fix(render-helper): strip the Lumen "Posted via" byline

[feruzm]

Closes #1727

Lumen (app lumen/1.0) appends ---\n*Posted via Lumen* to every post and comment, with a by <user> suffix for lite accounts posting through a container. This adds it to the app-footer filters in cleanReply and extends the separator lookahead introduced for scrobble.life so the --- directly before the byline is dropped too. Unrelated horizontal rules are untouched. The marker includes the leading asterisk so a body that merely mentions posting via Lumen is not stripped.

Test plan

• New specs: plain footer, by <user> footer, unrelated hr preserved, plain mention preserved.
• Mutation check: removing the line filter or the lookahead extension fails three specs each.
• Real bodies from @hbd-temp and @lumenpublisher rendered through markdown2Html contain no byline and no <hr>.
• vitest run in packages/render-helper: 24 files, 1332 tests green. eslint and tsc clean.

[qodo-code-review]

ⓘ Qodo reviews are paused because your trial has ended. Ask your workspace admin to add credits to resume reviews. Manage billing

[coderabbitai]

Important

• 🔍 Trigger review

This repository does not receive automatic reviews because it has fewer than 10 stars.

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Team

Run ID: 8630614b-e1c5-464a-b1f3-cd8cdf243786

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[qodo-free-for-open-source-projects]

PR Summary by Qodo

Strip Lumen bylines and adjacent separators from rendered replies

🐞 Bug fix 🧪 Tests 🕐 10-20 Minutes

AI Description

• Removes Lumen-generated bylines from posts and comments during reply cleaning.
• Drops only separators immediately preceding recognized Lumen or scrobble.life footers.
• Adds regression coverage for lite accounts, false positives, and horizontal-rule preservation.

Diagram

graph TD
  A["Reply Markdown"] --> B{"Known footer?"} -- Yes --> C["Remove separator"] --> D["Filter bylines"] --> E["Clean reply"]
  B -- No --> D

Loading

High-Level Assessment

Extending the existing cleanReply footer pipeline is appropriate because Lumen uses the same separator-plus-byline structure already handled for scrobble.life. Matching the leading Markdown asterisk and coupling separator removal to a lookahead avoids stripping ordinary Lumen mentions or unrelated horizontal rules; broader substring removal or a new parser would add risk and complexity without benefit.

Files changed (2) +45 / -4

Bug fix (1) +11 / -4

clean-reply.method.tsRemove Lumen bylines and their adjacent separators +11/-4

Remove Lumen bylines and their adjacent separators

• Extends separator lookahead detection to recognize Lumen footer markers alongside scrobble.life. Adds a targeted line filter for Lumen bylines, including optional lite-account attribution suffixes.

packages/render-helper/src/methods/clean-reply.method.ts

Tests (1) +34 / -0

clean-reply.method.spec.tsCover Lumen footer removal and false-positive safeguards +34/-0

Cover Lumen footer removal and false-positive safeguards

• Adds regression tests for standard and lite-account Lumen bylines. Verifies that the attached separator is removed while unrelated horizontal rules and ordinary Lumen mentions remain intact.

packages/render-helper/src/methods/clean-reply.method.spec.ts

[qodo-free-for-open-source-projects]

Code Review by Qodo

🐞 Bugs (0) 📘 Rule violations (0) 📎 Requirement gaps (1) 🎨 UX issues (0) 🔗 Cross-repo conflicts (0) 📜 Skill insights (0)

1. Lumen filter strips inline prose 📎 Requirement gap ≡ Correctness

Description

The new Lumen cleanup filter removes every line containing *posted via lumen, including ordinary
user-authored prose such as I *posted via Lumen yesterday*, instead of only footer lines beginning
with that marker. Because cleanReply processes complete post bodies and summaries, this can
silently delete legitimate content and violate the requirement to preserve non-footer mentions.

Code

packages/render-helper/src/methods/clean-reply.method.ts[44]

+    .filter(item => item.toLowerCase().includes('*posted via lumen') === false)

Evidence

The changed filter matches *posted via lumen anywhere in a line, while the compliance rule
requires matching footer content that begins with the marker. cleanReply applies these
unconditional substring filters to every newline-separated body line, and rendering paths pass
complete post bodies and summaries through it; existing preservation coverage checks only an
unitalicized mention and does not cover the exact marker substring or a preceding horizontal rule.

Strip Lumen footer bylines in cleanReply
packages/render-helper/src/methods/clean-reply.method.ts[44-44]
packages/render-helper/src/methods/clean-reply.method.ts[12-47]
packages/render-helper/src/markdown-2-html.ts[53-75]
packages/render-helper/src/post-body-summary.ts[60-67]
packages/render-helper/src/methods/clean-reply.method.spec.ts[280-285]

Agent prompt

The issue below was found during a code review. Follow the provided context and guidance below and implement a solution

## Issue description
Update the Lumen cleanup filter so it removes only actual footer lines whose content begins with `*posted via lumen`, while preserving ordinary prose that quotes, discusses, or otherwise contains the italicized phrase. Ensure the separator lookahead uses the same footer-specific matching rule so a preceding horizontal rule is not removed for ordinary content.
## Issue Context
Compliance ID 1 requires plain and proxied-user Lumen footers to be removed while ordinary mentions remain intact. `cleanReply` processes complete post bodies and post summaries, so an unrestricted substring match can cause user content to disappear. Add coverage for prose containing the italicized phrase and for a horizontal rule before such prose, while preserving removal of the actual footer and its optional `by <user>` suffix.
## Fix Focus Areas
- packages/render-helper/src/methods/clean-reply.method.ts[8-10]
- packages/render-helper/src/methods/clean-reply.method.ts[44-44]
- packages/render-helper/src/methods/clean-reply.method.spec.ts[280-285]

ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools

---

Tip of the day

💡 Did you know, you can show, collapse, or hide each part of a finding: code, evidence, and all

More tips ↗ | Customize Qodo ↗ | Qodo docs ↗

[qodo-free-for-open-source-projects]

1. Lumen filter strips inline prose 📎 Requirement gap ≡ Correctness

The new Lumen cleanup filter removes every line containing *posted via lumen, including ordinary
user-authored prose such as I *posted via Lumen yesterday*, instead of only footer lines beginning
with that marker. Because cleanReply processes complete post bodies and summaries, this can
silently delete legitimate content and violate the requirement to preserve non-footer mentions.

Agent Prompt

## Issue description
Update the Lumen cleanup filter so it removes only actual footer lines whose content begins with `*posted via lumen`, while preserving ordinary prose that quotes, discusses, or otherwise contains the italicized phrase. Ensure the separator lookahead uses the same footer-specific matching rule so a preceding horizontal rule is not removed for ordinary content.

## Issue Context
Compliance ID 1 requires plain and proxied-user Lumen footers to be removed while ordinary mentions remain intact. `cleanReply` processes complete post bodies and post summaries, so an unrestricted substring match can cause user content to disappear. Add coverage for prose containing the italicized phrase and for a horizontal rule before such prose, while preserving removal of the actual footer and its optional `by <user>` suffix.

## Fix Focus Areas
- packages/render-helper/src/methods/clean-reply.method.ts[8-10]
- packages/render-helper/src/methods/clean-reply.method.ts[44-44]
- packages/render-helper/src/methods/clean-reply.method.spec.ts[280-285]

ⓘ Copy this prompt and use it to remediate the issue with your preferred AI generation tools

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 35edb74227

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Restrict the Lumen filter to the actual footer

When ordinary prose contains an italicized phrase such as I *posted via Lumen* yesterday, this substring predicate removes the entire line. Because cleanReply processes all rendered bodies rather than only entries tagged lumen/1.0, legitimate non-footer content is lost; anchor the match to the complete terminal byline, including its closing * and optional by ... suffix.

Useful? React with 👍 / 👎.