CHANGES¶
v0.7.0 (2017/04/18)¶
Changes (from v0.6.4)
Seqparse.scan_path()
can now accept list-like collections of input search paths and deals with them recursively.- Moved all stat-related methods and properties from the
FrameSequence
to theFileSequence
class. FileExtension
now holdFileSequence
instances (which makes more sense if you think about it).FileExtension.output()
now correctly handles pad consolidation.FileSequence
instances can now clone otherFileSequence
instances.FileSequence
instances now have two new properties:frames
(which iterates over the contained padded frames) andpretty_frames
(which returns the string representation of the contained frame sequence).FileSequence.update()
can now accept FileSequence instances as input.- Coverage tests are pretty much at 100%.
v0.6.4 (2017/04/08)¶
Changes (from v0.6.3)
- Rebuilt setup files – (hopefully) ready for initial release on pypi!
v0.6.1 (2017/04/06)¶
Changes (from v0.6.0)
- Added support for python 3.4+.
- Packaging changes in preparation for pushing module to PyPi.
v0.6.0 (2017/04/05)¶
Changes (from v0.5.0)
- DOCUMENTED ALL THE THINGS!
- Some slight parameter refactoring (ie, iterable -> frame/frames where necessary), but no changes in functionality.
v0.5.0 (2017/04/01)¶
Changes (from v0.4.0)
- Added ability to query/report disk footprint for all scanned files.
seqls
: Implemented new –mindepth and –maxdepth to help limit the depth you’ll scan while searching for file sequences. They should have exactly the same functionality as find’s similarly named flags.seqls
: Added ``-a/--all
and-l/--long-format
options to maximize your file sequence scanning pleasure. The options should mimic the behaviour of the similarly named flags for the ls command. Also, paring up the--long-format
option with the new-H/--human-readable
option will causeseqls
to report file sizes in a much more legible (gnu-style) fashion.Seqparse
: Now returns all singleton files asFile
instances. As withFileSequence
instances, casting aFile
instance to a string will result in the name of the file.FileSequence
,File
: Addedsize
andmtime
properties that report disk footprint and latest file modification time, respectively, for the referenced files.- Renamed all flags with underscores to use dashes instead (hitting the shift key can be a drag).
v0.4.0 (2017/03/22)¶
Changes (from v0.3.0)
- Added a couple new options to the
seqls
command line tool: -S/--seqsonly
: Allows the user to filter output to only contain discovered file sequences.-m/--missing
: Allows the user to display the missing files in a sequence in the typical file sequence-style output.
- Added a couple new options to the
- Made output types consistent for all classes – every
output()
method returns class instances, all of which can be safely turned into a displayable string via thestr()
method. - Added
FrameChunk
,FrameSequence
, andFileSequence
equality/inequality methods. - Consolidated a lot of the sequence parsing activities in the
Seqparse
class. - Added the
FileSequence
class to represent individual padded file sequences. Containment behaviour handles both frame numbers and file names. - Added the
regex
submodule, home to the newSeqparseRegexMixin
class. This class adds a number of file name/sequence parsing methods that I’m using to great effect in a number of classes. FrameSequence
instances can now be instantiated with string frame sequences, and their containment behaviour will handle both padded and non-padded string frame values as well as integers.FrameSequence.add
will now handle string frame sequences as input as well.- Refactored all sequence-related classes (
FrameChunk
,FrameSequence
,FileSequence
) to their own module. - Removed the
seqparse.get_chunk()
method asFrameChunks
are really just for internal use only. - Private class attributes were turning out to be a drag, so now they’re all just protected.
- Overall module coverage is now hovering at around 98%. Yay.
v0.3.0 (2017/03/17)¶
Changes (from v0.2.0)
- Bugfix:
FrameSequence.discard()
now handles poorly padded frames correctly. - Bugfix: Messages thrown when an incorrect frame padding has been specified should now be more correct.
FrameChunk
andFrameSequence
instances may now be reversed via thereversed()
command.FrameChunk
andFrameSequence
may now be inverted (ie, report back the missing frames as aFrameSequence
instance) via their newinvert
method..coveragerc
has been updated to excludeseqparse.test*
.- Coverage tests for
FrameChunk
,FrameSequence
, andSeqparse
are now at 100%!
v0.2.0 (2017/03/17)¶
Changes (from v0.1.3)
FrameChunk
andFrameSequence
instances now have similar behaviour when used as iterators.- Testing on Linux-based machines now accepts executable test files (as I’m spending half my time programming on a Windows box).
- Yes, we do need some stinkin’ badges.
README.md
now indicates build status (travis-ci), coverage percentage (coveralls.io), and code health (landscape.io).
v0.1.3 (2017/03/16)¶
Changes (from v0.1.1)
- Added coveralls support! Now you can look at code coverage for the project on
coverage.io. And I get to add a pretty badge to the
README.md
.
v0.1.2 (2017/03/16)¶
Changes (from v0.1.1)
- Frame sequences may now be directly added to any
Seqparse
instance via the add_file method. FrameChunk
instances are now iterable and have proper containment tests for both (padded) string and integer frames.FileSequence
andSingletons
classes have been refactored toFileSequenceContainer
andSingletonContainer
, respectively, to more accurately reflect their functionality.
v0.1.1 (2017/03/15)¶
Test/Bugfix Release
Changes (from v.0.1)
seqls
: It’s accessed the same old way (ie,seqls
from the command line), but it’s no longer a dedicated script; it’s been moved toseqparse.cli.seqls
to ease installation.- The test suite has been expanded to cover pretty much everything as it currently stands.
- Fixed a bug where the padding on single frames wasn’t resolving properly.
README.md
updated with a laundry list of stuff I’d like to do before I’m moderately satisfied with my coding endeavours.
v0.1 (2017/03/14)¶
Initial release with basic functionality.
- Primary usage is via the included
seqls
script. - May be installed via setuptools-supplied
setup.py
. - No real documentation (yet).