September 1999 - Zope

Re: [Zope-dev] unpiclkleable errors
by Evan Simpson 12 Sep '99

12 Sep '99

Robin Becker wrote: > >Is this reproduceable? Were you trying 0.01 before? > > > OK I checked further and it comes back when I install clean and goes > away when I delete the dir. I'm running Zope2 + CVS under Win95 osr2. I recommend that you delete the PythonMethod directory, restart Zope, delete any PythonMethod instances you have created, re-install PM 0.1.1, and re-create them. Don't use Zope's export facility or copy & paste to transfer them, just copy your code into an editor window or something. Version 0.0.1 stored hard references to functions and classes in its pickled representation, and when these go away in 0.1.1, Zope gets really unhappy. This *shouldn't* be a problem in the future upgrades, since one of the many improvements in 0.1.1 is a much cleaner pickle containing only basic types. That said, this *is* still alpha code. beware-the-toe-gremlins-ly y'rs Evan Simpson

1 0

DCOracle Compile Problem
by Andrew 12 Sep '99

12 Sep '99

When trying to compile the DCOracle package, I get the error below. I am using Oracle8i, version 8.1.5.0.0. I have checked ORACLE_HOME, LD_LIBRARY_PATH. I can connect with sqlplus. ProC is installed and configured. I know I am missing something simple, but its going right past me. TIA Drew gcc -fPIC -I/usr/local/oracle/product/8.1.5/rdbms/demo -I/usr/local/oracle/product/8.1.5/rdbms/public -I/usr/local/oracle/product/8.1.5/network/public -I/usr/local/oracle/product/8.1.5/plsql/public -DDCORACLE8 -g -O2 -I/usr/include/python1.5 -I/usr/include/python1.5 -DHAVE_CONFIG_H -c ./oci_.c In file included from ./oci_.c:562: Buffer.h:79: warning: static declaration for `PyImport_Import' follows non-static oci_8.c: In function `Logon': In file included from ./oci_.c:693: oci_8.c:95: warning: passing arg 4 of `OCIHandleAlloc' makes integer from pointer without a cast oci_8.c: In function `newLob': oci_8.c:446: warning: passing arg 4 of `OCIDescriptorAlloc' makes integer from pointer without a cast oci_8.c: In function `initoci_8': oci_8.c:483: warning: passing arg 3 of `OCIEnvInit' makes integer from pointer without a cast ./oci_.c: In function `_wrap_odescr': ./oci_.c:2595: warning: assignment from incompatible pointer type gcc -shared oci_.o -L/usr/local/oracle/product/8.1.5/lib/ -L/usr/local/oracle/product/8.1.5/rdbms/lib /usr/local/oracle/product/8.1.5/rdbms/lib/defopt.o /usr/local/oracle/product/8.1.5/lib/sscoreed.o /usr/local/oracle/product/8.1.5/lib/nautab.o /usr/local/oracle/product/8.1.5/lib/naeet.o /usr/local/oracle/product/8.1.5/lib/naect.o /usr/local/oracle/product/8.1.5/lib/naedhs.o -lnetv2 -lnttcp -lnetwork -lncr -lnetv2 -lnttcp -lnetwork -lclient -lvsn -lcommon -lgeneric -lmm -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lnetv2 -lnttcp -lnetwork -lncr -lnetv2 -lnttcp -lnetwork -lclient -lvsn -lcommon -lgeneric -lepc -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lclient -lvsn -lcommon -lgeneric -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lsocket -lnsl -lgen -ldl -lelf -laio -lposix4 -lsql -lm -lm -lnetv2 -lnttcp -lnetwork -lncr -lnetv2 -lnttcp -lnetwork -lclient -lvsn -lcommon -lgeneric -lmm -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lnetv2 -lnttcp -lnetwork -lncr -lnetv2 -lnttcp -lnetwork -lclient -lvsn -lcommon -lgeneric -lepc -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lclient -lvsn -lcommon -lgeneric -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lsocket -lnsl -lgen -ldl -lelf -laio -lposix4 -lsql -lm -lm -o oci_.so /usr/bin/ld: cannot open -lnetv2: No such file or directory collect2: ld returned 1 exit status make: *** [oci_.so] Error 1 $ cat Setup *shared* # Setup file for Oracle 8.0.4 (Tested for Solarus on Intel and Sparc). # Unfortunately, the libraries needed to link Oracle change from revision # to revision. # Oracle provides a make file for building "demo" programs in # $ORACLE_HOME/rdbms/demo/demo_rdbms.mk. I have been unable to # use this make file to build Python extensions. Adapting it is # not adequate, because it may change for new Oracle revs. # My approach is to use demo_rdbms.mk to build one of the # demo programs, look at the libraries linked, and start from there. :-( ORACLE_INCLUDES=$(ORACLE_HOME)/rdbms/demo -I$(ORACLE_HOME)/rdbms/public -I$(ORACLE_HOME)/network/public -I$(ORACLE_HOME)/plsql/public -I$(ORACLE_HOME)/precomp/demo ORACLE_L=-lnetv2 -lnttcp -lnetwork -lncr -lnetv2 -lnttcp -lnetwork -lclient -lvsn -lcommon -lgeneric -lmm -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lnetv2 -lnttcp -lnetwork -lncr -lnetv2 -lnttcp -lnetwork -lclient -lvsn -lcommon -lgeneric -lepc -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lclient -lvsn -lcommon -lgeneric -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lsocket -lnsl -lgen -ldl -lelf -laio -lposix4 -lsql -lm -lm ORACLE_LIBS=$(ORACLE_HOME)/lib/ -L$(ORACLE_HOME)/rdbms/lib -L$(ORACLE_HOME)/network/lib $(ORACLE_HOME)/rdbms/lib/defopt.o $(ORACLE_HOME)/lib/sscoreed.o $(ORACLE_HOME)/lib/nautab.o $(ORACLE_HOME)/lib/naeet.o $(ORACLE_HOME)/lib/naect.o $(ORACLE_HOME)/lib/naedhs.o $(ORACLE_L) $(ORACLE_L) oci_ oci_.c -I$(ORACLE_INCLUDES) -L$(ORACLE_LIBS) -DDCORACLE8 Buffer Buffer.c

2 1

RE: [Zope] etcUserFolder and multiple roles
by Chris McDonough 12 Sep '99

12 Sep '99

Oof. I knew you were going to come up with something elegant and totally embarrassing. I'll just take my ball and go home now. That works. Thanks! :) -----Original Message----- From: Rob Page [mailto:rob.page@digicool.com] Sent: Sunday, September 12, 1999 9:57 AM To: 'mcdonc(a)iqgroup.com' Cc: 'zope(a)zope.org' Subject: RE: [Zope] etcUserFolder and multiple roles > I have a folder which does not acquire any of its parents permissions, > we'll call it "Private". Private's parent is the root folder. <sniiiip> Wow, I had to sketch this out on a breakfast napkin... <whew> Why not define the role PrivateUsers in the root folder and only assign it in the Private folder? Then, let PrivateUsers see everything in the root folder. --Rob

1 0

RE: [Zope] etcUserFolder and multiple roles
by Rob Page 12 Sep '99

12 Sep '99

> I have a folder which does not acquire any of its parents permissions, > we'll call it "Private". Private's parent is the root folder. <sniiiip> Wow, I had to sketch this out on a breakfast napkin... <whew> Why not define the role PrivateUsers in the root folder and only assign it in the Private folder? Then, let PrivateUsers see everything in the root folder. --Rob

1 0

global upgrade of dtml tags from version 1 to 2
by Darran Edmundson 12 Sep '99

12 Sep '99

I wonder, has anyone written an external method to traverse the entire Zope database upgrading DTML tags from version 1 to version 2 syntax?  and <dtml-var title> are functionally equivalent but the latter is much more esthetically pleasing. Cheers, Darran. Darran Edmundson [Darran.Edmundson(a)anu.edu.au]

2 1

can't delete most recently added key from a Persistent dictionary
by Roman Milner 12 Sep '99

12 Sep '99

Hello. I have a problem trying to use the Persistent mixin class. I have a dictionary attribute that holds configurable fields. The fields are added like this: def store_new_fields(self, REQUEST): f = self.fields if hasattr(REQUEST, 'abstract_p'): f[string.strip(REQUEST.new_item_name)] = {'type': REQUEST.dtype, 'abstract_p': REQUEST.abstract_p} else: f[string.strip(REQUEST.new_item_name)] = {'type': REQUEST.dtype, 'abstract_p': 0} self.fields = f I then delete items thusly: def del_fields(self, the_fields): f = self.fields if type(the_fields) == type(""): del(f[the_fields]) else: for fname in the_fields: del(f[fname]) self.fields = f This mostly works fine except the most recently added new field can not be deleted. The del() call does not raise an error but when I look at self.fields.keys() after calling del_fields() on the most recently added field - it still exists. I'm sure there is something with the persistence stuff that I am missing. Any help would be greatly appreciated. Thanks, ^Roman

2 1

Pluggable brains question
by Sascha Matzke 12 Sep '99

12 Sep '99

Hello, are there any examples of extensionclasses for SQL-Methods... Which classes do I need the inherit. Is it possible to manage the permissions for the methods... Sascha -- .-> Sascha Matzke - sascha(a)bespin.de - http://www.bespin.de -. | Das Kapital ist die Flamme, der Faschismus das Benzin | | Anarchist Academy | `-- On this earth for 23 years, 333 days <----------------'

1 0

Re: [Zope] Some questions about ZCatalog
by Itai Tavor 12 Sep '99

12 Sep '99

Martijn Pieters wrote: > At 03:36 10/09/99 , Itai Tavor wrote: >>Just thought of a possible solution to the last problem... add a method to >>the object, named title_of_parent, which returns the parent's title. Tell >>the catalog to index title_of_parent. Take a vacation. >> >>Makes sense? Good solution? Comments? > > That is the recommended way. Good, thanks. Did that, working great, and now I know I'm not doing something stupid :) -- Itai Tavor -- "Je sautille, donc je suis." -- itavor(a)vic.bigpond.net.au -- - Kermit the Frog -- -- "What he needs now is understanding... and a confederate victory" -- -- Dr. Jacobi, Twin Peaks --

1 0

etcUserFolder and multiple roles
by mcdonc＠iqgroup.com 12 Sep '99

12 Sep '99

Zope is about the best thing to come around since nail clippers, but it is slowly driving me insane, nuts -- like.. daytrader. This message is about an etcUserFolder-based product. We all know etcUserFolder isn't a supported product. But what the heck, might as well ask. The worst that can happen is that I don't get awarded my Zope "cool users" pledge pin. Situation: I have a folder which does not acquire any of its parents permissions, we'll call it "Private". Private's parent is the root folder. Within the root folder, I have placed an HTTP-basic-auth-driven etcUserFolder containing two users, which retrieves its user information from a file named "etcUsers". We'll call the two users defined within this file "testuser-privileged" and "testuser-unprivileged." When creating the etcUserFolder, I assigned a default role to these users of "Normalusers." I have granted to the Normalusers role all permissions in the root directory. Now, on to the Private folder. I have placed within the Private folder a different HTTP-basic-auth-driven etcUserFolder containing (by virtue of its use of a different passwd-style file named "etcUsers-Private") a single user "testuser-privileged." Now, you'll note that this user's name matches that of the one defined within Root's etcUserFolder. This user's password (as defined in the "etcUsers-Private" file) matches the one given to the "testuser-privileged" account in the top-level etcUserFolder product. When creating the etcUserFolder in "Private", I assigned a default role to these users of "Privateusers." I have granted to the Privateusers role all permissions in the Private directory. The Private directory, otherwise, has all acquired permissions revoked and none granted to "Normaluser". Being a rational sort of guy, and knowing that HTTP basic auth compares the username and password on every browser-server transaction, I sort of figured that Zope would cope with this sort of setup by re-comparing the username and password when the user "testuser-privileged" switches between the Root and Private folders and would adjust the user's role and privileges accordingly when displaying elements and methods that make up a page. But this is not the case. What happens is this: when testuser-privileged hits the root folder, and enters in his password he can see index_html no problem. However, when he tries to visit the "Private" folder, Zope tries to re-authenticate him. It pops up another login request, and *even if he enters the correct password it will not let him properly view the page*. Now this in itself drove me completely *fruity* until I figured out *why* it wouldn't show him the page. It's trying to show him the index_html of the Private directory. That file is made up of DTML containing the standard_html_header & footer tags, which are acquired from DTML methods in the root folder. Because he has presumably "switched" roles from "Normalusers" to "Privateusers", and "Privateusers" has no permissions to standard_html_header and standard_html_footer, it does not allow him to see the page. In other words, it's not an authentication problem, it's a permissions problem. It seems that once he authenticates to the Private folder, he's stuck in the "Privateusers" role with no way out. And because the "Privateusers" role doesn't exist at the root folder, there's no way to assign it privileges to view items at this level. So its at this point that I have to admit I've been lying. It's not etcUserFolder that I've been using, it's a hacked-up etcUserFolder (smbUserFolder) that authenticates "over the wire" from the Linux box that this thing runs on to an NT domain. But it's materially the same as the etcUserFolder product, the roles logic hasn't changed (actually, only nine lines or so of code are different from product-to-product, and those are only the ones which actually go out to check if the login is OK). This shouldn't make a difference. So far, you're probably saying, "That's nice, Chris, but what's your point?" The problem is this. I *do not* want to maintain YAUAD (yet another user authentication database). I have one, we use it for everything, it's NT domain-based, it works, people don't have to remember seventeen different usernames and passwords thank you. Therefore, I don't want to use the normal "User Folder" product that ships with Zope. I *do not* want to move the site to NT and use the NTUserFolder Product (which claims to handle multiple roles per-user), it defeats the purpose of this project (which is to develop an intranet product that does not run on NT but can use NT authentication). I *do not* want to maintain local roles on a per-object basis (although this sounds a little more feasible, and could be done as a last resort) as long as I had a clear strategy to be able to do it. OK, I've realized now that I'm in way over my head here. I'm not even going to try to explain this problem any more, it's giving me a headache. Maybe I'm just doing it all wrong. Can anyone lend me a hand with a strategy to use the etcUserFolder product or similar to maintain nontrivial security requirements within Zope?

1 0

DCOracle Compile Problem
by Andrew 12 Sep '99

12 Sep '99

When trying to compile the DCOracle package, I get the following error. I am using Oracle8i, version 8.1.5.0.0. I have checked ORACLE_HOME, LD_LIBRARY_PATH. ProC is installed and configured. I know I am missing something simple, but its going right past me this last at night. TIA Drew gcc -fPIC -I/usr/local/oracle/product/8.1.5/rdbms/demo -I/usr/local/oracle/product/8.1.5/rdbms/public -I/usr/local/oracle/product/8.1.5/network/public -I/usr/local/oracle/product/8.1.5/plsql/public -DDCORACLE8 -g -O2 -I/usr/include/python1.5 -I/usr/include/python1.5 -DHAVE_CONFIG_H -c ./oci_.c In file included from ./oci_.c:562: Buffer.h:79: warning: static declaration for `PyImport_Import' follows non-static oci_8.c: In function `Logon': In file included from ./oci_.c:693: oci_8.c:95: warning: passing arg 4 of `OCIHandleAlloc' makes integer from pointer without a cast oci_8.c: In function `newLob': oci_8.c:446: warning: passing arg 4 of `OCIDescriptorAlloc' makes integer from pointer without a cast oci_8.c: In function `initoci_8': oci_8.c:483: warning: passing arg 3 of `OCIEnvInit' makes integer from pointer without a cast ./oci_.c: In function `_wrap_odescr': ./oci_.c:2595: warning: assignment from incompatible pointer type gcc -shared oci_.o -L/usr/local/oracle/product/8.1.5/lib/ -L/usr/local/oracle/product/8.1.5/rdbms/lib /usr/local/oracle/product/8.1.5/rdbms/lib/defopt.o /usr/local/oracle/product/8.1.5/lib/sscoreed.o /usr/local/oracle/product/8.1.5/lib/nautab.o /usr/local/oracle/product/8.1.5/lib/naeet.o /usr/local/oracle/product/8.1.5/lib/naect.o /usr/local/oracle/product/8.1.5/lib/naedhs.o -lnetv2 -lnttcp -lnetwork -lncr -lnetv2 -lnttcp -lnetwork -lclient -lvsn -lcommon -lgeneric -lmm -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lnetv2 -lnttcp -lnetwork -lncr -lnetv2 -lnttcp -lnetwork -lclient -lvsn -lcommon -lgeneric -lepc -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lclient -lvsn -lcommon -lgeneric -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lsocket -lnsl -lgen -ldl -lelf -laio -lposix4 -lsql -lm -lm -lnetv2 -lnttcp -lnetwork -lncr -lnetv2 -lnttcp -lnetwork -lclient -lvsn -lcommon -lgeneric -lmm -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lnetv2 -lnttcp -lnetwork -lncr -lnetv2 -lnttcp -lnetwork -lclient -lvsn -lcommon -lgeneric -lepc -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lclient -lvsn -lcommon -lgeneric -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lsocket -lnsl -lgen -ldl -lelf -laio -lposix4 -lsql -lm -lm -o oci_.so /usr/bin/ld: cannot open -lnetv2: No such file or directory collect2: ld returned 1 exit status make: *** [oci_.so] Error 1 $ cat Setup *shared* # Setup file for Oracle 8.0.4 (Tested for Solarus on Intel and Sparc). # Unfortunately, the libraries needed to link Oracle change from revision # to revision. # Oracle provides a make file for building "demo" programs in # $ORACLE_HOME/rdbms/demo/demo_rdbms.mk. I have been unable to # use this make file to build Python extensions. Adapting it is # not adequate, because it may change for new Oracle revs. # My approach is to use demo_rdbms.mk to build one of the # demo programs, look at the libraries linked, and start from there. :-( ORACLE_INCLUDES=$(ORACLE_HOME)/rdbms/demo -I$(ORACLE_HOME)/rdbms/public -I$(ORACLE_HOME)/network/public -I$(ORACLE_HOME)/plsql/public -I$(ORACLE_HOME)/precomp/demo ORACLE_L=-lnetv2 -lnttcp -lnetwork -lncr -lnetv2 -lnttcp -lnetwork -lclient -lvsn -lcommon -lgeneric -lmm -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lnetv2 -lnttcp -lnetwork -lncr -lnetv2 -lnttcp -lnetwork -lclient -lvsn -lcommon -lgeneric -lepc -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lclient -lvsn -lcommon -lgeneric -lnlsrtl3 -lcore4 -lnlsrtl3 -lcore4 -lnlsrtl3 -lsocket -lnsl -lgen -ldl -lelf -laio -lposix4 -lsql -lm -lm ORACLE_LIBS=$(ORACLE_HOME)/lib/ -L$(ORACLE_HOME)/rdbms/lib -L$(ORACLE_HOME)/network/lib $(ORACLE_HOME)/rdbms/lib/defopt.o $(ORACLE_HOME)/lib/sscoreed.o $(ORACLE_HOME)/lib/nautab.o $(ORACLE_HOME)/lib/naeet.o $(ORACLE_HOME)/lib/naect.o $(ORACLE_HOME)/lib/naedhs.o $(ORACLE_L) $(ORACLE_L) oci_ oci_.c -I$(ORACLE_INCLUDES) -L$(ORACLE_LIBS) -DDCORACLE8 Buffer Buffer.c

1 0