We can now rip CDs without gstreamer.
This is not the most clean attempt, but I have tried to remove most of the
code that depends on gstreamer. I hope there is not a lot of code left that
depends on code that I have removed - I can at least rip a CD fully.
* Add encoding using Xiph.org 'flac' program.
This adds a FlacEncodeTask that encodes wave files to flac files.
This commit also replaces morituri's EncodeTask with FlacEncodeTask, however, in
morituri, EncodeTask also does the tagging.
FlacEncodeTask will not perform the tagging.
So we will need an extra task for the tagging - this will be added soon.
Meanwhile, do not merge this commit to master yet.
* Add tagging using mutagen.
Replace the gstreamer tagging code with mutagen tagging code.
getTagList is rewritten to return a dictionary of tags, which are then simply
passed to mutagen.
The way it is set up right now is not the best - I don't think it makes sense
for tagging to take place in program/cdparanoia.py ; but this is how the current
code did it.
I suggest that, when we rip out all the gstreamer code, we also move the tagging
to a more sensible place; and then also make the tagging not be an actual
'task.Task'.
* Add gstreamer-less CRC32 version
Only works on wave files at this point. Should not be a problem, I think.
* Use proper musicbrainz tags and ALBUM tag.
* Add mutagen to .travis.yml
* introduce logcommand.Lager, Whipper(); use argparse for whipper image commands, stub logging
* update Lager docstring to mention config.Config()
* make incorrect subcommand and --version work on toplevel command
* migrate accurip show, expand Lager, do not attempt to return from Lager.__init__.
* migrate offset find, add Lager.error
* correct offset find drive symlink handling
* migrate drive
* change Lager.__init__(prog) to arg from kwarg
* but actually
* remove Whipper.usage
* add and use Lager.device_option() context manager
* help I married an axe murderer
* use unified options namespace for entire command tree
* migrate whipper cd without comprehensive config loading
* switch to logging module
- use logging instead of flog for non-extern modules
- use WHIPPER_DEBUG and WHIPPER_LOGFILE env variables
* convert self.log calls to logger.debug
* convert self.error calls to logger.error
* remove log.Loggable, use logger not logging
* Logging conversion continues
- Convert log.* calls to logger.*
- Remove morituri.common.log imports
* remove morituri.common.log from tests
* remove extern/flog, bare minimum Debug conversion
* update README for logging changes
* update soxi to use logging
* refactor Lager for more declarative subcommands
* Refactor Lager.device_option:
- inline into __init__
- throw IOError instead of Exception for missing drives
- remove CommandError checking in rip/main
* rename rip to whipper in rip.main
* convert rip.debug commands
* Rename logcommand.Lager to command.BaseCommand
- remove command.CommandError occurrences
- remove python-command external module
* remove submodules from README, update rclog formatter
* update minor ambiguity in readme for command invocation
* update version number to match setup.py
* remove gitmodules
* update version number in tests as well (boo)
* convert logger.error to logger.critical
* Change morituri.rip to morituri.command
- mv common.command to command.basecommand
- move TEMPLATES used only by rip.cd out of rip.common
- update entry point for command to command.main
* update basecommand documentation
* go pyflaking: import fixing
* replace self.stdout with sys.stdout
* remove BaseCommand.config, alphabetise imports
* convert self.stdXXX leftovers
* convert last getRootCommand to config.Config
* convert last getExceptionMessage's to str
* change musicbrainz useragent to whipper
* update README to force python2
* add support for PRE_EMPHASIS detection; mark pre-emphasis as FLAGS PRE in .cue
* correct typo, s/ivar/type/ for pre_emphasis element in track table
* add SoxPeakTask, program/sox module, sox.peak_level()
* use sox peak detection in lieu of gstreamer
Strip bitrate tags too for comparison of dicts.
Add a way to show us the different keys between dicts.
* morituri/common/encode.py:
Debug different keys.
add task argument to TaskRunner.schedule so we can get exceptions.
Add .schedule method to Task to pass self.
* morituri/common/checksum.py:
* morituri/common/encode.py:
* morituri/common/gstreamer.py:
* morituri/program/cdparanoia.py:
* morituri/program/cdrdao.py:
Adapt.
Generate an actual file by spawning gst-launch; otherwise
with proper error handling we get an error from wavparse that
there is not enough data to typefind.
* morituri/common/gstreamer.py:
Move the gst import to start() and set it as a class attribute.
Document methods.
* morituri/common/encode.py:
Convert EncodeTask to a GstPipelineTask.
wavenc does not have merge_tags, it seems.
So don't call an element a tagger, don't merge tags if there is
no tagger, and complain if there is no merge_tags when we think
there should be.
Add some debug.
* morituri/common/encode.py:
Add more debug.
Handle the case where peak is full scale, and peakdB thus 0,
which triggered not setting self.peak.
