helpers: rework normalize_operator() to use str_to_utf8()

Instead of blindly assuming that we can take whatever string given as
valid UTF-8, we'll always attempt to convert from the current modem
charset into UTF-8. Before we were doing this for hex-encoded UCS2,
but not for example for GSM-7.

And due to the now applied GSM-7 conversion, the mf627a/mf627b +COPS
parsing unit tests are updated accordingly, because when converting
from an input string that contains byte 0x40 ('@' in UTF-8) as if it
were GSM-7, the 0x40 is taken as character '¡', encoded as 0xc2,0xa1
in UTF-8).

1 files changed, 1 insertions, 1 deletions

diff --git a/plugins/altair/mm-broadband-modem-altair-lte.c b/plugins/altair/mm-broadband-modem-altair-lte.c
index d26335f7..6aebbd93 100644
--- a/plugins/altair/mm-broadband-modem-altair-lte.c
+++ b/plugins/altair/mm-broadband-modem-altair-lte.c
@@ -1067,7 +1067,7 @@ modem_3gpp_load_operator_name_finish (MMIfaceModem3gpp *self,
                                            error))
         return NULL;
 
-    mm_3gpp_normalize_operator (&operator_name, MM_MODEM_CHARSET_UNKNOWN);
+    mm_3gpp_normalize_operator (&operator_name, MM_MODEM_CHARSET_UNKNOWN, self);
     return operator_name;
 }