Recent changes to patches

mtsystems Lifelines - Non-commercial open-source translation c2java

Dale Athanasias — Tue, 12 May 2015 03:43:35 -0000

Courtesy of Marco Trudel from mtSystems a C Source Code to Java Source Code Translation of Lifelines as requested by myself after reading the following Reddit post.

https://www.reddit.com/r/programming/comments/349vfo/c_to_java_translation_automatic_complete_correct/

https://mtsystems.ch
https://www.reddit.com/domain/mtsystems.ch/
https://www.reddit.com/user/marco2357

To: Dale
From: Marco Trudel
Sent: Mon 5/11/15 7:52 AM
To: Dale

Attachments: lifelines.zip (3.4 MB)

Hi Dale

Please find attached the translation of lifelines. Sorry it took so
long. We are very busy these days :)

Please note that you need Java 8 to run it.

But we actually didn't run/test it ourselves since it needs a database or some data or something.

From your message it sounded like you want it to help understand the C version. So in that case you probably won't run it and there's no point in us thoroughly testing the Java version.

Have fun! Make some advertising for us if you like the translation ;)
Marco

Message from: Dale

Hi,

Saw a post on Reddit and was wondering if you'd have time (no rush) to translate the lifelines genealogy software from C (which I do not under stand) to Java?

Thank you in advance
Dale

http://lifelines.sourceforge.net/

Add UTF-8 trail fix to manip_name

elsapo — Sat, 04 Oct 2008 02:53:43 -0000

manip_name uses trim, which is not UTF-8 aware. It would better call limit_width after trim, to fix any broken trailing UTF-8 multibyte sequences.

Update of German translation

Andreas Kirschbaum — Thu, 09 Aug 2007 18:02:56 -0000

Attached is an updated German translation file. It is based on current CVS.

Fix crash in allfam10c report

Andreas Kirschbaum — Thu, 18 May 2006 21:42:40 -0000

The report allfam10c crashes with the following error
message if it encounters a family with an unknown
husband and/or wife:

Runtime Error at line 54: fullname: the arg #1 must be
a person.

The attached patch fixes this problem by not calling
fullname() for missing individuals.

Allow to compile in Ansi C89 mode

Andreas Kirschbaum — Thu, 18 May 2006 21:28:27 -0000

The attached patch converts a //-style comment to
/*...*/. This makes the program to compile in Ansi C89
mode.

Fix buffer overflow in browse list

Andreas Kirschbaum — Thu, 18 May 2006 21:24:12 -0000

After selecting "Browse the persons in the database"
and getting a list of matching individuals, the program
crashes due to a buffer overflow if one individual has
a very long name/events string and also contains a REFN
field.

The problem is that the function person_display() in
src/liflines/show.c contains at least two bugs when
calculating "keyspace = max_keywidth() + 3; /*
parentheses & leading space */": the format string can
be " (i%s)", thus the constant must be 4 instead of 3.
Also, "key_of_record(indi)" seems to return the REFN
field (if present) which can be longer than
"max_keywidth()". Thus the " (i%s)" format can overflow
the scratch1 buffer for lone name/events strings.

The attached patch fixes the buffer overflow condition,
but it can truncate the NROF value in some cases.

wxLifeLines-0.0.2.tar.gz

elsapo — Sat, 22 Oct 2005 20:07:36 -0000

Tarball of Kevin Knaus' prototype wxWindows frontend
for lifelines.

Check result values of file operations

Andreas Kirschbaum — Sat, 22 Oct 2005 12:21:04 -0000

This patch adds checks to various file operations to
make sure the operation actually did succeed.

In many cases the result of fclose() (for files being
written to) was not checked. But this function actually
can fail: streams normally do buffering, therefore the
(buffered) data will not be written to disk until the
file is closed. Thus a possible I/O error is not
detected until fclose() is called.

Fix resource leaks

Andreas Kirschbaum — Sat, 22 Oct 2005 12:19:32 -0000

This patch fixes two (unrelated) resource leaks.

Fix off-by-one access to signal names list

Andreas Kirschbaum — Sat, 22 Oct 2005 12:18:18 -0000

This patch fixes two off-by-one errors accessing the
signal names list
in sig_msgs[].

The problem fixed is that without this patch sig_msgs
was declared as
sig_msgs[20], thus sig_msgs[0]..sig_msgs[19] was
allocated. But the
initialization code did assign values up to
sig_msgs[20]. Later on, while
printing the signal name, the last "valid" index was
i==ARRSIZE(sig_msgs)
(==20) which also did access the invalid entry
sig_msgs[20].