102 lines
5.4 KiB
Python
102 lines
5.4 KiB
Python
import logging
|
|
import pprint
|
|
import uuid
|
|
|
|
from pelican import signals
|
|
|
|
log = logging.getLogger(__name__)
|
|
|
|
def modify_feed(context, feed):
|
|
articles = {}
|
|
|
|
for article in context['articles']:
|
|
if article.title in articles:
|
|
raise Exception(f"Duplicate article title found: {article.title}")
|
|
articles[article.title] = article
|
|
|
|
for item in feed.items:
|
|
item_title = item['title']
|
|
article = articles.get(item_title)
|
|
if not article:
|
|
raise Exception(f"Article not found for title: {item_title}")
|
|
|
|
if not hasattr(article, 'guid') or not article.guid:
|
|
log.info(f"Article '{article.title}' ({article.source_path}) is missing a guid. Generating and embedding one.")
|
|
new_guid_str = uuid.uuid4().hex
|
|
# The string to embed in the paragraph. Note the leading space.
|
|
guid_text_to_embed = f" Guid: {new_guid_str}"
|
|
|
|
source_path = article.source_path
|
|
|
|
# Ensure article object has the _content attribute
|
|
if not hasattr(article, '_content'):
|
|
log.error(f"Article '{article.title}' does not have '_content' attribute. Cannot embed Guid into source file.")
|
|
raise Exception(f"Cannot find raw content for article '{article.title}' to embed Guid.")
|
|
|
|
# Read the original file content.
|
|
# Python's open() in text mode uses universal newlines by default, converting \r\n and \r to \n.
|
|
# Pelican's MarkdownReader also provides article._content with \n newlines.
|
|
try:
|
|
with open(source_path, 'r', encoding='utf-8') as f:
|
|
original_file_content_universal_newlines = f.read()
|
|
except Exception as e:
|
|
log.error(f"Failed to read original content from '{source_path}': {e}")
|
|
raise
|
|
|
|
# Sanity check: the article's body content should be a suffix of the read file content.
|
|
if not original_file_content_universal_newlines.endswith(article._content):
|
|
log.error(f"Content mismatch for '{article.title}' in '{source_path}'. "
|
|
"The article's parsed content (article._content) does not match the "
|
|
"ending of the raw file (read with universal newlines). This is unexpected "
|
|
"and may indicate issues with file parsing or concurrent modifications.")
|
|
# For debugging, one might log tails of both strings here.
|
|
# log.debug(f"Tail of original file content: '{original_file_content_universal_newlines[-200:]}'")
|
|
# log.debug(f"Tail of article._content: '{article._content[-200:]}'")
|
|
raise Exception(f"Content boundary determination error for article '{article.title}'.")
|
|
|
|
# Determine the metadata part by subtracting the length of the content body.
|
|
metadata_section_length = len(original_file_content_universal_newlines) - len(article._content)
|
|
metadata_part_from_file = original_file_content_universal_newlines[:metadata_section_length]
|
|
|
|
# current_body_content is what Pelican parsed as the article's body.
|
|
current_body_content = article._content
|
|
|
|
# Split this body content to find its first paragraph.
|
|
# Paragraphs in Markdown are separated by one or more blank lines (\n\n).
|
|
body_parts = current_body_content.split('\n\n', 1)
|
|
first_paragraph_of_body = body_parts[0]
|
|
rest_of_body_content = body_parts[1] if len(body_parts) > 1 else ""
|
|
|
|
# Append the Guid text to the end of the first paragraph of the body.
|
|
# .rstrip() removes any trailing whitespace/newlines from the paragraph itself before appending.
|
|
modified_first_paragraph_of_body = first_paragraph_of_body.rstrip() + guid_text_to_embed
|
|
|
|
# Reconstruct the new body content with the embedded Guid.
|
|
new_body_content_with_guid = modified_first_paragraph_of_body
|
|
if rest_of_body_content: # Add back the rest of the body if it existed.
|
|
new_body_content_with_guid += '\n\n' + rest_of_body_content
|
|
|
|
# Construct the full new file content by combining the original metadata part and the new body.
|
|
# This preserves the original metadata block verbatim (including comments, formatting, and original newline characters if any within it,
|
|
# as metadata_part_from_file is a direct slice from original_file_content_universal_newlines which has \n newlines).
|
|
full_new_content = metadata_part_from_file + new_body_content_with_guid
|
|
|
|
try:
|
|
with open(source_path, 'w', encoding='utf-8') as f:
|
|
f.write(full_new_content)
|
|
log.info(f"Successfully wrote updated content with embedded Guid to '{source_path}'.")
|
|
except Exception as e:
|
|
log.error(f"Failed to write updated content to '{source_path}': {e}")
|
|
raise # Re-raise the exception to halt processing if file write fails
|
|
|
|
# Set article.guid for the current Pelican run, so it's used for the feed item
|
|
article.guid = new_guid_str
|
|
log.debug(f"Set in-memory article.guid = '{new_guid_str}' for '{article.title}'.")
|
|
|
|
item['unique_id'] = article.guid
|
|
|
|
|
|
def register():
|
|
signals.feed_generated.connect(modify_feed)
|
|
|