pike issueshttps://git.lysator.liu.se/pikelang/pike/-/issues2009-04-16T14:11:39Zhttps://git.lysator.liu.se/pikelang/pike/-/issues/1903Multiple levels of <type> generated by AutoDoc.2009-04-16T14:11:39ZPeter BortasMultiple levels of <type> generated by AutoDoc.Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1903
Reported by Martin Nilsson, IDA <nilsson@pike.ida.liu.se>
The autodoc
*! @mixed @[arr]
*! @type array
becomes
<mixed name='@[arr]'><group><type><type><a...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1903
Reported by Martin Nilsson, IDA <nilsson@pike.ida.liu.se>
The autodoc
*! @mixed @[arr]
*! @type array
becomes
<mixed name='@[arr]'><group><type><type><array/></type></type>
which appears to by one level too many <type> tags. BTW, should autodoc
references persists in the refdoc XML format as above (or is the autodoc
syntax illegal and the extractor just didn't complain)?Pike 7.4https://git.lysator.liu.se/pikelang/pike/-/issues/1902Multiple levels of <value> generated by AutoDoc.2009-04-16T14:11:39ZPeter BortasMultiple levels of <value> generated by AutoDoc.Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1902
Reported by Martin Nilsson, IDA <nilsson@pike.ida.liu.se>
The autodoc
*! @int
*! @value 0
*! Throw an error.
gets converted into
<int><group><value><valu...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1902
Reported by Martin Nilsson, IDA <nilsson@pike.ida.liu.se>
The autodoc
*! @int
*! @value 0
*! Throw an error.
gets converted into
<int><group><value><value>0</value></value><text><p>Throw an error.
</p>
</text></group>
which appears to be one level of <value> too much.Pike 7.4https://git.lysator.liu.se/pikelang/pike/-/issues/1901Autodoc should complain when methods are declared in foreign modules.2013-12-11T18:45:40ZPeter BortasAutodoc should complain when methods are declared in foreign modules.Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1901
Reported by Martin Nilsson, IDA <nilsson@pike.ida.liu.se>
The following comment in builtin.c, generated from builtin.cmod
/*! @decl string String.trim_whites (strin...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1901
Reported by Martin Nilsson, IDA <nilsson@pike.ida.liu.se>
The following comment in builtin.c, generated from builtin.cmod
/*! @decl string String.trim_whites (string s)
*!
*! Trim leading and trailing spaces and tabs from the string @[s].
*/
gets extracted to
<docgroup homogen-name='String' homogen-type='variable'><doc><text><p>Trim
leading and trailing white spaces characters (space, tab, newline and
carriage return) from the string <ref>s</ref>.</p>
</text></doc>
<variable name='String'><type><string/></type></variable>
</docgroup>
which is clearly wrong.Pike 7.4https://git.lysator.liu.se/pikelang/pike/-/issues/1900Enable addition of top-level functions in autodoc.2009-04-16T14:11:39ZPeter BortasEnable addition of top-level functions in autodoc.Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1900
Reported by Martin Nilsson, IDA <nilsson@pike.ida.liu.se>
It is currently not possible to parse the master.pike.in for autodoc. The
main reason is that is must be po...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1900
Reported by Martin Nilsson, IDA <nilsson@pike.ida.liu.se>
It is currently not possible to parse the master.pike.in for autodoc. The
main reason is that is must be possible to add functions on top level.
These top level functions is currently marked with the unknown keyword
@global.Pike 7.4https://git.lysator.liu.se/pikelang/pike/-/issues/1894Indexing mappings with arrays fail2009-04-16T14:11:39ZPeter BortasIndexing mappings with arrays failImported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1894
Reported by Marcus Agehall <agehall@roxen.com>
```
When creating a mapping m=([ ({1,1}):1 ]) I am unable to index it.
m[({1,1})] returns 0 while m[indices(m)[0]] ret...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1894
Reported by Marcus Agehall <agehall@roxen.com>
```
When creating a mapping m=([ ({1,1}):1 ]) I am unable to index it.
m[({1,1})] returns 0 while m[indices(m)[0]] returns ({1,1}).
```Pike 7.2Henrik (Grubba) GrubbströmHenrik (Grubba) Grubbströmhttps://git.lysator.liu.se/pikelang/pike/-/issues/1890both 7.2.110 and 7.3.9 postgres is broken2009-04-16T14:11:39ZPeter Bortasboth 7.2.110 and 7.3.9 postgres is brokenImported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1890
Reported by Randall Randall, 3PSecure.com, Ltd. <randall@3psecure.com>
```
when attempting
object db = Sql.sql("postgres://user:@127.0.0.1/db");
this error occurs:
/...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1890
Reported by Randall Randall, 3PSecure.com, Ltd. <randall@3psecure.com>
```
when attempting
object db = Sql.sql("postgres://user:@127.0.0.1/db");
this error occurs:
/usr/local/pike/7.2.110/lib/modules/Sql.pmod/postgres.pike:129:Too many
arguments to big_query.
/usr/local/pike/7.2.110/lib/modules/Sql.pmod/postgres.pike:129:Expected:
function(string : int | object)
/usr/local/pike/7.2.110/lib/modules/Sql.pmod/postgres.pike:129:Got :
function(string, object(implements 65607) : void | mixed)
```
The error is exactly the same for 7.3.9, except for the version number.Pike 7.2https://git.lysator.liu.se/pikelang/pike/-/issues/1887Backtrace in Tools.AutoDoc.ProcessXML.moveImages.2009-04-16T14:11:39ZPeter BortasBacktrace in Tools.AutoDoc.ProcessXML.moveImages.Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1887
Reported by Martin Nilsson, IDA <nilsson@pike.ida.liu.se>
(Well, not really, since the error handler in AutoDoc hides backtraces...)
moveImages throws when processi...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1887
Reported by Martin Nilsson, IDA <nilsson@pike.ida.liu.se>
(Well, not really, since the error handler in AutoDoc hides backtraces...)
moveImages throws when processing autodoc/build/src/modules/Mird/module.xml
due to a docgroup lacking a child with name attribute (contains inherit
elements).Pike 7.4https://git.lysator.liu.se/pikelang/pike/-/issues/1885ODBC fails in Solaris 72009-04-16T14:11:39ZPeter BortasODBC fails in Solaris 7Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1885
Reported by Stefan Berndtsson, GU <stefan@nocrew.org>
Used software:
iODBC v3.0.5
Pike v7.0.345
FreeTDS v0.52-cvssnap (2001-07-17)
other end has Microsoft SQL-Serve...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1885
Reported by Stefan Berndtsson, GU <stefan@nocrew.org>
Used software:
iODBC v3.0.5
Pike v7.0.345
FreeTDS v0.52-cvssnap (2001-07-17)
other end has Microsoft SQL-Server 6.5.
iODBC and FreeTDS seems to work fine. I can get data from the very same query using the sample program in iODBC.
When I try to use Pike in Solaris 7, with the same ODBC, I get:
```
bash-2.01# pike
Pike v7.0 release 345 running Hilfe v2.0 (Incremental Pike Frontend)
> object db=Sql.sql("odbc://foga/foga");
Result: object
> array result=db->query("select count(*) from regis");
query = select count(*) from regis
exit_res_struct(): Freeing of HSTMT failed:
SQLError failed (3:SQL_NO_DATA_FOUND)
Unknown program: unknown function()
-:1: ___Foo4711()
>
```
The very same versions of all parts in Linux on a PowerPC:
```
$ pike
Pike v7.0 release 345 running Hilfe v2.0 (Incremental Pike Frontend)
> object db=Sql.sql("odbc://foga/foga");
Result: object
> array result=db->query("select count(*) from regis");
query = select count(*) from regis
Result: ({ /* 1 element */
([ /* 2 elements */
"":"11540",
".":"11540"
])
})
>
```
The odbc.ini and odbcinst.ini are identical between the systems (copied over).Pike 7.0https://git.lysator.liu.se/pikelang/pike/-/issues/1884Timezone modified Calendar equivalence error2009-04-16T14:11:39ZPeter BortasTimezone modified Calendar equivalence errorImported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1884
Reported by Fredrik Noring, Roxen Internet Software <noring@roxen.com>
Both equal calls are supposed to return 1:
```
Pike v7.2 release 75 running Hilfe v2.0 (Inc...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1884
Reported by Fredrik Noring, Roxen Internet Software <noring@roxen.com>
Both equal calls are supposed to return 1:
```
Pike v7.2 release 75 running Hilfe v2.0 (Incremental Pike Frontend)
> string fmt = "%dT%h:%m:%s", tm = "19980717T14:08:55";
Result: "%dT%h:%m:%s"
Result: "19980717T14:08:55"
> equal(Calendar.ISO.parse(fmt, tm)->set_timezone("UTC"),
>> Calendar.ISO.parse(fmt, tm)->set_timezone("UTC"));
Result: 1
> equal(Calendar.ISO.set_timezone("UTC")->parse(fmt, tm),
>> Calendar.ISO.set_timezone("UTC")->parse(fmt, tm));
Result: 0
```Pike 7.2https://git.lysator.liu.se/pikelang/pike/-/issues/1879pike/src/gc.c:73: conflicting types for `my_log2'2009-04-16T14:11:39ZPeter Bortaspike/src/gc.c:73: conflicting types for `my_log2'Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1879
Reported by Fredrik Noring, Roxen Internet Software <noring@roxen.com>
This change in Pike/7.0/src/block_alloc.h 2001-07-04:
```
$ cvs di -r 1.19 -r 1.20 src/bloc...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1879
Reported by Fredrik Noring, Roxen Internet Software <noring@roxen.com>
This change in Pike/7.0/src/block_alloc.h 2001-07-04:
```
$ cvs di -r 1.19 -r 1.20 src/block_alloc.h | grep my_log2
< extern int my_log2(unsigned INT32 x);
> extern int my_log2(size_t x);
```
gives the compilation error below because my_log2 is declared as
int my_log2(unsigned INT32 x) is stuff.[ch].
```
make[1]: Entering directory
`/home/distmaker/Distmaker/2.1/build/pike.du4.0/build'
Compiling /home/distmaker/Distmaker/2.1/build/pike.du4.0/pike/src/gc.c
/home/distmaker/Distmaker/2.1/build/pike.du4.0/pike/src/gc.c: In function
`init_marker_hash':
/home/distmaker/Distmaker/2.1/build/pike.du4.0/pike/src/gc.c:73:
conflicting types for `my_log2'
/home/distmaker/Distmaker/2.1/build/pike.du4.0/pike/src/stuff.h:16:
previous declaration of `my_log2'
/home/distmaker/Distmaker/2.1/build/pike.du4.0/pike/src/gc.c:73: warning:
extern declaration of `my_log2' doesn't match global one
WARNING: Compiler failure! Trying without optimization!
echo /home/distmaker/Distmaker/2.1/build/pike.du4.0/build/smartlink gcc -I.
-I/home/distmaker/Distmaker/2.1/build/pike.du4.0/pike/src -DORACLE_DEBUG
-I/home/distmaker/Distmaker/2.1/link/du4.0/include
-I/home/distmaker/Distmaker/2.1/build/pike.du4.0/pike/src
-I/home/distmaker/Distmaker/2.1/build/pike.du4.0/build -mieee
-DORACLE_DEBUG -I/home/distmaker/Distmaker/2.1/link/du4.0/include
-I/home/distmaker/Distmaker/2.1/build/pike.du4.0/pike/src
-I/home/distmaker/Distmaker/2.1/build/pike.du4.0/build -W -Wall
-Wpointer-arith -Wno-unused -c
/home/distmaker/Distmaker/2.1/build/pike.du4.0/pike/src/gc.c -o gc.o
/home/distmaker/Distmaker/2.1/build/pike.du4.0/pike/src/gc.c: In function
`init_marker_hash':
/home/distmaker/Distmaker/2.1/build/pike.du4.0/pike/src/gc.c:73:
conflicting types for `my_log2'
/home/distmaker/Distmaker/2.1/build/pike.du4.0/pike/src/stuff.h:16:
previous declaration of `my_log2'
/home/distmaker/Distmaker/2.1/build/pike.du4.0/pike/src/gc.c:73: warning:
extern declaration of `my_log2' doesn't match global one
make[1]: *** [gc.o] Error 1
make[1]: Leaving directory
`/home/distmaker/Distmaker/2.1/build/pike.du4.0/build'
make: *** [all] Error 2
```Pike 7.0Henrik (Grubba) GrubbströmHenrik (Grubba) Grubbströmhttps://git.lysator.liu.se/pikelang/pike/-/issues/1875Bug when using specific type2009-04-16T14:11:39ZPeter BortasBug when using specific typeImported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1875
Reported by David Hedbor, Idonex / Real Networks <david@hedbor.org>
When using an object declared as Image.Color obj calling obj->rgb() fails
at compile time. This d...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1875
Reported by David Hedbor, Idonex / Real Networks <david@hedbor.org>
When using an object declared as Image.Color obj calling obj->rgb() fails
at compile time. This does not occur with Pike 7.3. See attached program.Pike 7.0https://git.lysator.liu.se/pikelang/pike/-/issues/1861Hilfe returns escaped backslash as "\\"2009-04-16T14:11:39ZPeter BortasHilfe returns escaped backslash as "\\"Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1861
Reported by Diane Murray <dsm@muenster.de>
```
Pike v7.0 release 345 running Hilfe v2.0 (Incremental Pike Frontend)
> string dot = "\.";
Result: dot = "."
> string b...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1861
Reported by Diane Murray <dsm@muenster.de>
```
Pike v7.0 release 345 running Hilfe v2.0 (Incremental Pike Frontend)
> string dot = "\.";
Result: dot = "."
> string backslash_dot = "\\.";
Result: backslash_dot = "\\."
```Pike 7.0Henrik (Grubba) GrubbströmHenrik (Grubba) Grubbströmhttps://git.lysator.liu.se/pikelang/pike/-/issues/1859src/modules/_Crypto/lib/desdata.c must not define printf()2009-04-16T14:11:39ZPeter Bortassrc/modules/_Crypto/lib/desdata.c must not define printf()Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1859
Reported by Kai Voigt, 123.org <k@123.org>
[11:54 Grendel] "gcc 3.0 defines (as allowed by the ANSI C standard) the
printf function as a macro
[11:54...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1859
Reported by Kai Voigt, 123.org <k@123.org>
[11:54 Grendel] "gcc 3.0 defines (as allowed by the ANSI C standard) the
printf function as a macro
[11:54 Grendel] the declaration of this function after including stdio.h
breaks the compile"
[11:52 Grendel] all 7.x versions
[11:52 Grendel] in the src/modules/_Crypto/lib/desdata.c
[11:53 Grendel] line 58 needs to be removed for Pike to work with gcc 3.0Pike 7.0Henrik (Grubba) GrubbströmHenrik (Grubba) Grubbströmhttps://git.lysator.liu.se/pikelang/pike/-/issues/1858this_object() in constants2009-04-16T14:11:39ZPeter Bortasthis_object() in constantsImported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1858
Reported by Martin Stjernholm <mast@roxen.com>
Consider:
```
constant foo = ({this_object()});
int main()
{
werror ("%O %O\n", foo, foo[0] == thi...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1858
Reported by Martin Stjernholm <mast@roxen.com>
Consider:
```
constant foo = ({this_object()});
int main()
{
werror ("%O %O\n", foo, foo[0] == this_object());
}
```
This prints "({0})" in 7.2 and later. I expected either an error due to a
nonconstant expression used in constant, or, if it somehow works anyway, an
array with this_object() in it.
Interestingly enough, the array contains an object in 7.0, but it's not the
same as this_object().Pike 7.2https://git.lysator.liu.se/pikelang/pike/-/issues/1856format_mtime not consistent2009-04-16T14:11:39ZPeter Bortasformat_mtime not consistentImported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1856
Reported by Dan Nelson <dnelson_1901@yahoo.com>
format_mtime() should return a date in the format "YYYY-MM-DD hh:mm", but
YMD.pmod doesn't:
```
(dan@dan) /home/dan>...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1856
Reported by Dan Nelson <dnelson_1901@yahoo.com>
format_mtime() should return a date in the format "YYYY-MM-DD hh:mm", but
YMD.pmod doesn't:
```
(dan@dan) /home/dan> pike
Pike v7.2 release 107 running Hilfe v2.0 (Incremental Pike Frontend)
> Calendar.now()->format_mtime();
Result: "2001-07-09 12:31"
> Calendar.now()->day()->format_mtime();
Result: "20010709 00:00"
>
```
Fix:
```
RCS file: /cvs/Pike/7.2/lib/modules/Calendar.pmod/YMD.pmod,v
retrieving revision 1.29
diff -b -u -r1.29 YMD.pmod
--- YMD.pmod 2001/01/12 15:26:38 1.29
+++ YMD.pmod 2001/07/09 17:35:18
@@ -510,7 +510,7 @@
string format_mtime()
{
- return format_ymd_short()+" 00:00";
+ return format_ymd()+" 00:00";
}
```
string format_xtime()Pike 7.2https://git.lysator.liu.se/pikelang/pike/-/issues/1843Freeze if Gz.deflate->deflate used twice2009-04-16T14:11:39ZPeter BortasFreeze if Gz.deflate->deflate used twiceImported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1843
Reported by Andreas Lange, Roxen IS <lange@roxen.com>
```
megalon:~$ gdb ~/pike/7.2.101/bin/pike
GNU gdb 5.0
Copyright 2000 Free Software Foundation, Inc.
GDB is fre...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1843
Reported by Andreas Lange, Roxen IS <lange@roxen.com>
```
megalon:~$ gdb ~/pike/7.2.101/bin/pike
GNU gdb 5.0
Copyright 2000 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you
are
welcome to change it and/or distribute copies of it under certain
conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB. Type "show warranty" for details.
This GDB was configured as "sparc-sun-solaris2.8"...
(gdb) r
Starting program: /home/lange/pike/7.2.101/bin/pike
[New LWP 1]
[New LWP 2]
[New LWP 3]
[New LWP 4]
[New LWP 5]
Pike v7.2 release 101 running Hilfe v2.0 (Incremental Pike Frontend)
> function g = Gz.deflate()->deflate;
Result: deflate
> g("1");
Result: "x\234""3\4\0\0""2\0""2"
> g("2");
```
[Here it stops]
```
Program received signal SIGTERM, Terminated.
0xff36fff0 in thr_self () from /usr/lib/libthread.so.1
(gdb) bt
#0 0xff36fff0 in thr_self () from /usr/lib/libthread.so.1
#1 0xfe7d16f0 in do_deflate (buf=0xffbee968, this=0x35e828, flush=4)
at /home/lange/Pike/7.2/src/modules/Gz/zlibmod.c:156
#2 0xfe7d1920 in gz_deflate (args=1)
at /home/lange/Pike/7.2/src/modules/Gz/zlibmod.c:237
#3 0x49d34 in mega_apply (type=APPLY_STACK, args=1, arg1=0x2e2444,
arg2=0x1)
at /home/lange/Pike/7.2/src/interpret.c:1228
#4 0x48164 in eval_instruction_without_debug (pc=0x307172 "\232\"\177-")
at /home/lange/Pike/7.2/src/interpret_functions.h:1489
#5 0x4a140 in mega_apply (type=APPLY_STACK, args=0, arg1=0x0, arg2=0x2)
at /home/lange/Pike/7.2/src/interpret.c:799
The code around this is
154: THREADS_ALLOW();
155: ret=deflate(& this->gz, flush);
156: THREADS_DISALLOW();
```Pike 7.2Henrik (Grubba) GrubbströmHenrik (Grubba) Grubbströmhttps://git.lysator.liu.se/pikelang/pike/-/issues/1840postgresql 7 detection2009-04-16T14:11:39ZPeter Bortaspostgresql 7 detectionImported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1840
Reported by Kai Voigt, 123.org <k@123.org>
Hi,
Pike's configure requires postgres.h to be installed to detect
PostgreSQL 7 successfully. postgres.h only gets insta...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1840
Reported by Kai Voigt, 123.org <k@123.org>
Hi,
Pike's configure requires postgres.h to be installed to detect
PostgreSQL 7 successfully. postgres.h only gets installed when
doing a "make install-all-headers" in the postgresql source.
But this is not required for linking with the libraries.
Here's some mail from the FreeBSD port maintainer of PostgreSQL.
```
----------------------------------------------------------------
Date: Tue, 26 Jun 2001 17:47:21 +0200
From: Palle Girgensohn <girgen@partitur.se>
To: Kai Voigt <k@123.org>
Subject: Re: postgresql7 port of FreeBSD: make install-all-headers
```
Kai Voigt wrote:
>
> Hi,
>
> I wanted to install Roxen Webserver, a swedisch webserver with postgresql
> support. But its configure script was not able to find anything
important
> of your postgresql port on my FreeBSD system.
>
> postgres.h was not installed, which is the case when only "make install"
> is called, not "make install-all-headers".
postgres.h is not needed for anything else than building
postgresql itself. The dependency on postgres.h in Roxen is
false. Jhust remove any occurrence of #include <postgres.h> and
it will compile fine.
```
postgres.h was previously installed by postgres, but for no
good at all. some packages, like Roxen, have just added
#include <postgres.h> without checking if this is really
needed. It isn't.
```
/Palle
----------------------------------------------------------------Pike 7.0Henrik (Grubba) GrubbströmHenrik (Grubba) Grubbströmhttps://git.lysator.liu.se/pikelang/pike/-/issues/1839Type mixup between classes2009-04-16T14:11:39ZPeter BortasType mixup between classesImported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1839
Reported by Martin Stjernholm <mast@roxen.com>
The following program should prints "string" instead of the expected
"mixed":
```
class Foo
{
mixed m;
...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1839
Reported by Martin Stjernholm <mast@roxen.com>
The following program should prints "string" instead of the expected
"mixed":
```
class Foo
{
mixed m;
}
class Bar
{
inherit Foo;
string m;
Foo b;
mixed f() {return typeof (b->m);}
}
int main()
{
werror ("%O\n", Bar()->f());
}
```
(Test added to the testsuite in 7.3, but note that the problem exists in
7.2 too.)Pike 7.2Henrik (Grubba) GrubbströmHenrik (Grubba) Grubbströmhttps://git.lysator.liu.se/pikelang/pike/-/issues/1831missing dependencies in post_modules/Unicode2009-04-16T14:11:39ZPeter Bortasmissing dependencies in post_modules/UnicodeImported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1831
Reported by Dan Nelson <dnelson_1901@yahoo.com>
The Pike build fails in Unicode if started with make -j2. The following
patch fixes this.
```
cvs server: Diffing ...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1831
Reported by Dan Nelson <dnelson_1901@yahoo.com>
The Pike build fails in Unicode if started with make -j2. The following
patch fixes this.
```
cvs server: Diffing .
Index: Makefile.in
===================================================================
RCS file: /cvs/Pike/7.2/src/post_modules/Unicode/Makefile.in,v
retrieving revision 1.2
diff -b -u -r1.2 Makefile.in
--- Makefile.in 2001/06/21 12:42:18 1.2
+++ Makefile.in 2001/07/01 03:27:15
@@ -27,5 +27,8 @@
$(SRCDIR)/canonicals.h \
$(SRCDIR)/decompositions.h
```
+split.o: $(SRCDIR)/wordbits.h
+
+normalize.o: $(SRCDIR)/canonicals.h $(SRCDIR)/decompositions.h
@dependencies@Pike 7.2Henrik (Grubba) GrubbströmHenrik (Grubba) Grubbströmhttps://git.lysator.liu.se/pikelang/pike/-/issues/1824Hard links in config and master.pike2009-04-16T14:11:39ZPeter BortasHard links in config and master.pikeImported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1824
Reported by Thomas Svenson, Idonex AB / Sales Engineer <thomass@roxen.com>
Found a few hard links that is not so good to have around in Roxen.
In config files:
confi...Imported from http://bugzilla.roxen.com/bugzilla/show_bug.cgi?id=1824
Reported by Thomas Svenson, Idonex AB / Sales Engineer <thomass@roxen.com>
Found a few hard links that is not so good to have around in Roxen.
In config files:
configurations/Test: <var name='db_filename'>
<str>/http/roxen/roxen/sites/Test.sb/acdb</str> </var>
```
This is probably not a good idea since it then means you can't move roxen
to another dir...
server-2.2.120/pike/lib/pike/master.pike:#if
"/http/roxen/roxen/server-2.2.120/pike/lib/pike/"[0]!='¤'
server-2.2.120/pike/lib/pike/master.pike:
add_include_path("/http/roxen/roxen/server-2.2.120/pike/lib/pike//include");
server-2.2.120/pike/lib/pike/master.pike:
add_module_path("/http/roxen/roxen/server-2.2.120/pike/lib/pike//modules");
server-2.2.120/pike/lib/pike/master.pike:#if
"/http/roxen/roxen/server-2.2.120/pike/lib/pike/"[0]!='¤'
server-2.2.120/pike/lib/pike/master.pike:
foreach(get_dir("/http/roxen/roxen/server-2.2.120/pike/lib/pike/"), string
ver)
server-2.2.120/pike/lib/pike/master.pike:#if
"/http/roxen/roxen/server-2.2.120/pike/lib/pike/"[0]!='¤'
server-2.2.120/pike/lib/pike/master.pike:
base=combine_path("/http/roxen/roxen/server-2.2.120/pike/lib/pike/",sprintf("%s",tmp));
```Pike 7.2Henrik (Grubba) GrubbströmHenrik (Grubba) Grubbström