Read rows in text file based on different column headers using Python

If text number is same in all file, and text code too, you can use read_csv and then filter columns rows by subset by contains and isnull, notnull:

import pandas as pd
import numpy as np
import io

temp=u""" number amount
    10 34
    23 65
    54 07

    code volume dim
    1 56 34
    23 57 565

    number amount
    40 674
    73 2365

    code volume dim
    341 3456 6534
    23 0957 908565

    number amount
    210 4534
    2343 4565"""

#after testing replace io.StringIO(temp) to filename
df = pd.read_csv(io.StringIO(temp), sep="\s+", index_col=None, skip_blank_lines=True, names=['a', 'b', 'c'])
print df
         a       b       c
0   number  amount     NaN
1       10      34     NaN
2       23      65     NaN
3       54      07     NaN
4     code  volume     dim
5        1      56      34
6       23      57     565
7   number  amount     NaN
8       40     674     NaN
9       73    2365     NaN
10    code  volume     dim
11     341    3456    6534
12      23    0957  908565
13  number  amount     NaN
14     210    4534     NaN
15    2343    4565     NaN

df = df[~((df.a.str.contains('number')) | (df.a.str.contains('code')))]

df1 = df[df.c.isnull()]
df1 = df1[['a', 'b']].reset_index(drop=True)
df1.columns = ['number','amount']

#if all numbers are integers
df1 = df1.astype(int)
print df1
   number  amount
0      10      34
1      23      65
2      54       7
3      40     674
4      73    2365
5     210    4534
6    2343    4565

df2 = df[df.c.notnull()].reset_index(drop=True)
df2.columns = ['code','volume', 'dim']
#if all numbers are integers
df2 = df2.astype(int)
print df2
   code  volume     dim
0     1      56      34
1    23      57     565
2   341    3456    6534
3    23     957  908565