Aluode/PerceptionLabPortable

1import os2import zipfile3 4import fsspec5from fsspec.archive import AbstractArchiveFileSystem6 7 8class ZipFileSystem(AbstractArchiveFileSystem):9    """Read/Write contents of ZIP archive as a file-system10 11    Keeps file object open while instance lives.12 13    This class is pickleable, but not necessarily thread-safe14    """15 16    root_marker = ""17    protocol = "zip"18    cachable = False19 20    def __init__(21        self,22        fo="",23        mode="r",24        target_protocol=None,25        target_options=None,26        compression=zipfile.ZIP_STORED,27        allowZip64=True,28        compresslevel=None,29        **kwargs,30    ):31        """32        Parameters33        ----------34        fo: str or file-like35            Contains ZIP, and must exist. If a str, will fetch file using36            :meth:`~fsspec.open_files`, which must return one file exactly.37        mode: str38            Accept: "r", "w", "a"39        target_protocol: str (optional)40            If ``fo`` is a string, this value can be used to override the41            FS protocol inferred from a URL42        target_options: dict (optional)43            Kwargs passed when instantiating the target FS, if ``fo`` is44            a string.45        compression, allowZip64, compresslevel: passed to ZipFile46            Only relevant when creating a ZIP47        """48        super().__init__(self, **kwargs)49        if mode not in set("rwa"):50            raise ValueError(f"mode '{mode}' no understood")51        self.mode = mode52        if isinstance(fo, (str, os.PathLike)):53            if mode == "a":54                m = "r+b"55            else:56                m = mode + "b"57            fo = fsspec.open(58                fo, mode=m, protocol=target_protocol, **(target_options or {})59            )60        self.force_zip_64 = allowZip6461        self.of = fo62        self.fo = fo.__enter__()  # the whole instance is a context63        self.zip = zipfile.ZipFile(64            self.fo,65            mode=mode,66            compression=compression,67            allowZip64=allowZip64,68            compresslevel=compresslevel,69        )70        self.dir_cache = None71 72    @classmethod73    def _strip_protocol(cls, path):74        # zip file paths are always relative to the archive root75        return super()._strip_protocol(path).lstrip("/")76 77    def __del__(self):78        if hasattr(self, "zip"):79            self.close()80            del self.zip81 82    def close(self):83        """Commits any write changes to the file. Done on ``del`` too."""84        self.zip.close()85 86    def _get_dirs(self):87        if self.dir_cache is None or self.mode in set("wa"):88            # when writing, dir_cache is always in the ZipFile's attributes,89            # not read from the file.90            files = self.zip.infolist()91            self.dir_cache = {92                dirname.rstrip("/"): {93                    "name": dirname.rstrip("/"),94                    "size": 0,95                    "type": "directory",96                }97                for dirname in self._all_dirnames(self.zip.namelist())98            }99            for z in files:100                f = {s: getattr(z, s, None) for s in zipfile.ZipInfo.__slots__}101                f.update(102                    {103                        "name": z.filename.rstrip("/"),104                        "size": z.file_size,105                        "type": ("directory" if z.is_dir() else "file"),106                    }107                )108                self.dir_cache[f["name"]] = f109 110    def pipe_file(self, path, value, **kwargs):111        # override upstream, because we know the exact file size in this case112        self.zip.writestr(path, value, **kwargs)113 114    def _open(115        self,116        path,117        mode="rb",118        block_size=None,119        autocommit=True,120        cache_options=None,121        **kwargs,122    ):123        path = self._strip_protocol(path)124        if "r" in mode and self.mode in set("wa"):125            if self.exists(path):126                raise OSError("ZipFS can only be open for reading or writing, not both")127            raise FileNotFoundError(path)128        if "r" in self.mode and "w" in mode:129            raise OSError("ZipFS can only be open for reading or writing, not both")130        out = self.zip.open(path, mode.strip("b"), force_zip64=self.force_zip_64)131        if "r" in mode:132            info = self.info(path)133            out.size = info["size"]134            out.name = info["name"]135        return out136 137    def find(self, path, maxdepth=None, withdirs=False, detail=False, **kwargs):138        if maxdepth is not None and maxdepth < 1:139            raise ValueError("maxdepth must be at least 1")140 141        # Remove the leading slash, as the zip file paths are always142        # given without a leading slash143        path = path.lstrip("/")144        path_parts = list(filter(lambda s: bool(s), path.split("/")))145 146        def _matching_starts(file_path):147            file_parts = filter(lambda s: bool(s), file_path.split("/"))148            return all(a == b for a, b in zip(path_parts, file_parts))149 150        self._get_dirs()151 152        result = {}153        # To match posix find, if an exact file name is given, we should154        # return only that file155        if path in self.dir_cache and self.dir_cache[path]["type"] == "file":156            result[path] = self.dir_cache[path]157            return result if detail else [path]158 159        for file_path, file_info in self.dir_cache.items():160            if not (path == "" or _matching_starts(file_path)):161                continue162 163            if file_info["type"] == "directory":164                if withdirs:165                    if file_path not in result:166                        result[file_path.strip("/")] = file_info167                continue168 169            if file_path not in result:170                result[file_path] = file_info if detail else None171 172        if maxdepth:173            path_depth = path.count("/")174            result = {175                k: v for k, v in result.items() if k.count("/") - path_depth < maxdepth176            }177        return result if detail else sorted(result)178