accessanalyser/core.py

'''


@author: olivier.massot
'''
import re

from _regex_core import MULTILINE
from path import Path


# TODO: gérer les cas où des objets de types différents portent le même nom. Ex: une table et un formulaire... Note: une requete et une table ne peuvent pas porter le même nom.
# TODO: ignorer les commentaires dans les modules
# TODO: ignorer les labels dans les formulaires et états
# TODO: Gérer les références circulaires
# TODO: Stocker un aperçu du contexte de la ou des références dans le code source, pour contrôle ultérieur
# TODO: Permettre de supprimer / ajouter des réferences
# TODO: Verifier que la recherche puisse être Case sensitive?
def recurse(acc_obj):
    deptree = []
    for dep in acc_obj.deps:
        deptree.append(dep)
        if dep.deps:
            deptree += recurse(dep)
    return deptree

class InvalidFileExt(IOError):
    pass

class Mention():
    def __init__(self, line, objname, quote):
        self.line = line
        self.objname = objname
        self.quote = quote
        self.obj = None

class AccessObject():
    type_ = "<unknown>"
    _valid_file_exts = (".bas")

    def __init__(self, name_):
        self.name_ = name_
        self.functions = []
        self.sourcefile = ""
        self._sourcecode = ""

        self.mentions = []
        self.deps = {}

    def __repr__(self):
        return "<{}: {}>".format(self.type_, self.name_)

    @classmethod
    def from_file(cls, file):
        file = Path(file)
        if file.ext not in cls._valid_file_exts:
            raise InvalidFileExt("Format de fichier d'entrée non valide ({})".format(file.name))
        obj = cls(AccessObject.path_to_name(file))
        obj.sourcefile = file
        obj._sourcecode = file.text()
        obj._sourcecode.replace("\r\n", "\n")
        return obj

    @property
    def sourcecode(self):
        if not self._sourcecode:
            self._sourcecode = self.sourcefile.text()
        return self._sourcecode

    SUBSTR = {92: "\\", 47: "/", 58: ":", 42: "*", 63:"?", 34:"\"", 60:"<", 62:">", 124:"|" }
    @staticmethod
    def path_to_name(path):
        name_ = path.name.stripext()
        for ascii_code, char in AccessObject.SUBSTR.items():
            name_ = name_.replace("[{}]".format(ascii_code), char)
        return name_

class TableObject(AccessObject):
    type_ = "Table"
    _valid_file_exts = (".xml", ".lnkd")

class QueryObject(AccessObject):
    type_ = "Query"

class FormObject(AccessObject):
    type_ = "Form"

class ReportObject(AccessObject):
    type_ = "Report"

class MacroObject(AccessObject):
    type_ = "Macro"

class ModuleObject(AccessObject):
    type_ = "Module"

    @classmethod
    def from_file(cls, file):
        obj = super(ModuleObject, cls).from_file(file)
        rx = re.compile(r"Sub|Function ([^(]+)\(")
        obj.functions = [fname for fname in rx.findall(file.text()) if fname]
        return obj

class RelationObject(AccessObject):
    type_ = "Relation"
    _valid_file_exts = (".txt")

class Analyse():
    objects = []
    duplicated_names = []

    @classmethod
    def report(cls, current, total, msg=""):
        pass

    @classmethod
    def ended(cls):
        pass

    @classmethod
    def load_objects(cls, source_dir):
        source_dir = Path(source_dir)

        cls.objects = []
        cls.duplicated_names = []

        sourcemap = {
                    "forms": FormObject,
                    "reports": ReportObject,
                    "relations": RelationObject,
                    "scripts": MacroObject,
                    "queries": QueryObject,
                    "tables": TableObject,
                    "modules": ModuleObject,
                     }

        for dirname, accobj in sourcemap.items():
            for file in Path(source_dir / dirname).files():
                try:
                    obj = accobj.from_file(file)
                    if obj.name_ in [other.name_ for other in cls.objects] and not obj.name_ in cls.duplicated_names:
                        cls.duplicated_names.append(obj.name_)
                    cls.objects.append(obj)
                except InvalidFileExt:
                    print("Ignored unrecognized file: {}".format(file))

    @classmethod
    def find_deps(cls, subject):

        # On cherche le nom de chaque autre objet, ainsi que le nom des fonctions issues des modules
        look_for = [obj.name_ for obj in cls.objects if obj is not subject] + sum([obj.functions for obj in cls.objects if obj is not subject])

        names = "|".join(list(set(look_for)))

        rx = re.compile("""(.*(?:^|\t| |\[|\]|&|\(|\)|\.|!|"|')({})(?:$|\t| |\[|\]|&|\(|\)|\.|!|"|').*)""".format(names), MULTILINE)
        mentions = []

        # Indexe la position des lignes
        line_ends = [m.end() for m in re.finditer('.*\n', subject.sourcecode)]

        for match in rx.finditer(subject.sourcecode):
            line = next(i for i in range(len(line_ends)) if line_ends[i] > match.start(1)) + 1
            quote = match.group(1).strip()
            objname = match.group(2)
            mentions.append(Mention(line, objname, quote))
        subject.mentions = mentions

#             if mentions:
#                 subject.deps[candidate] = mentions

    @classmethod
    def analyse_all(cls):
        # Mise à jour des dépendances:
        # # parcourt les objets, et recherche dans le code source de chacun des mentions du nom des autres objets.
        total = len(cls.objects)
        for index, subject in enumerate(cls.objects):
            cls.report(index, total, "* {}: {}".format(subject.type_, subject.name_))
            cls.find_deps(subject)

    @classmethod
    def run(cls, source_dir):
        # Liste les objets à partir de l'arborescence du repertoire des sources
        cls.report(0, 100, "Analyse du répertoire")
        cls.load_objects(source_dir)

        cls.report(0, 100, "> {} objets trouvés".format(len(cls.objects)))

        cls.analyse_all()

        cls.report(100, 100, "Analyse terminée")
        cls.ended()

        return cls.objects

if __name__ == "__main__":

    here = Path(__file__).parent.abspath()
    source_dir = here / r"test\source"
    resultfile = here / r"test\analyse.txt"
    resultfile.remove_p()

    def print_(i, total, msg):
        print("({}/{}) {}".format(i, total, msg))

    Analyse.report = print_

    Analyse.run(source_dir)

    with open(resultfile, "w+") as f:
        for obj in Analyse.objects:
            msg = "# L'objet [{}] '{}' mentionne dans son code-source:".format(obj.type_, obj.name_)
            for mention in obj.mentions:
                msg += "\n\t\t'{}'\t\tLine: {}\t>>\t{}".format(mention.objname, mention.line, mention.quote)
#             for dep, mentions in obj.deps.items():
#                 msg += "\n\t* [{}] '{}'".format(dep.type_, dep.name_)
#                 for mention in mentions:
#                     msg += "\n\t\tLine: {} >> {}".format(mention.line, mention.quote)
#             if not obj.deps:
#                 msg += "\n\t (pas de dépendances)"

            msg += "\n"
            f.write(msg)

    print("# Terminé")