[Due for payment 2025-08-26] [$250] Chat - Paste the copied message having quote markdown displays wrong line with original

[jponikarchuk]

If you haven’t already, check out our contributing guidelines for onboarding and email contributors@expensify.com to request to join our Slack channel!

---

Version Number: v9.1.73-0
Reproducible in staging?: Yes
Reproducible in production?: Yes
If this was caught during regression testing, add the test name, ID and link from TestRail: Exploratory
Email or phone of affected tester (no customers): N/A
Issue reported by: Applause Internal Team
Device used: Win 11/Chrome
App Component: Chat Report View

Action Performed:

1. Open https://staging.new.expensify.com
2. Go to any chat
3. Type ">" and space for creating quote markdown
4. Press Shift Enter
5. Type anything on the second line and send message
6. Copy this message
7. Paste this message to composer box

Expected Result:

The pasted message should be same as original

Actual Result:

The second line is displayed at first line in pasted message.

Workaround:

Unknown

Platforms:

• Android: App
• Android: mWeb Chrome
• iOS: App
• iOS: mWeb Safari
• iOS: mWeb Chrome
• Windows: Chrome
• MacOS: Chrome / Safari
• MacOS: Desktop

Screenshots/Videos

1.mp4

View all open jobs on GitHub

Upwork Automation - Do Not Edit

• Upwork Job URL: https://www.upwork.com/jobs/~021940992903128891410
• Upwork Job ID: 1940992903128891410
• Last Price Increase: 2025-07-11

Issue Owner

Current Issue Owner: @lydiabarclay

[melvin-bot]

Triggered auto assignment to @lydiabarclay (Bug), see https://stackoverflow.com/c/expensify/questions/14418 for more details. Please add this bug to a GH project, as outlined in the SO.

[nabi-ebrahimi]

Proposal

Please re-state the problem that we are trying to solve in this issue.

Chat - Paste the copied message having quote markdown displays wrong line with original

What is the root cause of that problem?

The root cause lies in the ExpensiMark parser’s handling of empty <blockquote> tags within the htmlToMarkdown method. The parser likely trims or collapses whitespace-only content within <blockquote> tags, resulting in a markdown output where the > marker is immediately followed by the next non-empty content (3) without a newline. This behavior disregards the structural intent of an empty blockquote, which should produce a > followed by a newline (\n) to indicate an empty quoted line in markdown.

The ExpensiMarkWithContext class inherits this behavior from ExpensiMark’s htmlToMarkdown method.

App/src/libs/Parser.ts

Lines 47 to 58 in 11ff1d3

	class ExpensiMarkWithContext extends ExpensiMark {
	htmlToMarkdown(htmlString: string, extras?: Extras, maxBodyLength?: number): string {
	return super.htmlToMarkdown(
	htmlString,
	{
	reportIDToName: extras?.reportIDToName ?? reportIDToNameMap,
	accountIDToName: extras?.accountIDToName ?? accountIDToNameMap,
	cacheVideoAttributes: extras?.cacheVideoAttributes,
	},
	maxBodyLength,
	);
	}

What changes do you think we should make in order to solve the problem?

To resolve the issue, we should implement a custom parsing logic in the ExpensiMarkWithContext class’s htmlToMarkdown method to explicitly handle empty <blockquote> tags before delegating the rest of the HTML to the parent ExpensiMark parser. The successful workaround splits the input HTML around empty <blockquote> tags, manually adds >\n for each empty blockquote, and processes the remaining HTML with super.htmlToMarkdown. The proposed change is to adopt this workaround as the permanent solution, with the following implementation:

 htmlToMarkdown(htmlString: string, extras?: Extras, maxBodyLength?: number): string {
        let markdownResult = '';

        const blockquoteRegex = /(<blockquote>\s*<\/blockquote>)/g;
        const parts = htmlString.split(blockquoteRegex);

        for (const part of parts) {
            if (part.match(blockquoteRegex)) {
                markdownResult += '>\n';
            } else if (part.trim()) {
                const processedPart = super.htmlToMarkdown(
                    part,
                    {
                        reportIDToName: extras?.reportIDToName ?? reportIDToNameMap,
                        accountIDToName: extras?.accountIDToName ?? accountIDToNameMap,
                        cacheVideoAttributes: extras?.cacheVideoAttributes,
                    },
                    maxBodyLength,
                );
                markdownResult += processedPart;
            }
        }

        if (markdownResult.endsWith('\n')) {
            markdownResult = markdownResult.replace(/\n+$/, '\n');
        }

        return markdownResult;
    }

Key Changes:

1. Split HTML Input: Use a regex (/<blockquote>\s*<\/blockquote>/g) to split the HTML string into parts, isolating empty <blockquote> tags.
2. Manual Blockquote Handling: For each empty <blockquote> tag, append >\n to the markdown output to ensure an empty quoted line.
3. Delegate to Parent Parser: Pass non-empty HTML parts to super.htmlToMarkdown for standard processing (e.g., converting <br /> to newlines).
4. Normalize Output: Trim redundant trailing newlines to ensure consistent markdown formatting.
5. Preserve Existing Functionality: Maintain the original behavior for htmlToText and truncateHTML methods, as they are unaffected by this issue.

This solution bypasses the ExpensiMark parser’s problematic handling of empty blockquotes while leveraging its functionality for other HTML tags, ensuring the desired markdown output.

What specific scenarios should we cover in automated tests to prevent reintroducing this issue in the future?

What alternative solutions did you explore? (Optional)

The alternative approaches can be fixed in the expensify-common package ExpensiMark class.

Workaround demo

Screen.Recording.1404-04-10.at.4.52.11.PM.mov

[melvin-bot]

Job added to Upwork: https://www.upwork.com/jobs/~021940992903128891410

[melvin-bot]

Triggered auto assignment to Contributor-plus team member for initial proposal review - @eVoloshchak (External)