Source code for bokeh.embed.standalone

#-----------------------------------------------------------------------------
# Copyright (c) 2012 - 2019, Anaconda, Inc., and Bokeh Contributors.
# All rights reserved.
#
# The full license is in the file LICENSE.txt, distributed with this software.
#-----------------------------------------------------------------------------
'''

'''

#-----------------------------------------------------------------------------
# Boilerplate
#-----------------------------------------------------------------------------
from __future__ import absolute_import, division, print_function, unicode_literals

import logging
log = logging.getLogger(__name__)

#-----------------------------------------------------------------------------
# Imports
#-----------------------------------------------------------------------------

# Standard library imports
from bokeh.util.future import collections_abc # goes away with py2

# External imports
from six import string_types

# Bokeh imports
from ..core.templates import AUTOLOAD_JS, AUTOLOAD_TAG, FILE, ROOT_DIV, MACROS
from ..document.document import DEFAULT_TITLE, Document
from ..model import Model
from ..util.compiler import bundle_all_models
from ..util.string import encode_utf8
from .bundle import bundle_for_objs_and_resources
from .elements import html_page_for_render_items, script_for_render_items
from .util import FromCurdoc, OutputDocumentFor, standalone_docs_json, standalone_docs_json_and_render_items
from .wrappers import wrap_in_onload, wrap_in_script_tag

#-----------------------------------------------------------------------------
# Globals and constants
#-----------------------------------------------------------------------------

__all__ = (
    'autoload_static',
    'components',
    'file_html',
    'json_item',
)

#-----------------------------------------------------------------------------
# General API
#-----------------------------------------------------------------------------

[docs]def autoload_static(model, resources, script_path): ''' Return JavaScript code and a script tag that can be used to embed Bokeh Plots. The data for the plot is stored directly in the returned JavaScript code. Args: model (Model or Document) : resources (Resources) : script_path (str) : Returns: (js, tag) : JavaScript code to be saved at ``script_path`` and a ``<script>`` tag to load it Raises: ValueError ''' # TODO: maybe warn that it's not exactly useful, but technically possible # if resources.mode == 'inline': # raise ValueError("autoload_static() requires non-inline resources") if isinstance(model, Model): models = [model] elif isinstance (model, Document): models = model.roots else: raise ValueError("autoload_static expects a single Model or Document") with OutputDocumentFor(models): (docs_json, [render_item]) = standalone_docs_json_and_render_items([model]) bundle = bundle_all_models() or "" script = script_for_render_items(docs_json, [render_item]) (modelid, elementid) = list(render_item.roots.to_json().items())[0] js = wrap_in_onload(AUTOLOAD_JS.render( js_urls = resources.js_files, css_urls = resources.css_files, js_raw = resources.js_raw + [bundle, script], css_raw = resources.css_raw_str, elementid = elementid, )) tag = AUTOLOAD_TAG.render( src_path = script_path, elementid = elementid, ) return encode_utf8(js), encode_utf8(tag)
[docs]def components(models, wrap_script=True, wrap_plot_info=True, theme=FromCurdoc): ''' Return HTML components to embed a Bokeh plot. The data for the plot is stored directly in the returned HTML. An example can be found in examples/embed/embed_multiple.py The returned components assume that BokehJS resources are **already loaded**. The html template in which they will be embedded needs to include the following scripts tags. The widgets and tables resources are only necessary if the components make use of widgets and tables. .. code-block:: html <script src="http://cdn.pydata.org/bokeh/release/bokeh-x.y.z.min.js"></script> <script src="http://cdn.pydata.org/bokeh/release/bokeh-widgets-x.y.z.min.js"></script> <script src="http://cdn.pydata.org/bokeh/release/bokeh-tables-x.y.z.min.js"></script> Note that in Jupyter Notebooks, it is not possible to use components and show in the same notebook cell. Args: models (Model|list|dict|tuple) : A single Model, a list/tuple of Models, or a dictionary of keys and Models. wrap_script (boolean, optional) : If True, the returned javascript is wrapped in a script tag. (default: True) wrap_plot_info (boolean, optional) : If True, returns ``<div>`` strings. Otherwise, return dicts that can be used to build your own divs. (default: True) If False, the returned dictionary contains the following information: .. code-block:: python { 'modelid': 'The model ID, used with Document.get_model_by_id', 'elementid': 'The css identifier the BokehJS will look for to target the plot', 'docid': 'Used by Bokeh to find the doc embedded in the returned script', } theme (Theme, optional) : Defaults to the ``Theme`` instance in the current document. Setting this to ``None`` uses the default theme or the theme already specified in the document. Any other value must be an instance of the ``Theme`` class. Returns: UTF-8 encoded *(script, div[s])* or *(raw_script, plot_info[s])* Examples: With default wrapping parameter values: .. code-block:: python components(plot) # => (script, plot_div) components((plot1, plot2)) # => (script, (plot1_div, plot2_div)) components({"Plot 1": plot1, "Plot 2": plot2}) # => (script, {"Plot 1": plot1_div, "Plot 2": plot2_div}) Examples: With wrapping parameters set to ``False``: .. code-block:: python components(plot, wrap_script=False, wrap_plot_info=False) # => (javascript, plot_dict) components((plot1, plot2), wrap_script=False, wrap_plot_info=False) # => (javascript, (plot1_dict, plot2_dict)) components({"Plot 1": plot1, "Plot 2": plot2}, wrap_script=False, wrap_plot_info=False) # => (javascript, {"Plot 1": plot1_dict, "Plot 2": plot2_dict}) ''' # 1) Convert single items and dicts into list was_single_object = isinstance(models, Model) or isinstance(models, Document) models = _check_models_or_docs(models) # now convert dict to list, saving keys in the same order model_keys = None dict_type = None if isinstance(models, dict): model_keys = models.keys() dict_type = models.__class__ values = [] # don't just use .values() to ensure we are in the same order as key list for k in model_keys: values.append(models[k]) models = values # 2) Append models to one document. Either pre-existing or new and render with OutputDocumentFor(models, apply_theme=theme): (docs_json, [render_item]) = standalone_docs_json_and_render_items(models) script = bundle_all_models() or "" script += script_for_render_items(docs_json, [render_item]) if wrap_script: script = wrap_in_script_tag(script) script = encode_utf8(script) def div_for_root(root): return ROOT_DIV.render(root=root, macros=MACROS) if wrap_plot_info: results = list(div_for_root(root) for root in render_item.roots) else: results = render_item.roots # 3) convert back to the input shape if was_single_object: result = results[0] elif model_keys is not None: result = dict_type(zip(model_keys, results)) else: result = tuple(results) return script, result
[docs]def file_html(models, resources, title=None, template=FILE, template_variables={}, theme=FromCurdoc, suppress_callback_warning=False, _always_new=False): ''' Return an HTML document that embeds Bokeh Model or Document objects. The data for the plot is stored directly in the returned HTML, with support for customizing the JS/CSS resources independently and customizing the jinja2 template. Args: models (Model or Document or seq[Model]) : Bokeh object or objects to render typically a Model or Document resources (Resources or tuple(JSResources or None, CSSResources or None)) : A resource configuration for Bokeh JS & CSS assets. title (str, optional) : A title for the HTML document ``<title>`` tags or None. (default: None) If None, attempt to automatically find the Document title from the given plot objects. template (Template, optional) : HTML document template (default: FILE) A Jinja2 Template, see bokeh.core.templates.FILE for the required template parameters template_variables (dict, optional) : variables to be used in the Jinja2 template. If used, the following variable names will be overwritten: title, bokeh_js, bokeh_css, plot_script, plot_div theme (Theme, optional) : Defaults to the ``Theme`` instance in the current document. Setting this to ``None`` uses the default theme or the theme already specified in the document. Any other value must be an instance of the ``Theme`` class. suppress_callback_warning (bool, optional) : Normally generating standalone HTML from a Bokeh Document that has Python callbacks will result in a warning stating that the callbacks cannot function. However, this warning can be suppressed by setting this value to True (default: False) Returns: UTF-8 encoded HTML ''' if isinstance(models, Model): models = [models] if isinstance(models, Document): models = models.roots with OutputDocumentFor(models, apply_theme=theme, always_new=_always_new) as doc: (docs_json, render_items) = standalone_docs_json_and_render_items(models, suppress_callback_warning=suppress_callback_warning) title = _title_from_models(models, title) bundle = bundle_for_objs_and_resources([doc], resources) return html_page_for_render_items(bundle, docs_json, render_items, title=title, template=template, template_variables=template_variables)
[docs]def json_item(model, target=None, theme=FromCurdoc): ''' Return a JSON block that can be used to embed standalone Bokeh content. Args: model (Model) : The Bokeh object to embed target (string, optional) A div id to embed the model into. If None, the target id must be supplied in the JavaScript call. theme (Theme, optional) : Defaults to the ``Theme`` instance in the current document. Setting this to ``None`` uses the default theme or the theme already specified in the document. Any other value must be an instance of the ``Theme`` class. Returns: JSON-like This function returns a JSON block that can be consumed by the BokehJS function ``Bokeh.embed.embed_item``. As an example, a Flask endpoint for ``/plot`` might return the following content to embed a Bokeh plot into a div with id *"myplot"*: .. code-block:: python @app.route('/plot') def plot(): p = make_plot('petal_width', 'petal_length') return json.dumps(json_item(p, "myplot")) Then a web page can retrieve this JSON and embed the plot by calling ``Bokeh.embed.embed_item``: .. code-block:: html <script> fetch('/plot') .then(function(response) { return response.json(); }) .then(function(item) { Bokeh.embed.embed_item(item); }) </script> Alternatively, if is more convenient to supply the target div id directly in the page source, that is also possible. If `target_id` is omitted in the call to this function: .. code-block:: python return json.dumps(json_item(p)) Then the value passed to ``embed_item`` is used: .. code-block:: javascript Bokeh.embed.embed_item(item, "myplot"); ''' with OutputDocumentFor([model], apply_theme=theme) as doc: doc.title = "" docs_json = standalone_docs_json([model]) doc = list(docs_json.values())[0] root_id = doc['roots']['root_ids'][0] return { 'target_id' : target, 'root_id' : root_id, 'doc' : doc, }
#----------------------------------------------------------------------------- # Dev API #----------------------------------------------------------------------------- #----------------------------------------------------------------------------- # Private API #----------------------------------------------------------------------------- def _check_models_or_docs(models): ''' ''' input_type_valid = False # Check for single item if isinstance(models, (Model, Document)): models = [models] # Check for sequence if isinstance(models, collections_abc.Sequence) and all(isinstance(x, (Model, Document)) for x in models): input_type_valid = True if isinstance(models, dict) and \ all(isinstance(x, string_types) for x in models.keys()) and \ all(isinstance(x, (Model, Document)) for x in models.values()): input_type_valid = True if not input_type_valid: raise ValueError( 'Input must be a Model, a Document, a Sequence of Models and Document, or a dictionary from string to Model and Document' ) return models def _title_from_models(models, title): # use override title if title is not None: return title # use title from any listed document for p in models: if isinstance(p, Document): return p.title # use title from any model's document for p in models: if p.document is not None: return p.document.title # use default title return DEFAULT_TITLE #----------------------------------------------------------------------------- # Code #-----------------------------------------------------------------------------