accessanalyser/core.py

'''


@author: olivier.massot
'''
import re

from path import Path


objects = []



# TODO: gérer les cas où des objets de types différents portent le même nom. Ex: une table et un formulaire...
# TODO: ignorer les commentaires dans le modules
# TODO: ignorer les labels dans les formulaires et états

# NB: une requete et une table ne peuvent pas porter le même nom.



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 AccessObject():
    type_ = "<unknown>"
    _valid_file_exts = (".bas")
    _re_w_sep = """^|\s|\[|\]|&|\(|\)|\.|!|"|'"""

    def __init__(self, nom):
        self.nom = nom
        self.functions = []
        self.sourcefile = ""
        self.deps = []
        self.refs = []
        self._sourcecode = ""

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

    @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()
        return obj

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

    def add_dep(self, obj):
        if not obj in self.deps:
            self.deps.append(obj)

    def add_ref(self, obj):
        if not obj in self.refs:
            self.refs.append(obj)

    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_

    def search_me_regex(self):
        return re.compile(r"""({sep}){nom}({sep})""".format(p=self._re_w_sep, nom=self.nom))

    def rawsearch_me_regex(self):
        return self.search_me_regex()

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")

REFS_ONLY = 1
DEPS_ONLY = 2
DEPS_AND_REFS = 3

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

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

    @staticmethod
    def ended():
        pass

    @classmethod
    def register(cls, obj):
        if obj.nom in [other.nom for other in cls.objects] and not obj.nom in cls.duplicated_names:
            cls.duplicated_names.append(obj.nom)
        cls.objects.append(obj)

    @staticmethod
    def containsRefsTo(source, target):
        if Analyse.RAW_SEARCH:
            rx = target.rawsearch_me_regex()
            return rx.search(source.sourcecode)

        if type(target) is ModuleObject:  # L'objet peut contenir des references aux fonctions de l'objet module
            for fname in target.functions:
                rx = re.compile(r"(?:^|\W)({})(?:$|\W)".format(fname))
                if rx.search(source.sourcecode):
                    return True
        else:
            rx = re.compile(r"(?:^|\W)({})(?:$|\W)".format(target.nom))
            return rx.search(source.sourcecode)


    @classmethod
    def run(cls, source_dir, mode=DEPS_AND_REFS):
        source_dir = Path(source_dir)

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

        # Liste les objets à partir de l'arborescence du repertoire des sources
        cls.report(0, 100, "Analyse du répertoire")

        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)
                    cls.register(obj)
                except InvalidFileExt:
                    print("Ignored unrecognized file: {}".format(file))

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

        # met à jour les dependances en listant les noms d'objets mentionnés dans le fichier subject
        for index, subject in enumerate(cls.objects):

            cls.report(index, total, "* {}: {}".format(subject.type_, subject.nom))

            for candidate in cls.objects[:index] + cls.objects[index + 1:]:

                if Analyse.containsRefsTo(subject, candidate):

                    if mode in (DEPS_AND_REFS, DEPS_ONLY):
                        subject.add_dep(candidate)

                    if mode in (DEPS_AND_REFS, REFS_ONLY):
                        candidate.add_ref(subject)

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

        return cls.objects


if __name__ == "__main__":
    source_dir = Path(r"c:\dev\access\Analytique\source")

    here = Path(__file__).parent.abspath()
    source_dir = here / "source"
    datafile = here / "access_analyser.txt"
    datafile.remove_p()

    def main_report(*_, msg=""):
        print(msg)
    Analyse.report = main_report

    Analyse.run(source_dir)

    with open("data.txt", "w+") as f:
        for o in Analyse.objects:
            msg = "# {}: '{}'".format(o.type_, o.nom)
            msg += "\n\t Utilise: {}".format(", ".join(["'{}'".format(d) for d in o.deps]))
            msg += "\n\t Est utilisé par: {}".format(", ".join(["'{}'".format(d) for d in o.refs]))
            msg += "\n"
            f.write(msg)

    print("# Terminé")