Normalizes SQL IN(?, ?, ...) statements to "in(?)" to reduce cardinality of span metrics using db.statement as an attribute

Author: swar8080

instrumentation-api-incubator/src/main/jflex/SqlSanitizer.jflex

@@ -5,6 +5,8 @@
 
 package io.opentelemetry.instrumentation.api.incubator.semconv.db;
 
+import java.util.regex.Pattern;
+
 %%
 
 %final
@@ -52,6 +54,9 @@ WHITESPACE          = [ \t\r\n]+
   // max length of the sanitized statement - SQLs longer than this will be trimmed
   static final int LIMIT = 32 * 1024;
 
+  private static final Pattern IN_STATEMENT_PATTERN = Pattern.compile("\\sin\\s*\\((?:\\s*\\?\\s*,?\\s*)+\\)", Pattern.CASE_INSENSITIVE);
+  private static final String IN_STATEMENT_NORMALIZED = " in(?)";
+
   private final StringBuilder builder = new StringBuilder();
 
   private void appendCurrentFragment() {
@@ -278,7 +283,8 @@ WHITESPACE          = [ \t\r\n]+
       builder.delete(LIMIT, builder.length());
     }
     String fullStatement = builder.toString();
-    return operation.getResult(fullStatement);
+    String normalizedStatement = IN_STATEMENT_PATTERN.matcher(fullStatement).replaceAll(IN_STATEMENT_NORMALIZED);
+    return operation.getResult(normalizedStatement);
   }
 
 %}

instrumentation-api-incubator/src/test/java/io/opentelemetry/instrumentation/api/incubator/semconv/db/SqlStatementSanitizerTest.java

@@ -271,7 +271,11 @@ public Stream<? extends Arguments> provideArguments(ExtensionContext context) th
           Arguments.of("select col from table1 as t1, table2 as t2", expect("SELECT", null)),
           Arguments.of(
               "select col from table where col in (1, 2, 3)",
-              expect("select col from table where col in (?, ?, ?)", "SELECT", "table")),
+              expect("select col from table where col in(?)", "SELECT", "table")),
+          Arguments.of(
+              "select 'a' IN(x, 'b') from table where col in(1) and z IN( '3', '4' )",
+              expect(
+                  "select ? IN(x, ?) from table where col in(?) and z in(?)", "SELECT", "table")),
           Arguments.of("select col from table order by col, col2", expect("SELECT", "table")),
           Arguments.of("select ąś∂ń© from źćļńĶ order by col, col2", expect("SELECT", "źćļńĶ")),
           Arguments.of("select 12345678", expect("select ?", "SELECT", null)),
@@ -298,6 +302,9 @@ public Stream<? extends Arguments> provideArguments(ExtensionContext context) th
               "delete from `my table` where something something", expect("DELETE", "my table")),
           Arguments.of(
               "delete from \"my table\" where something something", expect("DELETE", "my table")),
+          Arguments.of(
+              "delete from foo where x IN(1, 2, 3)",
+              expect("delete from foo where x in(?)", "DELETE", "foo")),
           Arguments.of("delete from 12345678", expect("delete from ?", "DELETE", null)),
           Arguments.of("delete   (((", expect("delete (((", "DELETE", null)),
 
@@ -307,6 +314,9 @@ public Stream<? extends Arguments> provideArguments(ExtensionContext context) th
           Arguments.of(
               "update `my table` set answer=42",
               expect("update `my table` set answer=?", "UPDATE", "my table")),
+          Arguments.of(
+              "update `my table` set answer=42 where x IN('a', 'b')",
+              expect("update `my table` set answer=? where x in(?)", "UPDATE", "my table")),
           Arguments.of(
               "update \"my table\" set answer=42",
               expect("update \"my table\" set answer=?", "UPDATE", "my table")),