Code now relies on `format()` for items and albums displaying/logging.
`ui.print_()` calls `unicode()` or `str()` on the strings so for most
usages calling `ui.print_(obj)` replaces `ui.print_(obj, lib, None)`.
Where there is a special format `ui.print_(format(obj, fmt))` is fine,
but when `fmt` can be None then one has to call
`ui.print_(ui.format_(obj, fmt))` -- which is what `ui.print_obj` now
does.
Include import of __future__ features division, absolute_imports and
print_function everywhere. Don't add unicode_literals yet for it is
harder to convert.
Goal is smoothing the transition to python 3.
This was well-intentioned but ended up being more confusing than it was worth.
It's always confused me when one digit gets un-highlighted in one of these
displays. The straw that broke the camel's back was when I got a "#1 -> #16"
change where the numeral "1" was un-highlighted. To fix this right would be
way more trouble than it's worth; I'm glad to be rid of this detail.
Conflicts:
docs/changelog.rst
This makes sure we store the updated file `mtime`. By providing the
same interface on `Album` and `item` we can also reduce some code
duplication in the `modify` command.
Items might have the `album` field set without belonging to an album in
the beets database. We only count the albums that are represented in
the database.
Forces a write of tags to file even if the file's tags match the database.
This is useful to force plugins that respond to write (e.g., Scrub and Zero) to run on those tags.
This may also make the TODO comment in zero.py less important since creates a way to manually run the zero plugin on a file imported as-is.
Added a one-line summary of each album (in lib, and import target) so
you can easily tell if (for example) you are about to overwrite your
FLAC copy with a low-bitrate mp3 from somewhere else.
* Control flow and implementation of help command is now
similar to the other commands.
* Simplifies and flattens some code and removes unused method.
* Makes SubcommandOptionParser agnostic of Subcommand.parser.
This groups together all of the optparse setup calls separately from the
functions. This comes at the expense of showing the command-line options above
the code that interprets them.
Making me wish for a more declarative CLI setup style...
Many commands and plugins use `item.write()` to update tags. Since the success
of the call is not critical to the functionality of most consumers we want to
catch any exceptions, log an error and continue with our task. The new method
encapsulates this logic.
This fixes#675.
This puts the OrderedEnum generic class next to where it is actually used. It
also refers to the recipe it is taken from on docs.python.org. I also took the
opportunity to give this a capitalized name (since it's a proper type).
Following the convention of the other field sets and such. This helps avoid
any confusion with user-specified fields (although it's unlikely people will
want to name a flexible field "media_fields" :).
This makes the errors fully self-descriptive, which simplifies logging them as
errors (you just have to `log.error(exc)` to get a reasonable message). We
also now handle these at the top level in case someone forgets to add a
handler. But in this case, we also send the full traceback to the debug log.
This makes sure to print out the user's configuration path location even if no
user config file exists. (This makes "BEETSCONFIG=xxx beet config -p" behave
as expected, for instance.) It's a little hacky. Does the approach make sense
to you, @geigerzaehler?
We actually added more full-blown YAML dumping to the Confit library a while
back but it looks like it never made it into beets. It offers a few benefits
over the hand-rolled flattening that the `config` command was previously
using, including printing ordered dicts in the right order. But it also
appears to have broken logic when attempting to hide defaults. I'll fix this
right quick.
This new alternative to _showdiff takes care of formatting and is better at
highlighting differences for non-string fields. This takes care of the issue
where "True -> False" would have everything but the "e" highlighted.
The command prints a shell script that provides completion for the `beet`
command. To test it run `eval "$(beet completion)"` in your shell.
I also included some crude testing for this. The `test/test_completion.sh`
script runs tests in a shell and exit with a non-zero status code if the tests
fail. It assumes that the completion script is already loaded in the executing
shell.
As of now the completion only works for bash 4.1 and newer.
Perhaps we should use this for other commands also. One outstanding issue is
strange highlighting (e.g., "True -> False" helpfully shows that the "e" did
not change).
If a directory contains multiple albums we can select the ALBUMS action to group
the tracks by album artist and album name and import those seperately.
This way, _showdiff returns the information that _different used to provide
(since _showdiff needs to calculate it anyway). Using a `changed` set also
makes it easier to avoid unnecessary work.
- don't sanitize paths (this is already done separately)
- album.path (or album['path']) is now an alias for album.item_dir(), which
restores the formatting of $path in templates
A second base class, LibModel, maintains a reference to the Library and should
take care of database-related tasks like load and store. This is the beginning
of the end of the terrible incongruity between Item and Album objects (only
the latter had a library reference). More refactoring to come.
One large side effect: Album objects no longer automatically store
modifications. You have to call album.store(). Several places in the code
assume otherwise; they need cleaning up.
ResultIterator is now polymorphic (it takes a type parameter, which must be a
subclass of LibModel).
Namespaces were a worthy idea, but they added a lot of complexity to both the
library code itself and every client of the flexattrs interfaces. Getting rid
of them, and having one flat namespace of both traditional fields and
flexattrs, has one huge benefit: we can "promote" flexattrs to real attributes
(and vice versa) without code changes in every client.
This frees us to have a somewhat less efficient implementation of flexattrs
because we have a smooth upgrade path for making attributes more efficient via
promotion.
This is an effort to make the distance object feel slightly more dict-like.
The name changed and order of tuples is reversed: we now yield (key, value)
instead of (value, key), which I think is a little more intuitive.
Saves paranoid and interested users from having to either force all max
recommendations to none or constantly go back to candidate selection
from a recommendation to see if there is another slightly less similar
but more preferred (by the user) candidate.
