pdf backend: Ignore diacritics when searching text

Ignore diacritics (accents, umlauts, diaeresis, etc.)
when searching in case insensitive mode (default mode), as
that would broaden the search giving more expected results.

Relevant flag was added in Poppler 0.73 and was implemented
in https://gitlab.freedesktop.org/poppler/poppler/issues/637

Issue #58

backend/pdf/ev-poppler.cc

@@ -1960,6 +1960,11 @@ pdf_document_find_find_text_with_options (EvDocumentFind *document_find,
 
 	if (options & EV_FIND_CASE_SENSITIVE)
 		find_flags |= POPPLER_FIND_CASE_SENSITIVE;
+#if POPPLER_CHECK_VERSION(0, 73, 0)
+	else    /* When search is not case sensitive, do also ignore diacritics
+	        to broaden our search in order to match on more expected results */
+		find_flags |= POPPLER_FIND_IGNORE_DIACRITICS;
+#endif
 	if (options & EV_FIND_WHOLE_WORDS_ONLY)
 		find_flags |= POPPLER_FIND_WHOLE_WORDS_ONLY;
 	matches = poppler_page_find_text_with_options (poppler_page, text, (PopplerFindFlags)find_flags);