February 2000 - Zope

Re. Scheduler product, anyone?
by Jerry.Spicklemire＠IFLYATA.COM 04 Feb '00

04 Feb '00

I had to do this just yesterday, and the urllib.py module made it absolutely painless. Here's the simplistic little script I came up with, whcih could be called by cron, or some other task manager. Later, Jerry S. --------------------------------------------------- """ urlgrabr.py is a simple script to hit an URL, retreive the generated output, and manipulate it, e.g. save it as a file, python urlgrabr.py http://server/folder/object file=filename.html based on parameters. Default is simply to request the URL, and exit, dumping all data. """ import urllib import sys import string c_url = sys.argv[1] c_action = None u_url = urllib.urlopen(c_url) if len(sys.argv) > 2 : c_action = sys.argv[2] if string.lower(c_action[:5]) == "file=" : f_action = open(c_action[5:], "w") # file name for l in u_url.readlines() : if string.strip(l[:-1]) : f_action.write(l) f_action.close()

1 0

Re: newbie questions about security
by danny shevitz 04 Feb '00

04 Feb '00

Thanks for the response. You were correct about the permission of /Root versus /Home. After making /Root hidden to Anonymouse I couldn't access /Home because standard_html_header and footer were being acquired. This again shows that permissions are determined by the permission of the parent in acquisition, not the child. I still have a question about setting up an /Home/img folder for images. I want to access the images from /Home but to not have them be directly viewable via a URL, e.g. /Home/img/image1. Is this possible? So far any combination of permissions that I have been able to come up with between /Home/img, /Home/img/image1, and /Home/img/index_html has the side effect that if /Home can access the image, so can the Anonymous URL /Home/img/image1. I can independently set the permission of /Home/img, just not the subimages. thanks, Danny __________________________________________________ Do You Yahoo!? Talk to your friends online with Yahoo! Messenger. http://im.yahoo.com

1 0

What's wrong with this code?
by Helge Tesdal 04 Feb '00

04 Feb '00

This is a part of the navbar on our site. When I try to access this directly through a folder (without the header), Zope seems to go into an seemingly infinite loop. Let's call the navbar 'test', and put it in the root folder. It contains the code below. Then make a folder testfolder, and call the navbar like this http://machine/testfolder/test On our machines, Zope goes into an infinite loop. By removing lines, it seems to recover from the "loop" more quickly. Seems like a n^x loop, where n is the number of lines. By removing the <dtml-in PARENTS skip_unauthorized> part it works. Is this somehow iterating all the lines in the dtml-method? Normally this is not a problem, as the header adds the <html> tags, which seems to make it work, but what worries me is that anybody can stop our site by trying to access this directly. Here's the code of the navbar: <dtml-unless expr="PARENTS[0]==PARENTS[-1]"> <dtml-if "meta_type == 'Artist'"> <dtml-call "REQUEST.set('artist_alias', id)"> <dtml-else> <dtml-in PARENTS skip_unauthorized> <dtml-if "meta_type == 'Artist'"> <dtml-call "REQUEST.set('artist_alias', id)"> </dtml-if> </dtml-in> </dtml-if> </dtml-unless>  <table border="0" width="100%" bgcolor="#000000" cellspacing="0" cellpadding="0"> <tr> <td width="100%"> <table border="0" width="100%" cellspacing=1 cellpadding=3> <tr bgcolor="#7D8DAF" align="center" valign="center"> <td width="100%">Manager-meny</td> -- Helge Tesdal tesdal(a)stud.ntnu.no

2 4

Renaming Properties
by James W. Howe 04 Feb '00

04 Feb '00

Is there some way to rename a property of a ZClass other than deleting the property and adding an new one of a different name? I've been working on a ZClass and have several instances in my database. I've now learned some more things about my domain and realize that some of the names of my properties should be different. Is there an easy (or moderately easy) way to accomplish this? Thanks. James W. Howe mailto:jwh@allencreek.com Allen Creek Software, Inc. pgpkey: http://ic.net/~jwh/pgpkey.html Ann Arbor, MI 48103

1 0

Newbie: Create items in Zope from external form?
by Chip Vanek 04 Feb '00

04 Feb '00

Hi, I have a simple question (I hope;) How do you create a DTML method item in Zope from an external form? I would like to allow users to create items in their home Zope folder using a post form from without logging into Zope first. The form itself could be served from the Zope server. I know that you can pass the login information in the URL but, what should the the DTML method look like to take variables from the posting form and writing a new DTML method in the users home folder? I would like to create a form and post to URL like below and get back a confirmation message from Zope that it worked OK. http://host.dom.com:8080/Members/user/createItem?__ac_name=uname&__ac_passwo rd=pwd Thanks for any pointers you provide. I have read all of the documentation on Zope.org, understand a fraction of it and have a site up and running. The mailing lists have been great to work through other problems I had but specific pointer would help now. Chip Vanek e-Mail: chip(a)upcast.com Office: 650-625-8929

1 0

Re: [Zope] How to list the attributes of an object?
by Tres Seaver 04 Feb '00

04 Feb '00

Robb Shecter wrote: > > Hi, > > Do you know offhand if your method shows proxy roles? Nope -- proxy roles are attributes of documents/methods, so my "method" object won't see whether they are present in other methods. > I'm debugging a problem > where proxy roles don't seem to be having any effect. I have to say that I find > all the various getRole methods confusing. Heh, failing to understand the ways in which proxy roles, normal roles, local roles, permission mappings, and acquisition play together is the motivation for this week's Zope Guru of the Week question. Jim Fulton (jim(a)digicool.com) gave me a hint about the topic at IPC8, but I haven't digested it yet: he pointed out that the "real place" of an object, rather than the location into which it is acquired, drives security. So, for instance, to use a method acquired from a parent/sibling folder in which you have no rights (even though you _do_ have them in the folder into which you acquire it), the _calling_ method/document must have a proxy role. I have messed this up at least twice! > It'd be nice to have a real > hyperlinked API with all classes, and all their methods explained... The Zope Quick Reference (ZQR) is a partial stab at that -- see http://www.zope.org/Members/ZQR. The Zope Documentation Project (ZDP) has other types of useful stuff: http://zdp.zope.org. Hope this helps! Tres. -- ========================================================= Tres Seaver tseaver(a)palladion.com 713-523-6582 Palladion Software http://www.palladion.com

1 0

Re: [Zope] Hooking Zope into Apache - please read
by Tres Seaver 04 Feb '00

04 Feb '00

Tom Deprez <tom.deprez(a)uz.kuleuven.ac.be> wrote: > > Hi all, > > No, this is not a question on how to do it. All the how-to's available > clearly explain this... it just an informational question and I hope > somebody can give me a straight answer. > > In reading Jeff Shelton's Python conference summaries (by the way, thanks > Jeff for doing this! Now, people who're not able (because of their job, > ...) to go to this conference can still pick something up! Great!) I read > the following : > > In Jim's opinion, the "preferred way" to hook into Apache will soon become > FastCGI rather than PCGI. > > Well, I can understand that FastCGI is better than PCGI, but I read (and > implemented) a how-to of anser (Using Apache with ZServer (Not Zope cgi)) > in which he explains the use of ProxyPass and ProxyPassReverse. > Well, this mehtod looks to me really handy. You use all the positive > things, because you can use Apache, but also the fast ZServer. > > So, now, you'll understand that I'm a little bit confused. What is now the > best way ... using FastCGI or ProxyPass....? What are the differences > between the two? What are the pro's and con's? Would you mind if I "borrowed" your question for a Zope Guru of the Week topic? (assuming it doesn't get answered quickly here). Or perhaps I should be "harvesting" stuff like this (including answers) from the list archives into "prefab" ZGotW questions, just as a searchable reference. -- ========================================================= Tres Seaver tseaver(a)palladion.com 713-523-6582 Palladion Software http://www.palladion.com

2 1

Hooking Zope into Apache - please read
by Tres Seaver 04 Feb '00

04 Feb '00

I have "harvested" Tom's question and Anthony's response from the list into a "synthetic" Zope Guru of the Week issue, at URL: http://zope.palladion.com/demos/ZGotW/synth_1 I made it "Synthetic" so that I could track and capture responses to the thread in in the ZGotW catalog, while keeping the "one question per week" idea stable. I will experiment with harvesting other "guru-like" threads over the next few weeks. Thanks! Tres -- ========================================================= Tres Seaver tseaver(a)palladion.com 713-523-6582 Palladion Software http://www.palladion.com

1 0

XMLDocument Enhancement
by James W. Howe 04 Feb '00

04 Feb '00

I posted a question recently about whether an XMLDocument could render the entire content of a specific XML tag. The situation I encountered involved the inclusion of HTML-ish markup (<p>, <br>, etc.) contained inside of another tag. For example: <entry> <question><p>To be or not to be?</p></question> </entry> Using the text_content('question') method resulted in no content being returned. In looking at the Node object, I noticed a method called toXML. Turns out that toXML is almost what I want, except that it produces the XML for the entire object. Using toXML and text_content as examples, I added the following code to the Node class in Node.py to facilitate returning XML content for a specific child node (one level deep): def nodeToXML(self, name=None, RESPONSE=None): """ Answer the XML for the specified child node name. If no node is specified, answer the XML for the entire node. """ if name is None: return self.toXML(self, 1, RESPONSE) else: result = "" for child in self: if child.getNodeName() == name: result = result + child.toXML() return result Used on XML such as above, the result of doing nodeToXML('question') would be "<p>To be or not to be?</p>" James W. Howe mailto:jwh@allencreek.com Allen Creek Software, Inc. pgpkey: http://ic.net/~jwh/pgpkey.html Ann Arbor, MI 48103

1 0

load_site.py troubles
by Maarten Slaets 04 Feb '00

04 Feb '00

hi all, i'm doing a weekly import of a certain site. i'm using load_site.py but, since its a weekly update, the files i'm importing are already there. the load_site gives me a 'The id %s is invalid - it is already in use.' error ofcourse. can anyone help me to make loadsite overwrite the existing objects? thanks ps: i'm using this: Zope version: Zope 2.0.1 (source release, python 1.5.1, linux2) Python version: 1.5.2 (#1, Sep 21 1999, 19:05:55) [GCC 2.8.1] System Platform: sunos5 -- Maarten Slaets Web Developer KPNQwest Belgium http://www.fonky.com/vcard/MaartenSlaets/

3 3