Message-ID: <gh-pgjdbc-pgjdbc-3757-c3172974185@github.com>
From: "rlbdv (@rlbdv)" <noreply+rlbdv@github.com>
To: "pgjdbc/pgjdbc" <noreply+pgjdbc-pgjdbc@github.com>
Date: Sun, 10 Aug 2025 23:47:19 +0000
Subject: Re: [pgjdbc/pgjdbc] issue #3757: PreparedStatement.toString() fails for bytea parameters with at least 42.7.7
In-Reply-To: <gh-pgjdbc-pgjdbc-3757@github.com>
References: <gh-pgjdbc-pgjdbc-3757@github.com>
List-Id: <gh-pgjdbc-pgjdbc.github.com>
X-GitHub-Author-Login: rlbdv
X-GitHub-Comment-Id: 3172974185
X-GitHub-Comment-Type: issue_comment
X-GitHub-Issue: 3757
X-GitHub-Repo: pgjdbc/pgjdbc
X-GitHub-Type: comment
X-GitHub-Url: https://github.com/pgjdbc/pgjdbc/issues/3757#issuecomment-3172974185
Content-Type: text/plain; charset=utf-8

I don't know pgjdbc well enough to have a good sense whether this is a plausible fix, but it does appear to resolve the immediate concerns.

In previous pgjdbc releases, a PGobject of type "bytea" and with a "\xHEX" value worked fine. It looks like what's happening now is that the string value is being passed directly to `toPGLiteral`, so this patch just adds a clause to handle it.

```
--- a/pgjdbc/src/main/java/org/postgresql/util/PGbytea.java
+++ b/pgjdbc/src/main/java/org/postgresql/util/PGbytea.java
@@ -182,6 +182,16 @@ public class PGbytea {
    * @throws IOException in case there's underflow in the input value
    */
   public static String toPGLiteral(Object value, SqlSerializationContext context) throws IOException {
+
+    if (value instanceof String) {
+      String str = (String) value;
+      StringBuilder sb = new StringBuilder(str.length() + 9);
+      sb.append("'");
+      sb.append(str);
+      sb.append("'::bytea");
+      return sb.toString();
+    }
+
     if (value instanceof byte[]) {
       byte[] bytes = (byte[]) value;
       StringBuilder sb = new StringBuilder(bytes.length * 2 + 11);
```