abstract-utilities 0.2.2.540__py3-none-any.whl → 0.2.2.667__py3-none-any.whl

abstract_utilities/file_utils/src/reader_utils/file_reader.py

@@ -0,0 +1,543 @@
+# file_reader.py
+from .imports import *
+# -------- Public API drop-ins that mirror your originals --------
+
+from .pdf_utils import *
+# ---------------------------------------------------------------------------
+# NOTE: The following helper functions must be provided elsewhere:
+#   - convert_date_string(s: str) -> datetime
+#   - read_from_file(path: str) -> pd.DataFrame
+# ---------------------------------------------------------------------------
+def _should_skip_dir(dir_name: str, exclude_dirs: set[str]) -> bool:
+    """
+    Return True if dir_name match=self.exclude_types)es one of the excluded directory names exactly.
+    """
+    return dir_name in exclude_dirs
+
+
+def _should_skip_file(filename: str, exclude_patterns: set[str]) -> bool:
+    """
+    Return True if filename matches any pattern in exclude_patterns.
+    Uses fnmatch (Unix‐style wildcard matching).
+    """
+    for pat in exclude_patterns:
+        if fnmatch.fnmatch(filename, pat):
+            return True
+    return False
+
+def _should_skip_type(filename: str, exclude_types:set[str]) -> bool:
+    """
+    Return True if filename matches any pattern in exclude_patterns.
+    Uses fnmatch (Unix‐style wildcard matching).
+    """
+    return is_media_type(filename,media_types=exclude_types)
+class shoudSkipManager(metaclass=SingletonMeta):
+    def __init__(self,exclude_types=None,exclude_file_patterns=None,exclude_dirs=None):
+        if not hasattr(self, 'initialized') or self.initialized == False:
+            self.initialized = True
+            exclude_types = exclude_types or set()
+            exclude_file_patterns = exclude_file_patterns or set()
+            exclude_dirs = exclude_dirs or set()
+            self.exclude_dirs = exclude_dirs.copy()
+            self.exclude_file_patterns = exclude_file_patterns.copy()
+            self.exclude_types = exclude_types.copy()
+    def should_skip(self,exclude_item=None,exclude_types=None,exclude_file_patterns=None,exclude_dirs=None):
+        if (exclude_dirs==None and exclude_file_patterns == None and exclude_types == None) and exclude_item:
+            if isinstance(item,str):
+               if _should_skip_dir(dir_name=exclude_item, exclude_dirs=self.exclude_dirs):
+                   return True
+               if _should_skip_type(filename=exclude_item, exclude_types=self.exclude_types):
+                   return True
+               if _should_skip_file(filename=exclude_item, exclude_patterns=self.exclude_file_patterns):
+                   return True
+               return False
+        elif exclude_types or exclude_file_patterns:
+            if exclude_file_patterns != False:
+                if _should_skip_file(filename=exclude_item,
+                                     exclude_patterns=self.exclude_file_patterns):
+                    return True
+            if exclude_types != False:
+                if _should_skip_type(filename=exclude_item,
+                                     exclude_types=self.exclude_types):
+                    return True
+            
+        if exclude_dirs:
+            if _should_skip_dir(filename=exclude_item, exclude_patterns=self.exclude_dirs):
+                return True
+        return False
+
+SKIP_MGR = shoudSkipManager()
+def should_skip(
+    exclude_item=None,
+    exclude_types=None,
+    exclude_file_patterns=None,
+    exclude_dirs=None
+    ):
+    return shoudSkipManager().should_skip(
+        exclude_item=exclude_item,
+        exclude_types=exclude_types,
+        exclude_file_patterns=exclude_file_patterns,
+        exclude_dirs=exclude_dirs
+        )
+def re_initialize_skip_mgr(exclude_types=None,
+                           exclude_file_patterns=None,
+                           exclude_dirs=None):
+    shoudSkipManager().initialized = False
+    shoudSkipManager(
+        exclude_types=exclude_types,
+        exclude_file_patterns=exclude_file_patterns,
+        exclude_dirs=exclude_dirs
+        )
+
+def convert_date_string(s):
+    # … your existing stub or real implementation …
+    try:
+        return datetime.fromisoformat(s)
+    except ValueError:
+        return None
+# file_utils.py (below your existing imports)
+
+
+
+
+def source_engine_for_ext(ext: str) -> str:
+    ext = ext.lower()
+    mapping = {
+        '.parquet': 'pyarrow',
+        '.txt':    'python',
+        '.csv':    'python',
+        '.tsv':    'python',
+        '.xlsx':   'openpyxl',
+        '.xls':    'xlrd',
+        '.xlsb':   'pyxlsb',
+        '.ods':    'odf',
+        '.geojson':'GeoJSON',
+    }
+    return mapping.get(ext)
+
+def is_valid_file_path(path: str) -> Union[str, None]:
+    if not (isinstance(path, str) and path.strip()):
+        return None
+    if os.path.isfile(path):
+        return os.path.splitext(path)[1].lower()
+    return None
+
+def is_dataframe(obj) -> bool:
+    return isinstance(obj, (pd.DataFrame, gpd.GeoDataFrame))
+
+def create_dataframe(data=None, columns=None) -> pd.DataFrame:
+    # … unchanged …
+    if is_dataframe(data):
+        return data.copy()
+    data = data or {}
+    if isinstance(data, dict):
+        data = [data]
+        if columns is None:
+            all_keys = set()
+            for row in data:
+                if isinstance(row, dict):
+                    all_keys.update(row.keys())
+            columns = list(all_keys)
+        if columns is False:
+            columns = None
+    try:
+        return pd.DataFrame(data, columns=columns)
+    except Exception as e:
+        _logger.error(f"Failed to create DataFrame: {e}")
+        return pd.DataFrame([], columns=columns)
+
+def read_ods_file(path: str) -> dict[str, pd.DataFrame]:
+    # … unchanged …
+    if not is_valid_file_path(path):
+        _logger.error(f"File not found or invalid: {path}")
+        return {}
+    try:
+        doc = ezodf.opendoc(path)
+    except Exception as e:
+        _logger.error(f"Failed to open ODS document: {e}")
+        return {}
+    sheets: dict[str, pd.DataFrame] = {}
+    for sheet in doc.sheets:
+        table_rows = []
+        for row in sheet.rows():
+            row_data = []
+            for cell in row:
+                if cell.value_type == 'date':
+                    row_data.append(convert_date_string(str(cell.value)))
+                else:
+                    row_data.append(cell.value)
+            table_rows.append(row_data)
+        df = pd.DataFrame(table_rows)
+        sheets[sheet.name] = df
+        _logger.info(f"Processed sheet: {sheet.name}")
+    return sheets
+
+def read_ods_as_excel(path: str, xlsx_path: str | None = None) -> pd.DataFrame:
+    # … unchanged …
+    if not is_valid_file_path(path):
+        _logger.error(f"File not found or invalid: {path}")
+        return pd.DataFrame()
+    if xlsx_path is None:
+        tmp_dir = tempfile.mkdtemp()
+        xlsx_path = os.path.join(tmp_dir, os.path.basename(path) + '.xlsx')
+        cleanup_temp = True
+    else:
+        cleanup_temp = False
+    try:
+        # You must implement ods_to_xlsx(...) externally
+        ods_to_xlsx(path, xlsx_path)
+    except Exception as e:
+        _logger.error(f"ODS→XLSX conversion failed: {e}")
+        if cleanup_temp:
+            shutil.rmtree(tmp_dir)
+        return pd.DataFrame()
+    try:
+        df = pd.read_excel(xlsx_path, engine='openpyxl')
+    except Exception as e:
+        _logger.error(f"Failed to read converted XLSX: {e}")
+        df = pd.DataFrame()
+    finally:
+        if cleanup_temp:
+            shutil.rmtree(tmp_dir)
+    return df
+
+def filter_df(
+    df: pd.DataFrame,
+    nrows: int | None = None,
+    condition: pd.Series | None = None,
+    indices: list[int] | None = None
+) -> pd.DataFrame:
+    if nrows is not None:
+        df = df.head(nrows)
+    if condition is not None:
+        df = df[condition]
+    if indices is not None:
+        df = df.iloc[indices]
+    return df
+
+def read_shape_file(path: str) -> Union[gpd.GeoDataFrame, None]:
+    # … unchanged …
+    ext = is_valid_file_path(path)
+    if not ext:
+        _logger.error(f"Shape file not found: {path}")
+        return None
+    ext = ext.lower()
+    try:
+        if ext in ('.shp', '.cpg', '.dbf', '.shx'):
+            return gpd.read_file(path)
+        if ext == '.geojson':
+            return gpd.read_file(path, driver='GeoJSON')
+        if ext == '.prj':
+            return read_from_file(path)  # Must return GeoDataFrame
+    except Exception as e:
+        _logger.error(f"Failed to read spatial data ({path}): {e}")
+        return None
+    _logger.error(f"Unsupported spatial extension: {ext}")
+    return None
+
+
+
+
+
+def pdf_to_text(path, keep_page_breaks=True, ocr_if_empty=True):
+    """
+    Return the full text of *path* (str or Path) as a single string.
+
+    keep_page_breaks → insert "\f" between pages so you can split later.
+    ocr_if_empty     → any page with no text layer is rasterised & OCR'd.
+    """
+    path = Path(path)
+    if not path.exists():
+        raise FileNotFoundError(path)
+
+    all_pages = []
+
+    with pdfplumber.open(path) as pdf:
+        for i, page in enumerate(pdf.pages, start=1):
+            text = page.extract_text() or ""   # might be None
+            if (not text.strip()) and ocr_if_empty:
+                # rasterise at 300 dpi then Tesseract
+                img = convert_from_path(str(path), dpi=300, first_page=i, last_page=i)[0]
+                text = pytesseract.image_to_string(img, lang="eng")
+            all_pages.append(text)
+
+    sep = "\f" if keep_page_breaks else "\n"
+    return sep.join(all_pages)
+def get_df(
+    source: Union[
+        str,
+        pd.DataFrame,
+        gpd.GeoDataFrame,
+        dict,
+        list,
+        FileStorage
+    ],
+    nrows: int | None = None,
+    skiprows: list[int] | int | None = None,
+    condition: pd.Series | None = None,
+    indices: list[int] | None = None
+) -> Union[pd.DataFrame, gpd.GeoDataFrame, dict[str, Union[pd.DataFrame, str]], None]:
+    """
+    Load a DataFrame or GeoDataFrame from various sources, then apply optional filters.
+    If `source` is a directory, returns read_directory(source) instead (a dict).
+    """
+
+    # ─── Check for directory first ─────────────────────────────────────────────
+    if isinstance(source, str) and os.path.isdir(source):
+        return read_directory(root_path=source)
+
+    # ─── If already a DataFrame/GeoDataFrame, just filter and return ───────────
+    if is_dataframe(source):
+        _logger.info("Source is already a DataFrame/GeoDataFrame; applying filters.")
+        return filter_df(source, nrows=nrows, condition=condition, indices=indices)
+
+    if source is None:
+        _logger.error("No source provided to get_df().")
+        return None
+
+    # ─── Next: If source is a file path, read according to extension ───────────
+    if isinstance(source, str) and os.path.isfile(source):
+        ext = os.path.splitext(source)[1].lower()
+        try:
+            _logger.info(f"Loading file {source} with extension '{ext}'.")
+            if ext in ('.csv', '.tsv', '.txt'):
+                sep = {'.csv': ',', '.tsv': '\t', '.txt': None}.get(ext)
+                df = pd.read_csv(source, skiprows=skiprows, sep=sep, nrows=nrows)
+            elif ext in ('.ods', '.xlsx', '.xls', '.xlsb'):
+                engine = source_engine_for_ext(ext)
+                if ext == '.ods':
+                    df = read_ods_as_excel(source)
+                else:
+                    df = pd.read_excel(source, skiprows=skiprows, engine=engine, nrows=nrows)
+            elif ext == '.json':
+                df = safe_read_from_json(source)
+                return df
+            elif ext == '.parquet':
+                df = pd.read_parquet(source)
+            elif ext in ('.shp', '.cpg', '.dbf', '.shx', '.geojson', '.prj'):
+                return read_shape_file(source)
+            elif ext in ['.pdf']:
+                df = pdf_to_text(source)
+                return df
+            else:
+                df = read_from_file(source)
+                return df
+
+            if not isinstance(df, (dict, list, FileStorage)):
+                return filter_df(df, nrows=nrows, condition=condition, indices=indices)
+            source = df  # pass on to next block if needed
+
+        except Exception as e:
+            _logger.error(f"Failed to read '{source}': {e}")
+            return None
+
+    # ─── If source is FileStorage (uploaded) ───────────────────────────────────
+    if isinstance(source, FileStorage):
+        try:
+            filename = secure_filename(source.filename or "uploaded.xlsx")
+            _logger.info(f"Reading uploaded file: {filename}")
+            df = pd.read_excel(source.stream, nrows=nrows)
+            return filter_df(df, nrows=nrows, condition=condition, indices=indices)
+        except Exception as e:
+            _logger.error(f"Failed to read FileStorage: {e}")
+            return None
+
+    # ─── If source is dict or list, turn into DataFrame ────────────────────────
+    if isinstance(source, (dict, list)):
+        _logger.info("Creating DataFrame from in-memory data structure.")
+        df = pd.DataFrame(source)
+        return filter_df(df, nrows=nrows, condition=condition, indices=indices)
+
+    _logger.error(f"Unsupported source type: {type(source)}")
+    return None
+
+
+def read_file_as_text(paths: Union[str, List[str]]) -> List[str]:
+    """
+    Given one path or a list of paths, return a list of textual representations
+    for each “file” found.  If a given path is:
+    
+      1) A directory → we call read_directory(...) on it (which skips node_modules,
+         __pycache__, *.ini, etc.) and iterate over each (relative_path → content).
+      2) A plain‐text file (extension ∈ SUPPORTED_TEXT_EXTENSIONS) → we open it and return its raw text.
+      3) Anything else (e.g. .xlsx, .ods, .parquet, .shp, etc.) → we delegate to get_df(...) and then
+         convert whatever get_df(...) gives us into CSV or “to_string()” as appropriate.
+    
+    Returns:
+        A list of strings—each string is the “file’s contents” for one actual file.  
+        (Ordering is “filesystem walk order” for directories, and “in order of the input list” for files.)
+    
+    Raises:
+        FileNotFoundError if any path in `paths` does not exist.
+        ValueError if a file cannot be parsed/read.
+    """
+    # Ensure we have a list to iterate
+    if isinstance(paths, str):
+        files_to_process = [paths]
+    else:
+        files_to_process = list(paths)
+
+    all_data: List[str] = []
+
+    for full_path in files_to_process:
+        if not os.path.exists(full_path):
+            raise FileNotFoundError(f"Not a valid path: {full_path!r}")
+
+        # ── If this is a directory, walk it via read_directory(...) ─────────────────
+        if os.path.isdir(full_path):
+            # read_directory returns a dict: { relative_path: (DataFrame or text) }
+            nested_dict: Dict[str, Union[pd.DataFrame, gpd.GeoDataFrame, str]] = read_directory(full_path)
+
+            for rel, content in nested_dict.items():
+                # `content` is either a DataFrame, GeoDataFrame, or a plain‐text string
+                if isinstance(content, (pd.DataFrame, gpd.GeoDataFrame)):
+                    # If GeoDataFrame, convert geometry column to WKT before CSV
+                    if isinstance(content, gpd.GeoDataFrame):
+                        gdf = content.copy()
+                        gdf["geometry"] = gdf["geometry"].apply(lambda g: g.wkt if g is not None else "")
+                        all_data.append(gdf.to_csv(index=False))
+                    else:
+                        all_data.append(content.to_csv(index=False))
+                else:
+                    # Already a text blob
+                    all_data.append(content)
+
+            continue  # move on to the next item in files_to_process
+
+        # ── At this point, full_path is guaranteed to be a file ───────────────────────
+        ext = os.path.splitext(full_path)[1].lower()
+
+        # 1) PURE TEXT EXTENSION?
+        #if ext in SUPPORTED_TEXT_EXTENSIONS:
+        try:
+            with open(full_path, "r", encoding="utf-8", errors="replace") as f:
+                raw = f.read()
+            all_data.append(raw)
+        except Exception as e:
+            raise ValueError(f"Error reading text file {full_path!r}: {e}")
+
+        continue
+
+        # 2) ANY OTHER FILETYPE → delegate to get_df(...) and convert result to text
+        try:
+            df_or = get_df(full_path)
+        except Exception as e:
+            raise ValueError(f"get_df() failed for {full_path!r}: {e}")
+
+        # 2a) If get_df returned a dict (e.g. an ODS with multiple sheets, or a directory)
+        if isinstance(df_or, dict):
+            # Join each sheet or sub‐file’s DataFrame into one big text block
+            for key, value in df_or.items():
+                if isinstance(value, (pd.DataFrame, gpd.GeoDataFrame)):
+                    if isinstance(value, gpd.GeoDataFrame):
+                        gdf = value.copy()
+                        gdf["geometry"] = gdf["geometry"].apply(lambda g: g.wkt if g is not None else "")
+                        block = f"=== {key} ===\n" + gdf.to_csv(index=False)
+                    else:
+                        block = f"=== {key} ===\n" + value.to_csv(index=False)
+                else:
+                    # It was already plain‐text under that key
+                    block = f"=== {key} ===\n" + str(value)
+                all_data.append(block)
+
+            continue
+
+        # 2b) If get_df returned a DataFrame or GeoDataFrame directly
+        if isinstance(df_or, (pd.DataFrame, gpd.GeoDataFrame)):
+            if isinstance(df_or, gpd.GeoDataFrame):
+                gdf = df_or.copy()
+                gdf["geometry"] = gdf["geometry"].apply(lambda g: g.wkt if g is not None else "")
+                all_data.append(gdf.to_csv(index=False))
+            else:
+                all_data.append(df_or.to_csv(index=False))
+
+            continue
+
+        # 2c) If get_df returned a list of dicts (rare, but possible)
+        if isinstance(df_or, list):
+            try:
+                temp_df = pd.DataFrame(df_or)
+                all_data.append(temp_df.to_csv(index=False))
+            except Exception:
+                all_data.append(repr(df_or))
+            continue
+
+        # 2d) Otherwise, fall back to repr()
+        all_data.append(repr(df_or))
+
+    return all_data
+def read_directory(
+    root_path: str,
+    exclude_dirs: set[str] = None,
+    exclude_file_patterns: set[str] = None,
+    exclude_types: set[str] = None,
+) -> Dict[str, Union[pd.DataFrame, str]]:
+    re_initialize_skip_mgr(exclude_types=exclude_types,
+                           exclude_file_patterns=exclude_file_patterns,
+                           exclude_dirs=exclude_dirs)
+    if not os.path.isdir(root_path):
+        raise FileNotFoundError(f"Not a valid directory: {root_path!r}")
+        collected: Dict[str, Union[pd.DataFrame, str]] = {}
+    root_path = os.path.abspath(root_path)
+    root_len = len(root_path.rstrip(os.sep)) + 1
+
+    for dirpath, dirnames, filenames in os.walk(root_path):
+        # 1) Skip excluded subfolders
+        dirnames[:] = [
+            d for d in dirnames if not should_skip(exclude_item=d,
+                                                   exclude_dirs=True)
+        ]
+
+        for fname in filenames:
+            # 2) Skip excluded filename patterns
+            if should_skip(exclude_types=fname,exclude_file_patterns=fname):
+                _logger.debug(f"Skipping file by pattern: {os.path.join(dirpath, fname)}")
+                continue
+
+            full_path = os.path.join(dirpath, fname)
+            rel_path = full_path[root_len:]  # e.g. "subdir/logs/vid_to_aud.log"
+            ext = os.path.splitext(fname)[1].lower()
+
+            # ── 2a) If it's one of our “plain‐text” extensions, read it as text right now:
+            if ext in {'.txt', '.md', '.csv', '.tsv', '.log'}:
+                try:
+                    with open(full_path, 'r', encoding='utf-8', errors='replace') as f:
+                        text = f.read()
+                    collected[rel_path] = text
+                    _logger.info(f"Read text file: {rel_path}")
+                except Exception as e:
+                    _logger.warning(f"Failed to read {rel_path} as text: {e}")
+                continue
+
+            # ── 2b) Otherwise, try to load via get_df(...) (DataFrame/GeoDataFrame/etc.)
+            try:
+                df_or_gdf = get_df(full_path)
+                if isinstance(df_or_gdf, (pd.DataFrame, gpd.GeoDataFrame)):
+                    collected[rel_path] = df_or_gdf
+                    _logger.info(f"Loaded DataFrame: {rel_path}")
+                    continue
+                # If get_df returned a dict (e.g. multi‐sheet ODS), merge as multiple entries
+                if isinstance(df_or_gdf, dict):
+                    for sheet_name, df in df_or_gdf.items():
+                        key = f"{rel_path}::[{sheet_name}]"
+                        collected[key] = df
+                        _logger.info(f"Loaded sheet DataFrame: {key}")
+                    continue
+                # If get_df returned something else (list, non‐DataFrame), fall through to text
+            except Exception as e:
+                _logger.debug(f"get_df failed for {rel_path}: {e}")
+
+            # ── 2c) Lastly, if it wasn’t a “pure text” file and get_df didn’t return a DataFrame,
+            # treat it as text via read_file_as_text(...) so you get at least something:
+            try:
+                text = read_file_as_text(full_path)
+                # read_file_as_text returns a List[str], but here we're in a single-file context,
+                # so just join on "\n\n" or take the first element. For simplicity:
+                combined = "\n\n".join(text)
+                collected[rel_path] = combined
+                _logger.info(f"Read fallback text for: {rel_path}")
+            except Exception as e:
+                _logger.warning(f"Could not read {rel_path} as text or DataFrame: {e}")
+
+    return collected
+