September 2002 - Zope

Re: [Zope] Diacriticals in ZMI __S_O_L_V_E_D__
by Jaroslav Lukesh 30 Sep '02

30 Sep '02

Hi all, I am very sorry for bad URL (just started with membership at zope), so please take look at http://www.zope.org/Members/lukesh/diacriticals/ Regards JL. ---------- | Odesílatel: Jaroslav Lukesh <lukesh(a)k-net.cz> | Datum: 24. záøí 2002 18:23 | | Hi all that you have problems with diacriticals in ZMI. | | Here are set of files that I was change. All<head> sections was replaced by | | <head><meta http-equiv="Content-Type" content="text/html; | charset=windows-1250"><meta http-equiv="Content-Language" content="cs"> | | Take look at | http://www.zope.org/Members/lukesh/ | (empty now, waiting for review, so be patient)

1 0

Which Accounting applications to choose?
by Amit Babaria 30 Sep '02

30 Sep '02

Hello Everyone: Do you recommend any of these accounting applications for a company in Travel Industry domain? www.stfb.com www.sql-ledger.com www.osas.com www.compiere.com www.clarisys.com Let me know. Thanks for your help. ---------------------------------------------------------------------------- --------------------------------- Amit Babaria Director of Sales / Co-Founder CIGNEX Technologies, Inc T: 408.244.8558 F: 408.273.6785 M: 408.656.6493 E: amit(a)cignex.com U: www.cignex.com "Digital Strategies. Defined. Implemented!"

1 0

PROBLEM: Redirect works properly with IIS but not with Apache
by Vitor Varalonga 30 Sep '02

30 Sep '02

Greetings! Something really weird is happening with my Zope application. Since I wanted to develop a kind of authentication process to validate the access to each DTML document, I just created a python script that is called at the beginning of each document. The only thing this script does it to check if a certain session variable exists and if it a consistent value. If so, then the script does nothing more and returns to the DTML doc. If not it cause a redirection to a login page. Something like that: General DTML document: --------------------- <dtml-call expr="pyscripts.pyCheckLogin()"> <dtml-var standard_html_header> (...) <dtml-var standard_html_footer> This worked properly when I ran Zope behind IIS. Now that I transferred my Zope application to another server, this authentication philosophy stopped working and the only thing that changed is that Zope is no longer running behind IIS but behind Apache 2.0. I've exhaustedly debugged the code and this what happens: - As far as what it is concerned with the Python Script itself, everything seems to work properly; so when redirection is needed the python script executes it (or at least it tries - see next comment); - No matter what happens with the Python script (redirection or just return), the control will always return to the DTML doc which didn't happen while Zope ran behind IIS. Now... I've managed to detour this problem by changing the general structure of each DTML doc to the following: <dtml-if expr="pyscripts.pyCheckLogin()"> <dtml-var standard_html_header> (...) <dtml-var standard_html_footer> </dtml-if> Which means I also change the script to return 1 when authentication is valid and O when it's not (redirection). Still, this intrigues me a lot since there should be a similar behaviour no matter what Web Server Zope is running behind. Any suggestion to enlighten me on this X-File? thanks in advance, Vitor Varalonga

1 0

Storing a ZopeTime() within an MS Access DB
by McDonnell, Larry 30 Sep '02

30 Sep '02

Hi, I am running into a wall when it comes to populating the variable arch_date with ZopeTime. I have a web form that I need to generate an audit trail. Before I write the updated record, I dtml-in the active table and write the "old" information in an archive table. I added a date field (arch_date) to the archive table but I do not know how to save the variable in my ZSQL Method. First I setup the variable: <input type="hidden" name="arch_date"> <dtml-call expr="REQUEST.set('arch_date', ZopeTime())"> <dtml-var arch_date> Next, I write the records: <dtml-in capa_report_no_search> <dtml-call create_arch_capa_record> </dtml-in> <dtml-if expr="evaluator==''"> <p>The form requires the "Evaluator" field must be filled in for the record to be saved. Use the Back Arrow on your Web Browser <img src="/intranet/back_arrow"> to enter the information.</p> <dtml-else> <dtml-call "REQUEST.SESSION.set('originator', originator)"> <dtml-call "REQUEST.SESSION.set('evaluator', evaluator)"> <dtml-call expr="update_capa_record(REQUEST)"> My SQL looks like this: <dtml-if arch_date><dtml-if "arch_date==''">NULL<dtml-else> <dtml-sqlvar arch_date type=nb></dtml-if><dtml-else>NULL</dtml-if>) I have tried type=string but I see no errors and the table is left blank. Any help would be appreciated, thanks. Larry McDonnell Proton Energy Systems 50 Inwood Rd. Rocky Hill, CT 06067 (860) 571-6533 ext. 531 Email:lmcdonnell@protonenergy.com www.protonenergy.com

2 1

dtml document or dtml method?
by Jaroslav Lukesh 30 Sep '02

30 Sep '02

Hi, I have trouble with linking dynamically generated images. So now I totally dont understand differences between DTML documents and methods: [cite] The most important of these differences is that while a DTML Document is an object with properties of its own, a DTML Method is an object which inherits properties from its container. [/cite] So I have specified as DTML document(!) page segment which is called from nested segments that are DTML methods. This DTML document builds image menu from TinyTablePlus. But images are downloaded at each subfolder again and again because it have that URL: http://localhost:8080/www/www.k-net.cz/cs2/info/mapa/images/buttonky/M2.gif instead of http://localhost:8080/www/www.k-net.cz/cs2/images/buttonky/M2.gif Button's URL is constructed from that code in DTML Document (simplified): <dtml-in buttonky_L_table> <a href="<dtml-var link>"><img src="images/buttonky/<dtml-var letter>.gif"></a><br> </dtml-in> "letter" and "link" are columns in TinyTabel called buttonky_L_table I think that construction of right buttons URL have somewhat with acquisition, but do not have any idea how to do that. Maybe some complicated construction with PARENT, but I think that is not right way. Many thanks, J. Lukesh

1 0

override cookie espacing?
by John Doe 30 Sep '02

30 Sep '02

Hi, Is it possible to override the way Zope escapes it's cookies? When a cookie is set by Zope (eg. when you consult the help or by setting it with REQUEST.RESPONSE.setCookie), the cookie-value is enclosed by escaped quotes (eg \"mycookievalue\"). I would like to use quote-less cookievalues because it confuses some log-analyzing programs. Anyone encountered this problem (and preferrably it's solution)? Grts, Ronny __________________________________________________ Do you Yahoo!? New DSL Internet Access from SBC & Yahoo! http://sbc.yahoo.com

4 4

Re: [Zope] Session and objects
by Jens Vagelpohl 30 Sep '02

30 Sep '02

well, you will not be able to directly store those connections in the session. i would store them somewhere else (a tool for it in the ZODB maybe?) and retrieve the right connection based on a simple key you store in the session. jens On Monday, Sep 30, 2002, at 07:48 US/Eastern, Bo M. Maryniuck wrote: > On Monday 30 September 2002 13:44, you wrote: >> i'm not sure what you are trying to achieve here. it is always the >> easiest to store nothing but simple python data types (strings, >> sequences, dictionaries) in a session data container. a much better >> solution might be to store something that acts as a "key" and lets you >> find what you need in the real ZODB when you need it. > Well, I would like to store a Oracle DA connection to the session, > because I > need _different_ connection at the same time for different user. The > problem > is to logout user, if he is just close browser (not press button > "Logout" or > so). Then I want somehow hook this exception. There is no way > (perhaps) to do > it with ZODB, so I need put this connection to the session and the > connection > will die as soon as user closes the browser, itn't it? > > -- > Regards, Bogdan > > Everyone has a photographic memory. Some don't have film. >

3 4

Using __import__ in python product to access a local module attribute
by Jeff Kowalczyk 30 Sep '02

30 Sep '02

I'm trying to use __import__ to dynamically import a module in a customized formulator python product, and thereby access a singleton instance attribute of the primary class wtih a call to getattr. I can get it to work with hardcoding the module name, but I don't see what I'm missing when I try to use __import__. The code below is just an early version of a means to an end, and I'm wide open to better ways accessing the instance attribute. Metaprogramming babysteps, I guess... The MyExtClassInstance attribute in MyExtModule.py that I'm trying to access is defined like this: MyExtClassInstance = MyExtClass() justl like the widget and validator formulator classes are. MyExtModule.py is in the same product directory with my customized formulator filesystem product, call it 'FormulatorMyExt'. --- FieldRegistry.py ------------------------------------------------------ def initializeFieldForm(field_class, components=('widget','validator')): """ """ from Form import BasicForm from DummyField import fields form = BasicForm() override_form = BasicForm() tales_form = BasicForm() for component in components: if component not in ('widget','validator'): # when the seq is ('widget','validator','myExtModule') import MyExtModule # This works ... # but what I intend is a working form of either of these next lines.... # __import__(component.capitalize(), globals(), locals()) #__import__(getattr(Products.FormulatorMyExt,component.capitalize()), globals(), locals()) setattr(field_class, component, getattr(MyExtModule,component.capitalize()+'Instance')) I get errors like: NameError: global name 'Layout' is not defined (in the first case, and) NameError: global name 'Products' is not defined (in the second case) Does anyone see my mistakes here? __import__ isn't restricted to product python code is it? And obviously, I want to get that last line of hardcoded MyExtModule name out of there, but it obfuscated the problem to add that here. I can do it with a setattr call, I think, contingent on solving the former problem. Thanks for any help with using __import__, including the from list, or suggestion of a completely different way to do it as appropriate.

1 0

patch to ZopePageTemplate.py
by Mark McEahern 30 Sep '02

30 Sep '02

I patched /zope/lib/python/Products/PageTemplates/ZopePageTemplate.py to allow adding a Page Template to Zope via ZPublisher.Client. The reason I had to patch it is that when you add a Page Template via ZPublisher.Client, the text and title arguments are ignored and so the resulting Page Template has the default content rather than the content you specified. I've appended my patch as well as a script that shows how I was using ZPublisher.Client to add a Page Template below. I'd be very interested to hear any feedback on this approach. If the patch makes sense, how do I contribute it to Zope (if merely posting it here isn't sufficient)? Thanks, // mark Here's the patch: *** ZopePageTemplate.py.orig Sat Sep 28 09:31:44 2002 --- ZopePageTemplate.py.fixed Sat Sep 28 09:32:09 2002 *************** *** 316,326 **** file = REQUEST.form.get('file') headers = getattr(file, 'headers', None) if headers is None or not file.filename: ! zpt = ZopePageTemplate(id) else: zpt = ZopePageTemplate(id, file, headers.get('content_type')) self._setObject(id, zpt) try: u = self.DestinationURL() except: u = REQUEST['URL1'] --- 316,338 ---- file = REQUEST.form.get('file') headers = getattr(file, 'headers', None) if headers is None or not file.filename: ! # 20020928 MMcEahern ! # Patch to allow adding PageTemplate content via ! # ZPublisher.Client interface. ! if text is None: ! zpt = ZopePageTemplate(id) ! else: ! zpt = ZopePageTemplate(id, text) else: zpt = ZopePageTemplate(id, file, headers.get('content_type')) self._setObject(id, zpt) + # 20020928 MMcEahern + # Patch to allow adding PageTemplate title via + # ZPublisher.Client interface. + if title: + ob = getattr(self, id) + ob.pt_setTitle(title) try: u = self.DestinationURL() except: u = REQUEST['URL1'] ************ End of patch ************** Here's how I'm adding a Page Template: #!/usr/local/zope/bin/python __doc__ = \ """ This is an experimental script for adding Page Templates. You may need to adjust the shebang line above to match your Zope--or, of course, just call this explicitly with your Zope's python. Other hard-coded stuff you may want to adjust is noted. """ # -------------------------------------------------------------------------- --- # Standard library imports # -------------------------------------------------------------------------- --- import sys import time import os # -------------------------------------------------------------------------- --- # BEGIN: Hard-coded stuff you may need to adjust. # -------------------------------------------------------------------------- --- python_path = '/usr/local/zope/lib/python' username = None password = None base_url = "http://www.foobar.com/somefolder" # -------------------------------------------------------------------------- --- # END: Hard-coded stuff you may need to adjust. # -------------------------------------------------------------------------- --- # -------------------------------------------------------------------------- --- # Zope imports # -------------------------------------------------------------------------- --- # We can't import this until we modify sys.path. sys.path.insert(0, python_path) import ZPublisher.Client # -------------------------------------------------------------------------- --- # Add a page template # -------------------------------------------------------------------------- --- pt_url = "manage_addProduct/PageTemplates" # 2-step, get the manage_addProduct object for PT, then add it. url = os.path.join(base_url, pt_url) z = ZPublisher.Client.Object(url, username=username, password=password) try: # Use time.time() as a quick and dirty way to get a unique id. id = str(time.time()) # Distinguish title from id. title = "%s - title" % id text = "<div class='content'></div>" z.manage_addPageTemplate(id=id, title=title, text=text) except ZPublisher.Client.ServerError, e: redirect_prefix = "3" if not str(e).startswith(redirect_prefix): raise -

2 2

Need help understanding when threds get in my way
by Dario Lopez-Kästen 30 Sep '02

30 Sep '02

Hello! I need help in undertanding when I get hit by the threading machinery in Zope. I asked a question on irc.openprojects.org and I suddenly realised that I am not sure of what I am trying to do. The problem: for various reasons I need to cache several pieces of data, like the results form ZSQL methods, other objects, etc. Now, I understand that trying to store objects that, among other things are threadbound, is not a good idea, since they loose their context (?) when recalled later. The questions: a) in what other circumstances will caching objects be a bad idea? b) Is there a way to test, for a given object, if it is threadbound, either directly or indirectly? This I need mostly for logging purposes, since it will let me identify the places where our app is malfunctioning. c) if I discover or suspect that an object might be problematic, is there a way to only access that objects "data" and nothing else? Thanks for any insight (including the ona that states that I make no sense at all ;). Sincerely, /dario - -------------------------------------------------------------------- Dario Lopez-Kästen, dario(a)ita.chalmers.se IT Systems & Services System Developer/System Administrator Chalmers University of Tech.

2 1