# Copyright 2019-2021 Huawei Technologies Co., Ltd
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
"""
This module is to write data into mindrecord.
"""
import os
import platform
import queue
import re
import shutil
import stat
import time
import multiprocessing as mp
import numpy as np
from mindspore import log as logger
from .shardwriter import ShardWriter
from .shardreader import ShardReader
from .shardheader import ShardHeader
from .shardindexgenerator import ShardIndexGenerator
from .shardutils import MIN_SHARD_COUNT, MAX_SHARD_COUNT, VALID_ATTRIBUTES, VALID_ARRAY_ATTRIBUTES, \
check_filename, VALUE_TYPE_MAP, SUCCESS
from .common.exceptions import ParamValueError, ParamTypeError, MRMInvalidSchemaError, MRMDefineIndexError
from .config import _get_enc_key, _get_enc_mode, _get_dec_mode, _get_hash_mode, encrypt, decrypt, append_hash_to_file, \
verify_file_hash
__all__ = ['FileWriter']
[文档]class FileWriter:
r"""
Class to write user defined raw data into MindRecord files.
Note:
After the MindRecord file is generated, if the file name is changed,
the file may fail to be read.
Args:
file_name (str): File name of MindRecord file.
shard_num (int, optional): The Number of MindRecord files.
It should be between [1, 1000]. Default: ``1`` .
overwrite (bool, optional): Whether to overwrite if the file already exists. Default: ``False`` .
Raises:
ParamValueError: If `file_name` or `shard_num` or `overwrite` is invalid.
Examples:
>>> from mindspore.mindrecord import FileWriter
>>>
>>> writer = FileWriter(file_name="test.mindrecord", shard_num=1, overwrite=True)
>>> schema_json = {"file_name": {"type": "string"}, "label": {"type": "int32"}, "data": {"type": "bytes"}}
>>> writer.add_schema(schema_json, "test_schema")
>>> indexes = ["file_name", "label"]
>>> writer.add_index(indexes)
>>> for i in range(10):
... data = [{"file_name": str(i) + ".jpg", "label": i,
... "data": b"\x10c\xb3w\xa8\xee$o&<q\x8c\x8e(\xa2\x90\x90\x96\xbc\xb1\x1e\xd4QER\x13?\xff"}]
... writer.write_raw_data(data)
>>> writer.commit()
"""
def __init__(self, file_name, shard_num=1, overwrite=False):
if platform.system().lower() == "windows":
file_name = file_name.replace("\\", "/")
check_filename(file_name)
self._file_name = file_name
if shard_num is not None:
if isinstance(shard_num, int):
if shard_num < MIN_SHARD_COUNT or shard_num > MAX_SHARD_COUNT:
raise ParamValueError("Parameter shard_num's value: {} should between {} and {}."
.format(shard_num, MIN_SHARD_COUNT, MAX_SHARD_COUNT))
else:
raise ParamValueError("Parameter shard_num's type is not int.")
else:
raise ParamValueError("Parameter shard_num is None.")
if not isinstance(overwrite, bool):
raise ParamValueError("Parameter overwrite's type is not bool.")
self._shard_num = shard_num
self._index_generator = True
suffix_shard_size = len(str(self._shard_num - 1))
if self._shard_num == 1:
self._paths = [self._file_name]
else:
if _get_enc_key() is not None or _get_hash_mode() is not None:
raise RuntimeError("When encode mode or hash check is enabled, " +
"the automatic sharding function is unavailable.")
self._paths = ["{}{}".format(self._file_name,
str(x).rjust(suffix_shard_size, '0'))
for x in range(self._shard_num)]
self._overwrite = overwrite
self._append = False
self._flush = False
self._header = ShardHeader()
self._writer = ShardWriter()
self._generator = None
# parallel write mode
self._parallel_writer = None
self._writers = None
self._queues = [] # the data queue from main process to worker process
self._msg_queues = [] # the err msg from worker process to main process
self._workers = None
self._workers_loop_index = 0
self._index_workers = None
[文档] @classmethod
def open_for_append(cls, file_name):
r"""
Open MindRecord file and get ready to append data.
Args:
file_name (str): String of MindRecord file name.
Returns:
FileWriter, file writer object for the opened MindRecord file.
Raises:
ParamValueError: If file_name is invalid.
FileNameError: If path contains invalid characters.
MRMOpenError: If failed to open MindRecord file.
MRMOpenForAppendError: If failed to open file for appending data.
Examples:
>>> from mindspore.mindrecord import FileWriter
>>>
>>> data = [{"file_name": "0.jpg", "label": 0,
... "data": b"\x10c\xb3w\xa8\xee$o&<q\x8c\x8e(\xa2\x90\x90\x96\xbc\xb1\x1e\xd4QER\x13?\xff"}]
>>> writer = FileWriter(file_name="test.mindrecord", shard_num=1, overwrite=True)
>>> schema_json = {"file_name": {"type": "string"}, "label": {"type": "int32"}, "data": {"type": "bytes"}}
>>> writer.add_schema(schema_json, "test_schema")
>>> writer.write_raw_data(data)
>>> writer.commit()
>>>
>>> write_append = FileWriter.open_for_append("test.mindrecord")
>>> for i in range(9):
... data = [{"file_name": str(i+1) + ".jpg", "label": i,
... "data": b"\x10c\xb3w\xa8\xee$o&<q\x8c\x8e(\xa2\x90\x90\x96\xbc\xb1\x1e\xd4QER\x13?\xff"}]
... write_append.write_raw_data(data)
>>> write_append.commit()
"""
if platform.system().lower() == "windows":
file_name = file_name.replace("\\", "/")
check_filename(file_name)
# decrypt the data file and index file
index_file_name = file_name + ".db"
decrypt_filename = decrypt(file_name, _get_enc_key(), _get_dec_mode())
decrypt_index_filename = decrypt(index_file_name, _get_enc_key(), _get_dec_mode())
# verify integrity check
verify_file_hash(decrypt_filename)
verify_file_hash(decrypt_index_filename)
# move after decrypt and hash check all success
if decrypt_filename != file_name:
shutil.move(decrypt_filename, file_name)
shutil.move(decrypt_index_filename, index_file_name)
# construct ShardHeader
reader = ShardReader()
reader.open(file_name, False)
header = ShardHeader(reader.get_header())
reader.close()
instance = cls("append")
instance.init_append(file_name, header)
return instance
# pylint: disable=missing-docstring
def init_append(self, file_name, header):
self._append = True
if platform.system().lower() == "windows":
self._file_name = file_name.replace("\\", "/")
else:
self._file_name = file_name
self._header = header
self._writer.open_for_append(self._file_name)
self._paths = [self._file_name]
[文档] def add_schema(self, content, desc=None):
"""
The schema is added to describe the raw data to be written.
Note:
Please refer to the Examples of :class:`mindspore.mindrecord.FileWriter` .
.. list-table:: The data types supported by MindRecord.
:widths: 25 25 50
:header-rows: 1
* - Data Type
- Data Shape
- Details
* - int32
- /
- integer number
* - int64
- /
- integer number
* - float32
- /
- real number
* - float64
- /
- real number
* - string
- /
- string data
* - bytes
- /
- binary data
* - int32
- [-1] / [-1, 32, 32] / [3, 224, 224]
- numpy ndarray
* - int64
- [-1] / [-1, 32, 32] / [3, 224, 224]
- numpy ndarray
* - float32
- [-1] / [-1, 32, 32] / [3, 224, 224]
- numpy ndarray
* - float64
- [-1] / [-1, 32, 32] / [3, 224, 224]
- numpy ndarray
Args:
content (dict): Dictionary of schema content.
desc (str, optional): String of schema description, Default: ``None`` .
Raises:
MRMInvalidSchemaError: If schema is invalid.
MRMBuildSchemaError: If failed to build schema.
MRMAddSchemaError: If failed to add schema.
Examples:
>>> # Examples of available schemas
>>> schema1 = {"file_name": {"type": "string"}, "label": {"type": "int32"}, "data": {"type": "bytes"}}
>>> schema2 = {"input_ids": {"type": "int32", "shape": [-1]},
... "input_masks": {"type": "int32", "shape": [-1]}}
"""
ret, error_msg = self._validate_schema(content)
if ret is False:
raise MRMInvalidSchemaError(error_msg)
schema = self._header.build_schema(content, desc)
self._header.add_schema(schema)
[文档] def add_index(self, index_fields):
"""
Select index fields from schema to accelerate reading.
schema is added through `add_schema` .
Note:
The index fields should be primitive type. e.g. int/float/str.
If the function is not called, the fields of the primitive type
in schema are set as indexes by default.
Please refer to the Examples of :class:`mindspore.mindrecord.FileWriter` .
Args:
index_fields (list[str]): fields from schema.
Raises:
ParamTypeError: If index field is invalid.
MRMDefineIndexError: If index field is not primitive type.
MRMAddIndexError: If failed to add index field.
MRMGetMetaError: If the schema is not set or failed to get meta.
"""
if not index_fields or not isinstance(index_fields, list):
raise ParamTypeError('index_fields', 'list')
for field in index_fields:
if field in self._header.blob_fields:
raise MRMDefineIndexError("Failed to set field {} since it's not primitive type.".format(field))
if not isinstance(field, str):
raise ParamTypeError('index field', 'str')
self._header.add_index_fields(index_fields)
[文档] def write_raw_data(self, raw_data, parallel_writer=False):
"""
Convert raw data into a series of consecutive MindRecord \
files after the raw data is verified against the schema.
Note:
Please refer to the Examples of :class:`mindspore.mindrecord.FileWriter` .
Args:
raw_data (list[dict]): List of raw data.
parallel_writer (bool, optional): Write raw data in parallel if it equals to True. Default: ``False`` .
Raises:
ParamTypeError: If index field is invalid.
MRMOpenError: If failed to open MindRecord file.
MRMValidateDataError: If data does not match blob fields.
MRMSetHeaderError: If failed to set header.
MRMWriteDatasetError: If failed to write dataset.
TypeError: If parallel_writer is not bool.
"""
if not isinstance(parallel_writer, bool):
raise TypeError("The parameter `parallel_writer` must be bool.")
if self._parallel_writer is None:
self._parallel_writer = parallel_writer
if self._parallel_writer != parallel_writer:
raise RuntimeError("The parameter `parallel_writer` must be consistent during use.")
if not self._parallel_writer:
if not self._writer.is_open:
self._writer.open(self._paths, self._overwrite)
if not self._writer.get_shard_header():
self._writer.set_shard_header(self._header)
if not isinstance(raw_data, list):
raise ParamTypeError('raw_data', 'list')
if self._flush and not self._append:
raise RuntimeError("Not allowed to call `write_raw_data` on flushed MindRecord files." \
"When creating new MindRecord files, please remove `commit` before " \
"`write_raw_data`. In other cases, when appending to existing MindRecord files, " \
"please call `open_for_append` first and then `write_raw_data`.")
for each_raw in raw_data:
if not isinstance(each_raw, dict):
raise ParamTypeError('raw_data item', 'dict')
self._verify_based_on_schema(raw_data)
self._writer.write_raw_data(raw_data, True, parallel_writer)
return
## parallel write mode
# init the _writers and launch the workers
if self._writers is None:
self._writers = [None] * len(self._paths) # writers used by worker
self._queues = [mp.Queue(2) for _ in range(len(self._paths))] # data queue for worker
self._msg_queues = [mp.Queue(2) for _ in range(len(self._paths))] # msg queue for worker
self._workers = [None] * len(self._paths) # worker process
worker_pid_list = []
for i, path in enumerate(self._paths):
self._writers[i] = ShardWriter()
self._writers[i].open([path], self._overwrite)
self._writers[i].set_shard_header(self._header)
# launch the workers for parallel write
self._queues[i]._joincancelled = True # pylint: disable=W0212
self._msg_queues[i]._joincancelled = True # pylint: disable=W0212
p = mp.Process(target=self._write_worker, name="WriterWorker" + str(i),
args=(i, self._queues[i], self._msg_queues[i]))
p.daemon = True
p.start()
logger.info("Start worker process(pid:{}) to parallel write.".format(p.pid))
self._workers[i] = p
worker_pid_list.append(self._workers[i].pid)
# fill the self._queues
check_interval = 0.5 # 0.5s
start_time = time.time()
while True:
try:
self._queues[self._workers_loop_index].put(raw_data, block=False)
except queue.Full:
if time.time() - start_time > check_interval:
start_time = time.time()
logger.warning("Because there are too few MindRecord file shards, the efficiency of parallel " \
"writing is too low. You can stop the current task and add the parameter " \
"`shard_num` of `FileWriter` to upgrade the task.")
self.check_worker_status(self._workers_loop_index)
continue
self._workers_loop_index = (self._workers_loop_index + 1) % len(self._paths)
return
def check_all_worker_status(self):
for index in range(len(self._msg_queues)):
self.check_worker_status(index)
def check_worker_status(self, index):
if not self._msg_queues[index].empty():
if self._msg_queues[index].get() == "Error":
raise RuntimeError("Worker process(pid:{}) has stopped abnormally. Please check " \
"the above log".format(self._workers[index].pid))
[文档] def set_page_size(self, page_size):
"""
Set the size of page that represents the area where data is stored, \
and the areas are divided into two types: raw page and blob page. \
The larger a page, the more data the page can store. If the size of \
a sample is larger than the default size (32MB), users need to call the API \
to set a proper size.
Args:
page_size (int): Size of page, in bytes, which between 32*1024(32KB) and
256*1024*1024(256MB).
Raises:
MRMInvalidPageSizeError: If failed to set page size.
Examples:
>>> from mindspore.mindrecord import FileWriter
>>> writer = FileWriter(file_name="test.mindrecord", shard_num=1)
>>> writer.set_page_size(1 << 26) # 64MB
"""
self._writer.set_page_size(page_size)
[文档] def commit(self): # pylint: disable=W0212
"""
Flush data in memory to disk and generate the corresponding database files.
Note:
Please refer to the Examples of :class:`mindspore.mindrecord.FileWriter` .
Raises:
MRMOpenError: If failed to open MindRecord file.
MRMSetHeaderError: If failed to set header.
MRMIndexGeneratorError: If failed to create index generator.
MRMGenerateIndexError: If failed to write to database.
MRMCommitError: If failed to flush data to disk.
RuntimeError: Parallel write failed.
"""
if not self._parallel_writer:
self._flush = True
if not self._writer.is_open:
self._writer.open(self._paths, self._overwrite)
# permit commit without data
if not self._writer.get_shard_header():
self._writer.set_shard_header(self._header)
self._writer.commit()
if self._index_generator:
if self._append:
self._generator = ShardIndexGenerator(self._file_name, self._append)
elif len(self._paths) >= 1:
self._generator = ShardIndexGenerator(os.path.realpath(self._paths[0]), self._append)
self._generator.build()
self._generator.write_to_db()
else:
# maybe a empty mindrecord, so need check _writers
if self._writers is None:
self._writers = [None] * len(self._paths)
for i, path in enumerate(self._paths):
self._writers[i] = ShardWriter()
self._writers[i].open(path, self._overwrite)
self._writers[i].set_shard_header(self._header)
self._parallel_commit()
# change file mode first, because encrypt / hash check may failed
mindrecord_files = []
index_files = []
for item in self._paths:
if os.path.exists(item):
os.chmod(item, stat.S_IRUSR | stat.S_IWUSR)
mindrecord_files.append(item)
index_file = item + ".db"
if os.path.exists(index_file):
os.chmod(index_file, stat.S_IRUSR | stat.S_IWUSR)
index_files.append(index_file)
for item in self._paths:
if os.path.exists(item):
# add the integrity check string
if _get_hash_mode() is not None:
append_hash_to_file(item)
append_hash_to_file(item + ".db")
# encrypt the mindrecord file
if _get_enc_key() is not None:
encrypt(item, _get_enc_key(), _get_enc_mode())
encrypt(item + ".db", _get_enc_key(), _get_enc_mode())
logger.info("The list of mindrecord files created are: {}, and the list of index files are: {}".format(
mindrecord_files, index_files))
def _index_worker(self, i):
"""The worker do the index generator"""
generator = ShardIndexGenerator(os.path.realpath(self._paths[i]), False)
generator.build()
generator.write_to_db()
def _parallel_commit(self):
"""Parallel commit"""
# waiting for all the self._queues had been writed by workers
for index in range(len(self._paths)):
while not self._queues[index].empty():
logger.info("Waiting for the data is writed by worker: {}.".format(self._workers[index].pid))
time.sleep(1)
# check the workers status
self.check_worker_status(index)
# send EOF to worker process
for index in range(len(self._paths)):
logger.info("Send EOF flag to worker: {}.".format(self._workers[index].pid))
while True:
try:
self._queues[index].put("EOF", block=False)
except queue.Full:
# check the workers status
self.check_worker_status(index)
time.sleep(1)
continue
break
worker_sucess = 0
for index in range(len(self._msg_queues)):
while True:
logger.info("Waiting for the worker: {} to exit.".format(self._workers[index].pid))
if not self._msg_queues[index].empty():
ret = self._msg_queues[index].get()
if ret == "Success":
worker_sucess += 1
break
else:
raise RuntimeError("Worker process(pid:{}) has stopped abnormally. Please check " \
"the above log".format(self._workers[index].pid))
time.sleep(1)
del self._queues
del self._msg_queues
if self._index_generator:
# use parallel index workers to generator index
self._index_workers = [None] * len(self._paths)
for index in range(len(self._paths)):
p = mp.Process(target=self._index_worker, name="IndexWorker" + str(index), args=(index,))
p.daemon = True
p.start()
logger.info("Start worker process(pid:{}) to generate index.".format(p.pid))
self._index_workers[index] = p
# wait the index workers stop
for index in range(len(self._paths)):
self._index_workers[index].join()
def _validate_array(self, k, v):
"""
Validate array item in schema
Args:
k (str): Key in dict.
v (dict): Sub dict in schema
Returns:
bool, whether the array item is valid.
str, error message.
"""
if v['type'] not in VALID_ARRAY_ATTRIBUTES:
error = "Field '{}' contain illegal " \
"attribute '{}'.".format(k, v['type'])
return False, error
if 'shape' in v:
if isinstance(v['shape'], list) is False:
error = "Field '{}' contain illegal " \
"attribute '{}'.".format(k, v['shape'])
return False, error
else:
error = "Field '{}' contains illegal attributes.".format(v)
return False, error
return True, ''
def _verify_based_on_schema(self, raw_data):
"""
Verify data according to schema and remove invalid data if validation failed.
1) allowed data type contains: "int32", "int64", "float32", "float64", "string", "bytes".
Args:
raw_data (list[dict]): List of raw data.
"""
error_data_dic = {}
schema_content = self._header.schema
for field in schema_content:
for i, v in enumerate(raw_data):
if i in error_data_dic:
continue
if field not in v:
error_data_dic[i] = "for schema, {} th data is wrong, " \
"there is not field: '{}' in the raw data.".format(i, field)
continue
field_type = type(v[field]).__name__
if field_type not in VALUE_TYPE_MAP:
error_data_dic[i] = "for schema, {} th data is wrong, " \
"data type: '{}' for field: '{}' is not matched.".format(i, field_type, field)
continue
if schema_content[field]["type"] not in VALUE_TYPE_MAP[field_type]:
error_data_dic[i] = "for schema, {} th data is wrong, " \
"data type: '{}' for field: '{}' is not matched." \
.format(i, schema_content[field]["type"], field)
continue
if field_type == 'ndarray':
if 'shape' not in schema_content[field]:
error_data_dic[i] = "for schema, {} th data is wrong, " \
"data shape for field: '{}' is not specified.".format(i, field)
elif 'type' not in schema_content[field]:
error_data_dic[i] = "for schema, {} th data is wrong, " \
"data type for field: '{}' is not specified.".format(i, field)
elif schema_content[field]['type'] != str(v[field].dtype):
error_data_dic[i] = "for schema, {} th data is wrong, " \
"data type: '{}' for field: '{}' is not matched." \
.format(i, str(v[field].dtype), field)
else:
try:
np.reshape(v[field], schema_content[field]['shape'])
except ValueError:
error_data_dic[i] = "for schema, {} th data is wrong, " \
"data shape: '{}' for field: '{}' is not matched." \
.format(i, str(v[field].shape), field)
error_data_dic = sorted(error_data_dic.items(), reverse=True)
for i, v in error_data_dic:
raw_data.pop(i)
logger.warning(v)
def _validate_schema(self, content):
"""
Validate schema and return validation result and error message.
Args:
content (dict): Dict of raw schema.
Returns:
bool, whether the schema is valid.
str, error message.
"""
error = ''
if not content:
error = 'Schema content is empty.'
return False, error
if isinstance(content, dict) is False:
error = 'Schema content should be dict.'
return False, error
for k, v in content.items():
if not re.match(r'^[0-9a-zA-Z\_]+$', k):
error = "Field '{}' should be composed of " \
"'0-9' or 'a-z' or 'A-Z' or '_'.".format(k)
return False, error
if v and isinstance(v, dict):
if len(v) == 1 and 'type' in v:
if v['type'] not in VALID_ATTRIBUTES:
error = "Field '{}' contain illegal " \
"attribute '{}'.".format(k, v['type'])
return False, error
elif len(v) == 2 and 'type' in v:
res_1, res_2 = self._validate_array(k, v)
if not res_1:
return res_1, res_2
else:
error = "Field '{}' contains illegal attributes.".format(v)
return False, error
else:
error = "Field '{}' should be dict.".format(k)
return False, error
return True, error
def _write_worker(self, i, in_queue, msg_queue):
"""The worker do the data check and write to disk for parallel mode"""
while True:
# try to get new raw_data from master
try:
raw_data = in_queue.get(block=False)
except queue.Empty:
continue
# get EOF from master, worker should commit and stop
if raw_data == "EOF":
ret = self._writers[i].commit()
if ret != SUCCESS:
msg_queue.put("Error")
raise RuntimeError("Commit the {}th shard of MindRecord file failed.".format(i))
logger.info("Send Success flag from worker: {} to master: {}.".format(os.getpid(), os.getppid()))
msg_queue.put("Success")
break
# check the raw_data
if not isinstance(raw_data, list):
msg_queue.put("Error")
raise ParamTypeError('raw_data', 'list')
for each_raw in raw_data:
if not isinstance(each_raw, dict):
msg_queue.put("Error")
raise ParamTypeError('raw_data item', 'dict')
try:
self._verify_based_on_schema(raw_data)
self._writers[i].write_raw_data(raw_data, True, False)
except Exception as e:
msg_queue.put("Error")
raise e