- Zope-Dev - Zope lists

Got "expand_all" to work in a tree tag with leaves
by Andrew Wilcox 18 Nov '99

18 Nov '99

I've been battling the tree tag all day to get expand_all to work, and it almost won, but then I hacked the source. {insert evil manical laughter here} Simple use of the tree tag, listing the titles of objects within folders:  <p><a href="<dtml-var URL0>?expand_all=1">Expand all</a></p>     Where listObjects is a simple DTML Method that lists the titles of the objects within the folder, with a little indenting:  <p><dtml-in "objectValues()"> <dtml-if "_['sequence-index'] != 0"><br></dtml-if>     <dtml-var title_or_id> </dtml-in>  Now, if you hit the "expand all", it expands all the folders except those at the lowest level, the ones that just contain (non-folder) objects. My first thought was that maybe it wasn't really considering the lowest level folders to be a branch, so I spent a while trying different functions for the "branches" attribute. That didn't work. Take a look at tpValuesId inside of TreeTag.py, which builds a list of all the expandable subitems for expand_all: try: items=getattr(self, branches)() except AttributeError: items=() for item in items: try: if getattr(item, branches)(): If I'm reading this right, the first call on branches gets the subitems which are considered a branch (defaulting to isPrincipiaFolderish over in ObjectManager). But then the second call on branches checks if the subitem itself has a non-zero number of branches. I commented out the second call there, and lo and behold! Suddenly my tree was fully expanding. So, what did I break? :) Andrew

1 0

Re: [Zope-dev] more on sessions in general.
by Anthony Baxter 18 Nov '99

18 Nov '99

>>> "Stuart 'Zen' Bishop" wrote > There is a need for someone to write ZSession or DumbSession that simply > stores the information in a in-memory dictionary (or var/session for ZEO > compatibility?), which would then define the API. Yep. Don't look at me - SQL Sessions do what I need. > How would you 'slot in' a session? > > Perhaps it just calls getSession(this(),REQUEST), which would be aquired > from higher-on-up. This method could defined in SQLSessionFolder, or just > a DTML/external method the site manager left in the root folder, or we could > patch OFS/Folder.py to give it a 'Session' property defining the class of the Right now, you simply <dtml-call Session>. The Session is the SQLSession object - it contains the SQL methods used by the session. It creates REQUEST.SESSION, which is a dictionary-like thing. > It would be nice if REQUEST.SESSION was filled in for you if the session > class has been defined, but someone else will need to comment on its > feasibility :-) That's what I do now. It seemed appropriate - it felt like it should be on an equivalent level to RESPONSE. Anthony -- Anthony Baxter <anthony(a)interlink.com.au> It's never too late to have a happy childhood.

1 0

ZServer FastCGI patch for IRIX (and others?)
by Phillip J. Eby 18 Nov '99

18 Nov '99

Below is a patch to let ZServer's FastCGI support work with FastCGI client implementations that don't properly close the STDIN sub-stream. On the IRIX/Apache/cgi-fcgi combo used where I work, the client side doesn't seem to know when it's done sending STDIN data, and thus the ZServer side simply hangs forever waiting for it. The patch below makes ZServer ignore STDIN altogether if there is no Content-Length, and also stop looking for more STDIN data if Content-Length or more bytes have already been read. --- FCGIServer.py Thu Nov 18 17:08:24 1999 +++ Zope2/ZServer/FCGIServer.py Thu Nov 18 17:05:02 1999 @@ -429,6 +429,8 @@ elif rec.recType == FCGI_PARAMS: if rec.contentLength == 0: # end of the stream self.remainingRecs = self.remainingRecs - 1 + self.content_length=string.atoi(self.env.get('CONTENT_LENGTH','0')) + if self.content_length==0: self.remainingRecs = self.remainingRecs - 1 else: self.env.update(rec.values) @@ -438,6 +440,8 @@ self.remainingRecs = self.remainingRecs - 1 else: self.stdin.write(rec.content) + self.content_length = self.content_length - rec.contentLength + if self.content_length<=0: self.remainingRecs = self.remainingRecs - 1 # read some filter data elif rec.recType == FCGI_DATA: @@ -448,7 +452,7 @@ # We've processed the record. Now what do we do? - if self.remainingRecs: + if self.remainingRecs > 0: # prepare to get the next record self.setInitialState()

1 0

HTMLFile() question
by Martijn Faassen 18 Nov '99

18 Nov '99

Hi there, In my Product code I do more or less the following: .. from Globals import HTMLFile form1 = HTMLFile('addForm', globals(), formname='form1') form2 = HTMLFile('addForm', globals(), formname='form2') form3 = HTMLFile('addForm', globals(), formname='form3') .. addForm.dtml contains something like this: .. The name of the form is:<dtml-var formname> .. After this I register several product classes with Zope, product1, product2 and product3 each given their own form (form1, form2 and form3) as their addForm. Now, upon trying to add product1, I'd like it to have 'form1' appear in its addForm. But instead 'form3' appears, the HTMLFile that was created last. This happens for form2 as well. Somehow all the earlier values are forgotten. Why does this happen? What am I doing wrong? Should I use globals() as an argument here? What other arguments could work in its place? If I don't use globals() but None instead, Zope won't find the addForm.dtml file at all, looking for it in 'lib/python' for some reason. I suppose I should dig more into the code, but could anybody enlighten me on what's going on? Regards, Martijn

2 2

more on sessions in general.
by Anthony Baxter 18 Nov '99

18 Nov '99

It seems to me that the Session type object as used in SQLSession is a useful general class of object (regardless of the backend storage). Is it worth making this a more general piece of Zope? Right now, you have to do a <dtml-call Session> at the start of the transaction to get the REQUEST.SESSION object created - it would be nice if this was done automagically (and done in a way that allowed you to slot in a different Session object if you wanted it...) Any thoughts? Anthony

4 3

Re: [Zope-dev] Easy DTML question
by Hannu Krosing 18 Nov '99

18 Nov '99

John Fohrman wrote: > > Thanks for your response. Yes, it works with a fixed string, that is why I'm > trying to do that. try something like   It _may_ be the same thing I encountered when setting cookies (in ver 1.0.3, not sure it is still in 2.x), namely that in some places variables are not evaluated In my case I had a rcursive loop that constructed a path from SQL database, and when setting a cookie with RESPONSE.setCookie('path',make_path) the source code of the method was set in the cookie resulting in broken page ;( using the trick with REQUEST.set solved the problem --------- Hannu

4 5

RE:Javascript help
by Anthony Pfrunder 18 Nov '99

18 Nov '99

Wow! Thanks to all you kind souls. I've got enough information to go from here and will let you know of my progress. I'll release the jsZDOM in a separate product as I think it is a generally useful thing to be able to remotely cause updating of Zope without causing a reload of an expensive (processing wise) page. (Never ceases to amaze me the speed and quality of the work that goes on in the Zope world - back from lunch and I have a written step by step set of instructions ;) Thanks all, Anthony Pfrunder PS: Monitor the Zwiki site for the scratch-pads of other interesting projects. I'm sure DC will move Zwiki onto the main Zope site fairly soon (hint hint) as it is a GREAT way to discuss projects.

3 2

XMLDocument shortcomings
by Melis G�bor 18 Nov '99

18 Nov '99

Having retrieved the current code from the CVS I see no difference from my old version. No matter, it was a useful exercise. I realize there are problems with the current code: - Special characters (&,<,>) are not escaped by the toxml() method - there is no way to specify encoding for toxml or for display purpose in general The first is a trivial issue, but I am not sure how (or where) to solve the second one. It gets pretty annoying if at the management interface one types: <?xml version="1.0" encoding="iso-8859-1"?> <root> öäå </root> but after clicking on the Change button gets: <?xml version="1.0"?> <root> Ã¶Ã€Ã¥ </root> As I understand it, the result is UTF-16. In most cases though, I would like to output it in 'iso-8859-1'. I think that XMLDocument should remember the encoding used to create the DOM and use that encoding to produce xml. Perhaps toXML() can take an encoding parameter and there is a getEncoding() method, too. So one can write: mydoc.toXML(encoding=mydoc.getEncoding()) I can imagine making this the default, as well. What do you guys think? And what is the policy on bringing new library dependencies? I am thinking about using more of xml-sig's stuff. Regards, Gabor

1 0

Interoperability between Zope and Fast.Serv FastCGI solution
by kamon.ayeva＠bureauveritas.com 18 Nov '99

18 Nov '99

Hi, I would like to know if DC or someone from the community has done or is planning to do some work for using Fast.Serv from Fast Engines (www.fastengines.com) as a FastCGI solution for running Zope with Microsoft IIS and/or Netscape servers. A while ago, I got a feedback from Fast Engines people themselves saying they were interested in working with DC for this interoperability. Just to know if it is on a TODO list and to have your feeling on this solution. Regards, Kamon

2 1

A few modifications to the tree tag
by Anthony Baxter 18 Nov '99

18 Nov '99

A few modifications, or proposed modifications, to the tree tag: I wanted to make the tree tag support my SQLSession product, so that it could store the tree state in the session, rather than in a cookie. I've created a patch against the current CVS that adds this - there's a new tree tag option, 'store_session', which when set will use the session to store the tree state. Yay. no more annoying cookie warning dialog boxes. Patch is at http://www.zope.org/Members/anthony/patches/ The second point, would a patch to allow a 'prefix' for the tree-[ecs] options be a useful thing? That way you can have multiple trees on the same page. So if prefix=foobar, the cookies/form values would be of the type 'foobartree-e' &c. Third point: is it worth allowing the tree icons +/- to be overridden? I thought I remembered someone had done this, but I can't remember it for sure, and I can't find it in my zope archives... ta, Anthony

5 5