Plugins

How to use plugins

Plugins must be specified with the plugins setting:

plugins = ['sigal.plugins.adjust', 'sigal.plugins.copyright']

You can either specify the name of the module which contains the plugin, or import the plugin before adding it to the list:

from sigal.plugins import copyright
plugins = ['sigal.plugins.adjust', copyright]

Note

Using an import like this will break the multiprocessing feature, because the settings dict must be serializable. So in most cases you should prefer the first option.

The plugin_paths setting can be used to specify paths to search for plugins (if they are not in the python path).

Write a new plugin

Plugins are based on signals with the blinker library. A plugin must subscribe to a signal (the list is given below). New signals can be added if need. See an example with the copyright plugin:

"""Plugin which add a copyright to the image.

Settings:

- ``copyright``: the copyright text.
- ``copyright_text_font``: the copyright text font - either system/user
  font-name or absolute path to font.ttf file.  If no font is specified, or
  specified font is not found, the default font is used.
- ``copyright_text_font_size``: the copyright text font-size. If no font is
  specified, this setting is ignored.
- ``copyright_text_color``: the copyright text color in a tuple (R, G, B)
  with decimal RGB code, e.g. ``(255, 255, 255)`` is white.
- ``copyright_text_position``: the copyright text position in 2 tuple (left,
  top).  By default text would be positioned at bottom-left corner.

"""

import logging

from PIL import ImageDraw, ImageFont

from sigal import signals

logger = logging.getLogger(__name__)


def add_copyright(img, settings=None):
    logger.debug("Adding copyright to %r", img)
    draw = ImageDraw.Draw(img)
    text = settings["copyright"]
    font = settings.get("copyright_text_font", None)
    font_size = settings.get("copyright_text_font_size", 10)
    assert font_size >= 0
    color = settings.get("copyright_text_color", (0, 0, 0))
    bottom_margin = 3  # bottom margin for text
    text_height = bottom_margin + 12  # default text height (of 15)
    if font:
        try:
            font = ImageFont.truetype(font, font_size)
            text_height = font.getsize(text)[1] + bottom_margin
        except Exception:  # load default font in case of any exception
            logger.debug("Exception: Couldn't locate font %s, using default font", font)
            font = ImageFont.load_default()
    else:
        font = ImageFont.load_default()
    left, top = settings.get("copyright_text_position", (5, img.size[1] - text_height))
    draw.text((left, top), text, fill=color, font=font)
    return img


def register(settings):
    if settings.get("copyright"):
        signals.img_resized.connect(add_copyright)
    else:
        logger.warning("Copyright text is not set")

Signals

sigal.signals.album_initialized(album)

Called after the Album is initialized.

Parameters:

album – the Album object.

sigal.signals.gallery_initialized(gallery)

Called after the gallery is initialized.

Parameters:

gallery – the Gallery object.

sigal.signals.media_initialized(media)

Called after the Media (Image or Video) is initialized.

Parameters:

media – the media object.

sigal.signals.gallery_build(gallery)

Called after the gallery is build (after media are resized and html files are written).

Parameters:

gallery – the Gallery object.

sigal.signals.img_resized(img, settings=settings)

Called after the image is resized. This signal work differently, the modified image object must be returned by the registered function.

Parameters:
  • img – the PIL image object.

  • settings – the settings dict.

List of plugins

Adjust plugin

Plugin which adjust the image after resizing.

Based on pilkit’s Adjust processor.

Settings:

adjust_options = {
    'color': 1.0,
    'brightness': 1.0,
    'contrast': 1.0,
    'sharpness': 1.0
}

Compress assets plugin

Plugin to compress static files for faster HTTP transfer.

Your web server must be configured in order to select pre-compressed files instead of the required one, to save CPU. For example, in nginx:

server {
    gzip_static;
}

Currently, 3 methods are supported:

  • gzip: No dependency required. This is the fastest, but also largest output.

  • zopfli: Requires zopfli. gzip compatible output with optimized size.

  • brotli: Requires brotli. Brotli is the best compressor for web usage.

Settings available as dictionary in compress_assets_options:

  • method: One of gzip, zopfli or brotli.

  • suffixes: List of file suffixes eligible to compression. Default to ['htm', 'html', 'css', 'js', 'svg']

Copyright plugin

Plugin which add a copyright to the image.

Settings:

  • copyright: the copyright text.

  • copyright_text_font: the copyright text font - either system/user font-name or absolute path to font.ttf file. If no font is specified, or specified font is not found, the default font is used.

  • copyright_text_font_size: the copyright text font-size. If no font is specified, this setting is ignored.

  • copyright_text_color: the copyright text color in a tuple (R, G, B) with decimal RGB code, e.g. (255, 255, 255) is white.

  • copyright_text_position: the copyright text position in 2 tuple (left, top). By default text would be positioned at bottom-left corner.

Encrypt plugin

Plugin to protect gallery by encrypting image files using a password.

Options:

encrypt_options = {
    'password': 'password',
    'ask_password': False,
    'gcm_tag': 'randomly_generated',
    'kdf_salt': 'randomly_generated',
    'kdf_iters': 10000,
}
  • password: The password used to encrypt the images on gallery build, and decrypt them when viewers access the gallery. No default value. You must specify a password.

  • ask_password: Whether or not viewers are asked for the password to view the gallery. If set to False, the password will be present in the HTML files so the images are decrypted automatically. Defaults to False.

  • gcm_tag, kdf_salt, kdf_iters: Cryptographic parameters used when encrypting the files. gcm_tag, kdf_salt are meant to be randomly generated, kdf_iters defaults to 10000. Do not specify them in the config file unless you have good reasons to do so.

Note: The plugin caches the cryptographic parameters (but not the password) after the first build, so that incremental builds can share the same credentials. DO NOT CHANGE THE PASSWORD OR OTHER CRYPTOGRAPHIC PARAMETERS ONCE A GALLERY IS BUILT, or there will be inconsistency in encrypted files and viewers will not be able to see some of the images any more.

Compatibility with other plugins:

  • zip_gallery: if you enable both this plugin and the zip_gallery plugin, the generated zip archives will contain encrypted images, which is generally meaningless since viewers cannot easily decrypt them outside a browser.

Extended caching plugin

Decreases the time needed to build large galleries (e.g.: 25k images in 2.5s instead of 30s)

This plugin allows extended caching, which is useful for large galleries. Once a gallery has been built it caches all metadata for all media (markdown, exif, itpc) in the gallery target folder. Before the next run it restores them so that the image and metadata files do not have to be parsed again. For large galleries this can speed up the creation of index files dramatically.

Feeds plugin

Plugin which add RSS/ATOM feeds.

This plugin requires feedgenerator. It uses all the images and videos of the gallery, sorted by date, to show the most recent ones.

Settings:

  • rss_feed and atom_feed, see below.

Example:

rss_feed = {'feed_url': 'http://example.org/feed.rss', 'nb_items': 10}
atom_feed = {'feed_url': 'http://example.org/feed.atom', 'nb_items': 10}

Media page plugin

Plugin which generates HTML pages for each image.

Currently this plugin can be used only with the colorbox theme, the other themes have to be adapted.

For themes, the previous_media and next_media variables contain the previous/next Media objects.

Nomedia plugin

This plugin offers more fine-grained control over exluded images and folders, similarly to how it’s handled on Android.

To ignore a folder or image put a .nomedia file next to it in its parent folder and put its name into the file. E.g.:

content of folder:
    IMG_3425.JPG, IMG_2426.JPG, IMG_2427.JPG, subfolder, .nomedia
content of .nomedia:
    IMG_2426.JPG
    IMG_2427.JPG
    subfolder

will ignore all images but IMG_3425.JPG and the subfolder.

Alternatively, if you put a .nomedia file into a folder and leave it blank (i.e. an empty file called .nomedia in a folder containing images), this ignores the whole folder it’s located in (like on Android).

WARNING: When you have a pre-existing gallery from a previous run of sigal adding a new .nomedia file will not remove the newly ignored images/albums from the existing gallery (only the entries in the parent gallery pointing to it). They might still be reachable thereafter. Either remove the whole gallery to be sure or remove the ignored files/folders inside the gallery to remove them for good.

Non-media Files plugin

Plugin to index non-media files.

This plugin will copy the files into the build tree and generate generic thumbnails for the files. In-browser previews will likely fail, and it is up to the theme to provide correct support for downloads.

Settings available as dictionary in nonmedia_files_options:

  • ext_as_thumb: Enable simple thumbnail showing ext. Default to True

  • ignore_ext: List of file extensions to ignore. Default to [".md"]

  • thumb_bg_color: Background color for thumbnail. Accepts (r, g, b) tuple. Default to white (255, 255, 255).

  • thumb_font: Name or path to font file. Default to None to use the PIL built-in font.

  • thumb_font_color: Font color for thumbnail. Accepts (r, g, b) tuple. Default to black (0, 0, 0).

  • thumb_font_size: Font size for thumbnail. Must select a font to apply. Default to 40.

Note

Thumbnails are generated using the file extension text on a background color. It is highly recommended to select a font (such as "FreeMono.ttf"), since the default PIL font does not respect font size.

Watermark plugin

Plugin which adds a watermark to the image.

Based on http://code.activestate.com/recipes/362879-watermark-with-pil/ (Licensed under the PSF License).

Settings:

  • watermark: path to the watermark image.

  • watermark_position: the watermark position either ‘scale’ or ‘tile’ or a 2-tuple giving the upper left corner, or a 4-tuple defining the left, upper, right, and lower pixel coordinate, or None (same as (0, 0)). If a 4-tuple is given, the size of the pasted image must match the size of the region.

  • watermark_opacity: the watermark opacity (0.0 to 1.0).

Titleregexp plugin

This plugin modifies titles of galleries by using regular-expressions and simple string or character replacements. It is acting in two phases: First, all regular-expression-based modifications are carried out, second, string/character replacements are done.

The first phase may be interrupted if a match occurs by the ‘break’-setting individually per regular-expression part. Additionally, if a match occurs, string/character replacements may be added.

The second phase is done even if the first phase had been interrupted.

Settings:

  • titleregexp with the following keys:
    • regexp, which is an array of dicts with ‘search’, ‘replace’, ‘break’ ‘substitute’ and ‘count’ keys. All but ‘break’ and ‘substitute’ are the arguments for re.subn, without given ‘count’ all matches are replaced. If ‘break’ is anything but an empty string, the rest of the regexp-array is being skipped. The ‘substitute’-key contains an array following the substitute format explained in the next sentence and does string-replacement if the regular-expression matched.

    • substitute, which is an array of 2-element-arrays, of which the occurences of the first element will be replaced by the second element by the replace-method of strings.

Example:

titleregexp = {
    'regexp' : [
        { 'search': r"^([0-9]*)-(.*)$", 'replace': r"\2 (\1)", 'count': 1,
          'break': 1, substitute: [ ['ae','ä'] ] },
        { 'search': r"([a-z][a-z])([A-Z][a-z])", 'replace': r"\1 \2" }
        ],
    'substitute' : [ [ '_', ' ' ] ]
}