Fix incorrect self harm alarm warning on server while no self harm is detected

[dengcn]

Fixes #2333 (issue)

Description

This is to fix self harm alarm warning incorrectly logged on server when no self harm is detected

Please include a summary of the change.

Any new packages installed?

Give details How to test: 1) Use harmful words in text description inputs, and submit report, check server side log, you should see something like: "Self harm words detected: suicide,kill,knife,depressed"

2) Do not use any harmful words in text description inputs, and submit report, check server side log, you should not see any thing like "Self harm words detected"

Required followup work

Is there anything related to this that still needs to be done (ex: documentation changes).

Checklist:

• [ ] I have updated the azurescript.sh with any new environment variables and added them to the appsettings
• [X ] I have looked at my code on GitHub and it all looks good (ex: no random commented out code or console.logs)
• [ ] I have added and needed tests for my changes (in particular for new screens)
• [ ] I have added a comment to any confusing code
• [ ] I have added documentation to any modified front-end code. (Or added missing documentation)