In python, is there an easy way to set up a file-like object for writing that is actually backed by multiple output streams? For instance, I want something like this:
file1 = open("file1.txt", "w")
file2 = open("file2.txt", "w")
ostream = OStreamWrapper(file1, file2, sys.stdout)
#Write to both files and stdout at once:
ostream.write("ECHO!")
So what I'm looking for is OStreamWrapper. I know it'd be pretty easy to write my own, but if there's an existing one, I'd rather use that and not have to worry about finding and covering edge cases.
class OStreamWrapper(object):
def __init__(self, *streams):
self.streams = list(streams)
def write(self, string):
for stream in self.streams:
stream.write(string)
def writelines(self, lines):
# If you want to use stream.writelines(), you have
# to convert lines into a list/tuple as it could be
# a generator.
for line in lines:
for stream in self.streams:
stream.write(line)
def flush(self):
for stream in self.streams:
stream.flush()
Way to wrap all public file functions:
import sys
def _call_for_all_streams(func_name):
def wrapper(self, *args, **kwargs):
result = []
for stream in self._streams:
func = getattr(stream, func_name)
result.append(func(*args, **kwargs))
return result
return wrapper
class OStreamWrapper(object):
def __init__(self, *streams):
self._streams = streams
for method in filter(lambda x: not x.startswith('_'), dir(file)):
setattr(OStreamWrapper, method, _call_for_all_streams(method))
if __name__ == '__main__':
file1 = open("file1.txt", "w")
file2 = open("file2.txt", "w")
ostream = OStreamWrapper(file1, file2, sys.stdout)
ostream.write("ECHO!")
ostream.close()
But it's kinda dirty.
Logbook is another option although it is more than that. Its handlers are more powerful and you can combine whatever you like.
Related
So I am working on a script to generate serialnumbers for a product. I want to make a txt file where I the script prints the generated key. somehow it cant print in there but I don't know what I need to changes about it.
key = Key('aaaa-bbbb-cccc-dddd-1111')
fh = open('key.txt')
fh.write(Key)
Ok, based on your response, I've mocked up the Key class as follows. Without more information, it's not possible to give you a definitive answer, but hopefully this helps!
class Key:
def __init__(self, serial):
self.serial = serial
def process_serial(self):
# Your processing here
...
return processed_serial # This should be a string
Then to write to file, you can do:
key = Key('aaaa-bbbb-cccc-dddd-1111')
with open('key.txt', 'w') as f:
f.write(key.process_serial())
Alternatively, you can add a __str__ method to your class, which will specify what happens when you call the Python builtin str on your object.
class Key:
def __init__(self, serial):
self.serial = serial
def __str__(self):
out = ... # construct what you want to write to file
return out
Giving:
key = Key('aaaa-bbbb-cccc-dddd-1111')
with open('key.txt', 'w') as f:
f.write(str(key))
You might also consider adding this as a method to your Key class
class Key:
__init__(self, serial):
self.serial = serial
def process_serial(self):
# Your processing here
...
return processed_serial # This should be a string
def write(self, file_name):
with open(file_name, 'w') as f:
f.write(self.process_serial)
Try:
key = "Key('aaaa-bbbb-cccc-dddd-1111')"
fh = open('key.txt', "w")
fh.write(key)
To generate a text file that doesn't already exist you need to use "w" .
Try doing:
key = Key('aaaa-bbbb-cccc-dddd-1111')
with open('key.txt', 'w') as fh:
fh.write(key)
Hope that Helps!
Note: it must be in the with ... so it writes, if its not there the file is considered as closed.
I have csv file having contents below
101,item_1
101,item_1
if it is csv my below code will execute
import csv
fName = input()
def read_csv(fName):
try:
with open(fName, 'r') as f:
reader = csv.reader(f)
for row in reader:
print (row)
read_csv(fName)
Here how to write the exception in decorator function and call on the top of that.
first decorator
if fName not endswith .txt or .csv then it has to generate output not accept
Second decorator
if fName = file.txt text file then below operations has to taken care
def read_txt(fName):
f = open(fName, "r")
print(f.readline())
if csv then first function to execute and if txt next function to execute. How to achieve using decorator. I can put if conditon to achieve the situation, but that is not the case
My whole code without decorator is below
fName = input()
def read_csv(fName):
if fName.endswith('.csv'):
#print ('hi')
try:
with open(fName, 'r') as f:
reader = csv.reader(f)
for row in reader:
print (row)
except IOError:
print ("Could not read file:", fName)
#SECOND DECORATOR
if fName.endswith('.txt'):
f = open(fName, "r")
print(f.readline())
#FIRST DECORATOR
if not(fName.endswith('.csv')) and not(fName.endswith('.txt')):
print ('not accept')
read_csv(fName)
You can do it like this with decorators:
import functools
def check_arguments(func):
#functools.wraps(func)
def wrapper(*args, **kwargs):
fname = kwargs['fname']
if not fname.endswith('.csv') and not fname.endswith('.txt'):
print('not accept')
return func(*args, **kwargs)
return wrapper
def set_file_processor(func):
def read_csv(fname):
print('read_csv', fname)
def read_txt(fname):
print('read_txt', fname)
#functools.wraps(func)
def wrapper(*args, **kwargs):
fname = kwargs['fname']
if fname.endswith('.csv'):
read_csv(fname)
elif fname.endswith('.txt'):
read_txt(fname)
return func(*args, **kwargs)
return wrapper
#check_arguments
#set_file_processor
def process(fname):
pass
process(fname='input.csv')
Your problem doesn't seem to come under decorator but under factory pattern i.e. process differently based on the input file.
The below code is a very simple and basic Factory pattern solution to your problem, this should be modified accordingly as per your need,
import os
from abc import ABC, abstractmethod
class FileProcessor(ABC):
#abstractmethod
def process():
pass
class TextFileProcessor(FileProcessor):
def process(self, file_path):
print("Text file processing goes here")
class CsvFileProcessor(FileProcessor):
def process(self, file_path):
print("CSV file processing goes here")
class DefaultFileProcessor(FileProcessor):
def process(self, file_path):
raise ValueError("File %s is not valid" % file_path)
class FileFactory:
processors = {
'txt': TextFileProcessor,
'csv': CsvFileProcessor,
'default': DefaultFileProcessor
}
def __init__(self, file_path):
if not os.path.exists(file_path):
raise IOError("File not found")
self.file_path = file_path
def process(self):
dot_splits = self.file_path.split(".")
ext = dot_splits[-1] if len(dot_splits) > 1 else "default"
ext = ext if ext in self.processors else "default"
processor_class = self.processors.get(ext)
return processor_class().process(self.file_path)
FileFactory(file_path).process()
In later stage if you would like to add json processor then it can also be done easily by adding
processors = {
'txt': TextFileProcessor,
'csv': CsvFileProcessor,
'json': JsonFileProcessor,
'default': DefaultFileProcessor
}
and creating new Json processor class,
class JsonFileProcessor(FileProcessor):
def process(self, file_path):
print("JSON file processing goes here")
Based on your code and this very useful guide, here is a possible solution:
def read_file_decorator(fName):
def read_csv():
print('read_csv')
with open(fName, 'r') as f:
reader = csv.reader(f)
for row in reader:
print(row)
def read_txt():
print('read_txt')
f = open(fName, 'r')
for row in f:
print(row)
if fName.endswith('.csv'):
return read_csv
elif fName.endswith('.txt'):
return read_txt
else:
return None
reader_function = read_file_decorator(fileName)
if reader_function != None:
reader_function()
else:
print('not accept')
I use a stateful decorator remembering the file name inside the reader function before actually executing it (in order not to pass it twice); and I use the fixed value None for invalid file types.
Based on the requirements use of decorator would be an overkill of decorators. But if it's mandatory to implement this using decorator, this is how we can implement:
We can create a dummy function called read_file and a decorator function called reader
User will always call read_file with filename as argument and decorator function reader will check passed filename extension and call the required function - read_csv or read_text
def reader(fun):
def wrapper(*args):
fname = args[0]
if fname.endswith('.csv'):
read_csv(fname)
elif fname.endswith('.txt'):
read_text(fname)
else:
print('not accepted')
return wrapper
def read_csv(fname):
print('In read_csv()')
def read_text(fname):
print('In read_text()')
#reader
def read_file(fname):
pass
read_file('a.csv')
read_file('a.txt')
read_file('filename.py')
Output
In read_csv()
In read_text()
not accepted
I've made a function for a flask application to create a decorator and a function and then write them to a file but when I run it, it doesn't create a file and write to it and it doesn't return any errors.
def make_route(title):
route = "#app.route(/%s)" %(title)
def welcome():
return render_template("%s.html" %(title))
return welcome
f = open('test1.txt', 'w')
f.write(route, '/n', welcome, '/n')
f.close()
make_route('Hi')
A return statement terminates execution of the function, so any code after it is ignored. Also, write writes a string, not random objects. You want:
def make_route(title):
route = "#app.route(/%s)" %(title)
def welcome():
return render_template("%s.html" %(title))
with open('test1.txt', 'w') as f:
f.write('%r\n%r\n' % (route, welcome))
return welcome
make_route('Hi')
I would use philhag answer but use %s instead of %r or you'll write a string, and you could use .name if you want to use the function more than once(Which you probably do).
def make_route(title):
route = "#app.route('/%s')" %(title)
def welcome():
return render_template("%s.html" %(title))
with open('test2.py', 'w') as f:
f.write('%s\n%s\n' % (route, welcome))
welcome.__name__ = title
return welcome
make_route('Hi')
I'm getting the following error when trying to read the row and column count of a CSV:
> coercing to Unicode: need string or buffer, S3BotoStorageFile found
import csv
class CSV:
def __init__(self, file=None):
self.file = file
def read_file(self):
data = []
file_read = read_file(self.file)
return file_read
def get_row_count(self):
return len(self.read_file())
def get_column_count(self):
new_data = self.read_file()
return len(new_data[0])
def get_data(self, rows=1):
data = self.read_file()
return data[:rows]
def read_file(self):
with open(self.file, 'r') as f:
data = [row for row in csv.reader(f.read().splitlines())]
return data
How do I resolve?
well, after reading your code my first reaction was OMG! How many does he open that poor file?
Here's a new version of your class
class CSV:
def __init__(self, file=None):
self.file = file
with open(self.file, 'r') as f:
self.data = [row for row in csv.reader(f)]
def get_row_count(self):
return len(self.data)
def get_column_count(self):
return len(self.data[0])
def get_data(self, rows=1):
return self.data
I also fixed your csv.reader() handling. It accepts a file object, no need to .read() or .read().splitlines(), it can only lead to errors. Which may be the reason why it failed.
Ok, given from what you say, you're working on AWS, and your file is not a string path to a file, but already a file object. So you don't need the open() part as is. You may want to modify your code so it is as follows:
class CSV:
def __init__(self, f=None):
self.file = f
if isinstance(self.file, str): # if the file is a string, it's a path that has to be opened
with open(self.file, 'r') as f:
self.data = [row for row in csv.reader(f)]
elif isinstance(self.file, File) or isinstance(self.file, file): # if that's a file object, no need to open
self.data = [row for row in csv.reader(self.file)]
else: # otherwise, I don't know what to do, so aaaaaaaargh!
raise Exception("File object type unknown: %s %s" % (type(file), file,))
def get_row_count(self):
return len(self.data)
def get_column_count(self):
return len(self.data[0])
def get_data(self, rows=1):
return self.data
Reading the S3BotoStorage.py, the S3BotoStorage class inherits from django.core.files.base.File, which inherits from django.core.files.utils.FileProxyMixin, which is a composition of attributes of the global python file class.
So a File object is not an instance of file, but it has a compatible interface. Therefore, in the previous code I have tested whether the self.file is a str, then it shall be a path that we open() so we get a file() and parse it. Otherwise, self.file is a File object or a file() object, and we just need to parse it. If it's neither of those, then it's an error, and we shall except.
I have a (recursive) function which I would like to accept either a string or an opened file object. If the argument is a string, then the function opens a file and uses that file object. It seems best to close this opened file object explicitly when I return from the function, but only if a string was passed in. (Imagine the surprise from the user when they pass in an opened file object and find that their file object was closed somewhere). Here's what I'm currently using:
def read_file(f, param):
do_close = isinstance(f,basestring)
f = open(f, 'rb') if do_close else f
try:
info = f.read(4)
#check info here
if info == Info_I_Want(param):
return f.read(get_data(info))
else:
f.seek(goto_new_position(info))
return read_file(f,param)
except IKnowThisError:
return None
finally:
if do_close:
f.close()
You can assume that IKnowThisError will be raised at some point if I don't find the info I want.
This feels very kludgy. Is there a better way?
Why not wrapping your recursive function with a wrapper to avoid overhead ?
def read_file(f, param):
if isinstance(f, basestring):
with open(f, 'rb') as real_f:
return read_file2(real_f, param)
else:
return read_file2(real_f, param)
def read_file2(f, param):
# Now f should be a file object
...
How about calling your function recursively?
def read_file(f, param):
if isinstance(f, basestring):
with open(f, 'rb') as real_f:
return read_file(real_f, param)
else:
# normal path
The upcoming Python 3.3 offers a more general solution for this kind of problem, namely contextlib.ExitStack. This allow to conditionally add context managers to the current with-block:
def read_file(f, param):
with ExitStack() as stack:
if isinstance(f, basestring):
f = stack.enter_context(open(f, 'rb'))
# Your code here