handling name without leading /

pypdf/generic/_base.py

@@ -42,6 +42,7 @@
     read_non_whitespace,
     read_until_regex,
     str_,
+    WHITESPACES,
 )
 from ..errors import STREAM_TRUNCATED_PREMATURELY, PdfReadError, PdfStreamError
 
@@ -557,6 +558,7 @@ def write_to_stream(
 class NameObject(str, PdfObject):  # noqa: SLOT000
     delimiter_pattern = re.compile(rb"\s+|[\(\)<>\[\]{}/%]")
     surfix = b"/"
+    NUMBER = b"#"
     renumber_table: ClassVar[Dict[str, bytes]] = {
         "#": b"#23",
         "(": b"#28",
@@ -588,18 +590,24 @@ def write_to_stream(
         stream.write(self.renumber())
 
     def renumber(self) -> bytes:
-        out = self[0].encode("utf-8")
-        if out != b"/":
+        out = self.surfix
+        try:
+            val = self[:].encode("latin1")
+            # a sequence of any characters (8-bit values)
+        except UnicodeEncodeError as e:
+            logger_warning(repr(e), __name__)
+            val = self[:].encode("utf8")
+        if val[:1] != self.surfix:
             logger_warning(f"Incorrect first char in NameObject:({self})", __name__)
-        for c in self[1:]:
-            if c > "~":
-                for x in c.encode("utf-8"):
-                    out += f"#{x:02X}".encode()
+        else:
+            val = val[1:]
+        escaped = b"".join(WHITESPACES) + self.NUMBER
+        for c in val:
+            assert c  # null char is not allowed
+            if 0x21 <= c <= 0x7E and c not in escaped:
+                out += bytes([c])
             else:
-                try:
-                    out += self.renumber_table[c]
-                except KeyError:
-                    out += c.encode("utf-8")
+                out += self.NUMBER + f"{c:02X}".encode()
         return out
 
     @staticmethod

tests/test_generic.py

@@ -219,7 +219,7 @@ def test_name_object(caplog):
     caplog.clear()
     b = BytesIO()
     NameObject("hello").write_to_stream(b)
-    assert bytes(b.getbuffer()) == b"hello"
+    assert bytes(b.getbuffer()) == b"/hello"
     assert "Incorrect first char" in caplog.text
 
     caplog.clear()
@@ -228,10 +228,14 @@ def test_name_object(caplog):
     assert bytes(b.getbuffer()) == b"/DIJMAC+Arial#20Black#231"
     assert caplog.text == ""
 
+    b = BytesIO()
+    NameObject("#42").write_to_stream(b)
+    assert bytes(b.getbuffer()) == b"/#2342"
+
     b = BytesIO()
     NameObject("/你好世界 (%)").write_to_stream(b)
-    assert bytes(b.getbuffer()) == b"/#E4#BD#A0#E5#A5#BD#E4#B8#96#E7#95#8C#20#28#25#29"
-    assert caplog.text == ""
+    assert bytes(b.getbuffer()) == b"/#E4#BD#A0#E5#A5#BD#E4#B8#96#E7#95#8C#20(%)"
+    assert "UnicodeEncodeError" in caplog.text
 
 
 def test_destination_fit_r():