Thanks for testing and for the report about ftplibftpt.cpp. I hadn't tried a non-curl build, but should have. Committed your fix. Thank you!

On 09/15/2013 06:02 PM, John Austin wrote:
I just checked the implementation with xulsword and it works great in Linux. Search results for modules with non-Latin keys are returning correct results now. I'm glad that the code was handy.

I did find that the recent change to ftplibftpt.cpp caused this compiler error with gcc 4.6.3:

make[1]: Entering directory `/home/dale/xulsword/mk-dev/Cpp/sword-svn/lib' /bin/bash ../libtool --tag=CXX --mode=compile g++ -DHAVE_CONFIG_H -I. -I../include -I../include -I/usr/local/include -DUSE_AUTOTOOLS -DUNIX -Dunix -D__unix__ -DGLOBCONFPATH=\"/usr/local/etc/sword.conf\" -D_FTPLIB_NO_COMPAT -O3 -fPIC -ftemplate-depth-25 -I/usr/local/include -I/usr/local/lib -DUSELUCENE -I/usr/local/lib -MT ftplibftpt.lo -MD -MP -MF .deps/ftplibftpt.Tpo -c -o ftplibftpt.lo `test -f '../src/mgr/ftplibftpt.cpp' || echo './'`../src/mgr/ftplibftpt.cpp libtool: compile: g++ -DHAVE_CONFIG_H -I. -I../include -I../include -I/usr/local/include -DUSE_AUTOTOOLS -DUNIX -Dunix -D__unix__ -DGLOBCONFPATH=\"/usr/local/etc/sword.conf\" -D_FTPLIB_NO_COMPAT -O3 -fPIC -ftemplate-depth-25 -I/usr/local/include -I/usr/local/lib -DUSELUCENE -I/usr/local/lib -MT ftplibftpt.lo -MD -MP -MF .deps/ftplibftpt.Tpo -c ../src/mgr/ftplibftpt.cpp -fPIC -DPIC -o .libs/ftplibftpt.o ../src/mgr/ftplibftpt.cpp:41:2: error: a storage class can only be specified for objects and functions
make[1]: *** [ftplibftpt.lo] Error 1
make[1]: Leaving directory `/home/dale/xulsword/mk-dev/Cpp/sword-svn/lib'
make: *** [all-recursive] Error 1

I had to change line 37 from "static struct MyProgressData" to just "struct MyProgressData" to get it to compile.

-John


On 09/15/2013 06:05 AM, Troy A. Griffitts wrote:
Dear John, thanks for the report.
Dear Jeff, thanks for the suggestion.

Looking through our code we had an old swunicod.[h,cpp] unit sitting
around which had a method to convert from UTF32 to UTF8. This unit was
only used in one obscure filter (RTFHTML).  I've removed this unit and
moved the code over as the implementation of the empty method which John
pointed out.

I didn't evaluate the logic of the actual implementation.  I am going
off the assumption that it worked :)  If we have any troubles with it
(John could you test your use case), then we can change the
implementation to use Jeff's suggestion.

Troy




On 09/14/2013 07:18 AM, John Austin wrote:
I've been testing the latest SWORD code for some time now, with
xulsword, on both Windows and Linux, and it's working great. But I did
just discover that Lucene searches in some non-versekey modules fail
because utilstr.cpp's wcharToUTF8() has not yet been coded. It's got
"TODO: finish this logic" in the code. So search returns incorrect
results in some modules (non-Latin dictionary and general book
modules). Maybe somebody's got a quick solution? Sorry to post another
issue so late- but I just discovered the problem yesterday and figured
it's better to mention it than not to.

But other than this, xulsword has been working very happily without
issue using SWORD 1.7, which is terrific. A new version of xulsword is
also to be released soon.

-John

_______________________________________________
sword-devel mailing list: [email protected]
http://www.crosswire.org/mailman/listinfo/sword-devel
Instructions to unsubscribe/change your settings at above page


_______________________________________________
sword-devel mailing list: [email protected]
http://www.crosswire.org/mailman/listinfo/sword-devel
Instructions to unsubscribe/change your settings at above page


_______________________________________________
sword-devel mailing list: [email protected]
http://www.crosswire.org/mailman/listinfo/sword-devel
Instructions to unsubscribe/change your settings at above page

Reply via email to