A framework to read and write OFX files Version 1.
EDOFX is a python framework / DSL to manipulate OFX Version 1 files.
EDOFX has no dependencies other than Python standard library.
EDOFX has been tested with python 2.5 to 2.7.
Python 3 is not supported yet.
Content of this file
- Parse an OFX file or string
- Inspect and manipulate OFX file
- Navigate OFX tree
- Delete Nodes
- Modify Nodes
- OFX tree display and export
- Package Description
- test directory description
EDOFX is licensed under MIT License. See LICENCE.txt
Use pip or easy_install:
$ pip install inouk.edofx or $ easy_install inouk.edofx
Git clone from the official repository then install with:
python setup.py install
Parse an OFX file or string
To experiment, create a virtualenv, install inouk.edofx then launch python and use:
>>> from inouk import edofx >>> f = open('test/fixtures/multi_account_file.ofx', 'U') # Files must be opened using 'U' mode as the parser do not manage \r\n EOL >>> OFX = edofx.OFXParser(f.read()).parse() >>> OFX <OFX>...</OFX>
OFX now contains an OFXNode’s tree representation of test/fixtures/multi_account_file.ofx.
Inspect and manipulate the OFX tree
To navigate OFX tree, use :
>>> OFX.children [<SIGNONMSGSRSV1>...</SIGNONMSGSRSV1>, <BANKMSGSRSV1>...</BANKMSGSRSV1>, <CREDITCARDMSGSRSV1>...</CREDITCARDMSGSRSV1>] >>> OFX.SIGNONMSGSRSV1.children [<SONRS>...</SONRS>] >>> OFX.SIGNONMSGSRSV1.SONRS.children [<STATUS>...</STATUS>, <DTSERVER>20100305094649, <LANGUAGE>FRA] >>> OFX.SIGNONMSGSRSV1.SONRS.DTSERVER <DTSERVER>20100305094649 >>> OFX.SIGNONMSGSRSV1.SONRS.DTSERVER.value '20100305094649' # This is the raw value as a string >>> OFX.SIGNONMSGSRSV1.SONRS.DTSERVER.val # returns date or float when appropriate datetime.date(2010, 3, 5)
Iterators are ok:
>>> # To list account numbers present in file >>> for account in OFX.BANKMSGSRSV1.STMTTRNRS: ... print 'account #', account.TRNUID.val ... account # 34995678052 account # 85576218690 account # 44498008728 account # 42025380697 account # 62631064788
Indexers are ok:
>>> OFX.BANKMSGSRSV1.STMTTRNRS.TRNUID.val '44498008728' >>> len(OFX.BANKMSGSRSV1.STMTTRNRS) 5
To delete nodes:
>>> OFX.children [<SIGNONMSGSRSV1>...</SIGNONMSGSRSV1>, <BANKMSGSRSV1>...</BANKMSGSRSV1>, <CREDITCARDMSGSRSV1>...</CREDITCARDMSGSRSV1>] >>> del OFX.CREDITCARDMSGSRSV1 # Warning does not work (yet) on indexed items >>> OFX.children [<SIGNONMSGSRSV1>...</SIGNONMSGSRSV1>, <BANKMSGSRSV1>...</BANKMSGSRSV1>]
To modify nodes:
For that purpose, you must let ‘.value’ with a string. Note that ‘.val’ attribute is not ( yet ) writable.
>>> OFX.SIGNONMSGSRSV1.SONRS.LANGUAGE <LANGUAGE>FRA >>> OFX.SIGNONMSGSRSV1.SONRS.LANGUAGE.value = 'Italian' >>> OFX.SIGNONMSGSRSV1.SONRS.LANGUAGE <LANGUAGE>Italian >>>
To alter Tree structure:
>>> OFX.SIGNONMSGSRSV1.SONRS.children.append(edofx.OFXNode(name='COLOR', value='blue')) # insert is ok too >>> OFX.SIGNONMSGSRSV1.SONRS.children [<STATUS>...</STATUS>, <DTSERVER>20100305094649, <LANGUAGE>Italian, <COLOR>blue]
OFX tree display and export
OFXNode supports 3 output / dump formats :
- OFX ; to re-export as OFX after tree/nodes modifications
- XML ; easier to read
- Obfuscated ; re-export as OFX with all sensitive information jammed.
Important information: Nodes ‘ACCTTYPE’, ‘CODE’, ‘STATUS’, ‘SEVERITY’, ‘LANGUAGE’, ‘CURDEF’, ‘TRNTYPE’ are not obfuscated.
>>> print OFX.ofx_repr() <OFX> <SIGNONMSGSRSV1> <SONRS> <STATUS> <CODE>0 <SEVERITY>INFO </STATUS> <DTSERVER>20100305094649 <LANGUAGE>Italian <COLOR>blue </SONRS> ...
>>> print OFX.obfuscated_ofx_repr() <OFX> <SIGNONMSGSRSV1> <SONRS> <STATUS> <CODE>0 <SEVERITY>INFO </STATUS> <DTSERVER>20100305094649 <LANGUAGE>Italian <COLOR>BHVB </SONRS>
>>> print OFX.xml_repr() <OFX> <SIGNONMSGSRSV1> <SONRS> <STATUS> <CODE>0</CODE> <SEVERITY>INFO</SEVERITY> </STATUS> <DTSERVER>20100305094649</DTSERVER> <SIZE>XXL</SIZE> <LANGUAGE>Italian</LANGUAGE> <COLOR>blue</COLOR> </SONRS>
OFX Headers are available as a dict.
Headers can be parsed in two ways:
- if the parse() has been called, they can be retrieved using attribute OFX_headers of the parser object.
- OFXParser can de asked to parse only the headers (not the content) with the parse_headers() method (Cf. tests 7,8 and 9 of test_01_parser.py for an example).
This package contains :
- edofx/__init__.py ; the framework
- tests ; contains Python unit-test classes
- edofx_integration.py ; an example demonstrating how to parse an OFX source into specific classes.
The framework by itself is structured in 3 classes:
Use python help for method level description. (Even if I’m slightly optimistic about the python help content quality…)
OFXNode is used to store parsed OFX sources.
OFXParser builds OFXNode tree from a string containing some OFX content.
OFXObfuscator is a naive obfuscator based on lexical analysis. OFXObfuscator is able to obfuscate OFX sources OFXParser has been unable to parse. Use this if you want to sent me some OFX files that OFXParser fails to process.
There you will find 3 examples and some unittests:
- edofx_integration.py (see below)
- edofx2csv.py ; shows how to export andofx file to csv in
- ofxplode.py ; shows how to generate a single account OFX file from a multi account one
- unit tests
This file shows how to parse a multi account type / multi account OFX file into Statement and StatementTransaction classes and export everything into a set of per account csv files.
Most interesting part is the render_as_DOT() function.
Following render_as_DOT() code snippet shows how OFXNode makes it easy to load OFX content into arbitrary python objects:
if OFX.BANKMSGSRSV1 : # For each account statement... for aSTMTTRNRS in OFX.BANKMSGSRSV1.STMTTRNRS: stmt = Statement('CHECKING') stmt.currency = aSTMTTRNRS.STMTRS.CURDEF.val stmt.bank_id = aSTMTTRNRS.STMTRS.BANKACCTFROM.BANKID.val stmt.branch_id = aSTMTTRNRS.STMTRS.BANKACCTFROM.BRANCHID.val stmt.account_id = aSTMTTRNRS.STMTRS.BANKACCTFROM.ACCTID.val stmt.start_date = aSTMTTRNRS.STMTRS.BANKTRANLIST.DTSTART.val # returned as date stmt.end_date = aSTMTTRNRS.STMTRS.BANKTRANLIST.DTEND.val # returned as date # for each transaction in statement for s in aSTMTTRNRS.STMTRS.BANKTRANLIST.STMTTRN: st = StatementTransaction() st.fitid = s.FITID.val st.type = s.TRNTYPE.val st.date = s.DTPOSTED.val # returned as date st.amount = s.TRNAMT.val # returned as float st.name = s.NAME.val st.memo = s.MEMO.val stmt.transaction_list.append(st) stmt.balance = aSTMTTRNRS.STMTRS.LEDGERBAL.BALAMT.val # returned as float stmt.balance_date = aSTMTTRNRS.STMTRS.LEDGERBAL.DTASOF.val # returned as date statement_list.append(stmt)
Shows how to export an edofx file to csv format.
Takes a multi account OFX file and generates:
- a single account OFX file
- an XML representation of the single OFX file
- Parser tests (test_01_parser.py)
- OFXTree export tests (test_02_OFXNode_export.py)
- Obfuscation tests (test_03_obfuscation.py)
Directory tests/fixtures contains a set of OFX file.real_file.ofx and multi_account_file.ofx have been built from real files using OFXNode.obfuscated_ofx_repr()
Disclaimer: EDOFX was my very first python code !! So many things I would have not written like this today.
- Cyril MORISSE
0.3.8 -> 0.3.9
- Fix import of formatted OFX files (ie. with indented tags)
0.3.6 -> 0.3.8
- Improve PEP8 Compliance
- Manage file encoding based on OFX Header content
- Fix problem with file names in uppercase (eg. README.rst)
- Port from mercurial to git.
- Move from bitbucket to github.
- Publish on Pypi.
- Rework some code (4 years after)
- Pypi and classifiers setup