Das deutsche Python-Forum

Code: Alles auswählen

query   | subject     | query  start | query end | subject start | Subject end
Locus_1 | tig00007234 | 1            | 1054      | 140511        | 137383
Locus_1 | tig00007234 | 1            | 1057      | 77478         | 74353
...

Code: Alles auswählen

def read_relationship(filename):
    with open(filename) as f:
        for line in f:
            try:
                parts = line.rstrip().split('|')
                query_name = parts[0]
                subject_name = parts[1]
                query_start = parts[2]
                query_end = parts[3]
                subject_start = parts[4]
                subject_end = parts[5]
            except ValueError:
                continue

Code: Alles auswählen

var data = [
  [
    source: {
      id: 'Locus_1',
      start: 1,
      end: 1054
    },
    target: {
      id: 'tig00007234',
      start: 140511,
      end: 137383
    }
  ],
  [
    source: {
      id: 'Locus_1',
      start: 1,
      end: 1057
    },
    target: {
      id: 'tig00007234',
      start: 77478,
      end: 74353
    }
  ],
  ...
];

Code: Alles auswählen

#!/usr/bin/env python3
# coding: utf-8

import csv
import io
import json

data = '''
query   | subject     | query  start | query end | subject start | Subject end
Locus_1 | tig00007234 | 1            | 1054      | 140511        | 137383
Locus_1 | tig00007234 | 1            | 1057      | 77478         | 74353
'''

def identity(x):
    return x

transmogrifiers = {
    'query': identity,
    'subject': identity,
    'query  start': int,
    'query end': int,
    'subject start': int,
    'Subject end': int,
}

def massage(a_dict):
    return {k.strip(): transmogrifiers[k.strip()](v.strip())
            for k, v in a_dict.items()}

def read_data(data_file):
    for row in csv.DictReader(data_file, delimiter='|', skipinitialspace=True):
        yield massage(row)

def main():
    print(json.dumps(list(read_data(io.StringIO(data.strip())))))

if __name__ == '__main__':
    main()

Code: Alles auswählen

Traceback (most recent call last):
  File "/projects/circos/test.py", line 44, in <module>
    main()
  File "/projects/circos/test.py", line 40, in main
    print(json.dumps(list(read_data(io.StringIO(data.strip())))))
TypeError: initial_value must be unicode or None, not str

Code: Alles auswählen

...
data = u'''
query   | subject     | query  start | query end | subject start | Subject end
Locus_1 | tig00007234 | 1            | 1054      | 140511        | 137383
Locus_1 | tig00007234 | 1            | 1057      | 77478         | 74353
'''
...
transmogrifers = {
    u'query': identity,
    u'subject': identity,
    u'query  start': int,
    u'query end': int,
    u'subject start': int,
    u'Subject end': int,
}
...