pdfinfo: improve the regex

src/ocrmypdf/pdfinfo.py

@@ -36,7 +36,16 @@ Encoding = Enum('Encoding',
                 'ccitt jpeg jpeg2000 jbig2 asciihex ascii85 lzw flate ' + \
                 'runlength')
 
-regex_remove_char_tags = re.compile(br"<char[^\/]+\/>")
+# Forgive me for I have sinned
+# I am using regular expressions to parse XML. However the XML in this case,
+# generated by Ghostscript, is self-consistent enough to be parseable.
+regex_remove_char_tags = re.compile(br"""
+    <char\b
+    (?:   [^>]   # anything single character but >
+        | \">\"  # special case: trap ">"
+    )*
+    />           # terminate with '/>'
+""", re.VERBOSE)
 
 FRIENDLY_COLORSPACE = {
     '/DeviceGray': Colorspace.gray,