| 1 | """ |
|---|
| 2 | This is the base file reader class most file readers should inherit from. |
|---|
| 3 | All generic functionality required for a file loader/reader is built into this |
|---|
| 4 | class |
|---|
| 5 | """ |
|---|
| 6 | |
|---|
| 7 | import os |
|---|
| 8 | import sys |
|---|
| 9 | import math |
|---|
| 10 | import logging |
|---|
| 11 | from abc import abstractmethod |
|---|
| 12 | |
|---|
| 13 | import numpy as np |
|---|
| 14 | from .loader_exceptions import NoKnownLoaderException, FileContentsException,\ |
|---|
| 15 | DataReaderException, DefaultReaderException |
|---|
| 16 | from .data_info import Data1D, Data2D, DataInfo, plottable_1D, plottable_2D,\ |
|---|
| 17 | combine_data_info_with_plottable |
|---|
| 18 | |
|---|
| 19 | logger = logging.getLogger(__name__) |
|---|
| 20 | |
|---|
| 21 | if sys.version_info[0] < 3: |
|---|
| 22 | def decode(s): |
|---|
| 23 | return s |
|---|
| 24 | else: |
|---|
| 25 | def decode(s): |
|---|
| 26 | return s.decode() if isinstance(s, bytes) else s |
|---|
| 27 | |
|---|
| 28 | class FileReader(object): |
|---|
| 29 | # String to describe the type of data this reader can load |
|---|
| 30 | type_name = "ASCII" |
|---|
| 31 | # Wildcards to display |
|---|
| 32 | type = ["Text files (*.txt|*.TXT)"] |
|---|
| 33 | # List of allowed extensions |
|---|
| 34 | ext = ['.txt'] |
|---|
| 35 | # Bypass extension check and try to load anyway |
|---|
| 36 | allow_all = False |
|---|
| 37 | # Able to import the unit converter |
|---|
| 38 | has_converter = True |
|---|
| 39 | # Default value of zero |
|---|
| 40 | _ZERO = 1e-16 |
|---|
| 41 | |
|---|
| 42 | def __init__(self): |
|---|
| 43 | # List of Data1D and Data2D objects to be sent back to data_loader |
|---|
| 44 | self.output = [] |
|---|
| 45 | # Current plottable_(1D/2D) object being loaded in |
|---|
| 46 | self.current_dataset = None |
|---|
| 47 | # Current DataInfo object being loaded in |
|---|
| 48 | self.current_datainfo = None |
|---|
| 49 | # File path sent to reader |
|---|
| 50 | self.filepath = None |
|---|
| 51 | # Open file handle |
|---|
| 52 | self.f_open = None |
|---|
| 53 | |
|---|
| 54 | def read(self, filepath): |
|---|
| 55 | """ |
|---|
| 56 | Basic file reader |
|---|
| 57 | |
|---|
| 58 | :param filepath: The full or relative path to a file to be loaded |
|---|
| 59 | """ |
|---|
| 60 | self.filepath = filepath |
|---|
| 61 | if os.path.isfile(filepath): |
|---|
| 62 | basename, extension = os.path.splitext(os.path.basename(filepath)) |
|---|
| 63 | self.extension = extension.lower() |
|---|
| 64 | # If the file type is not allowed, return nothing |
|---|
| 65 | if self.extension in self.ext or self.allow_all: |
|---|
| 66 | # Try to load the file, but raise an error if unable to. |
|---|
| 67 | try: |
|---|
| 68 | self.f_open = open(filepath, 'rb') |
|---|
| 69 | self.get_file_contents() |
|---|
| 70 | |
|---|
| 71 | except DataReaderException as e: |
|---|
| 72 | self.handle_error_message(e.message) |
|---|
| 73 | except OSError as e: |
|---|
| 74 | # If the file cannot be opened |
|---|
| 75 | msg = "Unable to open file: {}\n".format(filepath) |
|---|
| 76 | msg += e.message |
|---|
| 77 | self.handle_error_message(msg) |
|---|
| 78 | finally: |
|---|
| 79 | # Close the file handle if it is open |
|---|
| 80 | if not self.f_open.closed: |
|---|
| 81 | self.f_open.close() |
|---|
| 82 | if len(self.output) > 0: |
|---|
| 83 | # Sort the data that's been loaded |
|---|
| 84 | self.sort_one_d_data() |
|---|
| 85 | self.sort_two_d_data() |
|---|
| 86 | else: |
|---|
| 87 | msg = "Unable to find file at: {}\n".format(filepath) |
|---|
| 88 | msg += "Please check your file path and try again." |
|---|
| 89 | self.handle_error_message(msg) |
|---|
| 90 | |
|---|
| 91 | # Return a list of parsed entries that data_loader can manage |
|---|
| 92 | final_data = self.output |
|---|
| 93 | self.reset_state() |
|---|
| 94 | return final_data |
|---|
| 95 | |
|---|
| 96 | def reset_state(self): |
|---|
| 97 | """ |
|---|
| 98 | Resets the class state to a base case when loading a new data file so previous |
|---|
| 99 | data files do not appear a second time |
|---|
| 100 | """ |
|---|
| 101 | self.current_datainfo = None |
|---|
| 102 | self.current_dataset = None |
|---|
| 103 | self.filepath = None |
|---|
| 104 | self.ind = None |
|---|
| 105 | self.output = [] |
|---|
| 106 | |
|---|
| 107 | def nextline(self): |
|---|
| 108 | """ |
|---|
| 109 | Returns the next line in the file as a string. |
|---|
| 110 | """ |
|---|
| 111 | #return self.f_open.readline() |
|---|
| 112 | return decode(self.f_open.readline()) |
|---|
| 113 | |
|---|
| 114 | def nextlines(self): |
|---|
| 115 | """ |
|---|
| 116 | Returns the next line in the file as a string. |
|---|
| 117 | """ |
|---|
| 118 | for line in self.f_open: |
|---|
| 119 | #yield line |
|---|
| 120 | yield decode(line) |
|---|
| 121 | |
|---|
| 122 | def readall(self): |
|---|
| 123 | """ |
|---|
| 124 | Returns the entire file as a string. |
|---|
| 125 | """ |
|---|
| 126 | #return self.f_open.read() |
|---|
| 127 | return decode(self.f_open.read()) |
|---|
| 128 | |
|---|
| 129 | def handle_error_message(self, msg): |
|---|
| 130 | """ |
|---|
| 131 | Generic error handler to add an error to the current datainfo to |
|---|
| 132 | propagate the error up the error chain. |
|---|
| 133 | :param msg: Error message |
|---|
| 134 | """ |
|---|
| 135 | if len(self.output) > 0: |
|---|
| 136 | self.output[-1].errors.append(msg) |
|---|
| 137 | elif isinstance(self.current_datainfo, DataInfo): |
|---|
| 138 | self.current_datainfo.errors.append(msg) |
|---|
| 139 | else: |
|---|
| 140 | logger.warning(msg) |
|---|
| 141 | |
|---|
| 142 | def send_to_output(self): |
|---|
| 143 | """ |
|---|
| 144 | Helper that automatically combines the info and set and then appends it |
|---|
| 145 | to output |
|---|
| 146 | """ |
|---|
| 147 | data_obj = combine_data_info_with_plottable(self.current_dataset, |
|---|
| 148 | self.current_datainfo) |
|---|
| 149 | self.output.append(data_obj) |
|---|
| 150 | |
|---|
| 151 | def sort_one_d_data(self): |
|---|
| 152 | """ |
|---|
| 153 | Sort 1D data along the X axis for consistency |
|---|
| 154 | """ |
|---|
| 155 | for data in self.output: |
|---|
| 156 | if isinstance(data, Data1D): |
|---|
| 157 | # Normalize the units for |
|---|
| 158 | data.x_unit = self.format_unit(data.x_unit) |
|---|
| 159 | data.y_unit = self.format_unit(data.y_unit) |
|---|
| 160 | # Sort data by increasing x and remove 1st point |
|---|
| 161 | ind = np.lexsort((data.y, data.x)) |
|---|
| 162 | data.x = self._reorder_1d_array(data.x, ind) |
|---|
| 163 | data.y = self._reorder_1d_array(data.y, ind) |
|---|
| 164 | if data.dx is not None: |
|---|
| 165 | if len(data.dx) == 0: |
|---|
| 166 | data.dx = None |
|---|
| 167 | continue |
|---|
| 168 | data.dx = self._reorder_1d_array(data.dx, ind) |
|---|
| 169 | if data.dxl is not None: |
|---|
| 170 | data.dxl = self._reorder_1d_array(data.dxl, ind) |
|---|
| 171 | if data.dxw is not None: |
|---|
| 172 | data.dxw = self._reorder_1d_array(data.dxw, ind) |
|---|
| 173 | if data.dy is not None: |
|---|
| 174 | if len(data.dy) == 0: |
|---|
| 175 | data.dy = None |
|---|
| 176 | continue |
|---|
| 177 | data.dy = self._reorder_1d_array(data.dy, ind) |
|---|
| 178 | if data.lam is not None: |
|---|
| 179 | data.lam = self._reorder_1d_array(data.lam, ind) |
|---|
| 180 | if data.dlam is not None: |
|---|
| 181 | data.dlam = self._reorder_1d_array(data.dlam, ind) |
|---|
| 182 | data = self._remove_nans_in_1d_data(data) |
|---|
| 183 | if len(data.x) > 0: |
|---|
| 184 | data.xmin = np.min(data.x) |
|---|
| 185 | data.xmax = np.max(data.x) |
|---|
| 186 | data.ymin = np.min(data.y) |
|---|
| 187 | data.ymax = np.max(data.y) |
|---|
| 188 | |
|---|
| 189 | @staticmethod |
|---|
| 190 | def _reorder_1d_array(array, ind): |
|---|
| 191 | """ |
|---|
| 192 | Reorders a 1D array based on the indices passed as ind |
|---|
| 193 | :param array: Array to be reordered |
|---|
| 194 | :param ind: Indices used to reorder array |
|---|
| 195 | :return: reordered array |
|---|
| 196 | """ |
|---|
| 197 | array = np.asarray(array, dtype=np.float64) |
|---|
| 198 | return array[ind] |
|---|
| 199 | |
|---|
| 200 | @staticmethod |
|---|
| 201 | def _remove_nans_in_1d_data(data): |
|---|
| 202 | """ |
|---|
| 203 | Remove data points where nan is loaded |
|---|
| 204 | :param data: 1D data set |
|---|
| 205 | :return: data with mask=0 for any value of nan in data .x, .y, .dx, .dy |
|---|
| 206 | """ |
|---|
| 207 | mask = np.ones(len(data.x)) |
|---|
| 208 | data_list = [data.x, data.y, data.dx, data.dy, data.dxl, data.dxw] |
|---|
| 209 | for array in data_list: |
|---|
| 210 | if array is not None: |
|---|
| 211 | # Set mask[i] to 0 when data.<param> is nan |
|---|
| 212 | mask[np.isnan(array)] = 0 |
|---|
| 213 | # Data indices to mask/remove from the data |
|---|
| 214 | nans = np.where(mask == 0)[0] |
|---|
| 215 | if len(nans) > 0: |
|---|
| 216 | data.x = np.delete(data.x, nans) |
|---|
| 217 | data.y = np.delete(data.y, nans) |
|---|
| 218 | if data.dx is not None: |
|---|
| 219 | data.dx = np.delete(data.dx, nans) |
|---|
| 220 | if data.dxl is not None: |
|---|
| 221 | data.dxl = np.delete(data.dxl, nans) |
|---|
| 222 | if data.dxw is not None: |
|---|
| 223 | data.dxw = np.delete(data.dxw, nans) |
|---|
| 224 | if data.dy is not None: |
|---|
| 225 | data.dy = np.delete(data.dy, nans) |
|---|
| 226 | return data |
|---|
| 227 | |
|---|
| 228 | def sort_two_d_data(self): |
|---|
| 229 | for dataset in self.output: |
|---|
| 230 | if isinstance(dataset, Data2D): |
|---|
| 231 | # Normalize the units for |
|---|
| 232 | dataset.x_unit = self.format_unit(dataset.Q_unit) |
|---|
| 233 | dataset.y_unit = self.format_unit(dataset.I_unit) |
|---|
| 234 | dataset.data = dataset.data.astype(np.float64) |
|---|
| 235 | dataset.qx_data = dataset.qx_data.astype(np.float64) |
|---|
| 236 | dataset.xmin = np.min(dataset.qx_data) |
|---|
| 237 | dataset.xmax = np.max(dataset.qx_data) |
|---|
| 238 | dataset.qy_data = dataset.qy_data.astype(np.float64) |
|---|
| 239 | dataset.ymin = np.min(dataset.qy_data) |
|---|
| 240 | dataset.ymax = np.max(dataset.qy_data) |
|---|
| 241 | dataset.q_data = np.sqrt(dataset.qx_data * dataset.qx_data |
|---|
| 242 | + dataset.qy_data * dataset.qy_data) |
|---|
| 243 | if dataset.err_data is not None: |
|---|
| 244 | dataset.err_data = dataset.err_data.astype(np.float64) |
|---|
| 245 | if dataset.dqx_data is not None: |
|---|
| 246 | dataset.dqx_data = dataset.dqx_data.astype(np.float64) |
|---|
| 247 | if dataset.dqy_data is not None: |
|---|
| 248 | dataset.dqy_data = dataset.dqy_data.astype(np.float64) |
|---|
| 249 | if dataset.mask is not None: |
|---|
| 250 | dataset.mask = dataset.mask.astype(dtype=bool) |
|---|
| 251 | |
|---|
| 252 | if len(dataset.data.shape) == 2: |
|---|
| 253 | n_rows, n_cols = dataset.data.shape |
|---|
| 254 | dataset.y_bins = dataset.qy_data[0::int(n_cols)] |
|---|
| 255 | dataset.x_bins = dataset.qx_data[:int(n_cols)] |
|---|
| 256 | dataset.data = dataset.data.flatten() |
|---|
| 257 | if len(dataset.data) > 0: |
|---|
| 258 | dataset.xmin = np.min(dataset.qx_data) |
|---|
| 259 | dataset.xmax = np.max(dataset.qx_data) |
|---|
| 260 | dataset.ymin = np.min(dataset.qy_data) |
|---|
| 261 | dataset.ymax = np.max(dataset.qx_data) |
|---|
| 262 | |
|---|
| 263 | def format_unit(self, unit=None): |
|---|
| 264 | """ |
|---|
| 265 | Format units a common way |
|---|
| 266 | :param unit: |
|---|
| 267 | :return: |
|---|
| 268 | """ |
|---|
| 269 | if unit: |
|---|
| 270 | split = unit.split("/") |
|---|
| 271 | if len(split) == 1: |
|---|
| 272 | return unit |
|---|
| 273 | elif split[0] == '1': |
|---|
| 274 | return "{0}^".format(split[1]) + "{-1}" |
|---|
| 275 | else: |
|---|
| 276 | return "{0}*{1}^".format(split[0], split[1]) + "{-1}" |
|---|
| 277 | |
|---|
| 278 | def set_all_to_none(self): |
|---|
| 279 | """ |
|---|
| 280 | Set all mutable values to None for error handling purposes |
|---|
| 281 | """ |
|---|
| 282 | self.current_dataset = None |
|---|
| 283 | self.current_datainfo = None |
|---|
| 284 | self.output = [] |
|---|
| 285 | |
|---|
| 286 | def data_cleanup(self): |
|---|
| 287 | """ |
|---|
| 288 | Clean up the data sets and refresh everything |
|---|
| 289 | :return: None |
|---|
| 290 | """ |
|---|
| 291 | self.remove_empty_q_values() |
|---|
| 292 | self.send_to_output() # Combine datasets with DataInfo |
|---|
| 293 | self.current_datainfo = DataInfo() # Reset DataInfo |
|---|
| 294 | |
|---|
| 295 | def remove_empty_q_values(self): |
|---|
| 296 | """ |
|---|
| 297 | Remove any point where Q == 0 |
|---|
| 298 | """ |
|---|
| 299 | if isinstance(self.current_dataset, plottable_1D): |
|---|
| 300 | # Booleans for resolutions |
|---|
| 301 | has_error_dx = self.current_dataset.dx is not None |
|---|
| 302 | has_error_dxl = self.current_dataset.dxl is not None |
|---|
| 303 | has_error_dxw = self.current_dataset.dxw is not None |
|---|
| 304 | has_error_dy = self.current_dataset.dy is not None |
|---|
| 305 | # Create arrays of zeros for non-existent resolutions |
|---|
| 306 | if has_error_dxw and not has_error_dxl: |
|---|
| 307 | array_size = self.current_dataset.dxw.size - 1 |
|---|
| 308 | self.current_dataset.dxl = np.append(self.current_dataset.dxl, |
|---|
| 309 | np.zeros([array_size])) |
|---|
| 310 | has_error_dxl = True |
|---|
| 311 | elif has_error_dxl and not has_error_dxw: |
|---|
| 312 | array_size = self.current_dataset.dxl.size - 1 |
|---|
| 313 | self.current_dataset.dxw = np.append(self.current_dataset.dxw, |
|---|
| 314 | np.zeros([array_size])) |
|---|
| 315 | has_error_dxw = True |
|---|
| 316 | elif not has_error_dxl and not has_error_dxw and not has_error_dx: |
|---|
| 317 | array_size = self.current_dataset.x.size - 1 |
|---|
| 318 | self.current_dataset.dx = np.append(self.current_dataset.dx, |
|---|
| 319 | np.zeros([array_size])) |
|---|
| 320 | has_error_dx = True |
|---|
| 321 | if not has_error_dy: |
|---|
| 322 | array_size = self.current_dataset.y.size - 1 |
|---|
| 323 | self.current_dataset.dy = np.append(self.current_dataset.dy, |
|---|
| 324 | np.zeros([array_size])) |
|---|
| 325 | has_error_dy = True |
|---|
| 326 | |
|---|
| 327 | # Remove points where q = 0 |
|---|
| 328 | x = self.current_dataset.x |
|---|
| 329 | self.current_dataset.x = self.current_dataset.x[x != 0] |
|---|
| 330 | self.current_dataset.y = self.current_dataset.y[x != 0] |
|---|
| 331 | if has_error_dy: |
|---|
| 332 | self.current_dataset.dy = self.current_dataset.dy[x != 0] |
|---|
| 333 | if has_error_dx: |
|---|
| 334 | self.current_dataset.dx = self.current_dataset.dx[x != 0] |
|---|
| 335 | if has_error_dxl: |
|---|
| 336 | self.current_dataset.dxl = self.current_dataset.dxl[x != 0] |
|---|
| 337 | if has_error_dxw: |
|---|
| 338 | self.current_dataset.dxw = self.current_dataset.dxw[x != 0] |
|---|
| 339 | elif isinstance(self.current_dataset, plottable_2D): |
|---|
| 340 | has_error_dqx = self.current_dataset.dqx_data is not None |
|---|
| 341 | has_error_dqy = self.current_dataset.dqy_data is not None |
|---|
| 342 | has_error_dy = self.current_dataset.err_data is not None |
|---|
| 343 | has_mask = self.current_dataset.mask is not None |
|---|
| 344 | x = self.current_dataset.qx_data |
|---|
| 345 | self.current_dataset.data = self.current_dataset.data[x != 0] |
|---|
| 346 | self.current_dataset.qx_data = self.current_dataset.qx_data[x != 0] |
|---|
| 347 | self.current_dataset.qy_data = self.current_dataset.qy_data[x != 0] |
|---|
| 348 | self.current_dataset.q_data = np.sqrt( |
|---|
| 349 | np.square(self.current_dataset.qx_data) + np.square( |
|---|
| 350 | self.current_dataset.qy_data)) |
|---|
| 351 | if has_error_dy: |
|---|
| 352 | self.current_dataset.err_data = self.current_dataset.err_data[x != 0] |
|---|
| 353 | if has_error_dqx: |
|---|
| 354 | self.current_dataset.dqx_data = self.current_dataset.dqx_data[x != 0] |
|---|
| 355 | if has_error_dqy: |
|---|
| 356 | self.current_dataset.dqy_data = self.current_dataset.dqy_data[x != 0] |
|---|
| 357 | if has_mask: |
|---|
| 358 | self.current_dataset.mask = self.current_dataset.mask[x != 0] |
|---|
| 359 | |
|---|
| 360 | def reset_data_list(self, no_lines=0): |
|---|
| 361 | """ |
|---|
| 362 | Reset the plottable_1D object |
|---|
| 363 | """ |
|---|
| 364 | # Initialize data sets with arrays the maximum possible size |
|---|
| 365 | x = np.zeros(no_lines) |
|---|
| 366 | y = np.zeros(no_lines) |
|---|
| 367 | dx = np.zeros(no_lines) |
|---|
| 368 | dy = np.zeros(no_lines) |
|---|
| 369 | self.current_dataset = plottable_1D(x, y, dx, dy) |
|---|
| 370 | |
|---|
| 371 | @staticmethod |
|---|
| 372 | def splitline(line): |
|---|
| 373 | """ |
|---|
| 374 | Splits a line into pieces based on common delimiters |
|---|
| 375 | :param line: A single line of text |
|---|
| 376 | :return: list of values |
|---|
| 377 | """ |
|---|
| 378 | # Initial try for CSV (split on ,) |
|---|
| 379 | toks = line.split(',') |
|---|
| 380 | # Now try SCSV (split on ;) |
|---|
| 381 | if len(toks) < 2: |
|---|
| 382 | toks = line.split(';') |
|---|
| 383 | # Now go for whitespace |
|---|
| 384 | if len(toks) < 2: |
|---|
| 385 | toks = line.split() |
|---|
| 386 | return toks |
|---|
| 387 | |
|---|
| 388 | @abstractmethod |
|---|
| 389 | def get_file_contents(self): |
|---|
| 390 | """ |
|---|
| 391 | Reader specific class to access the contents of the file |
|---|
| 392 | All reader classes that inherit from FileReader must implement |
|---|
| 393 | """ |
|---|
| 394 | pass |
|---|
