解析以空格格式化的报告pyparsing中的多行标题

Question

我正在尝试分析表中具有多行标题的文件：

                        Categ_1   Categ_2   Categ_3    Categ_4
data1 Group             Data      Data      Data       Data     (     %)  Options
--------------------------------------------------------------------------------
param_group1            6.366e-03 6.644e-03 6.943e-05    0.0131 (57.42%)  i
param_group2            1.251e-05 7.253e-06 4.256e-04 4.454e-04 ( 1.96%)  
param_group3            2.205e-05 6.421e-05 2.352e-03 2.438e-03 (10.70%)  
param_group4            1.579e-07    0.0000 1.479e-05 1.495e-05 ( 0.07%)  
param_group5            3.985e-03 2.270e-07 2.789e-03 6.775e-03 (29.74%)  
param_group6            0.0000    0.0000    0.0000    0.0000 ( 0.00%)  
param_group7            -8.121e-09
                                     0.0000 1.896e-08 1.084e-08 ( 0.00%)  

过去我已经成功地使用pyparsing来解析这样的表，但是标头在一行中，而且标头字段中都没有空格( %)

这是我的做法：

def mustMatchCols(startloc,endloc):
    return lambda s,l,t: startloc <= col(l,s) <= endloc+1

def tableValue(expr, colstart, colend):
    return Optional(expr.copy().addCondition(mustMatchCols(colstart,colend), message="text not in expected columns"))

if header:
    column_lengths = determine_header_column_widths(header_line)

# Then run the tableValue function for each start,end pair.

是否有内置的构造/示例用于以pyparsing或任何其他方法来解析此类空间格式的表？

Answer 1

如果可以预先确定列的宽度，那么这里是将多个列标题缝合在一起的代码：

headers = """\
                        Categ_1   Categ_2   Categ_3    Categ_4
data1 Group             Data      Data      Data       Data     (     %)  Options
"""

col_widths = [24, 10, 10, 11, 9, 10, 10]

# convert widths to slices
col_slices = []
prev = 0
for cw in col_widths:
    col_slices.append(slice(prev, prev + cw))
    prev += cw

# verify slices
# for line in headers.splitlines():
#     for slc in col_slices:
#         print(line[slc])

def extract_line_parts(slices, line_string):
    return [line_string[slc].strip() for slc in slices]

# extract the different column header parts
parts = [extract_line_parts(col_slices, line) for line in headers.splitlines()]
for p in parts:
    print(p)

# use zip(*parts) to transpose list of row parts into list of column parts
header_cols = list(zip(*parts))
print(header_cols)

for header in header_cols:
    print(' '.join(filter(None, header)))

打印：

['', 'Categ_1', 'Categ_2', 'Categ_3', 'Categ_4', '', '']
['data1 Group', 'Data', 'Data', 'Data', 'Data', '(     %)', 'Options']

[('', 'data1 Group'), ('Categ_1', 'Data'), ('Categ_2', 'Data'), ('Categ_3', 'Data'), ('Categ_4', 'Data'), ('', '(     %)'), ('', 'Options')]

data1 Group
Categ_1 Data
Categ_2 Data
Categ_3 Data
Categ_4 Data
(     %)
Options