[Bug #19902] Update the coderange regarding the changed region

1 files changed, 27 insertions, 0 deletions

diff --git a/string.c b/string.c
index deeed4a12a..18d03dc72e 100644
--- a/string.c
+++ b/string.c
@@ -2985,6 +2985,33 @@ rb_str_set_len(VALUE str, long len)
     if (len > (capa = (long)str_capacity(str, termlen)) || len < 0) {
         rb_bug("probable buffer overflow: %ld for %ld", len, capa);
     }
+
+    int cr = ENC_CODERANGE(str);
+    if (cr == ENC_CODERANGE_UNKNOWN) {
+        /* Leave unknown. */
+    }
+    else if (len > RSTRING_LEN(str)) {
+        if (ENC_CODERANGE_CLEAN_P(cr)) {
+            /* Update the coderange regarding the extended part. */
+            const char *const prev_end = RSTRING_END(str);
+            const char *const new_end = RSTRING_PTR(str) + len;
+            rb_encoding *enc = rb_enc_get(str);
+            rb_str_coderange_scan_restartable(prev_end, new_end, enc, &cr);
+            ENC_CODERANGE_SET(str, cr);
+        }
+        else if (cr == ENC_CODERANGE_BROKEN) {
+            /* May be valid now, by appended part. */
+            ENC_CODERANGE_SET(str, ENC_CODERANGE_UNKNOWN);
+        }
+    }
+    else if (len < RSTRING_LEN(str)) {
+        if (cr != ENC_CODERANGE_7BIT) {
+            /* ASCII-only string is keeping after truncated.  Valid
+             * and broken may be invalid or valid, leave unknown. */
+            ENC_CODERANGE_SET(str, ENC_CODERANGE_UNKNOWN);
+        }
+    }
+
     STR_SET_LEN(str, len);
     TERM_FILL(&RSTRING_PTR(str)[len], termlen);
 }