Java中如何用List<String>匹配字符串敏感词？现有代码优化

Let's work through this — your current code has a couple of critical logic issues that are keeping it from detecting target words like "apple" and "hi". Let's break down what's wrong and build a working solution.

What's Wrong With the Current Code? #

1. Incorrect Stream Usage: You're calling words.stream() on a single String variable, which creates a stream of individual characters, not the words in your content. That's not what you want for word matching.
2. Misplaced Equality Check: anyMatch(badWordsList::equals) is checking if each character in your content equals the entire list of bad words — that's never going to be true, and it's not the match logic you need.

Solution 1: Basic Substring Matching (Case-Insensitive) #

If you just need to detect if the sensitive word appears anywhere in the content (even as part of a longer word, like "apple" in "pineapple"), use this straightforward approach:

import java.util.List;
import java.util.stream.Collectors;

private boolean checkWildcardFilter(String contentStripped, GuildTransformer guild) {
    // Convert content to lowercase for case-insensitive matching
    String lowerCaseContent = contentStripped.toLowerCase();
    
    // Get your bad words list and convert all to lowercase too
    List<String> badWordsList = replace(guild.getBadWordsWildcard());
    List<String> lowerCaseBadWords = badWordsList.stream()
                                                .map(String::toLowerCase)
                                                .collect(Collectors.toList());
    
    // Check if the content contains ANY of the bad words
    return lowerCaseBadWords.stream()
                            .anyMatch(lowerCaseContent::contains);
}

Solution 2: Exact Word Matching (Avoid Partial Hits) #

If you want to only match full words (so "apple" doesn't trigger on "pineapple", and "hi" doesn't trigger on "hiring"), use regular expressions with word boundaries:

import java.util.List;
import java.util.regex.Pattern;
import java.util.stream.Collectors;

private boolean checkWildcardFilter(String contentStripped, GuildTransformer guild) {
    String lowerCaseContent = contentStripped.toLowerCase();
    List<String> badWordsList = replace(guild.getBadWordsWildcard());
    
    // Preprocess bad words: lowercase + escape regex special chars (like *, ., ?)
    List<Pattern> badWordPatterns = badWordsList.stream()
                                                .map(String::toLowerCase)
                                                .map(badWord -> {
                                                    // Escape any regex-specific characters in the bad word
                                                    String escapedWord = Pattern.quote(badWord);
                                                    // Use \b to match full word boundaries
                                                    return Pattern.compile("\\b" + escapedWord + "\\b");
                                                })
                                                .collect(Collectors.toList());
    
    // Check if any pattern matches the content
    return badWordPatterns.stream()
                          .anyMatch(pattern -> pattern.matcher(lowerCaseContent).find());
}

Key Notes for Your Use Case #

• Case Insensitivity: Converting both the content and sensitive words to lowercase ensures matches regardless of how the words are capitalized (e.g., "Apple", "APPLE", "apple" all get caught).
• Regex Escaping: Pattern.quote() is crucial if your database has sensitive words with special characters (like "*" or "."), since those would otherwise be interpreted as regex syntax instead of literal characters.
• Performance Tip: If a guild's bad words don't change often, cache the compiled Pattern list for that guild instead of rebuilding it every time the method runs — this will speed up repeated checks.

Testing this with your sample sentence: "I will go and eat a grand apple with my new hi re." will correctly detect both "apple" and "hi" no matter where they appear, returning true to trigger your filtering.

内容的提问来源于stack exchange，提问作者Codixer