PD_del_word

PD_del_word (PD_del_word) removes every occurrence of a keyword from a string. Input a block of text and a keyword, get the text back with all instances of that keyword gone. It's a one-trick node, but the trick is one you keep needing: scrub a banned tag out of a caption, strip a stray word from a pasted prompt, clean trigger words off a text dump before feeding it somewhere it doesn't belong.

How it works

Plain string replacement under the hood - find the keyword and delete it everywhere it appears. No regex, no fuzzy matching, no case-folding magic. That means it's exact and predictable, which is exactly what you want from a text cleaner, but it also means you're at the mercy of literal matching: dog won't touch doggy or Dog unless you type them as the keyword.

Inputs and output

• input_string - the text to clean (multiline, wireable).
• keyword - the exact string to delete everywhere it appears.

Output is a single result_string. That's the whole interface - no options for case sensitivity or replace-with. If you need find-and-replace with a substitution, that's the pack's JSON caption editor or a general text-replace node; this one only deletes.

Installing

Part of the 7BEII/Comfyui_PDuse pack. ComfyUI Manager → search "Comfyui_PDuse" → install, or:

cd ComfyUI/custom_nodes
git clone https://github.com/7BEII/Comfyui_PDuse
cd Comfyui_PDuse
pip install -r requirements.txt

Restart after. No models, no extra dependencies.

Gotchas

The keyword match is literal and case-sensitive, and deletion is global - every occurrence goes, not just the first. If your keyword is a substring of a word you want to keep (art inside artist), you'll delete more than intended; include surrounding spaces or punctuation in the keyword to pin it down. And deleting leaves the surrounding whitespace behind, so a removed mid-sentence keyword can leave a double space - cosmetic, but it'll show up in your prompt. If that bothers you, the pack's PD_empty_word (whitespace collapser) is a natural companion in the same text-cleanup chain.