Hi Holger
On 19.03.2011 01:54, Holger Hans Peter Freyther wrote:
On 03/18/2011 11:01 PM, Dennis Wehrle wrote:
Hello
Hi Dennis,
thank you very much for your patch. There are some formal issues with the patch and it would be nice if you could resolve them.
sure ;-)
- In general we don't keep commented out code around, please remove it.
- We avoid the // comments.
Can you explain why? (just for my interest)
- When introducing a new function that has known static linkage please use a namespace, e.g. not get_octlet_length but at least gsm_get_octet_length or put in 7bit as well.
Besides that your patch looks nice, but I do have two concerns. Your 160 char limit assumes the default GSM alphabet? In the future we might be able to encode and decode from different (e.g. chinese/unicode) alphabet. The same concern is for the new code in db.c that handles things specially. It would be nice if all of this could be hidden in libosmocore or such.
I have moved the septet length lookup to the gsm_utils.c. This function looks for characters defined at the "GSM 7bit default alphabet extension table" (GSM 03.38 6.2.1.1) which needs two septets. These characters are: ^ | € { } [ ] ~ \
cheers holger
Best Regards Dennis