    <<- morphex(\2, ":[GNAY]", ":(?:Q|3s)|>(?:priori|post[eé]riori|contrario|capella|fortiori)/")
        or (\2 == "bientôt" and isEnd())
    -1>> À                                                                                          # S’il s’agit de la préposition « à », il faut accentuer la majuscule.
__<s>/typo(typo_À_début_phrase2)__
    ^ *(A) [ldnms]’  @@*  <<- -1>> À                                                                # S’il s’agit de la préposition « à », il faut accentuer la majuscule.
__<s>/typo(typo_À_début_phrase3)__
    ^ *(A) t’(?!il |elle |ont? )  @@*  <<- -1>> À                                                   # S’il s’agit de la préposition « à », il faut accentuer la majuscule.

TEST: {{A}} vaincre sans péril, on triomphe sans gloire.
TEST: « {{A}} partir de maintenant, ce ne sera plus comme avant.
TEST: — {{A}} n’en plus pouvoir
TEST: — {{A}} t’emmener loin de tout ceci.
TEST: {{A}} bientôt, mon ami.
TEST: A bientôt fini son devoir.
TEST: A priori, nul ne peut y parvenir sans une aide extérieure.
TEST: A devient notre meilleure chance d’y parvenir.


!!!! Accentuation des majuscules                                                                  !!

__[u]/maj(maj_accents)__
    E(?:tat|glise|co(?:le|nomie)|quipe|lectri(?:cité|que)|gal(?:ité|ement)|té)s? @@1
    <<- ->> ="É"+\0[1:]                                                                            # Accentuez les majuscules.
    <<- ~>> ="É"+\0[1:]

TEST: Le budget de l’{{Etat}}.



!!!
!!!
!!! Désambiguïsation                                                                              !!
!!!
!!!

# mots grammaticaux
__[i](d_dans)__
    dans
    <<- not morph(word(-1), ":D.*:p|>[a-z]+ièmes/", False, False) =>> select(\0, ":R")

__[i](d_ton_son)__
    (\w+) ([ts]on)  @@0,$
    <<- morph(\1, ">(?:le|ce[st]?|ton|mon|son|quel(?:que|)s?|[nv]otre|un|leur|ledit|dudit)/") =>> exclude(\2, ":D")

# Pronoms le/la/les
__[i](d_je_le_la_les)__
    je (l(?:e(?:ur|s|)|a)) @@$                  <<- not morph(word(-1), ":1s", False, False) =>> select(\1, ":Oo")
__[i](d_tu_le_la_les)__
    tu (l(?:e(?:ur|s|)|a)) @@$                  <<- not morph(word(-1), ":2s", False, False) =>> select(\1, ":Oo")
__[i](d_il_elle_on_le_la_les)__

__[i](d_tu_verbe)__
    tu +(?:l(?:es? +|la +|’)|[nv]ous +|)({w_1}) @@$
    <<- morph(word(-1), ":Cs", False, True) and not morph(\1, ":(?:Oo|X)", False) =>> select(\1, ":[123][sp]")
__[s](d_nom_propre_verbe)__
    ([A-ZÉÈ]{w_1}) +({w_1})  @@0,$
    <<- morph(\1, ":M") and \2.islower() and morphex(\2, ":[123][sg]", ":Q") and morph(\2, ":N", False) and morph(word(-1), ":Cs", False, True)
    =>> select(\2, ":[123][sp]")
    <<- morph(\1, ":M", False) and morphex(\2, ":[123]s|>(?:[nmts]e|nous|vous)/", ":A") and isStart() =>> =select(\1, ":M")
__[i](d_que_combien_pourquoi_en_y_verbe)__
    (?:que?|combien|pourquoi) +(?:en +|y +|)({w_3}) @@$
    <<- =>> exclude(\1, ":E")

# groupe nominal
__[i](d_aucun_non_verbe)__
    aucun +({w_4}) @@$

    <<- =>> exclude(\1, ":V")
__[i](d_par_non_verbe)__
    par +({w_3}) @@$
    <<- =>> exclude(\1, ":V[123]")
__[i](d_très_non_verbe)__
    très +({w_2}) @@$
    <<- =>> exclude(\1, ":[123][sp]")
__[i](d_en_tant_que)__
    en tant que +({w_2}) @@$
    <<- =>> exclude(\1, ":[123][sp]")

# divers
__[i](p_bac_plus_nombre)__
    bac ?([+] ?\d\d?) @@$
    <<- ~1>> *
    <<- =>> define(\0, [":N:e:i"])


TEST: il s’agit d’{{un}} {{anagramme}}
TEST: nul ne sait qui arriva à ce pauvre Paul surpris par la pluie.
TEST: elle finit par être très fière de son fils.
TEST: en tant que président du conseil




!!
!!
!!!! OCR                                                                                          !!
!!
!!

# ?
__<s]/ocr(ocr_point_interrogation)__
    [  ]7
    <<- after0("^(?: +[A-ZÉÈÂ(]|…|[.][.]+| *$)") ->> " ?"                                           # Erreur de numérisation ?

# cette / cotte
__[i]/ocr(ocr_cette2)__     cotte <<- not morph(word(-1), ":D.*:f:[si]") ->> cette                  # Erreur de numérisation ?

TEST: __ocr__ on poirautait, {{cotte}} mariée n’arrivait pas à se décider.


# Comme / Gomme
__[s]/ocr(ocr_comme)__      Gomme <<- not morph(word(1), ">(?:et|o[uù])/") ->> Comme                # Erreur de numérisation ?

TEST: __ocr__ {{Gomme}} il était sage à cette époque-là !


# Comment / Gomment
__[s]/ocr(ocr_comment)__    Gomment <<- ->> Comment                                                 # Erreur de numérisation ?

# Mais / Hais / Mats / niais
__[u]/ocr(ocr_mais1)__      Hais <<- ->> Mais                                                       # Erreur de numérisation ?
__[i]/ocr(ocr_mais2)__      mats <<- not morph(word(-1), ":D:[me]:p", False, False) ->> mais        # Erreur de numérisation ?
__[i]/ocr(ocr_mais3)__      maïs <<- not morph(word(-1), ":D:(?:m:s|e:p)", False, False) ->> mais   # Erreur de numérisation ?
__[s]/ocr(ocr_mais4)__
    niais <<- not morph(word(-1), ">(?:homme|ce|quel|être)/", False, False) ->> mais                # Erreur de numérisation ?

TEST: __ocr__ {{Hais}} il en sait trop.
TEST: __ocr__ c’était bien, {{mats}} quelle journée
TEST: __ocr__ c’est bien, {{niais}} trop subtil.
TEST: __ocr__ c’est parfait, {{maïs}} trop subtil.

## Casse
__[s]/ocr(ocr_casse1)__
    [A-ZÉÈÂÊÎÔ]{w_1}
    <<- \0.istitle() and before(r"(?i)\w") >>>
    <<- morphex(\0, ":G", ":M") ->> =\0.lower()                                                     # Erreur de numérisation ? Casse improbable.
    <<- __else__ and morphex(\0, ":[123][sp]", ":[MNA]|>Est/") ->> =\0.lower()                      # Erreur de numérisation ? Casse improbable.

TEST: __ocr__ votre ami la regarde, {{Vous}} ne l’avez pas achetée
TEST: __ocr__ pour accommoder son regard, {{La}} lourde forme demeure
TEST: __ocr__ parler de Nicole, {{Le}} sommeil ne vient pas.
TEST: __ocr__ a fait de toi, Charles, {{Tu}} étais beau quand
TEST: __ocr__ s’habituer à se faire servir, {{Au}} début ça
TEST: __ocr__ Tu as tué ce petit garçon, Henri, {{Et}} tu le sais.

TEST: __ocr__ trouve {{l£}} temps
TEST: __ocr__ elle s’{{avance*}} sur le seuil
TEST: __ocr__ par beaucoup d’argent ? {{{Il}} débouche le Jack Daniels


!!
!!
!!!! Incohérences de base                                                                         !!
!!
!!

### double négation
__[i](double_négation)__
    pas (personne|aucune?|jamais)  @@4
    <<- not morph(word(-1), ":D:[me]" ,False, False) ->> \1|pas, \1

TEST: {{Ces}} {{cette}} canaille qui nous a donné tant de fil à retordre.
TEST: Mon {{il}} est une merveille.


__[s](incohérence_globale_au_qqch)__
    ([aA]u) ({w2})  @@0,$
    <<- not \2.isupper() >>>
    <<- morph(\2, ">(?:[cdlmst]es|[nv]os|cettes?|[mts]a|mon|je|tu|ils?|elle?|[vn]ous|on|parce)/", False)
    -2>> =suggSimil(\2, ":[NA].*:[si]", True)                                                       # Incohérence : les mots “\1” et “\2” ne devraient pas se succéder.
    <<- __else__ and morph(\2, ">quelle/", False) ->> auquel|auxquels|auxquelles                    # Incohérence. Soudez les deux mots.|https://fr.wiktionary.org/wiki/auquel
    <<- __else__ and \2 == "combien" and morph(word(1), ":[AY]", False) -1>> ô                      # Incohérence probable.|https://fr.wiktionary.org/wiki/%C3%B4_combien

TEST: au {{nos}} enfants.
TEST: {{Au quel}} faut-il s’adresser ?
TEST: Au MES, rien de nouveau.


__[s](incohérence_globale_aux_qqch)__
    ([aA]ux) ({w2})  @@0,$
    <<- not \2.isupper() >>>
    <<- morph(\2, ">(?:[cdlmst]es|[nv]os|cettes?|[mts]a|mon|je|tu|ils?|elle?|[vn]ous|on|parce)/", False)
    -2>> =suggSimil(\2, ":[NA].*:[pi]", True)                                                       # Incohérence : les mots “\1” et “\2” ne devraient pas se succéder.
    <<- __else__ and morph(\2, ">quelle/", False) ->> auxquels|auxquelles                           # Incohérence. Soudez les deux mots.|https://fr.wiktionary.org/wiki/auquel
    <<- __else__ and \2 == "combien" and morph(word(1), ":[AY]", False) -1>> ô                      # Incohérence probable.|https://fr.wiktionary.org/wiki/%C3%B4_combien

TEST: ils jouent aux {{des}}.
TEST: {{Aux quels}} a-t-il adressé sa requête. ?
TEST: Des individus {{aux}} combien sensibles aux usages.


__[s](incohérences_globales3)__
    ([dD]es) ([cdlmst]es|[nv]os|cettes?|[mts]a|mon|je|tu|ils?|elle?|[vn]ous|on)  @@0,$
    <<- -2>> =suggSimil(\2, ":[NA].*:[pi]", True)                                                   # Incohérence : les mots “\1” et “\2” ne devraient pas se succéder.
    <<- -1>> de                                                                                     # Incohérence : les mots “\1” et “\2” ne devraient pas se succéder.

TEST: je ne sais {{des}} {{ses}} choses.



!!
!!
!!!! Style                                                                                        !!
!!
!!

#__bs__  Mr <<- ->> M.                          # M. est l’usage courant pour “Monsieur”. « Mr » est l’abréviation ancienne, française.

# à / en
__[i]/bs(bs_en_à_ville)__
    (en) A(?:gen|miens|ngers|jjacio|rles|vignon)  @@0
    <<- -1>> à       # On utilise la préposition “à” avant les villes (à Avignon, à Arles…), la préposition “en” avant les régions (en Amérique, en Afrique…).

TEST: {{En}} Avignon
TEST: {{En}} Agen


# avoir été
__[i]/bs(bs_avoir_été_chez)__
    (?<!l’)({avoir}) été chez  @@0
    <<- not re.search("(?i)^avoir$", \1) and morph(\1, ">avoir/", False)
    ->> _                                                                                           # Tournure familière. Utilisez « être allé ».

TEST: J’{{ai été chez}} le coiffeur.
TEST: Chez les intellectuels, le mot utopie n’a jamais été synonyme de folie, mais il l’a été pour l’homme de la rue.


# abyme / abîme
__[i]/bs(bs_mise_en_abyme)__
    mis\w+ en (ab[îi]mes?) @@$ <<- -1>> abyme
    # Classiquement, on écrit plutôt « abyme » dans cette expression.|https://fr.wiktionary.org/wiki/mise_en_abyme

TEST: La mise en {{abîme}}.


# à date / jusqu’à date
__[i]/bs(bs_à_date)__
    ({etre}|m\w+) ([aà] date)  @@0,$  <<- morph(\1, ">(?:être|mettre)/", False) -2>> à jour         # Anglicisme incompris hors du Québec.
__[i]/bs(bs_jusquà_date)__
    jusqu [àa] date <<- ->> jusqu’ici|jusqu’à maintenant|jusqu’à ce jour|à ce jour                  # Anglicisme incompris hors du Québec.

TEST: être {{à date}}
TEST: mettre {{a date}}
TEST: {{jusqu’à date}}

#Se rappelle de l’amour

#enjoindre à qqn de faire qqch


!!
!!
!!!! Pléonasmes                                                                                   !!
!!
!!

__[i]/pleo(pleo_abolir)__               (abol\w+) (?:absolument|entièrement|compl[èé]tement|totalement) @@0 <<- morph(\1, ">abolir/", False) ->> \1         # Pléonasme.
__[i]/pleo(pleo_acculer)__              (accul\w+) aux? pieds? du mur @@0 <<- morph(\1, ">acculer/", False) ->> \1                                          # Pléonasme.
__[i]/pleo(pleo_achever)__              (ach[eè]v\w+) (?:absolument|entièrement|compl[èé]tement|totalement) @@0 <<- morph(\1, ">achever/", False) ->> \1    # Pléonasme.
__[i]/pleo(pleo_en_cours)__             actuellement en cours <<- not after(r" +de?\b") ->> en cours                                            # Pléonasme.
__[i]/pleo(pleo_en_train_de)__          (actuellement en train) d(?:e(?! nuit)|’{w_2}) @@0 <<- -1>> en train                                    # Pléonasme.
__[i]/pleo(pleo_ajouter)__              (ajout\w+) en plus @@0 <<- ->> \1                                                                       # Pléonasme.
__[i]/pleo(pleo_apanage)__              (apanages?) exclusifs? @@0 <<- ->> \1                                                                   # Pléonasme.
__[i]/pleo(pleo_applaudir)__            (applaudi\w+) des deux mains @@0 <<- ->> \1                                                             # Pléonasme.
__[i]/pleo(pleo_aujourd_hui)__          au jour d’aujourd’hui <<- ->> aujourd’hui                                                               # Pléonasme.
__[i]/pleo(pleo_avancer)__              (avan[cç]\w+) en avant @@0 <<- morph(\1, ">avancer/", False) ->> \1                                     # Pléonasme.
__[i]/pleo(pleo_s_avérer)__             s’av([éè]r\w+) vrai(e?s?) @@4,$ <<- ->> s’av\1 exact\2                                                  # Pléonasme.
__[i]/pleo(pleo_avéré)__                (avérée?s?) vraie?s? @@0 <<- ->> \1                                                                     # Pléonasme.
__[i]/pleo(pleo_avenir)__               avenir devant (?:lui|[mts]oi|eux|[nv]ous) <<- morph(word(-1), ":A|>un", False) ->> avenir               # Pléonasme.
__[i]/pleo(pleo_bourrasque)__           (bourrasques?) de vent @@0 <<- ->> \1                                                                   # Pléonasme.
__[i]/pleo(pleo_car_en_effet)__         car en effet <<- ->> car|en effet                                                                       # Pléonasme.
__[i]/pleo(pleo_cirrhose)__             (cirrhoses?) du foie @@0 <<- ->> \1                                                                     # Pléonasme.
__[i]/pleo(pleo_collaborer)__           (collabor\w+) ensemble @@0 <<- morph(\1, ">collaborer/", False) ->> \1                                  # Pléonasme.
__[i]/pleo(pleo_comme_par_exemple)__    comme par exemple <<- ->> comme|par exemple                                                             # Pléonasme.
__[i]/pleo(pleo_comparer)__             (compar\w+) entre (?:eux|elles) @@0 <<- morph(\1, ">comparer/", False) ->> \1                           # Pléonasme.
__[i]/pleo(pleo_contraindre)__          (contrai\w+) malgré (?:soi|eux|lui|moi|elle|toi) @@0 <<- morph(\1, ">contraindre/", False) ->> \1       # Pléonasme.
__[i]/pleo(pleo_descendre)__            (descend\w+) en bas(?! de) @@0 <<- ->> \1                                                               # Pléonasme.
__[i]/pleo(pleo_dessiner)__             (dessin\w+) un dessin @@0 <<- ->> \1                                                                    # Pléonasme.
__[i]/pleo(pleo_dorénavant)__           à (?:partir|compter) de dorénavant <<- ->> dorénavant|à partir de maintenant                            # Pléonasme.
__[i]/pleo(pleo_donc_par_conséquent)__  donc par conséquent <<- ->> donc|par conséquent|c’est pourquoi                                          # Pléonasme.
__[i]/pleo(pleo_enchevêtrer)__          (enchevêtr\w+) les uns dans les autres @@0 <<- morph(\1, ">enchevêtrer/", False) ->> \1                 # Pléonasme.
__[i]/pleo(pleo_entraider)__            (entraid\w+) (?:mutuellement|les uns les autres) @@0 <<- morph(\1, ">entraider/", False) ->> \1         # Pléonasme.
__[i]/pleo(pleo_entraide)__             (entraides?) mutuelles? @@0 <<- ->> \1                                                                  # Pléonasme.
__[i]/pleo(pleo_erreur)__               (erreurs?) involontaires? @@0 <<- ->> \1                                                                # Pléonasme.
__[i]/pleo(pleo_étape)__                (étapes?) intermédiaires? @@0 <<- ->> \1                                                                # Pléonasme.
__[i]/pleo(pleo_hasard)__               (hasards?) imprévus? @@0 <<- ->> \1                                                                     # Pléonasme.
__[i]/pleo(pleo_hémorragie)__           (hémorragies?) de sang @@0 <<- ->> \1                                                                   # Pléonasme.
__[i]/pleo(pleo_joindre)__              (join\w+) ensemble @@0 <<- morph(\1, ">joindre/") ->> \1|mettre ensemble                                # Pléonasme.
__[i]/pleo(pleo_lever)__                lever debout <<- ->> lever                                                                              # Pléonasme.
__[i]/pleo(pleo_mais_qqch)__            mais (cependant|pourtant|toutefois) @@5 <<- ->> mais|cependant|pourtant|toutefois                       # Pléonasme.
__[i]/pleo(pleo_marche)__               (marches?) à pieds? @@0 <<- ->> \1                                                                      # Pléonasme.
__[i]/pleo(pleo_méandre)__              (méandres?) sinueux @@0 <<- ->> \1                                                                      # Pléonasme.
__[i]/pleo(pleo_media)__                (m[eé]dias?) d’informations? @@0 <<- ->> \1                                                             # Pléonasme.
__[i]/pleo(pleo_monopole)__             (monopoles?) exclusifs? @@0 <<- ->> \1                                                                  # Pléonasme.
__[i]/pleo(pleo_monter)__               (mont\w+) en haut(?! d[eu’]) @@0 <<- morph(\1, ">monter/", False) ->> \1                                # Pléonasme.
__[i]/pleo(pleo_opportunité)__          (opportunités?) à saisir @@0 <<- ->> \1                                                                 # Pléonasme.
__[i]/pleo(pleo_orage)__                (orages?) électriques? @@0 <<- ->> \1                                                                   # Pléonasme.
__[i]/pleo(pleo_jumelles)__             paires? de jumelles? <<- ->> jumelles                                                                   # Pléonasme.
__[i]/pleo(pleo_panacée)__              (panacées?) universelles? @@0 <<- ->> \1|remède universel                                               # Pléonasme.
__[i]/pleo(pleo_perspective)__          (perspectives?) d’avenir @@0 <<- ->> \1                                                                 # Pléonasme.
__[i]/pleo(pleo_balbutiement)__         premiers? (balbutiements?) @@$ <<- ->> \1                                                               # Pléonasme.
__[i]/pleo(pleo_priorité)__             premières? (priorités?) @@$ <<- ->> \1                                                                  # Pléonasme.
__[i]/pleo(pleo_projet1)__              (projets?) futurs? @@0 <<- ->> \1                                                                       # Pléonasme.
__[i]/pleo(pleo_projet2)__              futurs? (projets?) @@$ <<- ->> \1                                                                       # Pléonasme.
__[i]/pleo(pleo_prototype)__            (prototypes?) expérimenta(?:l|ux) @@0 <<- ->> \1                                                        # Pléonasme.
__[i]/pleo(pleo_rénover)__              (rénov\w+) à neuf @@0 <<- morph(\1, ">rénov(?:er|ation)/", False) ->> \1                                # Pléonasme.
__[i]/pleo(pleo_puis_qqch)__            puis (?:après|ensuite|alors) <<- ->> puis|après|ensuite|alors                                           # Pléonasme.
__[i]/pleo(pleo_réunir)__               (réuni\w*) ensemble @@0 <<- morph(\1, ">réunir/", False) ->> \1                                         # Pléonasme.
__[i]/pleo(pleo_reculer)__              (recul\w*) en arrière @@0 <<- morph(\1, ">recul(?:er|)/", False) ->> \1                                 # Pléonasme.
__[i]/pleo(pleo_risque)__               (risques?) (?:potentiels?|de menaces?) @@0 <<- ->> \1                                                   # Pléonasme.
__[i]/pleo(pleo_secousse)__             (secousses?) sé?ismiques? @@0 <<- ->> secousse tellurique|secousses telluriques|tremblement de terre    # Pléonasme.
__[i]/pleo(pleo_solidaire)__            (solidaires?) les uns des autres @@0 <<- ->> \1                                                         # Pléonasme.
__[i]/pleo(pleo_suffire)__              (suffi\w+) simplement @@0 <<- morph(\1, ">suffire/", False) ->> \1                                      # Pléonasme.
__[i]/pleo(pleo_talonner)__             (talonn\w+) de près @@0 <<- morph(\1, ">talonner/", False) ->> \1                                       # Pléonasme.
__[i]/pleo(pleo_taux_alcoolémie)__      taux d’alcoolémies? @@7 <<- ->> taux d’alcool|alcoolémie                                                # Pléonasme. L’alcoolémie est le taux d’alcool dans le sang.
__[i]/pleo(pleo_tunnel)__               (tunnels?) souterrains? @@0 <<- ->> \1                                                                  # Pléonasme.
__[i]/pleo(pleo_hardes)__               vieilles hardes <<- ->> hardes                                                                          # Pléonasme.
__[i]/pleo(pleo_voire_même)__           voire même <<- ->> voire|même                                                                           # Pléonasme.|https://fr.wiktionary.org/wiki/voire_m%C3%AAme

TEST: il faut {{abolir totalement}} ces pratiques   ->> abolir
TEST: il faut {{achever complètement}} ce projet    ->> achever

TEST: {{vieilles hardes}}                           ->> hardes
TEST: {{voire même}}                                ->> voire|même


# d’avance / à l’avance
__[i]/pleo(pleo_verbe_à_l_avance)__
    ((?:pré[pvds]|pressen|pronostiqu|réserv|dev(?:an[cç]|in)|avert)\w+) (?:d’avance|à l’avance)  @@0
    <<- morph(\1, ">(?:prévenir|prévoir|prédire|présager|préparer|pressentir|pronostiquer|avertir|devancer|deviner|réserver)/", False)
    ->> \1                                                                                                              # Pléonasme.

TEST: {{prédire à l’avance}}                  ->> prédire
TEST: {{pronostiquer d’avance}}               ->> pronostiquer
TEST: {{réserver d’avance}}                         ->> réserver


# plus tard / à une date ultérieure
__[i]/pleo(pleo_différer_ajourner_reporter)__
    ((?:diff|ajourn|report)\w+) à (?:plus tard|date ultérieure|une date ultérieure)  @@0
    <<- morph(\1, ">(?:ajourner|différer|reporter)/", False)
    ->> \1                                                                                                              # Pléonasme.

TEST: {{Ajourner à une date ultérieure}}      ->> Ajourner
TEST: {{différer à une date ultérieure}}      ->> différer
TEST: {{reporter à plus tard}}                ->> reporter

TEST: {{ta}} aimée                                            ->> ton
TEST: {{ma}} obligée                                          ->> mon
TEST: Ce couple va donner à la France sa très importante collection qui rejoindra le musée d’Orsay


!!
!!
!!!! Confusions                                                                                   !!
!!
!!

__[s>/conf(conf_ne_n)__     [nN]e n’                        <<- ->> ne m’|n’                        # Incohérence. Double négation.
__[s>/conf(conf_pronoms1)__ [mtMT]e ([nmst](?:’|e )) @@$    <<- ->> \1                              # Incohérence.
__[s>/conf(conf_pronoms2)__ [sS]e ([mst](?:’|e )) @@$       <<- ->> \1                              # Incohérence.
__[s>/conf(conf_de_d)__     [dD][eu] d’(?![A-ZÉÂÔÈ])        <<- ->> d’                              # Incohérence.

TEST: Il {{ne n’}}arrive jamais à l’heure.
TEST: Ça {{me te }}prend la tête, toutes ces complications vaines.
TEST: il {{se m’}}est difficile d’y parvenir.
TEST: Ça t’arrive {{de d’}}arriver à l’heure ?
TEST: Les écrits de d’Alembert.

TEST: toute chose pourrait finalement advenir indifféremment en n’importe quel point du globe


__[i]/conf(conf_malgré_le_la_les_leur)__
    malgré l(?:es? +|eurs? +|a +|’)({w_3})  @@$
    <<- morphex(\1, ":", ":[GNAWMB]") -1>> =suggSimil(\1, ":[NA]", True)                             # Incohérence : après “malgré”, on devrait trouver un groupe nominal.

TEST: malgré l’{{arrête}} qui interdisait le port
TEST: malgré les deux précédentes erreurs


__[i]/conf(conf_ma_ta_cette_verbe)__
    ([mt]a|cette) +({w_2})  @@0,$
    <<- morphex(\2, ":V.*:(?:Y|[123][sp])", ":[NAQ]") and \2[0].islower()
    -2>> =suggSimil(\2, ":[NA]:[fe]:[si]", True)                                                    # Incohérence avec « \1 » : « \2 » est un verbe.

TEST: d’une {{habille}} femme
TEST: plus d’un ont été traumatisés
TEST: plus d’une sont parties aussi vite qu’elles étaient venues


__[i]/conf(conf_il_on_pas_verbe)__
    (?<!t’)(?:il|on) (?:l’|l(?:es?|a|eur|ui) +|[nv]ous +|)({w_2}) @@$
    <<- morphex(\1, ":", ":(?:[123][sp]|O[onw]|X)|>ou/") and morphex(word(-1), ":", ":3s", True)
    -1>> =suggSimil(\1, ":(?:3s|Oo)", False)                                                        # Incohérence : « \1 » devrait être un verbe, un pronom objet, un adverbe de négation, etc.

TEST: il {{et}} parti.


__[i]/conf(conf_ils_pas_verbe)__
    (?<!t’)ils (?:l’|l(?:es?|a|eur|ui) +|[nv]ous +|)({w_2}) @@$
    <<- morphex(\1, ":", ":(?:[123][sp]|O[onw]|X)|>ou/") and morphex(word(-1), ":", ":3p", True)
    -1>> =suggSimil(\1, ":(?:3p|Oo)", False)                                                        # Incohérence avec « ils » : « \1 » devrait être un verbe, un pronom objet, un adverbe de négation, etc.

TEST: ils {{son}} du même bois.
TEST: Ils {{étai}} partie au {{restaurent}}


__[i]/conf(conf_je_pas_verbe)__

# Trop de faux positifs
#__[i]/conf__
#    très +(bien|\w+ent) +({w2})  @@w,$
#    <<- morph(\1, ":W", False) and morphex(\2, ":[123][sp]", ":[GAQW]") -2>> _
#    # Incohérence avec « très » : « \2 » n’est ni un adjectif, ni un participe passé, ni un adverbe.


__[i]/conf(conf_très_verbe)__
    très +(?!envie)({w_2})  @@$
    <<- morphex(\1, ":(?:Y|[123][sp])", ":[AQW]") -1>> =suggSimil(\1, ":[AW]", True)                # Incohérence avec « très » : « \1 » n’est ni un adjectif, ni un participe passé, ni un adverbe.
    <<- morph(\1, ">jeûne/", False) -1>> =\1.replace("û", "u")                                      # Confusion. Le jeûne est une privation de nourriture.|https://fr.wiktionary.org/wiki/jeune

TEST: Il est très {{cite}}.
TEST: très {{suivit}} par ce détective
TEST: il était très {{habille}}
TEST: Très {{jeûne}}, elle a su qu’elle ne voulait pas d’une vie ordinaire.

TEST: Ces gens sont si {{prit}} par leur travail qu’ils en oublient de vivre.
TEST: Ça ira mieux demain, surtout si émerge une demande forte de la part des consommateurs.


__[i]/conf(conf_de_plus_en_plus_verbe)__
    de plus en plus +({w_2})  @@$
    <<- morphex(\1, ":(?:[123][sp]|Y)", ":(?:[GAQW]|3p)") and not morph(word(-1), ":V[123].*:[123][sp]|>(?:pouvoir|vouloir|falloir)/", False, False)
    -1>> =suggVerbPpas(@)
    # Incohérence avec « de plus en plus » : « \1 » n’est ni un adjectif, ni un participe passé, ni un adverbe.

TEST: de plus en plus {{gagnait}} par la folie.
TEST: de plus en plus {{concerner}} par ce problème
TEST: avec quel zèle ils remplissent leurs devoirs de citoyens (de plus en plus deviennent enseignant·e·s, infirmier·e·s ou s’engagent dans l’armée, etc.).
TEST: il faut de plus en plus aider ces gens.

__[i]/conf(conf_a_à_tout_à_coup)__  tout h?ah? co[uû][pt]?                                          <<- ->> tout à coup # Confusion.
__[i]/conf(conf_a_à_coup_sûr)__     a coup s[uû]r                                                   <<- ->> à coup sûr  # Confusion.
__[i]/conf(conf_a_à_face_à)__       face (a) @@5    <<- not before(r"(?i)\b(?:[lmts]a|leur|une|en) +$") -1>> à          # Confusion.
__[i]/conf(conf_a_à_pas_à_pas)__    pas (a) pas @@4                                                 <<- -1>> à          # Confusion.
__[i]/conf(conf_a_à_par_rapport)__  par rapport (a) ({w_2}) @@12,$  <<- morph(\2, ":(?:D|Oo|M)", False) -1>> à          # Confusion.
__[i]/conf(conf_a_à_être_à)__
    ({etre}) (a)(?! priori| posteriori| fortiori)  @@0,$
    <<- morph(\1, ">être/:V") and not before(r"(?i)\bce que? ") -2>> à                                # Confusion. Utilisez la préposition « à ».
__[i]/conf(conf_a_à_peu_près)__
    (?:a peu[tx]? (?:près|prés?|prêts?)|à peu[tx] (?:près|prés?|prêts?)|à peu (?:prés?|prêts?))
    <<- ->> à peu près                                                                              # Confusion.
    <<- ~>> *
__[i]/conf(conf_a_à_pronoms1)__     ne +l(?:es?|a) +(?:l(?:eur|ui) +|)(à)  @@$  <<- -1>> a          # Confusion : “à” est une préposition. Pour le verbe avoir, écrivez :
__[i]/conf(conf_a_à_pronoms2)__     ne +[nv]ous +(?:les +|l’|)(à)  @@$          <<- -1>> a          # Confusion : “à” est une préposition. Pour le verbe avoir, écrivez :
__[i]/conf(conf_a_à_pronoms3)!6__   [mtsn]e l(?:es +|’)(à)  @@$                 <<- -1>> a          # Confusion : “à” est une préposition. Pour le verbe avoir, écrivez :

    <<- re.search("(?i)^(?:côtés?|coups?|peu-près|pics?|propos|valoir|plat-ventrismes?)", \2)
    ->> l’à-\2                                                                                      # Il faut un trait d’union.
__[i]/conf(conf_a_à_il_on_à)__
    (?:il|on) +(?:l(?:es +|’)|en +|y +(?:en +|)|[vn]ous +|)(à)  @@$
    <<- not morph(word(-1), ":3s", False, False) -1>> a                                             # Confusion probable : “à” est une préposition. Pour le verbe avoir, écrivez :
__[i]/conf(conf_a_à_elle_à)__
    elle +(?:l(?:es +|’)|en +|y +(?:en |)|[vn]ous +|)(à)  @@$
    <<- not morph(word(-1), ":(?:3s|R)", False, False) and not morph(word(1), ":Oo|>qui/", False, False)
    -1>> a                                                                                          # Confusion probable : “à” est une préposition. Pour le verbe avoir, écrivez :
__[i]/conf(conf_a_à_qui_pronom_à)__
    qui (?:l(?:ui|eur)(?: en|)|nous|vous|en|y) +(à)  @@$ <<- -1>> a                                 # Confusion : “à” est une préposition. Pour le verbe avoir, écrivez :
__[i]/conf(conf_a_à_qui_a)__
    qui (à) +({w_2})  @@4,$  <<- morphex(\2, ":Q", ":M[12P]") -1>> a                                # Confusion : “à” est une préposition. Pour le verbe avoir, écrivez :
__[i]/conf(conf_a_à_le)__
    (à le) ({w_2})  @@0,5
    <<- morphex(\2, ":[NAQ].*:[me]", ":(?:Y|Oo)") -1>> a le|au
    # Confusion. Ou vous confondez “a” (préposition) avec “a” (avoir), ou vous devez employer “au”. Exemple : _à_ midi, il _a_ pris son déjeuner _au_ restaurant.
    <<- morphex(\2, ":[NAQ].*:f", ":(?:Y|Oo)") -1>> à la
    # Confusion. Ou vous confondez “a” (préposition) avec “a” (avoir), ou vous devez employer “au”. Exemple : _à_ midi, il _a_ pris son déjeuner _au_ restaurant.
__[i]/conf(conf_a_à_les)__
    (à les) ({w_2})  @@0,6  <<- morphex(\2, ":[NAQ]", ":(?:Y|Oo)") -1>> a les|aux
    # Confusion. Ou vous confondez “a” (préposition) avec “a” (avoir), ou vous devez employer “aux”. Exemple : Elle _a_ donné du travail _aux_ employés _à_ profusion.
__[i]/conf(conf_a_à_au_aux)__
    (à) aux?(?! (?:moins|plus)) @@0 <<- -1>> a                                                      # Confusion : “à” est une préposition. Pour le verbe avoir, écrivez :
__[i]/conf(conf_a_à_base_cause)__
    (a) (?:base|cause) d(?:es?|u|) @@0 <<- not before(r"(?i)\bce que?\b") -1>> à                    # Confusion. Utilisez la préposition « à ».
__[i]/conf(conf_a_à_faim_peur_honte_soif)__
    (à) +(?:faim|peur|honte|soif) @@0 <<- -1>> a                                                    # Confusion : “à” est une préposition. Pour le verbe avoir, écrivez :
__[i]/conf(conf_a_à_part)__
    (a) part ({w1}) @@0,7 <<- morph(\2, ":(?:M[12]|D|Oo)") -1>> à                                   # Confusion probable.
__[i]/conf(conf_a_à_les_à)__
    les (à)(?! côtés| peu près| prioris?| post[eé]rioris?| valoirs?| pics?| propos) @@4 <<- -1>> a
    # Confusion : “à” est une préposition. Pour le verbe avoir, écrivez :
__[i]/conf(conf_a_à_avant_conj_prep)__
    (à) +(?:a(?:fin|lors|près|uprès|ux?(?! moins| plus| mieux)|vant|vec)|au-de(?:dans|hors|là|sso?us|vant)|chez|d(?:ans|evant|ès|déjà|onc|urant)|lorsque?|malgré|par(?:ce|mi|)|p(?:endant|our|uisque)|que?|sur|tandis)  @@0
    <<- -1>> a                                                                                      # Confusion probable : “à” est une préposition. Écrivez “a” pour la conjugaison de “avoir”.
__[i]/conf(conf_a_participe_passé_ou_vconj)__
    (à) +({w_2})  @@0,$
    <<- \2.islower() and \2 != "coté" >>>
    <<- morphex(\2, ":(?:V.......[_z][az].*:Q|V1.*:Ip:2p)", ":[MGWNY]") -1>> a                      # Confusion probable : “à” est une préposition. Pour le verbe “avoir”, écrivez :

    -1>> à                  # Confusion probable : “a” est la conjugaison du verbe “avoir”. Utilisez la préposition “à”.|http://fr.wiktionary.org/wiki/%C3%A0
__[s]/conf(conf_a_à_locutions4)__
    (a) (?:[mts](?:es|on|a)|[nv]o(?:s|tre)|leurs?) (?:avis|c(?:onnaissance|ôtés)|c(?:œur|orps) défendant|dé(?:pens|triment)|disposition|encontre|égard|grand(?: désarroi|e (?:surprise|tristesse))|guise|insu|portée|risques et périls|sujet|tour)  @@0
    <<- not before(r"(?i)(?:\bque? |[ln]’$|(?<!-)\b(?:il|elle|on|y|n’en) +$)")
    -1>> à                  # Confusion probable : “a” est la conjugaison du verbe “avoir”. Utilisez la préposition “à”.|http://fr.wiktionary.org/wiki/%C3%A0
__[s]/conf(conf_a_à_infi)__
    (?<![mtn]’)(a) (?:[mts](?:’(?:en +|y +|)|e +)|[nv]ous|)({w_2}(?:er|re|ir))  @@0,2
    <<- morph(\2, ":Y", False) and not before(r"(?i)\bque? |(?:il|elle|on|n’(?:en|y)) +$")
    -1>> à                  # Confusion probable : “a” est la conjugaison du verbe “avoir”. Utilisez la préposition “à”.|http://fr.wiktionary.org/wiki/%C3%A0
__[s]/conf(conf_a_à_après_interrogative)__
    \w+-(?:je|ils?|elles?|je|tu|on|vous|nous) (a)(?! priori| posteriori| fortiori)  @@$
    <<- -1>> à              # Confusion probable : “a” est la conjugaison du verbe “avoir”. Utilisez la préposition “à”.|http://fr.wiktionary.org/wiki/%C3%A0


TEST: Cette femme {{à}} été confondue avec une autre.

TEST: Je tenais à le lui faire savoir.
TEST: il va falloir songer à les leur donner
TEST: le coût est estimé à au moins 1000 milliards de dollars
TEST: elle à qui nous avons donné notre cœur et pour laquelle nous avons tout tenté pour la faire aimer de nos concitoyens.
TEST: jusqu’à parfois se mettre en danger
TEST: les méthodes de l’à peu près ne suffisent pas.

## substantifs
__[i](conf_arc_à_poulies)__
    arcs? (a) poulies @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_armes_à_feu)__
    armes? (a) feu @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_bombe)__
    bombes? (a) (?:eau|fragmentation|retardement) @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_canne_à)__
    (cann?es?) (a) (?:pêche|sucre)s? @@0,w
    <<- \1.find("ane") != -1 -1>> =\1.replace("ane", "anne")    # Confusion : la cane est la femelle du canard.|https://fr.wiktionary.org/wiki/canne
    <<- \2 == "a" -2>> à                                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_caisse_à_outils)__
    caisses? (a) outils? @@w
    <<- \1 == "a" -1>> à                                # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_chair_à)__
    chairs? (a) (?:pâté|canons?) @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_crayon_à_qqch)__
    crayons? (a) (?:lèvres|sourcils) @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_cuillère)__
    cuillères? (a) (?:café|soupe) @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_fard_à_paupières)__
    fards? (a) paupières @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_fils_fille_à)__
    fil(?:les?|s) (a) (?:papa|maman) @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_gaz_à_effet_de_serre)__
    gaz (a) effet de serre  @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_lime_à_ongles)__
    limes? (a) ongles? @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_machine_à_qqch)__
    machines? (a) (?:café|coudre|laver|remonter le temps) @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_moule_à_qqch)__
    moules? (a) (?:gaufres?|gâteaux?|brioche|manqué|savarin|charlotte) @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](p_papier_à)__
    papiers? (a) (?:lettres?|musique) @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_remire_à_plat)__
    remise?s? (a) plat @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_rouge_à_lèvres)__
    rouges? (a) lèvres @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_sac_à_qqch)__
    sacs? (a) (?:dos|main|langer|merde|foutre) @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_silo)__
    silos? (a) (?:grains?|blé) @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_soue_à_cochons)__
    soues? (a) cochons? @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_trou_à_rat)__
    trous? (a) rats? @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_tueur_à_gages)__
    tueu(?:r|se)s? (a) gages @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_vache_à_lait)__
    vaches? (a) lait @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_vente_à_domicile)__
    ventes? (a) domicile @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_vernis_à_ongles)__
    vernis (a) ongles @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_vol_à)__
    vols? (a) l(?:’étalage|a (?:sauvette|tire)) @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.
__[i](conf_usine_à_gaz)__
    usines? (a) gaz @@w
    <<- \1 == "a" -1>> à                        # Confusion : “a” est une conjugaison du verbe “avoir”. Pour la préposition, écrivez “à”.

TEST: C’est un arc {{a}} poulies
TEST: Ceci est une arme {{a}} feu.
TEST: Combien de bombes {{a}} fragmentation
TEST: C’est une nouvelle caisse {{a}} outils
TEST: Elle va en faire de la chair {{a}} pâté
TEST: Une cuillère {{a}} café.
TEST: Un crayon {{a}} sourcils
TEST: Elle a mis du fard {{a}} paupières.
TEST: ce fils {{a}} papa est d’un ennui mortel.
TEST: cette fille {{a}} papa nous pourrit la vie.
TEST: Les conséquences des gaz {{a}} effet de serre.
TEST: Où est la lime {{a}} ongles ?
TEST: devant la machine {{a}} café.
TEST: Achète un moule {{a}} gaufres.
TEST: Fais la mise {{a}} jour
TEST: Amenez-moi du papier {{a}} lettres.
TEST: Va acheter du papier {{a}} musique.
TEST: Elle mit du rouge {{a}} lèvres.
TEST: on a besoin d’une remise {{a}} plat.
TEST: passe-moi mon sac {{a}} dos.
TEST: dans le silo {{a}} grain.
TEST: sa chambre, c’est une vraie soue {{a}} cochons.
TEST: Un tueur {{a}} gages ne s’y prendrait jamais comme ça.
TEST: Ventes {{a}} domicile insatisfaisantes
TEST: Ce n’est que du vernis {{a}} ongles.
TEST: le vol {{a}} l’étalage est un fléau.
TEST: Il nous faut une {{voix}} de recours.
TEST: c’est une putain d’usine {{a}} gaz.
TEST: ce produit est notre vache {{a}} lait
TEST: ils vont et viennent, toujours {{cotes a cotes}}…
TEST: Nous irons {{tours à tours}} chercher du bois.


__[i]/conf(conf_celui_celle_à_qui)__
    ce(?:lui|lles?|ux) (a) qui  @@w
    <<- -1>> à                                  # Confusion. Utilisez la préposition “à”.|http://fr.wiktionary.org/wiki/%C3%A0

TEST: Ceux {{a}} qui nous avons notifié le problème sont partis.
TEST: Il y a qui au dîner ce soir ?


__[i]/conf(conf_mener_à_bien)__
    (m[eèé]n\w+) (a) bien  @@0,w
    <<- morph(\1, ">mener/", False) and ( not before(r"\bque? ") or morph(word(-1), ">(?:falloir|aller|pouvoir)/", False, True) )
    -2>> à                  # Confusion probable. Dans cette locution, utilisez la préposition « à ».|https://fr.wiktionary.org/wiki/mener_%C3%A0_bien
    <<- __also__ ~>> \1

TEST: Mener {{a}} bien cette guerre sera plus difficile qu’on le pense.
TEST: Je peux mener {{a}} bien cette opération.
TEST: Cette coalition que tu penses mener a bien l’intention de te trahir.


__[i]/conf(conf_mettre_à_profit)__
    (m(?:i[st]|ett)\w*).* (a) profit  @@0,w
    <<- morph(\1, ">mettre/", False) -2>> à     # Confusion probable. Dans « mettre à profit », utilisez la préposition « à ».|https://fr.wiktionary.org/wiki/mettre_%C3%A0_profit

TEST: Mettre {{a}} profit ses compétences
TEST: Il a mis son talent {{a}} profit.


__[s]/conf(conf_m_a_tuer)__
    m’a +(tuer) @@$

# ça / çà / sa
__[i]/conf(conf_ça_sa)__
    (ça) ({w_2}) @@0,3 <<- morph(\2, ":[NAQ].*:f") and not re.search("^seule?s?", \2) -1>> sa       # Confusion : “sa” (sa maison, sa passion) ≠ “ça” (ça vient, ça heurte).
__[i]/conf(conf_sa_ça1)__
    (sa) +({w_2}) @@0,$
    <<- morphex(\2, ":G", ">(?:tr(?:ès|op)|peu|bien|plus|moins|toute)/|:[NAQ].*:f") -1>> ça         # Confusion : “sa” (sa maison, sa passion) ≠ “ça” (ça vient, ça heurte).
__[i>/conf(conf_sa_ça2)__       (sa) +(?:[dnmtsjl]’|lorsqu |qu |puisqu )  @@0 <<- -1>> ça           # Confusion : “sa” (sa maison, sa passion) ≠ “ça” (ça vient, ça heurte).
__[i]/conf(conf_çà_ça)__        çà(?! et là) <<- not before(r"\b(?:[oO]h|[aA]h) +$") ->> ça         # Confusion : « çà » ne s’emploie plus guère que dans l’expression « çà et là ».
__[i]/conf(conf_çà_et_là)__     ça et là <<- not morph(word(-1), ":R") ->> çà et là                 # Confusion : « ça » équivaut à « cela ». Dans l’expression « çà et là », « çà » équivaut à « ici ».
__[s]/conf(conf_sa_fin)__       (sa) *$  @@0  <<- -1>> ça                           # Confusion probable : “sa” est un déterminant féminin singulier. Pour l’équivalent de “cela” ou “ceci”, écrivez :

TEST: Je prends {{sa}}…
TEST: {{ça}} patrie, il la détestait

TEST: sa toute nouvelle application de synchronisation de fichiers


# ce / se / ceux
__[s]/conf(conf_se_verbe)__
    ([cC]e) ({w_2})  @@0,3
    <<- \2[0].islower() and \2 != "faire"
        and ( morphex(\2, ":V[123].*:(?:Y|[123][sp])", ":[NAGM]|>(?:devoir|pouvoir|sembler)/") or re.search("-(?:ils?|elles?|on)$", \2) )
    -1>> se                                                                 # Confusion : « \2 » est un verbe. Exemples : ce bâtiment, se perdre.
__[i]/conf(conf_pour_ce_faire)__
    pour (se) faire,? ({w_2})  @@5,$
    <<- (\0.find(",") >= 0 or morphex(\2, ":G", ":[AYD]"))
    -1>> ce                                                                 # Confusion probable. Dans cette locution, il faut employer “ce”.|http://fr.wiktionary.org/wiki/pour_ce_faire
__[i]/conf(conf_ne_se1)!6__
    ne (ce) @@3 <<- -1>> se                                                 # Confusion. Ce chien, ce chat… Se demander, se croire…
__[i>/conf(conf_ne_se2)__
    ne ([cç]’) @@3 <<- -1>> s’                                              # Confusion. Ce chien, ce chat… Se demander, se croire…
__[i]/conf(conf_ce_conj_prep)__
    (se) (?:qu[ei]?|dont|malgré|pourquoi|avec|pour|par) @@0 <<- -1>> ce     # Confusion. Exemples : ce bâtiment, se perdre.|http://bdl.oqlf.gouv.qc.ca/bdl/gabarit_bdl.asp?id=2440
__[i]/conf(conf_qui_se_verbe)__
    qui (ce) ({w_2})  @@4,$
    <<- morphex(\2, ":V", ":[NAQ].*:[me]") or before(r"(?i)\b[cs]e +$")
    -1>> se                                                                 # Confusion probable. Exemples : ce bâtiment, se perdre.|http://bdl.oqlf.gouv.qc.ca/bdl/gabarit_bdl.asp?id=2440
__[i]/conf(conf_ceux_ce_être)__
    (ceux) (?:ne |)(?:sont|serai(?:en|)[ts]?|f[uû](?:ren|)t|n’(?!ayant|étant)\w+) @@0
    <<- -1>> ce                                                             # Confusion.|http://www.intellego.fr/soutien-scolaire-6eme/aide-scolaire-francais/ce-ceux-ou-se/3829
__[s]/conf(conf_ce_ne_être_doit)__
    ([sS]e) n(?:e |’)({être}|d[eouû]\w+|p[oeuû]\w+)  @@0,$
    <<- morph(\2, ">(?:être|pouvoir|devoir)/.*:3s", False)
    -1>> ce                                                                 # Confusion probable.|http://bdl.oqlf.gouv.qc.ca/bdl/gabarit_bdl.asp?id=2440
__[i]/conf(conf_ce_ne)__
    (ceux) ne ({w_2}) @@0,$
    <<- morphex(\2, ":[123]s", ":P") -1>> ce                                # Confusion.|http://www.intellego.fr/soutien-scolaire-6eme/aide-scolaire-francais/ce-ceux-ou-se/3829
__[i]/conf(conf_ce_nom1)__
    (se) ({w1}) @@0,3
    <<- morphex(\2, ":[NAQ]", ":([123][sp]|Y|P|Q)|>l[ea]?/") -1>> ce        # Confusion. Ce chien, ce chat… Se demander, se croire…
__[i]/conf(conf_ce_nom2)__
    (ceux) (?!l[aà] |qu[ie]? )({w_2}) @@0,$
    <<- morphex(\2, ":N.*:s", ":(?:A.*:[pi]|P|R)|>autour/") -1>> ce         # Confusion probable.|http://www.intellego.fr/soutien-scolaire-6eme/aide-scolaire-francais/ce-ceux-ou-se/3829

TEST: il ne {{ce}} compte pas parmi eux
TEST: il ne {{ç’}}avançait jamais sans avoir pesé toutes les conséquences
TEST: {{Se}} seraient des histoires.
TEST: {{se}} seraient des jours heureux.
TEST: {{Se}} sont de grands enfants.
TEST: {{Se}} sera une fille.
TEST: {{ceux}} seraient des jours heureux
TEST: Pour {{se}} faire, ils sont prêts à tout.
TEST: {{se}} ne peut être ainsi.
TEST: C’est tout {{se}} qu’il y a
TEST: Tout {{se}} que je fais
TEST: tout {{se}} qu’il entend
TEST: {{Ce}} {{promener}} est relaxant.
TEST: Il {{ce}} {{sent}} seul
TEST: {{se}} {{train}} est en retard
TEST: {{ceux}} train arrive trop vite.