shwenzhang/AndResGuard · error · AndrolibException

writeSpecNameStringBlock styleOffsetCount != 0 …

Error message

writeSpecNameStringBlock styleOffsetCount != 0  styleOffsetCount %d

What it means

writeSpecNameStringBlock re-reads the resource-table string pool while writing out a stripped spec-name block, and it only supports pools with no style spans. If styleOffsetCount (the style-offset array length in the chunk header) is non-zero, the writer cannot represent the styles and throws AndrolibException.

Solutions

1. Remove or simplify styled (spanned) strings from resource tables so the string pool has no styles.
2. Use a newer AndResGuard version that may handle style sections, or patch StringBlock to copy the styles section verbatim.
3. As a workaround, pre-process the arsc with androidwide tooling (aapt2) to strip spans before obfuscation.
4. If the styled strings are only in resources.arsc (not needed at runtime), rebuild resources without formatting spans.

// Inspect the string pool header before writing
int styleOffsetCount = readIntAt(chunkStart + 20);
if (styleOffsetCount != 0) {
  // skip the spec-name fast path or strip styles first
}

try { block.writeSpecNameStringBlock(...); } catch (AndrolibException e) { if (e.getMessage().contains("styleOffsetCount != 0")) { /* fall back to full (non-stripped) write path */ } else throw e; }

When it happens

Trigger: Calling writeSpecNameStringBlock on a resources.arsc string pool whose header declares styleOffsetCount != 0 (the chunk contains styled strings/span info).

Common situations: Running AndResGuard over APKs that contain styled strings in the resource table (HTML-like <b>/<i> spans in resource strings), typically from apps with rich text resources.

Understand the failure class

Background: UnsupportedOperationException and "is not supported" errors: when a library deliberately refuses a call — this error's family across 30 libraries.

Related errors

• Style data size is not multiple of 4
• writeSpecNameStringBlock
• writeSpecNameStringBlock
• writeTableNameStringBlock UTF-8 length is different name
• writeTableNameStringBlock UTF-16LE length is different name

AI-assisted analysis of shwenzhang/AndResGuard@e4df245d82 (2026-09-12). Data as JSON: /api/errors/74666f248693e6ea. Report an issue: GitHub.

Appendix: source

Thrown at AndResGuard-core/src/main/java/com/tencent/mm/androlib/res/decoder/StringBlock.java:115

      int size = (chunkSize - stylesOffset);
      if ((size % 4) != 0) {
        throw new IOException("Style data size is not multiple of 4 (" + size + ").");
      }
      block.m_styles = reader.readIntArray(size / 4);
    }
    return block;
  }

  public static int writeSpecNameStringBlock(
          ExtDataInput reader, ExtDataOutput out, Map<String, Set<String>> specNames, Map<String, Integer> curSpecNameToPos)
      throws IOException, AndrolibException {
    int type = reader.readInt();
    int chunkSize = reader.readInt();
    int stringCount = reader.readInt();
    int styleOffsetCount = reader.readInt();

    if (styleOffsetCount != 0) {
      throw new AndrolibException(String.format("writeSpecNameStringBlock styleOffsetCount != 0  styleOffsetCount %d",
          styleOffsetCount
      ));
    }

    int flags = reader.readInt();
    boolean isUTF8 = (flags & UTF8_FLAG) != 0;
    int stringsOffset = reader.readInt();
    int stylesOffset = reader.readInt();
    reader.readIntArray(stringCount);
    int size = ((stylesOffset == 0) ? chunkSize : stylesOffset) - stringsOffset;

    if ((size % 4) != 0) {
      throw new IOException("String data size is not multiple of 4 (" + size + ").");
    }
    byte[] temp_strings = new byte[size];
    reader.readFully(temp_strings);
    int totalSize = 0;
    out.writeCheckInt(type, CHUNK_STRINGPOOL_TYPE);

View on GitHub (pinned to e4df245d82)