Error-catch Non-UTF8 DST header

Corrects a defect that would cause a non-utf8 header to prevent the reading of the body.

pyembroidery/DstReader.py

@@ -52,10 +52,13 @@ def process_header_info(out, prefix, value):
 
 def dst_read_header(f, out):
     header = f.read(512)
-    header_string = header.decode('utf8')
-    for line in [x.strip() for x in header_string.split('\r')]:
-        if len(line) > 3:
-            process_header_info(out, line[0:2].strip(), line[3:].strip())
+    try:
+        header_string = header.decode('utf8')
+        for line in [x.strip() for x in header_string.split('\r')]:
+            if len(line) > 3:
+                process_header_info(out, line[0:2].strip(), line[3:].strip())
+    except UnicodeDecodeError:  # The header contains non-utf8 information and omitted. See #83
+        pass
 
 
 def dst_read_stitches(f, out, settings=None):
@@ -90,7 +93,7 @@ def dst_read_stitches(f, out, settings=None):
         trim_distance = settings.get("trim_distance", trim_distance)
         clipping = settings.get('clipping', clipping)
     if trim_distance is not None:
-        trim_distance *= 10 # Pixels per mm. Native units are 1/10 mm.
+        trim_distance *= 10  # Pixels per mm. Native units are 1/10 mm.
     out.interpolate_trims(count_max, trim_distance, clipping)
 
 

pyembroidery/EmbCompress.py

@@ -1,6 +1,11 @@
 def expand(data, uncompressed_size=None):
-    compress = EmbCompress()
-    return compress.decompress(data, uncompressed_size)
+    emb_compress = EmbCompress()
+    return emb_compress.decompress(data, uncompressed_size)
+
+
+def compress(data):
+    size = len(data)
+    return bytearray([(size >> 0) & 0xFF, (size >> 8) & 0xFF, 0x02, 0xA0, 0x01, 0xFE]) + data
 
 
 class Huffman: