If there aren't any situations where ';' goes after '}', such
situations should probably not be introduced. Syntax should be
consistent. nocatch{} is then probably better than try{};, but I still
like code attributes solution the most. The feature of pointing out
exceptions that should be ignored is IMO pure code-generation thing,
you don't change any logic by using nocatch {} instead of just ignoring
the exception handling, you change only the generated code.
best regards,
Maybe nocatch { ... } would make that case even clearer. Am Dienstag, den 01.11.2011, 00:43 +0100 schrieb rastersoft:I don't think that try { ... } is a good idea, but try { ... }; is. Just removing the catch can result in involuntary errors, but if you have to choose between a ";" or a "catch", the probability of "forgetting" it is greatly reduced. El 31/10/11 23:22, pancake escribió:Hi On 31/10/2011, at 20:16, Aleksander Wabik<alex wabik gmail com> wrote:My 2 pennies:What about adding a code attribute like [IgnoreException] ? that would perform better than trycatchingIt seems the most logical option to me. Vala is a language that allows ignoring exceptions, and (as far as I remember, it was a while ago when I was writing my vala code) it generates code printing warning about uncaught exception each time the exception is thrown, but not caught. I guess that we should have two solutions (not one or the other, but both implemented): - command line switch like -Wno-exceptions - it would disable all warnings about uncaught exceptions at the compile time, and in the non-debug builds it would also cause not generating code for printing exception information if the exception is thrown;Looks an ugly solution to me. Some of those exceptions are important :)- and some code attribute, that could be used in code, if the author is 100% sure that in this particular case exception will never be thrown, or that it can be ignored (compile and run time behaviour the same as above).The option of allowing syntax like Try { ... } Without catch looks good to me, and probably cleaner than adding a code attribute. But i dont know of any lang that does this already.. So maybe its inconsistentBoth these features have the advantage that no syntax changes in the language are needed. best regards,On 31/10/2011, at 10:06, Xavier Bestel<xavier bestel free fr> wrote:On Sun, 2011-10-30 at 11:04 -0400, Sam Wilson wrote:Perhaps a better way to do this is like this: string[] test = new string[3]; for (int i = 0; i< 3; i++) { try { test[i] = kf.get_string(group, key); } catch (KeyFile.Error error) { // Do nothing } } if (!test[0]&& !test[1]&& !test[3]) return false; What do you think?Won't that interrupt the execution flow, i.e. if the first g_key_file_get_string() throws an exception, the other ones won't be executed ? Xav _______________________________________________ vala-list mailing list vala-list gnome org http://mail.gnome.org/mailman/listinfo/vala-list_______________________________________________ vala-list mailing list vala-list gnome org http://mail.gnome.org/mailman/listinfo/vala-list-- Mój klucz publiczny o identyfikatorze 1024D/E12C5A4C znajduje się na serwerze hkp://keys.gnupg.net My public key with signature 1024D/E12C5A4C is on the server hkp://keys.gnupg.net_______________________________________________ vala-list mailing list vala-list gnome org http://mail.gnome.org/mailman/listinfo/vala-list_______________________________________________ vala-list mailing list vala-list gnome org http://mail.gnome.org/mailman/listinfo/vala-list
-- Mój klucz publiczny o identyfikatorze 1024D/E12C5A4C znajduje się na serwerze hkp://keys.gnupg.net My public key with signature 1024D/E12C5A4C is on the server hkp://keys.gnupg.net
Attachment:
signature.asc
Description: PGP signature