python numpy文本文件拆分

Question

I have a text file as formatted below with large number of lines,

[ABC] [text text] [1234]
[DEF] [text text: text] [2345]
....

I want to split the columns into arrays.

names = [ ABC , DEF]
text = [text text, text text: text]
values = [1234, 2345]

I am trying with numpy.genfromtxt but I am not sure how to set the delimiter as there are spaces within the text content. Is it possible to have a delimiter defined to be '[]' in some way.

Answer 1

Here is an example of transposing Columns to Rows.

>>> import numpy as np
>>> s = "[ABC] [text text] [1234]\n[DEF] [text text: text] [2345]"
>>> lines = s.split('\n')
# this is were the delimiters are split apart --vvv
>>> rows = [line.lstrip('[').rstrip(']').split('] [') for line in lines] # list comprehension
>>> rows
[['ABC', 'text text', '1234'], ['DEF', 'text text: text', '2345']]
>>> np_rows = np.array(rows)
>>> np_rows.T
array([['ABC', 'DEF'],
       ['text text', 'text text: text'],
       ['1234', '2345']],
      dtype='|S15')
>>> np_rows.transpose()
array([['ABC', 'DEF'],
       ['text text', 'text text: text'],
       ['1234', '2345']],
      dtype='|S15')