opentaxforms 0.4.17

OpenTaxForms opens and automates US tax forms–it reads PDF tax forms (currently from IRS.gov only, not state forms), converts them to more feature-full HTML5, and offers a database and API for developers to create their own tax applications. The converted forms will be available to test (and ultimately use) at OpenTaxForms.org.

• License

  GNU AGPLv3

• Install

  pip install opentaxforms

• Github

  • code

  • [issue tracker link forthcoming]

  • [milestones link forthcoming]

• Build status

  

• Form status

  The script reports a status for each form. Current status categories are:

  • layout means textboxes and checkboxes–they should not overlap.

  • refs are references to other forms–they should all be recognized (ie, in the list of all forms).

  • math is the computed fields and their dependencies–each computed field should have at least one dependency, or else what is it computed from?

  Each status error has a corresponding warning in the log file, so they’re easy to find. Each bugfix will likely reduce errors across many forms.

  [form status listing forthcoming]

• API

  The ReSTful API is read-only and provides a complete accounting of form fields: data type, size and position on page, and role in field groupings like dollars-and-cents fields, fields on the same line, fields in the same table, fields on the same page, and fields involved in the same formula. The API will also provide status information and tester feedback for each form.

  [API docs forthcoming, for now see examples in test/run_apiclient.sh]

• How it works

  Most of the IRS tax forms embed all the fillable field information in the XML Forms Architecture (XFA) format. The OpenTaxForms python script extracts the XFA from each PDF form, and parses out:

  • relationships among fields (such as dollar and cent fields; fields on the same line; columns and rows of a table).

  • math formulas, including which fields are computed vs user-entered (such as “Subtract line 37 from line 35. If line 37 is greaterthan line 35, enter -0-“).

  • references to other forms

  All this information is stored in a database (optionally PostgreSQL or the default (sqlite)[https://sqlite.org/]) and served according to a ReSTful API. For each tax form page, an html form (with javascript to express the formulas) is generated and overlaid on an svg rendering of the original PDF. The javascript saves all user inputs to local/web storage in the browser via basil.js. When the page is loaded, those values are retrieved. Values are keyed by tax year, form number (eg 1040), and XFA field id (and soon taxpayer name now that I do my kids’ taxes too). Testers will annotate the page image with boxes and comments via annotorious.js. A few of the 900+ IRS forms don’t have embedded XFA (such as 2016 Form 1040 Schedule A). Eventually those forms may be updated to contain XFA, but until then, the best automated approach is probably OCR (optical character recognition). OCR maybe a less fool-proof approach in general, especially for state (NJ, NY, etc) forms, which generally are not XFA-based.

• To do

  • Replace ToDo with a link to github/issues.

  • Refactor toward a less script-ish architecture that will scale to more developers. [architecturePlease]

  • Switch to a pdf-to-svg converter that preserves text (rather than converting text to paths), perhaps pdfjs, so that testers can easily copy and paste text from forms. [copyableText]

  • Should extractFillableFields.py be a separate project called xfadump? [xfadump] This might provide a cleaner target output interface for an OCR effort.

  • Replace allpdfnames.txt with a more detailed form dictionary via a preprocess step. [formDictionary]

  • Offer entire-form html interface (currently presenting each page separately). [formAsSingleHtmlPage]

  • Incorporate instructions and publications, especially extracting the worksheets from instructions. [worksheets]

  • Add the ability to process US state forms. [stateForms]

  • Fix countless bugs, especially in forms that contain tables (see [issues])

  • Don’t seek in a separate file a schedule that occurs within a form. [refsToEmbeddedSchedules]

  • Take any needed action regarding this from tox.readthedocs.io: Warning Integrating tox with setup.py test is as of October 2016 discouraged as it breaks packaging/testing approaches as used by downstream distributions which expect setup.py test to run tests with the invocation interpreter rather than setting up many virtualenvs and installing packages. [toxSansSetup]

  • separate dirName into pdfInputDir,htmlOutputDir [splitIoDirs]

0.4.16 (2017-01-13)

• linting via landscape.io, including shortening long lines, removing unused variables, and refactoring some long functions.

• fixed all broken doctests.

• both cross-module globals, log and cfg, are now importable.

0.4.15 (2017-01-09)

• bugfix: the ignoreCaches option now works.

• bugfix: skip forms setup if dirName is None.

• bugfix: ut.exists now insensitive to trailing slash.

• test: moved from relative to absolute paths so eg py.test works from outside of test/ dir.

• test: reorganize test/ dir: test_* subdirs are disposable, test-* arent.

• test/test_opentaxforms.py: command line options changed and expanded.

• test/test_opentaxforms.py: refactor TestOtfSteps, add test_run_1040_xfa, and finally add a script test to the [pytest] automated tests.

• test/forms/f1040.js,test/run_html5.sh: add casperjs script [to test f1040 html files] and shell script [for context].

• add test/README.md as guide to the test/ directory.

• probe access to phantomjs,casperjs in .travis.yml, hopefully for future integration of html5 tests.

• opentaxforms.py: in main, move form-summarizing code to try-else block.

• cache results of Form.pdfInfo for faster test run.

• add internal option to computeOverlap [for layout boxes; currently done inefficiently, thus costly].

• update 1040 target output file to fix broken test.

• cleanup opentaxforms.py/main.

0.4.14 (2016-12-31)

• mv math.py cmds.py: workaround for tox build failure [tox.Random imports py.math and gets our math.py instead].

• oops apparently gotta incl README for travis build to succeed.

• setup.cfg: git flow release already tags the release, so we dont want bumpversion to do so, otherwise git flow refuses to release.

0.4.13 (2016-12-30)

• gitignore release.sh temp files so they dont affect git status.

• move most of 1100-line main into four new files.

• add Form class.

• domain.py->irs.py.

• extractFillableFields.El derives from dict.

• combined rst format files into a single README file.

• noticed markdown readme is not rendered on pypi, fixing [part of fix is in release.sh].

0.4.12 (2016-12-23) - bind arrow keys to next/prev page links [for demo video]. - [release.sh remains untracked while it is being tested.]

0.4.11 (2016-12-21) - allow multiple ‘rootForms’ via call or commandline. - output form statuses for external processing. - clean up “import *”. - add cleanup script. - merge the missing and spurious categories into the form status message. - use bumpversion as cookiecutter does.