Get summary statistics of user's archive.

Extract the real client IP from a proxied request. Relies on the known proxy chain: Client → HAProxy → Nginx → Django. HAProxy's ``optio

Extract a plain text excerpt from a story's content.

Get previous messages in the conversation for context.

Return a deterministic seed that changes daily, for stable pagination of shuffled feeds.

Get full content of a specific feed story.

Batch-fetch trending read times for a list of feed IDs.

Retrieve previously read stories, optionally filtered.

Get most recently archived pages.

Get full content of a specific starred story.

Get summary of user's starred stories.

Return the set of column names for a table.

Return the set of column names for a table.

Get top items by reading each daily key and applying decay weights in Python. No ZUNIONSTORE — just pipelined ZREVRANGE reads.

Check if an image URL (or a variant with different scheme/params/CDN) already exists in content.

Build a compact intelligence indicator string showing all non-zero classifiers.

Check if a response indicates the request was blocked by anti-bot protection.

Check if the configured server URL points to a local dev instance.

List all folders with their feeds and unread counts.

Load briefing models, applying settings override if BRIEFING_MODELS is defined. Self-hosters can define BRIEFING_MODELS in settings as a dict of

Load cached Feedly results.

Load cached newsletter results.

Load cached results.

Load models, applying settings override if ASK_AI_MODELS is defined. Self-hosters can define ASK_AI_MODELS in settings as a dict of model configs

Return the canonical taxonomy for feed discovery.

View or configure push notifications for specific feeds.

Detect and fix entity-encoded HTML content from email providers. Some providers (e.g. ImprovMX) send HTML with entities encoded, e.g.

http://www.mongodb.org/display/DOCS/Mongo+Wire+Protocol

Convert ImprovMX JSON format to Mailgun-compatible params. apps/newsletters/views.py:11

Parse AI response, ensuring clean category names. Handles various response formats and cleans up the output.

Parse a Feedly search result into a newsletter feed dict.

Parse a single Feedly search result into a feed dict.

Parse a single Feedly search result into a feed dict.

Parse a Ghost Explore publication page to extract metadata.

Parse OPML XML and extract feed URLs.

Parse OPML XML and extract feed URLs.

Parse PayPal billing_info.next_billing_time (ISO 8601, UTC) into a naive-UTC datetime, or None (apps/profile/models.py).

Parse Substack subscriber count string like '1,100,000' or '459K+' to int.

Parse a Substack publication object into our feed dict format.

Parse a TopPub.xyz tag page to extract Medium publication data. HTML structure per publication: <tr> <td><a href="https://t

Print detailed gap analysis.

Print summary showing remaining gaps after merge.

Print a summary of feeds by type and category.

Report a meter event to Stripe for usage-based billing.

Reset prometheus mmap cache when corruption is detected

Walk up folder hierarchy to find auto_mark_read_days setting. Folder titles use " - " as hierarchy separator (e.g., "Parent - Child -

Retry linking unlinked PopularFeed records by directly creating Feed objects. Groups URLs by domain and rate-limits to avoid hammering any si

Sanitize a string for use as a Prometheus label value.

Helper function to save classifiers for a single feed. classifier_data is a dict like: { "like_author": ["author1"], "dis

Process the html after sanitization

Process the html before sanitization

Search user's archives with filters using Elasticsearch for full-text search.

Search user's feed stories using Elasticsearch.

Search Feedly for feeds using taxonomy keywords.

Search Feedly for feeds using generated search queries.

Search Feedly for newsletters on a specific platform. Feedly doesn't support site: filters, so we use domain name + keyword queries a

Search stories shared by people the user follows.

Search user's starred/saved RSS stories.

Show a sample of entries that would be deleted.

Build a feedparser-compatible result dict from extracted stories.

Convert internal feed dict to fixture JSON format.

Convert internal feed dict to fixture JSON format.

Convert internal feed dict to fixture JSON format.

Convert internal feed dict to fixture JSON format.

Verify a feed URL returns valid RSS/Atom content via HTTP.

Calculate weekly window for free users. Resets Sunday night at midnight (Saturday night into Sunday morning) in user's timezone.

Calculate weekly window for free users. Resets Sunday night at midnight (Saturday night into Sunday morning) in user's timezone.

Write to fixture file, replacing existing podcast entries.

Write to fixture file, preserving non-RSS feeds.

Write to fixture file, replacing existing newsletter entries.

Write to fixture file, replacing existing reddit entries.