def load_export_config(export_config_path): """load the export configuration""" if isinstance(export_config_path, string_types): export_config_path = pathlib.Path(export_config_path) data = read_file_from_directory( export_config_path.parent, export_config_path.name, "export configuration", logger, interp_ext=True, ) # validate against schema global _EXPORT_SCHEMA if _EXPORT_SCHEMA is None: # lazy load schema once _EXPORT_SCHEMA = read_file_from_directory( get_module_path(schema), _EXPORT_SCHEMA_FILE, "export configuration schema", logger, interp_ext=True, ) try: jsonschema.validate(data, _EXPORT_SCHEMA) except jsonschema.ValidationError as err: handle_error( "validation of export config {} failed against {}: {}".format( export_config_path, _EXPORT_SCHEMA_FILE, err.message), jsonschema.ValidationError, logger=logger, ) return data
def iter_all_export_paths(config_folder_paths=(), regex="*.json"): """we iterate through all json files in the supplied plugin_folder_paths, and then in the `export_plugins` folder """ for plugin_folder_path in config_folder_paths: for jsonpath in glob.glob(os.path.join(plugin_folder_path, regex)): name = os.path.splitext(os.path.basename(jsonpath))[0] yield name, pathlib.Path(jsonpath) module_path = get_module_path(export_plugins) for jsonpath in glob.glob(os.path.join(str(module_path), regex)): name = os.path.splitext(os.path.basename(jsonpath))[0] yield name, pathlib.Path(jsonpath)
def create_template(outline_template, outline_name, segment_datas, outpath=None): # type: (dict, Tuple[dict]) -> str """ build a latex jinja template from; - a jinja(2) template outline, which may contain segment placeholders, - and json segment files adhering to the segment.schema.json schema if a segment contains the key "overwrite", then its value should be a list of keys, such that these key values overwrite any entries before Parameters ---------- outline_template: str segment_datas: tuple or dict outpath: None or str if not None, output to path """ # get the placeholders @ipubreplace{above|below}{name} regex = re.compile("\\@ipubreplace\\{([^\\}]+)\\}\\{([^\\}]+)\\}", re.MULTILINE) placeholder_tuple = regex.findall(outline_template) if not placeholder_tuple: if segment_datas: handle_error( "the segment data is provided, " + "but the outline template contains no placeholders", KeyError, logger, ) if outpath: _output_to_file(outline_template, outpath) return outline_template placeholders = {name: append for append, name in placeholder_tuple} # TODO validate that placeholders to not exist multiple times, # with above and below replacements = {key: "" for key in placeholders.keys()} docstrings = ["outline: {}".format(outline_name)] if segment_datas: docstrings.append("with segments:") global _SEGMENT_SCHEMA if _SEGMENT_SCHEMA is None: # lazy segment schema once _SEGMENT_SCHEMA = read_file_from_directory( get_module_path(schema), _SEGMENT_SCHEMA_FILE, "segment configuration schema", logger, interp_ext=True, ) for seg_num, segment_data in enumerate(segment_datas): # validate segment try: jsonschema.validate(segment_data, _SEGMENT_SCHEMA) except jsonschema.ValidationError as err: handle_error( "validation of template segment {} failed: {}".format( seg_num, err.message), jsonschema.ValidationError, logger=logger, ) # get description of segment docstrings.append("- {0}: {1}".format(segment_data["identifier"], segment_data["description"])) # find what key to overwrite overwrite = segment_data.get("overwrite", []) logger.debug("overwrite keys: {}".format(overwrite)) for key, segtext in segment_data.get("segments").items(): if key not in placeholders: handle_error( "the segment key '{}' ".format(key) + "is not contained in the outline template", KeyError, logger, ) if not isinstance(segtext, string_types): segtext = "\n".join(segtext) if key in overwrite: replacements[key] = segtext elif placeholders[key] == "above": replacements[key] = segtext + "\n" + replacements[key] elif placeholders[key] == "below": replacements[key] = replacements[key] + "\n" + segtext else: handle_error( ("the placeholder @ipubreplace{{{0}}}{{{1}}} ".format( key, placeholders[key]) + "should specify 'above' or 'below' appending"), jsonschema.ValidationError, logger=logger, ) if "meta_docstring" in placeholders: docstring = "\n".join([s for s in docstrings if s]).replace("'", '"') replacements["meta_docstring"] = docstring if "ipypub_version" in placeholders: # TODO add option to include ipypub version in output file # not included by default, # since tests need to be changed to ignore version number replacements["ipypub_version"] = "" # str(__version__) prefix = "@ipubreplace{" replace_dict = { prefix + append + "}{" + name + "}": replacements.get(name, "") for append, name in placeholder_tuple } outline = multireplace(outline_template, replace_dict) if outpath: _output_to_file(outline, outpath) return outline
def publish(self, ipynb_path, nb_node=None): """ convert one or more Jupyter notebooks to a published format paths can be string of an existing file or folder, or a pathlib.Path like object all files linked in the documents are placed into a single files_folder Parameters ---------- ipynb_path: str or pathlib.Path notebook file or directory nb_node: None or nbformat.NotebookNode a pre-converted notebook Returns -------- outdata: dict containing keys; "outpath", "exporter", "stream", "main_filepath", "resources" """ # setup the input and output paths if isinstance(ipynb_path, string_types): ipynb_path = pathlib.Path(ipynb_path) ipynb_name, ipynb_ext = os.path.splitext(ipynb_path.name) outdir = (os.path.join(os.getcwd(), "converted") if self.outpath is None else str(self.outpath)) with self._log_handlers(ipynb_name, outdir): if not ipynb_path.exists() and not nb_node: handle_error( "the notebook path does not exist: {}".format(ipynb_path), IOError, self.logger, ) # log start of conversion self.logger.info("started ipypublish v{0} at {1}".format( ipypublish.__version__, time.strftime("%c"))) self.logger.info("logging to: {}".format( os.path.join(outdir, ipynb_name + ".nbpub.log"))) self.logger.info("running for ipynb(s) at: {0}".format(ipynb_path)) self.logger.info("with conversion configuration: {0}".format( self.conversion)) if nb_node is None and ipynb_ext in self.pre_conversion_funcs: func = self.pre_conversion_funcs[ipynb_ext] self.logger.info("running pre-conversion with: {}".format( inspect.getmodule(func))) try: nb_node = func(ipynb_path) except Exception as err: handle_error( "pre-conversion failed for {}: {}".format( ipynb_path, err), err, self.logger, ) # doesn't work with folders # if (ipynb_ext != ".ipynb" and nb_node is None): # handle_error( # 'the file extension is not associated with any ' # 'pre-converter: {}'.format(ipynb_ext), # TypeError, self.logger) if nb_node is None: # merge all notebooks # TODO allow notebooks to remain separate # (would require creating a main.tex with the preamble in etc ) # Could make everything a 'PyProcess', # with support for multiple streams final_nb, meta_path = merge_notebooks( ipynb_path, ignore_prefix=self.ignore_prefix) else: final_nb, meta_path = (nb_node, ipynb_path) # validate the notebook metadata against the schema if self.validate_nb_metadata: nb_metadata_schema = read_file_from_directory( get_module_path(schema), "doc_metadata.schema.json", "doc_metadata.schema", self.logger, interp_ext=True, ) try: jsonschema.validate(final_nb.metadata, nb_metadata_schema) except jsonschema.ValidationError as err: handle_error( "validation of notebook level metadata failed: {}\n" "see the doc_metadata.schema.json for full spec". format(err.message), jsonschema.ValidationError, logger=self.logger, ) # set text replacements for export configuration replacements = { self.meta_path_placeholder: str(meta_path), self.files_folder_placeholder: "{}{}".format(get_valid_filename(ipynb_name), self.folder_suffix), } self.logger.debug("notebooks meta path: {}".format(meta_path)) # load configuration file ( exporter_cls, jinja_template, econfig, pprocs, pconfig, ) = self._load_config_file(replacements) # run nbconvert self.logger.info("running nbconvert") exporter, stream, resources = self.export_notebook( final_nb, exporter_cls, econfig, jinja_template) # postprocess results main_filepath = os.path.join(outdir, ipynb_name + exporter.file_extension) for post_proc_name in pprocs: proc_class = find_entry_point( post_proc_name, "ipypublish.postprocessors", self.logger, "ipypublish", ) proc = proc_class(pconfig) stream, main_filepath, resources = proc.postprocess( stream, exporter.output_mimetype, main_filepath, resources) self.logger.info("process finished successfully") return { "outpath": outdir, "exporter": exporter, "stream": stream, "main_filepath": main_filepath, "resources": resources, }