Python CSV module, special split string

Question

i had an unique problem. I have code:

with open("test.csv", "r") as csvFile:
reader = csv.reader(csvFile, skipinitialspace=True)
for row in reader:
    for obj in row:
        print(obj)

and exemplary csv file:

anotherCommand, e=5, f=6, g=7, h=9, test="aaa, bbb, ggggg"

i want split this string in this way:

anotherCommand
e=5
f=6
g=7
h=9
test="aaa, bbb, ggggg"

but code which i was presented, split these string in this way:

anotherCommand
e=5
f=6
g=7
h=9
test="aaa
bbb
ggggg"

This is wrong solution this problem. I saw topic like: Why is the Python CSV reader ignoring double-quoted fields? or How can i parse a comma delimited string into a list (caveat)?

But this example is different, and these examples not come up my expectation. Someone have idea?

@cᴏʟᴅsᴘᴇᴇᴅ doesn't work, the same situation like in linked sites :) — Robert Pawlak, Aug 17 '17 at 08:31
Can you please add some details? What exactly did you do, that didn't work. — cs95, Aug 17 '17 at 08:32
i wrote: with open("test.csv", "r") as csvFile: for line in csv.reader(csvFile, quotechar='"', delimiter=',', quoting=csv.QUOTE_ALL, skipinitialspace=True): print(len(line)) ---- and i still get the same values like presented in my main answer — Robert Pawlak, Aug 17 '17 at 08:35
should i use regex? If yes, did someone could give me a few directions to solution? — Robert Pawlak, Aug 17 '17 at 08:43
@Robert is `anotherCommand, e=5, f=6, g=7, h=9, test="aaa, bbb, ggggg"` an example of a *raw* row in your CSV file? — Jon Clements, Aug 17 '17 at 08:47
@zezollo in first link, there is a pure string, i mean: 56, "some, pure, str" ---- in my case, i have: key="some, pure, string". Understand? — Robert Pawlak, Aug 17 '17 at 08:47
@JonClements real example, which i must process: queue, name="EQ_L2_PS_PEER_0, ESP, ON", id=0x1007, qgroup_addr=0x31f597c0 — Robert Pawlak, Aug 17 '17 at 08:49
Are they all in that format? eg: one column of something, then following columns of key=value or key="value" ? — Jon Clements, Aug 17 '17 at 08:50
no, other row example in my csv file: msg, type=0x0000000000000000, id=0x0304, dst=0x00001003 — Robert Pawlak, Aug 17 '17 at 08:51
Key to your problem is that your "csv" isn't a well-formed CSV file. If you can generate it so it is valid then python csv parsing will work. Seems like for your example you need to add quotes around the test="aaa, bbb, ggggg" so it looks like "test="aaa, bbb, ggggg"", but you might get away with simply adding quotes around every value. — DisappointedByUnaccountableMod, Aug 17 '17 at 09:03

score 1 · Accepted Answer · answered Aug 17 '17 at 08:53

1

You could possibly make use of shlex.split here:

import shlex

with open('test.csv') as fin:
    for line in fin:
        row = [col.rstrip(',') for col in shlex.split(line)]
        print(*row, sep='\n')

answered Aug 17 '17 at 08:53

Jon Clements

138,671
33
247
280

Great man! Works! Could You explain me more Your solution? Or link sites, that should i visit to learn? – Robert Pawlak Aug 17 '17 at 08:58
This is a advantage I planed delete them in further file process – Robert Pawlak Aug 17 '17 at 08:59
1

@Robert as to references - have a look at the documentation for the `shlex` module - it tries to parse text as a shell would when passing command line arguments to a program... (which is what your lines looked more like to me than actual CSV data - so just thought I'd give it a go and see if it produced something workable with for your use case and it mostly did minus retaining trailing commas - hence the `str.rsplit` there)... – Jon Clements Aug 17 '17 at 09:01
this is an output with different format, which was designed on csv. – Robert Pawlak Aug 17 '17 at 09:04
@Robert Might make sense if you do a `values = dict(col.partition('=')[::2] for col in row)` to get a `dict` of key/values to use if that's applicable for what you want to use it for... – Jon Clements Aug 17 '17 at 09:07
@Robert then you get something like: `{'id': '0x1007', 'name': 'EQ_L2_PS_PEER_0, ESP, ON', 'qgroup_addr': '0x31f597c0', 'queue': ''}` - so if the first column is always straight text with key=value, then maybe ignore that, but otherwise, might be a more useful structure... – Jon Clements Aug 17 '17 at 09:10

Python CSV module, special split string

1 Answers1