Fixed base_link

The `alt` contains the LaTeX formula rendered as Unicode characters, and
that may include non-standard characters that break the RSS feed encoding.

CHANGELOG.md

@@ -1,5 +1,15 @@
 # Changelog
 
+## 0.2.19
+
+- Added `short_feed` configuration flag to permanently disable returning the
+  full content of the articles in the RSS feed.
+
+## 0.2.16
+
+- Removed `alt` attribute from LaTeX rendered `<img>` tags. It may generate
+  non-standard Unicode characters that break the RSS feed.
+
 ## 0.2.14
 
 - Better support for PWA tags and added a default config-generated `/manifest.json`.

README.md

@@ -72,6 +72,8 @@ logo: /path/or/url/here
 language: en-US
 # Show/hide the header (default: true)
 header: true
+# Enable/disable the short RSS feed (default: false)
+short_feed: false
 
 categories:
   - category1
@@ -133,4 +135,9 @@ $$
 RSS feeds for the blog are provided under the `/rss` URL.
 
 By default, the whole HTML-rendered content of an article is returned under `rss.channel.item.description`.
+
 If you only want to include the short description of an article in the feed, use `/rss?short` instead.
+
+If you want the short feed (i.e. without the fully rendered article as a
+description) to be always returned, then you can specify `short_feed=true` in
+your configuration.

madblog/__init__.py

@@ -1 +1 @@
-__version__ = '0.2.16'
+__version__ = '0.2.23'

madblog/app.py

@@ -12,11 +12,11 @@ from ._sorters import PagesSorter, PagesSortByTime
 
 
 class BlogApp(Flask):
-    _title_header_regex = re.compile(r'^#\s*((\[(.*)\])|(.*))')
+    _title_header_regex = re.compile(r"^#\s*((\[(.*)\])|(.*))")
 
     def __init__(self, *args, **kwargs):
         super().__init__(*args, template_folder=config.templates_dir, **kwargs)
-        self.pages_dir = os.path.join(config.content_dir, 'markdown')
+        self.pages_dir = os.path.join(config.content_dir, "markdown")
         self.img_dir = config.default_img_dir
         self.css_dir = config.default_css_dir
         self.js_dir = config.default_js_dir
@@ -27,73 +27,77 @@ class BlogApp(Flask):
             # `config.content_dir` is treated as the root for markdown files.
             self.pages_dir = config.content_dir
 
-        img_dir = os.path.join(config.content_dir, 'img')
+        img_dir = os.path.join(config.content_dir, "img")
         if os.path.isdir(img_dir):
             self.img_dir = os.path.abspath(img_dir)
         else:
             self.img_dir = config.content_dir
 
-        css_dir = os.path.join(config.content_dir, 'css')
+        css_dir = os.path.join(config.content_dir, "css")
         if os.path.isdir(css_dir):
             self.css_dir = os.path.abspath(css_dir)
 
-        js_dir = os.path.join(config.content_dir, 'js')
+        js_dir = os.path.join(config.content_dir, "js")
         if os.path.isdir(js_dir):
             self.js_dir = os.path.abspath(js_dir)
 
-        fonts_dir = os.path.join(config.content_dir, 'fonts')
+        fonts_dir = os.path.join(config.content_dir, "fonts")
         if os.path.isdir(fonts_dir):
             self.fonts_dir = os.path.abspath(fonts_dir)
 
-        templates_dir = os.path.join(config.content_dir, 'templates')
+        templates_dir = os.path.join(config.content_dir, "templates")
         if os.path.isdir(templates_dir):
             self.template_folder = os.path.abspath(templates_dir)
 
     def get_page_metadata(self, page: str) -> dict:
-        if not page.endswith('.md'):
-            page = page + '.md'
+        if not page.endswith(".md"):
+            page = page + ".md"
 
         md_file = os.path.join(self.pages_dir, page)
         if not os.path.isfile(md_file):
             abort(404)
 
         metadata = {}
-        with open(md_file, 'r') as f:
-            metadata['uri'] = '/article/' + page[:-3]
+        with open(md_file, "r") as f:
+            metadata["uri"] = "/article/" + page[:-3]
 
-            for line in f.readlines():
+            for line in f:
                 if not line:
                     continue
 
-                if not (m := re.match(r'^\[//]: # \(([^:]+):\s*([^)]+)\)\s*$', line)):
+                if not (m := re.match(r"^\[//]: # \(([^:]+):\s*(.*)\)\s*$", line)):
                     break
 
-                if m.group(1) == 'published':
-                    metadata[m.group(1)] = datetime.date.fromisoformat(m.group(2))
+                if m.group(1) == "published":
+                    metadata[m.group(1)] = datetime.datetime.fromisoformat(
+                        m.group(2)
+                    ).date()
                 else:
                     metadata[m.group(1)] = m.group(2)
 
-        if not metadata.get('title'):
+        if not metadata.get("title"):
             # If the `title` header isn't available in the file,
             # infer it from the first line of the file
-            with open(md_file, 'r') as f:
-                header = ''
+            with open(md_file, "r") as f:
+                header = ""
                 for line in f.readlines():
                     header = line
                     break
 
-            metadata['title_inferred'] = True
+            metadata["title_inferred"] = True
             m = self._title_header_regex.search(header)
             if m:
-                metadata['title'] = m.group(3) or m.group(1)
+                metadata["title"] = m.group(3) or m.group(1)
             else:
-                metadata['title'] = os.path.basename(md_file)
+                metadata["title"] = os.path.basename(md_file)
 
-        if not metadata.get('published'):
+        if not metadata.get("published"):
             # If the `published` header isn't available in the file,
             # infer it from the file's creation date
-            metadata['published'] = datetime.date.fromtimestamp(os.stat(md_file).st_ctime)
-            metadata['published_inferred'] = True
+            metadata["published"] = datetime.date.fromtimestamp(
+                os.stat(md_file).st_ctime
+            )
+            metadata["published_inferred"] = True
 
         return metadata
 
@@ -102,37 +106,42 @@ class BlogApp(Flask):
         page: str,
         title: Optional[str] = None,
         skip_header: bool = False,
-        skip_html_head: bool = False
+        skip_html_head: bool = False,
     ):
-        if not page.endswith('.md'):
-            page = page + '.md'
+        if not page.endswith(".md"):
+            page = page + ".md"
 
         metadata = self.get_page_metadata(page)
         # Don't duplicate the page title if it's been inferred
-        if not (title or metadata.get('title_inferred')):
-            title = metadata.get('title', config.title)
+        if not (title or metadata.get("title_inferred")):
+            title = metadata.get("title", config.title)
 
-        with open(os.path.join(self.pages_dir, page), 'r') as f:
+        with open(os.path.join(self.pages_dir, page), "r") as f:
             return render_template(
-                'article.html',
+                "article.html",
                 config=config,
                 title=title,
-                image=metadata.get('image'),
-                description=metadata.get('description'),
+                image=metadata.get("image"),
+                description=metadata.get("description"),
                 author=(
-                    re.match(r'(.+?)\s+<([^>]+>)', metadata['author'])[1]
-                    if 'author' in metadata else None
-                ),
-                author_email=(
-                    re.match(r'(.+?)\s+<([^>]+)>', metadata['author'])[2]
-                    if 'author' in metadata else None
-                ),
-                published=(
-                    metadata['published'].strftime('%b %d, %Y')
-                    if metadata.get('published') and not metadata.get('published_inferred')
+                    re.match(r"(.+?)\s+<([^>]+>)", metadata["author"])[1]
+                    if "author" in metadata
                     else None
                 ),
-                content=markdown(f.read(), extensions=['fenced_code', 'codehilite', MarkdownLatex()]),
+                author_email=(
+                    re.match(r"(.+?)\s+<([^>]+)>", metadata["author"])[2]
+                    if "author" in metadata
+                    else None
+                ),
+                published=(
+                    metadata["published"].strftime("%b %d, %Y")
+                    if metadata.get("published")
+                    and not metadata.get("published_inferred")
+                    else None
+                ),
+                content=markdown(
+                    f.read(), extensions=["fenced_code", "codehilite", MarkdownLatex()]
+                ),
                 skip_header=skip_header,
                 skip_html_head=skip_html_head,
             )
@@ -145,26 +154,25 @@ class BlogApp(Flask):
         sorter: Type[PagesSorter] = PagesSortByTime,
         reverse: bool = True,
     ) -> List[Tuple[int, dict]]:
-        pages_dir = app.pages_dir.rstrip('/')
+        pages_dir = app.pages_dir.rstrip("/")
         pages = [
             {
-                'path': os.path.join(root[len(pages_dir)+1:], f),
-                'folder': root[len(pages_dir)+1:],
-                'content': (
+                "path": os.path.join(root[len(pages_dir) + 1 :], f),
+                "folder": root[len(pages_dir) + 1 :],
+                "content": (
                     self.get_page(
                         os.path.join(root, f),
                         skip_header=skip_header,
                         skip_html_head=skip_html_head,
                     )
-                    if with_content else ''
-                ),
-                **self.get_page_metadata(
-                    os.path.join(root[len(pages_dir)+1:], f)
+                    if with_content
+                    else ""
                 ),
+                **self.get_page_metadata(os.path.join(root[len(pages_dir) + 1 :], f)),
             }
             for root, _, files in os.walk(pages_dir, followlinks=True)
             for f in files
-            if f.endswith('.md')
+            if f.endswith(".md")
         ]
 
         sorter_func = sorter(pages)

madblog/config.py

@@ -1,57 +1,61 @@
 import os
+from typing import List
 import yaml
 
-from dataclasses import dataclass
+from dataclasses import dataclass, field
 
 
 @dataclass
 class Config:
-    title = 'Blog'
-    description = ''
-    link = '/'
-    home_link = '/'
-    language = 'en-US'
-    logo = '/img/icon.png'
+    title = "Blog"
+    description = ""
+    link = "/"
+    home_link = "/"
+    language = "en-US"
+    logo = "/img/icon.png"
     header = True
-    content_dir = '.'
-    categories = None
+    content_dir = "."
+    categories: List[str] = field(default_factory=list)
+    short_feed = False
 
     basedir = os.path.abspath(os.path.dirname(__file__))
-    templates_dir = os.path.join(basedir, 'templates')
-    static_dir = os.path.join(basedir, 'static')
-    default_css_dir = os.path.join(static_dir, 'css')
-    default_js_dir = os.path.join(static_dir, 'js')
-    default_fonts_dir = os.path.join(static_dir, 'fonts')
-    default_img_dir = os.path.join(static_dir, 'img')
+    templates_dir = os.path.join(basedir, "templates")
+    static_dir = os.path.join(basedir, "static")
+    default_css_dir = os.path.join(static_dir, "css")
+    default_js_dir = os.path.join(static_dir, "js")
+    default_fonts_dir = os.path.join(static_dir, "fonts")
+    default_img_dir = os.path.join(static_dir, "img")
 
 
 config = Config()
 
 
-def init_config(content_dir='.', config_file='config.yaml'):
+def init_config(content_dir=".", config_file="config.yaml"):
     cfg = {}
     config.content_dir = content_dir
 
     if os.path.isfile(config_file):
-        with open(config_file, 'r') as f:
+        with open(config_file, "r") as f:
             cfg = yaml.safe_load(f)
 
-    if cfg.get('title'):
-        config.title = cfg['title']
-    if cfg.get('description'):
-        config.description = cfg['description']
-    if cfg.get('link'):
-        config.link = cfg['link']
-    if cfg.get('home_link'):
-        config.home_link = cfg['home_link']
-    if cfg.get('logo') is not None:
-        config.logo = cfg['logo']
-    if cfg.get('language'):
-        config.language = cfg['language']
-    if cfg.get('header') is False:
+    if cfg.get("title"):
+        config.title = cfg["title"]
+    if cfg.get("description"):
+        config.description = cfg["description"]
+    if cfg.get("link"):
+        config.link = cfg["link"]
+    if cfg.get("home_link"):
+        config.home_link = cfg["home_link"]
+    if cfg.get("logo") is not None:
+        config.logo = cfg["logo"]
+    if cfg.get("language"):
+        config.language = cfg["language"]
+    if cfg.get("header") is False:
         config.header = False
+    if cfg.get("short_feed"):
+        config.short_feed = True
 
-    config.categories = cfg.get('categories', [])
+    config.categories = cfg.get("categories", [])
 
 
 # vim:sw=4:ts=4:et:

madblog/latex.py

@@ -31,11 +31,11 @@ def call(*args, **kwargs):
 
 
 # Defines our basic inline image
-img_expr = '<img class="latex inline math-%s" alt="%s" id="%s" src="data:image/png;base64,%s">'
+img_expr = '<img class="latex inline math-%s" id="%s" src="data:image/png;base64,%s">'
 
 # Defines multiline expression image
-multiline_img_expr = '''<div class="multiline-wrapper">
-<img class="latex multiline math-%s" alt="%s" id="%s" src="data:image/png;base64,%s"></div>'''
+multiline_img_expr = """<div class="multiline-wrapper">
+<img class="latex multiline math-%s" id="%s" src="data:image/png;base64,%s"></div>"""
 
 # Base CSS template
 img_css = """<style scoped>
@@ -55,8 +55,8 @@ img.latex.inline {
 </style>"""
 
 # Cache and temp file paths
-tmpdir = tempfile.gettempdir() + '/markdown-latex'
-cache_file = tmpdir + '/latex.cache'
+tmpdir = tempfile.gettempdir() + "/markdown-latex"
+cache_file = tmpdir + "/latex.cache"
 
 
 class LaTeXPreprocessor(markdown.preprocessors.Preprocessor):
@@ -75,16 +75,20 @@ class LaTeXPreprocessor(markdown.preprocessors.Preprocessor):
 """
 
     # Math TeX extraction regex
-    math_extract_regex = re.compile(r'(.+?)((\\\(.+?\\\))|(\$\$\n.+?\n\$\$\n))(.*)', re.MULTILINE | re.DOTALL)
+    math_extract_regex = re.compile(
+        r"(.+?)((\\\(.+?\\\))|(\$\$\n.+?\n\$\$\n))(.*)", re.MULTILINE | re.DOTALL
+    )
 
     # Math TeX matching regex
-    math_match_regex = re.compile(r'\s*(\\\(.+?\\\))|(\$\$\n.+?\n\$\$\n)\s*', re.MULTILINE | re.DOTALL)
+    math_match_regex = re.compile(
+        r"\s*(\\\(.+?\\\))|(\$\$\n.+?\n\$\$\n)\s*", re.MULTILINE | re.DOTALL
+    )
 
     def __init__(self, *_, **__):
         if not os.path.isdir(tmpdir):
             os.makedirs(tmpdir)
         try:
-            with open(cache_file, 'r') as f:
+            with open(cache_file, "r") as f:
                 self.cached = json.load(f)
         except (IOError, json.JSONDecodeError):
             self.cached = {}
@@ -94,7 +98,8 @@ class LaTeXPreprocessor(markdown.preprocessors.Preprocessor):
             ("dvipng", "args"): "-q -T tight -bg Transparent -z 9 -D 200",
             ("delimiters", "text"): "%",
             ("delimiters", "math"): "$",
-            ("delimiters", "preamble"): "%%"}
+            ("delimiters", "preamble"): "%%",
+        }
 
     def _latex_to_base64(self, tex):
         """Generates a base64 representation of TeX string"""
@@ -104,18 +109,24 @@ class LaTeXPreprocessor(markdown.preprocessors.Preprocessor):
         with os.fdopen(tmp_file_fd, "w") as tmp_file:
             tmp_file.write(self.tex_preamble)
             tmp_file.write(tex)
-            tmp_file.write('\n\\end{document}')
+            tmp_file.write("\n\\end{document}")
 
         # compile LaTeX document. A DVI file is created
-        status = call(('latex -halt-on-error -output-directory={:s} {:s}'
-                       .format(tmpdir, path)).split(),
-                      stdout=PIPE, timeout=10)
+        status = call(
+            (
+                "latex -halt-on-error -output-directory={:s} {:s}".format(tmpdir, path)
+            ).split(),
+            stdout=PIPE,
+            timeout=10,
+        )
 
         # clean up if the above failed
         if status:
             self._cleanup(path, err=True)
-            raise Exception("Couldn't compile LaTeX document." +
-                            "Please read '%s.log' for more detail." % path)
+            raise Exception(
+                "Couldn't compile LaTeX document."
+                + "Please read '%s.log' for more detail." % path
+            )
 
         # Run dvipng on the generated DVI file. Use tight bounding box.
         # Magnification is set to 1200
@@ -129,8 +140,10 @@ class LaTeXPreprocessor(markdown.preprocessors.Preprocessor):
         # clean up if we couldn't make the above work
         if status:
             self._cleanup(path, err=True)
-            raise Exception("Couldn't convert LaTeX to image." +
-                            "Please read '%s.log' for more detail." % path)
+            raise Exception(
+                "Couldn't convert LaTeX to image."
+                + "Please read '%s.log' for more detail." % path
+            )
 
         # Read the png and encode the data
         try:
@@ -157,7 +170,7 @@ class LaTeXPreprocessor(markdown.preprocessors.Preprocessor):
     def run(self, lines):
         """Parses the actual page"""
         # Checks for the LaTeX header
-        use_latex = any(line == '[//]: # (latex: 1)' for line in lines)
+        use_latex = any(line == "[//]: # (latex: 1)" for line in lines)
         if not use_latex:
             return lines
 
@@ -165,7 +178,9 @@ class LaTeXPreprocessor(markdown.preprocessors.Preprocessor):
         page = "\n".join(lines)
 
         # Adds a preamble mode
-        self.tex_preamble += self.config[("general", "preamble")] + "\n\\begin{document}\n"
+        self.tex_preamble += (
+            self.config[("general", "preamble")] + "\n\\begin{document}\n"
+        )
 
         # Figure out our text strings and math-mode strings
         tex_expr = self.math_extract_regex.findall(page)
@@ -176,7 +191,7 @@ class LaTeXPreprocessor(markdown.preprocessors.Preprocessor):
 
         # Parse the expressions
         new_cache = {}
-        new_page = ''
+        new_page = ""
         n_multiline_expressions = 0
 
         while page:
@@ -200,21 +215,25 @@ class LaTeXPreprocessor(markdown.preprocessors.Preprocessor):
                     new_cache[tex_hash] = data
 
                 if is_multiline and n_multiline_expressions > 0:
-                    new_page += '</p>'
-                new_page += (multiline_img_expr if is_multiline else img_expr) % ('true', expr, tex_hash, data)
+                    new_page += "</p>"
+                new_page += (multiline_img_expr if is_multiline else img_expr) % (
+                    "true",
+                    tex_hash,
+                    data,
+                )
 
                 if is_multiline:
-                    new_page += '<p>'
+                    new_page += "<p>"
                     n_multiline_expressions += 1
 
             page = m.group(5)
 
         if n_multiline_expressions > 0:
-            new_page += '</p>'
+            new_page += "</p>"
 
         # Cache our data
         self.cached.update(new_cache)
-        with open(cache_file, 'w') as f:
+        with open(cache_file, "w") as f:
             json.dump(self.cached, f)
 
         # Make sure to re-split the lines
@@ -230,7 +249,7 @@ class MarkdownLatex(markdown.Extension):
 
     def extendMarkdown(self, md):
         md.preprocessors.register(
-             LaTeXPreprocessor(self),
-            'latex',
+            LaTeXPreprocessor(self),
+            "latex",
             1,
         )

madblog/routes.py

@@ -1,5 +1,7 @@
 import os
+import re
 from typing import Optional
+from urllib.parse import urljoin
 
 from flask import (
     jsonify,
@@ -115,8 +117,12 @@ def article_route(article: str):
 
 @app.route("/rss", methods=["GET"])
 def rss_route():
-    pages = app.get_pages(with_content=True, skip_header=True, skip_html_head=True)
-    short_description = "short" in request.args
+    short_description = "short" in request.args or config.short_feed
+    pages = app.get_pages(
+        with_content=not short_description,
+        skip_header=True,
+        skip_html_head=True,
+    )
 
     return Response(
         """<?xml version="1.0" encoding="UTF-8" ?>
@@ -149,27 +155,36 @@ def rss_route():
             ),
             items="\n\n".join(
                 [
-                    """
+                    (
+                        """
             <item>
                 <title>{title}</title>
                 <link>{base_link}{link}</link>
                 <pubDate>{published}</pubDate>
                 <description><![CDATA[{content}]]></description>
-                <media:content medium="image" url="{base_link}{image}" width="200" height="150" />
+                <media:content medium="image" url="{image}" width="200" height="150" />
             </item>
-            """.format(
+                    """
+                    ).format(
                         base_link=config.link,
                         title=page.get("title", "[No Title]"),
                         link=page.get("uri", ""),
-                        published=page["published"].strftime(
-                            "%a, %d %b %Y %H:%M:%S GMT"
-                        )
-                        if "published" in page
-                        else "",
-                        content=page.get("description", "")
-                        if short_description
-                        else page.get("content", ""),
-                        image=page.get("image", ""),
+                        published=(
+                            page["published"].strftime("%a, %d %b %Y %H:%M:%S GMT")
+                            if "published" in page
+                            else ""
+                        ),
+                        content=(
+                            page.get("description", "")
+                            if short_description
+                            else page.get("content", "")
+                        ),
+                        image=(
+                            urljoin(config.link, page["image"])
+                            if page.get("image")
+                            and not re.search(r"^https?://", page["image"])
+                            else page.get("image", "")
+                        ),
                     )
                     for _, page in pages
                 ]

setup.cfg

@@ -1,7 +1,8 @@
 [bumpversion]
-current_version = 0.2.16
+current_version = 0.2.23
 commit = True
 tag = True
 
 [metadata]
 description-file = README.md
+

setup.py

@@ -11,7 +11,7 @@ def readfile(file):
 
 setup(
     name='madblog',
-    version='0.2.16',
+    version='0.2.23',
     author='Fabio Manganiello',
     author_email='info@fabiomanganiello.com',
     description='A minimal platform for Markdown-based blogs',