CSV With no header row

[robi...@e2eservices.co.uk]

Hello,

I have what's hopefully a really simple question, but I can't see anything obviously in the docs for it.

I want to ingest a CSV file that has no header row, and either be able to pre-set the headers for the table or ideally tell the `etl.fromcsv(..)` what the headers should be called..

At the moment I've only really tried the following:

>>> import petl as etl

>>> table = etl.fromcsv("test.csv")

>>> table

+------------+------------------------------------+-----+-------+-------+

| 11/02/17   | this is a string                   | 1   | 6     | asd   |

+============+====================================+=====+=======+=======+

| '12/02/17' | 'oo another string'                | '2' | '7'   | 'qwe' |

+------------+------------------------------------+-----+-------+-------+

| '24/03/17' | 'one more time lets have a string' | '3' | '5.5' | 'zxc' |

+------------+------------------------------------+-----+-------+-------+

It looks like the fromcsv, function would need a new args added to set the header on the underlying table object (much like the fromcolumns method).

Thanks,

Robin

[robi...@e2eservices.co.uk]

I've got the behaviour i wanted but by using the petl.io.json.fromdicts() and csv.DictReader

>>> fieldnames = ['1','2','3','4','5']
>>> sourcerows = []

>>> with open('test.csv') as csvfile:

...     reader = csv.DictReader(csvfile, fieldnames=fieldnames)

...     for row in reader:

...             sourcerows.append(row)

>>> table = etl.json.fromdicts(sourcerows)

>>> table

+------------+------------------------------------+-----+-------+-------+

| 1          | 2                                  | 3   | 4     | 5     |

+============+====================================+=====+=======+=======+

| '11/02/17' | 'this is a string'                 | '1' | '6'   | 'asd' |

+------------+------------------------------------+-----+-------+-------+

| '12/02/17' | 'oo another string'                | '2' | '7'   | 'qwe' |

+------------+------------------------------------+-----+-------+-------+

| '24/03/17' | 'one more time lets have a string' | '3' | '5.5' | 'zxc' |

+------------+------------------------------------+-----+-------+-------+

Would still like to know if there is a way to do it via fromcsv, or if there could be

[robi...@e2eservices.co.uk]

Wasn't too hard to just add the functionality myself

https://github.com/alimanfoo/petl/pull/421

[Alistair Miles]

Thanks Robin, that's a nice convenience feature, have merged.

FWIW there is also a pushheader() function which can be used to prepend a header row in any situation where you are without one. E.g., etl.fromcsv('headerless.csv').pushheader(['1', '2', '3', '4', '5']).

Cheers,

Alistair

On Thu, Apr 6, 2017 at 11:38 AM, <robi...@e2eservices.co.uk> wrote:

Wasn't too hard to just add the functionality myself

https://github.com/alimanfoo/petl/pull/421

--
You received this message because you are subscribed to the Google Groups "python-etl" group.
To unsubscribe from this group and stop receiving emails from it, send an email to python-etl+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Alistair Miles
Head of Epidemiological Informatics
Centre for Genomics and Global Health <http://cggh.org>
The Wellcome Trust Centre for Human Genetics
Roosevelt Drive
Oxford
OX3 7BN
United Kingdom
Email: alim...@googlemail.com
Web: http://purl.org/net/aliman
Twitter: https://twitter.com/alimanfoo
Tel: +44 (0)1865 287721