Fix: handle GBK-encoded interface names and descriptions in ports polling (librenms#19528)

* Fix: handle GBK-encoded interface names and descriptions in ports polling

When SNMP returns GBK-encoded Chinese interface names (ifName/ifDescr),
they fail to decode as UTF-8 and become empty or garbled. This patch
adds multi-tier encoding support:

1. StringHelpers::inferEncoding() now prioritizes GB18030/GBK/GB2312
   fallbacks before Western encodings (Windows-1252/CP850), preventing
   mojibake when devices return non-UTF-8 interface descriptions.

2. Port polling/discovery explicitly applies inferEncoding() to
   ifName and ifDescr fields (consistent with ifAlias handling).

Fixes interfaces appearing with empty names like "ifName:  -> "
when devices use GBK encoding for interface aliases.

Co-authored-by: Copilot <copilot@github.com>

* Fix: prevent encoding detection conflicts between Western and GB codecs

Co-authored-by: Copilot <copilot@github.com>

* Fix: Improve prevent encoding detection conflicts between Western and GB codecs

Co-authored-by: Copilot <copilot@github.com>

---------

Co-authored-by: Copilot <copilot@github.com>

Author: Yish1

LibreNMS/Util/StringHelpers.php

@@ -107,6 +107,21 @@ public static function inferEncoding(?string $string): ?string
             return (string) $converted;
         }
 
+        // Detect GB multi-byte pattern: strict GB2312 range (0xA1-0xF7, 0xA1-0xFE)
+        // or GBK extended range (0x81-0xA0, 0x40-0x7E/0x80-0xFE). Count occurrences to avoid
+        // false positives from Western encodings like CP850 which may have single high-byte pairs.
+        $gbPatternCount = preg_match_all('/[\xA1-\xF7][\xA1-\xFE]|[\x81-\xA0][\x40-\x7E\x80-\xFE]/s', $string);
+        $hasGbPattern = $gbPatternCount >= 2;
+
+        if ($hasGbPattern) {
+            // GB pattern detected, prioritize GB family encodings
+            foreach (['GB18030', 'GBK', 'GB2312'] as $encoding) {
+                if (($converted = @iconv($encoding, 'UTF-8', $string)) !== false) {
+                    return (string) $converted;
+                }
+            }
+        }
+
         if ($charset !== 'Windows-1252' && ($converted = @iconv('Windows-1252', 'UTF-8', $string)) !== false) {
             return (string) $converted;
         }

includes/discovery/ports.inc.php

@@ -126,6 +126,8 @@
 foreach ($port_stats as $ifIndex => $snmp_data) {
     $snmp_data['ifIndex'] = $ifIndex; // Store ifIndex in port entry
     $snmp_data['ifAlias'] = StringHelpers::inferEncoding($snmp_data['ifAlias'] ?? null);
+    $snmp_data['ifName'] = StringHelpers::inferEncoding($snmp_data['ifName'] ?? null);
+    $snmp_data['ifDescr'] = StringHelpers::inferEncoding($snmp_data['ifDescr'] ?? null);
 
     // Get port_id according to port_association_mode used for this device
     $port_id = get_port_id($ports_mapped, $snmp_data, $port_association_mode);

includes/polling/ports.inc.php

@@ -587,6 +587,9 @@
             $this_port['ifName'] = $matches[1];
         }
 
+        $this_port['ifName'] = StringHelpers::inferEncoding($this_port['ifName'] ?? null);
+        $this_port['ifDescr'] = StringHelpers::inferEncoding($this_port['ifDescr'] ?? null);
+
         $polled_period = max($polled - $port['poll_time'], 1);
 
         $port['update'] = [];

tests/Unit/Util/StringHelperTest.php

@@ -45,6 +45,7 @@ public function testInferEncoding(): void
 
         $this->assertEquals('Øverbyvegen', StringHelpers::inferEncoding(base64_decode('w5h2ZXJieXZlZ2Vu')));
         $this->assertEquals('Øverbyvegen', StringHelpers::inferEncoding(base64_decode('2HZlcmJ5dmVnZW4=')));
+        $this->assertEquals('教科网IPv4', StringHelpers::inferEncoding(base64_decode('vcy/xs34SVB2NA==')));
 
         config(['app.charset' => 'Shift_JIS']);
         $this->assertEquals('コンサート', StringHelpers::inferEncoding(base64_decode('g1KDk4NUgVuDZw==')));