All paths should be relative to Generator.path unless we're actively
accessing the filesystem. This makes the argument less ambiguous, so
we have less likelyhood of joining paths multiple times.
We no longer instantiate the Static object in the StaticGenerator, so
we can't set the save_as argument anymore. If you want to adjust the
output path, use the upcoming EXTRA_PATH_METADATA setting.
If a setting exists in DEFAULT_CONFIG, assume it will be there
(instead of checking and/or providing a local default). The earlier
code was split between the two idioms, which was confusing.
This cuts down on the remaining difference between static files and
articles/pages. The main difference is that path-based metadata is
now parsed for static content.
The old get_relative_path() implementation assumed os.sep == '/',
which doesn't hold on MS Windows. The new implementation uses
split_all() for a more general component count.
I added path_to_url(), because the:
'/'.join(split_all(path))
idiom was showing up in a number of cases, and it's easier to
understand what's going on when that reads:
path_to_url(path)
This will fix a number of places where I think paths and URLs were
conflated, and should improve MS Windows support.
From the Python docs for os.sep [1]:
Note that knowing this is not sufficient to be able to parse or
concatenate pathnames - use os.path.split() and os.path.join()...
Where I touched a line, I also changed double quoted string literals
to single quotes, since they are used more often in the source:
wking@mjolnir ~/src/pelican $ git grep "'" pelican/*.py | wc -l
683
wking@mjolnir ~/src/pelican $ git grep '"' pelican/*.py | wc -l
181
[1]: http://docs.python.org/3/library/os.html#os.sep
Allows users to have per-year, per-month, and per-day archives of posts
automatically generated. The feature is disabled by default; to enable
it a user must supply format strings for a period's respective
`_SAVE_AS` setting.
I think the conversion from native paths to URLs is best put off until
we are actually trying to generate the URL. The old handling
(introduced in 2692586, Fixes#645 - Making cross-content linking
windows compatible, 2012-12-19) converted the path at StaticContent
initialization, which left you with a bogus StaticContent.src.
Once we drop the 'static' subdirectory, we will be able to drop the
`dest` and `url` parts from the StaticGenerator.generate_context()
handling, which will leave things looking a good deal cleaner than
they do now.
Static needs a lot of the same handling as other pages, so make it a
subclass of Page. The rename from StaticContent to Static makes for
cleaner configuration settings (STATIC_URL instead of
STATICCONTENT_URL).
All currently generated Static instances override the save_as
attribute explicitly on initialization, but it isn't hard to imagine
wanting to adjust STATIC file output based on metadata (e.g. extracted
from their source filename). With this union, the framework for
manipulating URLs and filenames is shared between all source file
types.
This makes it easier for StaticGenerator to walk FILES_TO_COPY, where
the input may be a directory or a bare filename.
Non-traversable file types (e.g. everything but directories and
symlinks to directories) are not checked against the exclude list.
The user-level effect of this is that explicit entries in STATIC_PATHS
or FILES_TO_COPY will override a hypothetical STATIC_EXCLUDES setting,
which seems like a reasonable approach.
I also removed the Python 2.5 compatibility check for `followlinks` in
os.walk, since Pelican is now Python >=2.7.
I wrote a plugin that I'd like to also be able to run on pages in
addition to articles. Adding this signal will let me update the content
when a page is finished being generated.
Making everything consistent is a bit awkward, since this is a
commonly used attribute, but I've done my best.
Reasons for not consolidating on `filename`:
* It is often used for the "basename" (last component in the path).
Using `source_path` makes it clear that this attribute can contain
multiple components.
Reasons for not consolidating on `filepath`:
* It is barely used in the Pelican source, and therefore easy to
change.
* `path` is more Pythonic. The only place `filepath` ever show up in
the documentation for `os`, `os.path`, and `shutil` is in the
`os.path.relpath` documentation [1].
Reasons for not consolidating on `path`:
* The Page elements have both a source (this attribute) and a
destination (.save_as). To avoid confusion for developers not aware
of this, make it painfully obvious that this attribute is for the
source. Explicit is better than implicit ;).
Where I was touching the line, I also updated the string formatting in
StaticGenerator.generate_output to use the forward compatible
'{}'.format() syntax.
[1]: http://docs.python.org/2/library/os.path.html#os.path.relpath
For reasons that are unclear to me, StaticContent introduces the
`filepath` attribute rather than using the existing (and semantically
equivalent) Page.filename. This has caused confusion before [1], and
it's probably a good idea to merge the two.
While I was touching the line, I also updated the string formatting in
StaticGenerator.generate_output to use the forward compatible
'{}'.format() syntax.
[1]: https://github.com/getpelican/pelican/issues/162#issuecomment-3000363
There have been earlier attempts to sort categories and authors
[1,2,3], but they either sorted based on the object id [3], or only
sorted the main author and categories list.
This patch uses rich comparisons (keyed off URLWrapper.name, but
easily adjustable in subclasses) to make the objects sortable without
specifying a key for each sort. For example, now
{% for tag, articles in tags|sort %}
works as expected in a Jinja template.
The functools.total_ordering decorator fills in the missing rich
comparisons [4,5].
[1]: 877d454c8f
[2]: 7f36e0ed20
[3]: d0ec18f4db
[4]: http://docs.python.org/2/library/functools.html#functools.total_ordering
[5]: http://docs.python.org/3/library/functools.html#functools.total_ordering
FEED_(ATOM|RSS) generated feeds include the version in the default language of a translated article,
whereas FEED_ALL(ATOM|RSS) would include *really* all posts, regardless of their language.
