rename article_handler.py -> article_generator.py

1 files changed, 172 insertions, 0 deletions

diff --git a/src/article_generator.py b/src/article_generator.py
new file mode 100644
index 0000000..ba60eff
--- /dev/null
+++ b/src/article_generator.py
@@ -0,0 +1,172 @@
+import dataclasses
+import datetime
+import os
+import logging
+import pathlib
+import glob
+import typing
+import markdown
+import yaml
+
+
+# Known exceptions, these are raised when generating on program exiting error, and are displayed to the user nicely
+class ArticleHandlerException(Exception):
+    pass
+
+
+class ArticleMetaDataMalformed(ArticleHandlerException):
+    pass
+
+
+class ArticleNoMetaData(ArticleHandlerException):
+    pass
+
+
+@dataclasses.dataclass
+class WebPage:
+    url: str
+    name: str
+
+
+@dataclasses.dataclass
+class MetaData():
+    description: str
+    created: datetime.date
+
+    @property
+    def pretty_print(self) -> str:
+        return f'''
+    Description: {self.description}
+    Created: {self.created}'''
+
+
+def truncate(text: str, max_length: int = 50):
+    if len(text) < max_length:
+        return text
+    else:
+        return text[0:max_length - 1]
+
+
+@dataclasses.dataclass
+class Article(WebPage):
+    metadata: MetaData
+    web_dir: tuple
+    source_path: str
+    source: str
+    html: str
+    folder_path: typing.Union[None, str] = None
+
+    @property
+    def pretty_print(self) -> str:
+        return f'''
+Name:        {self.name}
+Metadata:    {self.metadata.pretty_print}
+Web dir:     {self.web_dir}
+URL:         {self.url}
+Source path: {self.source_path}
+Folder path: {self.folder_path}
+HTML:        {truncate(self.html)}
+Source:      {truncate(self.source)}'''
+
+
+@dataclasses.dataclass
+class Folder(WebPage):
+    articles: typing.List[Article] = dataclasses.field(default_factory=list)
+    sub_folders: typing.Dict[str, object] = dataclasses.field(default_factory=dict)
+
+
+def get_web_dir(path, name) -> tuple[str, ...]:
+    dir_structure = path.split(name)[0]
+    # Split into tuple, remove first part, assemble to path again, all to remove the first source dir agnostically 
+    dir_structure_prefix_striped = pathlib.Path(dir_structure).parts[2:]
+    return dir_structure_prefix_striped
+
+
+def parse_article_meta_data(source: str) -> typing.Tuple[str, MetaData]:
+    if source.startswith('---'):
+        meta_data_yml_end_char_index = source.find('---', 3)
+        meta_data_yml = source[3:meta_data_yml_end_char_index]
+        # Strip metadata text from source, before feeding it to the markdown reader
+        source = source[meta_data_yml_end_char_index + 3:]
+        meta_data = yaml.safe_load(meta_data_yml)
+
+        try:
+            meta_data = MetaData(**meta_data)
+        except TypeError as type_error_exception:
+            raise ArticleMetaDataMalformed(str(type_error_exception))
+        return source, meta_data
+    else:
+        raise ArticleNoMetaData(f'No metadata found')
+
+
+def get_article(path: str) -> Article:
+    logging.debug(path)
+    article_args = {}
+    filename, file_extension = os.path.splitext(path)
+
+    basename = os.path.basename(filename)
+    if basename == 'index':
+        # Article is the folder
+        article_folder_name = os.path.dirname(path)
+        article_args['folder_path'] = article_folder_name
+        dir_basename = os.path.basename(article_folder_name)
+        article_args['name'] = dir_basename
+    else:
+        # Article one file
+        article_args['name'] = os.path.basename(filename)
+    article_args['web_dir'] = get_web_dir(path, article_args['name'])
+    article_args['source_path'] = path
+
+    article_args[
+        'url'] = f'{"/" if article_args["web_dir"] else ""}{"/".join(article_args["web_dir"])}/{article_args["name"]}'
+
+    with open(path, 'r') as file:
+        source = file.read()
+    article_args['source'], article_args['metadata'] = parse_article_meta_data(source)
+    article_args['html'] = markdown.markdown(article_args['source'], extensions=
+    [
+        'fenced_code',
+        'codehilite',
+        'tables',
+        'toc'  # Automatically generates unique IDs for headers allowing for ID URL referral (Anchor)
+    ])
+
+    article = Article(**article_args)
+    return article
+
+
+def discover_folder_structure(article: Article, articles: Folder):
+    previous_folder = articles
+    for folder_name in article.web_dir:
+        logging.debug(folder_name)
+
+        if folder_name not in previous_folder.sub_folders:
+            logging.debug('new')
+            current_folder = Folder(url=f'{previous_folder.url}{folder_name}/', name=folder_name)
+            previous_folder.sub_folders[folder_name] = current_folder
+        else:
+            logging.debug('reuse')
+            current_folder = previous_folder.sub_folders[folder_name]
+        previous_folder = current_folder
+        logging.debug('')
+
+    previous_folder.articles.append(article)
+
+
+def sort_articles(folder: Folder):
+    folder.articles = sorted(folder.articles,
+                             key=lambda x: x.metadata.created if x.metadata.created is not None else datetime.date.min,
+                             reverse=True)
+    for folder in folder.sub_folders.values():
+        sort_articles(folder)
+
+
+def discover_articles(path):
+    articles_paths = glob.glob(f'{path}/**/*.md', recursive=True)  # Equivalent to ls ./**.md
+    articles = Folder(url='/', name='index')
+    for article_path in articles_paths:
+        article = get_article(article_path)
+        logging.debug(article.pretty_print)
+        discover_folder_structure(articles=articles, article=article)
+    sort_articles(articles)
+    return articles