dbfread 1.0.0

read data from DBF files

Requires Python 3.2 or 2.7.

License: MIT

Latest version of the source code: http://github.com/olemb/dbfread/

Example

>>> import dbfread
>>> for record in dbfread.open('people.dbf'):
...     print(record)
...
{'NAME': 'Alice', 'BIRTHDATE': datetime.date(1987, 3, 1)}
{'NAME': 'Bob', 'BIRTHDATE': datetime.date(1980, 11, 12)}

If you have enough memory you can load the whole table into a list:

>>> import dbfread
>>> people = dbfread.read('people.dbf')
>>> people[1]
{'NAME': 'Bob', 'BIRTHDATE': datetime.date(1980, 11, 12)}

Both functions return a Table object. If records are loaded it also behaves like a list of records. (It’s a subclass of list.) See below for attributes and methods.

Using the dataset package it’s easy to move your data into a more modern database:

import dataset
import dbfread

db = dataset.connect('sqlite:///:memory:')

table = db['people']
for record in dbfread.open('people.dbf', lowernames=True):
    table.insert(record)

print(table.find_one(name='Alice'))

(Pass recfactory=collections.OrderedDict to open() if you want to preserve field order.)

Installing

pip install dbfread

Status

The library has been used to read FoxPro files with a wide range of data types, but is not widely tested with other DBF formats. It should still work for most files.

Foxpro memo files (.FPT) are fully supported. The alternative .DBT memo files will be supported if I find any examples to test with.

I intend for dbfread to be able to read any DBF file. If you have a file it can’t read, or you find a bug, I’d love to hear from you.

Supported Field Types

:	Field type	Converted to
0	flags	int
C	text	unicode string
D	date	datetime.date or None
F	float	float or None
I	integer	int or None
L	logical	True, False or None
M	memo	unicode string (memo) or byte string (picture or object)
N	numeric	int, float or None
T	time	datetime.datetime

Options for open() and read()

load=True

Load all records into memory. The Table object will behave as a list of records, and the deleted attribute will be a list of deleted records. This defaults to False for open() and True for read().

encoding=’latin1’

By default, dbfread will try to guess the character encoding from the language_driver byte. If this fails it uses “latin1”. You can override this with the encoding argument.

lowernames=True

Field names in DBF files are usually in uppercase. This converts them to lowercase.

recfactory=OrderedDict

Takes any function that will be used to produce new records. The function should take a list of (name, value) tuples.

ignorecase=False

The default is to ignore case in filenames.

parserclass=MyFieldParser

You can add new field types by subclassing the included field parser:

import dbfread.field_parser

class MyFieldParser(dbfread.field_parser.FieldParser):
    """Parse X field (complex number)"""
    def parseX(self, field, data):
        real, imag = data.strip().split(',', 1)
        return complex(float(real), float(imag))

for record in open('math.dbf', parserclass=MyFieldParser):
    ...

(Please let me know if you have new field types that should be supported out of the box.)

write a subclass of dbfread.dbf.field_parser

This allows you to add new field types.

raw=True

Returns all data values as bytestrings. This can be used for debugging or for doing your own decoding.

All list methods are also available when records are loaded.

Table attributes

deleted

Deleted records. If records are in memory this is a list of records, if not it is a RecordIterator object. In any case you can iterate over it and call len() on it.

loaded

True if records are loaded into memory.

name

Name of the table. This is the lowercased stem of the filename, for example the file /home/me/SHOES.dbf will have the name shoes.

date

Date when the file was last written to (as datetime.datetime).

field_names

A list of field names in the order they appear in the file. This can for example be used to produce the header line in a CSV file.

encoding

Character encoding used in the file. This is determined by the language_driver byte in the header, and can be overriden with the encoding keyword argument.

ignorecase, lowernames, parserclass, recfactory, raw

These correspond to the keyword arguments below.

header

The file header. Example:

DBFHeader(dbversion=48, year=12, month=7, day=11, numrecords=555,
headerlen=2408, recordlen=632, reserved1=0, incomplete_transaction=0,
encryption_flag=0, free_record_thread=0, reserved2=0, reserved3=0,
mdx_flag=3, language_driver=3, reserved4=0)

fields

A list of field headers from the file. Example:

[DBFField(name=u'NAME', type=u'C', address=1, length=25, decimal_count=0,
reserved1=0, workarea_id=0, reserved2=0, reserved3=0, set_fields_flag=0,
reserved4='\x00\x00\x00\x00\x00\x00\x00', index_field_flag=0),
... etc. ...]

Methods

load()

Load records into memory.

unload()

Unload records from memory.

__len__()

Return number of records in the file. If records are not loaded this will scan the file to count records.

__iter__()

Iterate through records.

dbf2sqlite

(This does not require the dataset package.)

A tool is included in the examples directory to convert DBF into sqlite, for example:

dbf2sqlite -o example.sqlite table1.dbf table2.dbf

This will create one table for each DBF file. You can also omit the -o example.sqlite option to have the SQL printed directly to stdout.

If you get character encoding errors you can pass --encoding to override the encoding, for example:

dbf2sqlite --encoding=latin1 ...

Contact

Ole Martin Bjorndalen - [e-mail]