merge in archlinux pkgbuild

Arch Linux package for 2.1

a bit of updates to docs

better link to homepage

style improvement again

better docs style

docs: fix spelling

Added tag 2.1 for changeset a083e23ed554

2.1

docstrings

fix traceback depth determining when command doesn't get expected arguments

link to inner help doc

fix changelog to render to PyPI

Added tag 2.0 for changeset a8a8ae1a3fbe

version 2.0

forgotten tests file

every function now has attribute '.help()', which will print help on call

fix help display when there is no doc

update copyright year

dictionary type handler

make doctests work previously they would fail because sys.stdout object was captured by function write() much earlier than doctest had any possibility to monkey patch it

update pkgbuild

Added tag 1.2 for changeset b8f101500f1d

fix subcommands list docstring display

fix md5 in pkgbuild

links to files in tests

incorporate tests in docs

updated archlinux pkgbuild to 1.1

Added tag 1.1 for changeset cfcc54fe6d49

fix _completion handling when global options are defined

archlinux pkgbuild update

Added tag 1.0 for changeset 6dc40423b257

that's 1.0, why not after all? :)

Merged upstream

custom css to limit width

a bit of docs config cleanup

build_sphinx options

update archlinux pkgbuild

Added tag 0.9.13 for changeset e21e182229c1

0.9.13 ready

fix sysname to not display full path to command

fix exception handling

small docs update

README update

Arch Linux package for 0.9.12

Merged upstream

Added tag 0.9.12 for changeset 8d3e644647f8

0.9.12 release

better run function in tests

use utf-8 instead of ascii as default output encoding

merge arch linux pkgbuild from Andrey Vlasovskikh

Explicit use of Python 2 for Arch Linux

Corrected the license field in the Arch Linux package

Added a package file for Arch Linux

locale troubles fixed, test suite moved to cram

Merged upstream

Added tag 0.9.11 for changeset bf6908d12aae

0.9.11

Fixed catching unnececcary exceptions Exception Abort can't be thrown, so it is useless. Catching SystemExit and KeyboardInterrupt makes little sense as the thread can be terminated by KeyboardInterrupt before or after entering the wrapped command processing function. So one have to handle these exceptions in his main function anyway. Catching any exception in order to print a message and re-raise it changes the standard behaviour and might seem confusing.

fix some pep8 warnings

Merged upstream

merge completer improvements

Added tag 0.9.10 for changeset 95653a562a05

version 0.9.10

some cleanup for completion

handle floats along with ints and strings in default arguments; thx Jungle

revert changes from 0178ac: don't require argv=sys.argv[1:]

hide completion from help

fix completion

Autocompletion of options works for prefixes of full command names

Fixed a bug in help_options

Augmented Bash completion script to exclude : from word breaks

Added the ability to provide a completer for arguments to options

Added FloatType as a possible option type

_completion skips middleware

Initial attempt at autocompletion (borrowed from PIP)

Fix help for options with functions as their default arguments

index.rst: fix link to overview

fix help output

update tests

various py3k incompatibilities removed

automate test system

switch to setuptools (to distribute actually)

backward incompatible change: require argv=sys.argv[1:] for @wrapped command to parse command line

always call a function if it's default argument for an option

multiple arguments example

test_cmd.py update to reflect updates in opster

Added tag 0.9.9 for changeset 472a33d66076

release 0.9.9

support calling commands from python when options are defined in list

when calling command from python set not supplied options to proper defaults Earlier you would get full option spec (short, default, help) for every option, not supplied as argument (or keyword argument) in function call.

Fix bug introduced in previous changeset

changelog update

respect empty strings as usage

replace _ with - in command names, same as in options names

fix example ui middleware

globaloptions were simply dropped after parsing, fold them in regular options

Added tag 0.9.8 for changeset cc24c392fd45

release 0.9.8

fix bug with option names clashing with call_cmd argument names

Added tag 0.9.7 for changeset d4bdbbf7a500

release 0.9.7

rename from finaloption to opster

Added tag 0.9.6 for changeset d283be877fc5

0.9.6 changelog and version bump

check that option has a name

Check if short name is of length == 1

ability to decorate all subcommands and example of such decorator

Use actual options names when guessing usage

readme fix

not everybody knows Die Krupps, add link

Added tag 0.9.5 for changeset e9678adf8bb1

prepare to release 0.9.5

allow commands with arguments

Added tag 0.9.4 for changeset 8fe50bf4458b

bumping version to 0.9.4

prepend subcommands usage with program name

ability to hide commands from listing

basic usage guess

better help for subcommands list

remove unused exception

Added tag 0.9.3 for changeset e8866635cf2f

setup.py fix

Added tag 0.9.2 for changeset 1d0173e1b412

version bump and minor documentation fixes

changelog

fix 0.9 tag pointer

ability to use commands as usual functions

small documentation update

ignore dist/ dir

Added tag 0.9.1 for changeset 188325c309d8

fix behavior when adding --help option

more tests

manifest doesn't need to be included in repo

test_opts: fix usage string

setup.py: better description

no epigraph in readme

Added tag 0.9 for changeset 7dd1ddfe518c

pypi wants me to supply email

hint about adding option lists

simple example of usage for subcommands

setup.py

link to song text, just for fun

readme: link to documentation

documentation

fix usage generation

rename to finaloption, docstring for @command, few small change to behaviour

refactoring of api and internals to be more consistent

decorator to parse options for single command

better name for ui status method

examples in help

78 chars is good enough to wrap help

callable() is better way to determine callable than hasattr

better naming

slightly better api for fancyopts

initial version of readme

fancyopts merged in fancycmd to become single python module

fix wording in comments

better tests

support SystemExit in dispatcher

fix handling of 0 and 1 as int parameters They were converted to True/False in case of 'value in tuple' comparison, so switch to 'value is value' is necessary.

tests for UI and better handling of quiet/verbose

handle global --help option

help improvements, tests added

some upgrade of command handling

intermediate commit for subcommand dispatcher

initial hgignore

Introduce intermediate map to handle default values. This helps to resolve issues with lists and functions: latest supplied option should win, which was impossible earlier because function were replaced with it's return value

Better help

Better naming for variables in parse

slightly better usage description

Some documentation for actual option parsing.

usage for command should be separate string

Remove usage of single-character variables

make equations more clear

more clear comment about getopt parameter generation

fancyopts: testing facilities

fancyopts option parser, first iteration