Help needed finding entry point to debug custom CSVImporter for amazon transactions

[Moritz]

Hi,

I've exported my Amazon transactions into a `CSV` file using the `Amazon Order History Reporter` extension. The file looks like this:

```
order id,order date,quantity,description,price
300-1000000-1000000,2022-10-26,1,"Foo","19.99"
300-1000000-2000000,2022-11-18,1,"bar","39.95"
```

I came up with this config file:

```

from beancount.ingest.importers.csv import Importer as CSVImporter, Col

my_amazon_csv_importer = CSVImporter({
                        Col.DATE: 'order date',
                        Col.NARRATION1: 'description',
                        Col.AMOUNT: 'price'
                        },
                       'Expenses:Amazon', # account
                       'EUR', # currency
                        # regexps used by ImporterProtocol.identify() to identify the correct file
                       #('order id,order date,quantity,description,price'                      
                       #),
                       skip_lines=0
                       )


CONFIG = [my_amazon_csv_importer]
```
to extract the transactions with:

```
bean-extract config_amazon_csv.py amazon_order_history_2022.csv
```

But this fails with:
```
ERROR:root:Importer beancount.ingest.importers.csv.Importer: "Expenses:Amazon".extract() raised an unexpected error: Unknown string format: order date
Traceback (most recent call last):
  File "/home/user/beancount/.venv/lib/python3.10/site-packages/beancount/ingest/extract.py", line 182, in extract
    new_entries = extract_from_file(
  File "/home/user/beancount/.venv/lib/python3.10/site-packages/beancount/ingest/extract.py", line 67, in extract_from_file
    new_entries = importer.extract(file, **kwargs)
  File "/home/user/beancount/.venv/lib/python3.10/site-packages/beancount/ingest/importers/csv.py", line 294, in extract
    date = parse_date_liberally(date, self.dateutil_kwds)
  File "/home/user/beancount/.venv/lib/python3.10/site-packages/beancount/utils/date_utils.py", line 45, in parse_date_liberally
    return dateutil.parser.parse(string, **parse_kwargs_dict).date()
  File "/home/user/beancount/.venv/lib/python3.10/site-packages/dateutil/parser/_parser.py", line 1368, in parse
    return DEFAULTPARSER.parse(timestr, **kwargs)
  File "/home/user/beancount/.venv/lib/python3.10/site-packages/dateutil/parser/_parser.py", line 643, in parse
    raise ParserError("Unknown string format: %s", timestr)
dateutil.parser._parser.ParserError: Unknown string format: order date
;; -*- mode: beancount -*-
```

What annoys me the most here is, that I cannot find a way to just call the `bean-extract config_amazon_csv.py amazon_order_history_2022.csv` part just from the `config_amazon_csv.py` file or another python script.
That would make it easy to set a breakpoint in my editor to check what is happening exactly.
As I'm quite new to python: what is the proper way of calling bean-extract for debugging purposes?

So any pointers highly appreciated to get my amazon transactions to beancount format.

Moritz

[Moritz]

Orrr. Figured it out and made the importer working:
Turned out, the last line of the CSV contained the headers again. Deleting this line fixed it.

The debug question remains open ;)

[David Avraamides]

bean-extract is just a Python script so you can debug it like any other script:

    python -m pdb `which bean-extract` my_importer.py test.csv

or if you want to stop at an exception:

    python -m pdb -c continue `which bean-extract` my_importer.py test.csv