含撇号的Mapping搜索分析器无法正常工作

Got it, let's walk through why your quote-converting char filter isn't kicking in—this is a common gotcha, so we'll cover the most likely fixes:

1. You didn't link the char filter to an analyzer (the #1 mistake!) #

Defining the quotes char filter in your settings is only half the job. You need to wrap it into a custom analyzer and assign that analyzer to your text field(s). Without this, Elasticsearch ignores the char filter entirely.

Here's what a complete, working configuration looks like:

{
  "settings": {
    "number_of_shards": 1,
    "number_of_replicas": 0,
    "analysis": {
      "char_filter": {
        "quotes": {
          "type": "mapping",
          "mappings": [
            "\\u0091=>\\u0027",
            "\\u0092=>\\u0027",
            "\\u2018=>\\u0027",
            "\\u2019=>\\u0027",
            "\\u201B=>\\u0027"
          ]
        }
      },
      "analyzer": {
        "quote_normalized_analyzer": {
          "char_filter": ["quotes"], // Attach your char filter here
          "tokenizer": "standard",
          "filter": ["lowercase"]
        }
      }
    }
  },
  "mappings": {
    "properties": {
      "your_target_field": { // Replace with your actual field name
        "type": "text",
        "analyzer": "quote_normalized_analyzer" // Use the custom analyzer here
      }
    }
  }
}

2. Verify the analyzer is actually processing your text #

Don't guess—test it directly with the _analyze API to confirm the char filter is working. Run this request against your index:

POST /your_index_name/_analyze
{
  "analyzer": "quote_normalized_analyzer",
  "text": "It’s a test with curly apostrophes: ‘hello’ and ’world’"
}

Look at the returned tokens—you should see it's instead of it’s, and 'hello' instead of ‘hello’. If you don't, double-check your char filter mappings and analyzer setup.

3. You can't modify existing fields' analyzers (rebuild if needed) #

If you already created the index without linking the char filter to an analyzer, you can't retroactively change the field's analyzer. You'll need to:

• Create a new index with the correct settings/mapping
• Reindex your data from the old index to the new one

4. Double-check your Unicode escape sequences #

Your mapping uses \\u0091 (escaped backslash) which is correct for JSON—since JSON requires backslashes to be escaped. If you accidentally used a single backslash (\u0091), the mapping will break silently, so confirm that part of your config is intact.

Give these steps a go, and if you're still stuck, share your full index settings and mapping so we can troubleshoot further!

内容的提问来源于stack exchange，提问作者Adam Benson