Ver 0.9.1 uploaded

Skip to first unread message

Tim Peters

Mar 9, 1999, 3:00:00 AM3/9/99
I uploaded version 0.9.1 of

to's FTP incoming directory. If ftpmaster is awake <wink>, I
expect they'll show up within a few days, under

In the meantime, I'll attach the README and the change log.

really-glad-it-serves-as-its-own-test<wink>-ly y'rs - tim


# Module doctest version 0.9.1
# Released to the public domain 08-Mar-1999,
# by Tim Peters ( provides a flexible framework for automatically testing
examples in docstrings. It's very easy to use (well, at least to
use as intended ...):

+ Run some illuminating examples in an interactive session.

+ Paste a copy of that session into a docstring, interspered freely
among your docstring prose.

+ Add a few lines of boilerplate code to the end of the module, to
invoke the doctest magic when your module is run directly as a
script: the docstring examples will be found and interpreted,
verifying they still work exactly as advertised.

Extensive docs are in's docstrings, and some even serve
as non-trivial examples of doctest's intended use. Warning: thinking
too hard about how doctest manages to run docstring examples of its
own use in its own source code can make your head explode <kaboom>.

Change log:

# 0,0,1 06-Mar-1999
# initial version posted
# 0,0,2 06-Mar-1999
# loosened parsing:
# cater to stinkin' tabs
# don't insist on a blank after PS2 prefix
# so trailing "... " line from a compound stmt no longer
# breaks if the file gets whitespace-trimmed
# better error msgs for inconsistent leading whitespace
# 0,9,1 08-Mar-1999
# exposed the Tester class and added client methods
# plus docstring examples of their use (eww - head-twisting!)
# fixed logic error in reporting total # of tests & failures
# added __test__ support to testmod (a pale reflection of Christian
# Tismer's vision ...)
# removed the "deep" argument; fiddle __test__ instead
# simplified endcase logic for extracting tests, and running them
# before, if no output was expected but some was produced
# anyway via an eval'ed result, the discrepancy wasn't caught
# made TestClass private and used __test__ to get at it
# many doc updates
# speed _SpoofOut for long expected outputs

Reply all
Reply to author
0 new messages