> I've looked at the patch and have found the following issues/comments: > > - There are now several warnings during the build, some because there are > still places (e.g. DefaultCatalog.cpp) where fingerprint is used as an int32 > instead of uint32. Some other warnings (during generation of the build tools) > complain about deprecated headers being used. There should really be no > warnings during the build of our own sources (maybe except for some > unavoidable cases). I did not notice the int32 problems. I'll check again. About the deprecated header, it is the hash_map that is used everywhere. Unfortunately, the standardized equivalent will only be available as part of C++0x, the next C++ standard, and it will ahve a slightly different api and name. I guess gcc2 will never get updated for it, and I don't think gcc4 provide a version of it that is not deprecated. If someone knows of one, tell me. > > - The resulting catalogs on the image have none of the locale-attributes > set (IIRC because of limitations in libroot_build.so). While that can be > fixed later, I expected that once I started the Locale preflet (which opens > french.catalog), the attributes of french.catalog would all be set. Language > and signature are alright now, but the fingerprint is still empty. The update_attributes function is disabled in the buildtool flavour of the catalogs. I think I did tht because libroot_build was not able to do it. I'll check what the problem is with the fingerprint. > > - The locale kit tests no longer build, due to the fingerprint related > changes. I assume that you haven't noticed because you do not include them in > the build. Please change your UserBuildConfig such that these tests are > copied onto the image, which will trigger them being built, too. woops. will do that... > > - The new method 'void escapeQuotedString(BString stringToEscape)' in > PlainTextCatalog.cpp is a no-op, because it uses a by-value parameter. It > should be escapeQuotedString(BString& stringToEscape). Additionally, I'd > suggest using the BString-method CharacterEscape() instead of the three > invocations of ReplaceAll(). Strangely enough, the resulting english.catkeys > file still contains the characters in escaped form (and continues to do so > after the invocations to escapeQuotedString() has been removed). Something is > fishy, there ;-) mh, i accidentally removed the & from parseQuotedChars, so they were never removed. woops again. > > - Style hint: in PlainTextCatalog::WriteToFile(), the code fragment where > textContent is filled should be wrapped such that the '<<' operator is at the > beginning of a line, not at the end. I have just checked and this does not > seem to be mentioned in our styleguide, but nevertheless, I believe that's > how it is done in most of our code. > ok, this one is easy to fix. > cheers, > Oliver > > So, i'll work on all that. What can we do about the deprecated hash_map header ? -- Adrien Destugues / PulkoMandy Elève ingénieur ENSSAT EII1- www.enssat.fr GSoC student for Haiku - http://haiku-os.org GrafX2 project team - http://code.google.com/p/grafx2