Fixing very incorrect parsing of "signature headers" block

The first set of headers is actually "signature headers" which _have somewhat different and conflicting set of key values_ than the main headers block.
The current way of parsing results in the second set overwriting unrelated keys from the first one.

You can see the "authoritative" procedure for parsing them by following the code at:
https://github.com/rpm-software-management/rpm/blob/d1075106bb315913574e1acac921058d23dd5130/tools/rpmdump.c#L205

Additionally I added some newer main tags that were previously not handled. You can see the current set of tags conveniently at
http://ftp.rpm.org/api/4.19.0/rpmtag_8h.html

rpmfile/headers.py

@@ -9,6 +9,32 @@
 from pprint import pprint
 from .errors import RPMError
 
+sigtags = {
+    "headerimage": 61,
+    "headersignatures": 62,
+    "size": 1000,
+    "lemd5_1": 1001,
+    "pgp": 1002,
+    "lemd5_2": 1003,
+    "sigmd5": 1004,  #md5 in C code but collides with md5 here
+    "gpg": 1005,
+    "pgp5": 1006,
+    "payloadsize": 1007,
+    "reservedspace": 1008,
+    "badsha1_1": 264,
+    "badsha1_2": 265, 
+    "signature": 267,
+    "rsaheader": 268,
+    "md5": 269,  #sha1header in C code
+    "longsigsize": 270,
+    "longarchivesize": 271,
+    "sha256": 273,
+    "filesignatures": 274,
+    "filesignaturelength": 275,
+    "veritysignatures": 276,
+    "veritysignaturealgo": 277,
+}
+
 tags = {
     "headerimage": 61,
     "headersignatures": 62,
@@ -17,14 +43,23 @@
     "headeri18ntable": 100,
     "sig_base": 256,
     "sigsize": 257,
+    "siglemd5_1": 258,
     "sigpgp": 259,
+    "siglemd5_2": 260,
+    "sigmd5": 261,
     "siggpg": 262,
+    "sigpgp5": 263,
+    "badsha1_1": 264,
+    "badsha1_2": 265,
     "pubkeys": 266,
-    "signature": 267,
+    "signature": 267, #dsaheader in C code
     "rsaheader": 268,
-    "md5": 269,
+    "md5": 269,  #sha1header in C code
     "longsigsize": 270,
     "longarchivesize": 271,
+    "sha256": 273,
+    "veritysignatures": 276,
+    "veritysignaturealgo": 277,
     "name": 1000,
     "version": 1001,
     "release": 1002,
@@ -307,9 +342,23 @@
     "autoinstalled": 5094,
     "identity": 5095,
     "modularitylabel": 5096,
+    "loaddigestalt": 5097,
+    "archsuffix": 5098,
+    "spec": 5099,
+    "translationurl": 5100,
+    "upstreamleases": 5101,
+    "sourcelicense": 5102,
+    "preuntrans": 5103,
+    "postuntrans": 5104,
+    "preuntransprog": 5105,
+    "postuntransprog": 5106,
+    "preuntransflags": 5107,
+    "postuntransflags": 5108,
+    "sysusers": 5109,
 }
 
 rtags = dict([(value, key) for (key, value) in tags.items()])
+rsigtags = dict([(value, key) for (key, value) in sigtags.items()])
 
 
 def extract_string(offset, count, store):
@@ -367,7 +416,7 @@ def extract_data(ty, offset, count, store):
         return "could not extract %s" % ty
 
 
-def _readheader(fileobj):
+def _readheader(fileobj, is_signature):
     char = fileobj.read(1)
     while char != b"\x8e":
         char = fileobj.read(1)
@@ -396,11 +445,11 @@ def _readheader(fileobj):
         entries.append(entry)
 
     store = fileobj.read(header_structure_size)
-    store
 
     headers = {}
+    tagsdict = rsigtags if is_signature else rtags
     for tag, ty, offset, count in entries:
-        key = rtags.get(tag, tag)
+        key = tagsdict.get(tag, tag)
         value = extract_data(ty, offset, count, store)
         headers[key] = value
     header_end = fileobj.tell()
@@ -412,9 +461,10 @@ def get_headers(fileobj):
     data = fileobj.read(lead.size)
     value = lead.unpack(data)
 
-    # Not sure what the first set of headers are for
-    first_range, first_headers = _readheader(fileobj)
-    second_range, second_headers = _readheader(fileobj)
+    #signature header
+    first_range, first_headers = _readheader(fileobj, True)
+    #main header
+    second_range, second_headers = _readheader(fileobj, False)
 
     first_headers.update(second_headers)