- Zope - Zope lists

RE: [Zope] - catching DB errors
by Paul Everitt 25 Jan '99

25 Jan '99

Try tinkering with the standard_error_message document. It gets passed in the exception information so you can do an  to handle exceptions. --Paul > -----Original Message----- > From: Brian Hooper [mailto:brian@garage.co.jp] > Sent: Sunday, January 24, 1999 10:10 PM > To: zope(a)zope.org > Subject: [Zope] - catching DB errors > > > Hi - > > I was wondering what is the recommended way to catch database errors > before they get to where the user can see them - I am doing a lot of > stuff using DTML scripting w/ SQLMethods, which works fine, but if > possible I want to be able to shield the user from having to see > 'Zope Unavailable' and/or a big traceback. > > Is there any way to do this from DTML, or do I have to channel things > through a try: except in an ExternalMethod? > > Does anyone have a clever solution for this? > > (By the way, I'm kind of interested in if the error messages can be > collected and customized in one place - as I'm working in Japan, I > want to change all of the server messages so they come out in > Japanese. > I expect there are probably other people that are interested in > internationalization issues also - what have you all been doing?) > > --Brian Hooper > >

1 0

- ODBC troubles
by Jeff Nielsen 25 Jan '99

25 Jan '99

I'm rather new to Zope and am having trouble getting ODBC connectivity to work. I've downloaded the file "ZODBCDA-3.0.3-win32-x86-python-1.5.1.tar.gz" from the web site, uncompressed it into a file named "sql.pyd" and placed that file in the root of my Zope installation direction. When I try to add a database connection, I don't see a choice of "Aqueduct ODBC Database Connection" or anything similar. I'm not sure that I'm uncompressing the file correctly. I'm using WinZip and it prompts me for the name of the file to uncompress. Any feedback, help, or an uncompressed copy of sql.pyd would be greatly appreciated. Thanks, Jeff Nielsen jeffniel(a)ix.netcom.com

1 0

- catching DB errors
by Brian Hooper 25 Jan '99

25 Jan '99

Hi - I was wondering what is the recommended way to catch database errors before they get to where the user can see them - I am doing a lot of stuff using DTML scripting w/ SQLMethods, which works fine, but if possible I want to be able to shield the user from having to see 'Zope Unavailable' and/or a big traceback. Is there any way to do this from DTML, or do I have to channel things through a try: except in an ExternalMethod? Does anyone have a clever solution for this? (By the way, I'm kind of interested in if the error messages can be collected and customized in one place - as I'm working in Japan, I want to change all of the server messages so they come out in Japanese. I expect there are probably other people that are interested in internationalization issues also - what have you all been doing?) --Brian Hooper

1 0

- Problem evaluating  in DTML
by skip＠calendar.com 24 Jan '99

24 Jan '99

I can't seem to access  directly in my DTML file though  works just fine. I'm using the DocumentTemplate that came with Zope 1.9.0. I'm calling the document template instance with REQUEST as one of the parameters. Any suggestions? Thanks, Skip Montanaro | Mojam: "Uniting the World of Music" http://www.mojam.com/ skip(a)calendar.com | Musi-Cal: http://concerts.calendar.com/ 518-372-5583

1 0

- ANNOUNCE: TinyTable Beta 1
by Ty Sarna 24 Jan '99

24 Jan '99

ANNOUNCING TinyTables Beta 1 (0.7.3) TinyTable is a product designed to manage a small amount of tabular data. It's intended to fill the gap between a Tabula or an Z SQL Methods accessed SQL table, which are overkill for many tasks, and folder token properties, which allow only a single "column". TinyTable also makes it possible to look up an item within the list, or to return a subset of the list rows where columns equal particular values. Data within a TinyTable is directly editable within Zope, presented as rows of delimited text. The delimited character can be chosen to suit the data being used. Columns may be strings, ints, floats, or dates. TinyTable supports pluggable brains, much like Z SQL Methods do. This is the first Beta of TinyTable. It's believed to be stable, and the internal data format isn't expected to change incompatably before the final release. However, it's a Beta -- no guarantees. This version is "read-only", in the sense that the data can only be changed all at once, using the management interface. Future versions will add methods for setting individual rows and fields within a row, deleting rows, adding rows, etc. Have fun, and please report any problems you encounter!

2 3

- *IMPORTANT* Licensing Issues
by Hadar Pedhazur 24 Jan '99

24 Jan '99

I've just posted a FAQ to the License section. You can jump straight to it via: http://www.zope.org/License/FAQ. It's _really_ long, so I apologize in advance. The really meaty (meaning _important_) part is toward the end (though it probably is close to 40% by volume). The point is this, if we want wide distribution in the Linux community, via the standard CD's (RedHat, Debian, Caldera, Suse, etc.), we will likely have to drop our attribution clause. By dropping it, I mean at the very least making it voluntary rather than mandatory. The end of the license FAQ deals with this issue in some detail (without arriving at any good conclusion!). So, since _you_ are the community, and not the distributors, I'm asking for some help on this. Please don't send any comments before reading the FAQ. I don't need a rehash of points already discussed to death, especially if they are passionate :-(. What I'd like in particular, are any fresh thoughts on the subject that could help sway us one way or the other. Finally, I don't want to pollute this list with a license flame war. So please, send your responses to mailto:zope-license@zope.org.

1 0

Re: [Zope] - Re: XML-RPC
by skip＠calendar.com 24 Jan '99

24 Jan '99

Christopher> Why not use text/x-xml-rpc? XML has a billion uses, that Christopher> honestly, it will eventually end up as: xml/rpc instead, Christopher> most likely :-) Christopher> What's the status of text/xml as a standard MIME type? In reverse order: I suppose text/xml is either already approved or well on its way to approval. It's too hot for the approval folks to ignore. All text/xml says is that XML is the "syntax" of the message. Interpreting the semantics of the message is up to the receiver. Everybody who thinks it's okay to start annotating the media type with semantic information should take a quick look at their local mime.types file. Note that the types there don't say anything about how to interpret the contents, just what the format of the contents are. There is an image/gif type which we are all familiar with. There are not image/gif-car, image/gif-naked-bodies and image/gif-xy-plot types. The media type of a request or a response is just that, nothing more. Skip Montanaro | Mojam: "Uniting the World of Music" http://www.mojam.com/ skip(a)calendar.com | Musi-Cal: http://concerts.calendar.com/ 518-372-5583

2 1

- Mailhost...
by Andy Smith 23 Jan '99

23 Jan '99

Are there any working examples anywhere of how one uses a Mailhost object? Cheers, Andy.

1 0

- RESPONSE.headers Help needed
by BScherer＠mobile.bam.com 23 Jan '99

23 Jan '99

I'm using the following: RESPONSE.headers['Content-Type'] = "text/csv" RESPONSE.headers['Content-Disposition'] = "inline; \ filename=salesdat.csv" RESPONSE.flush() in an attempt to get the browser to prompt the user to save the file to disk instead of displaying it in the browser. Problem is, the browser always displays the file. I've checked that the expected headers are coming through with a sniffer, and they are. I expect I'm not producing the headers right, or what I want to do isn't possible. Any ideas? TIA Bill Scherer Bell Atlantic Mobile PS. this is not Zope'd yet, it's vanilla Bobo.

7 7

- xml marshaling experiment
by Amos Latteier 23 Jan '99

23 Jan '99

With all this talk of xml-rpc integration into ZPublisher, I just had a small idea. Why not add simple xml marshalling as a Zope input marshalling format. As an experiment I chose Andrew K's xml.marshal, though of course, we could add any or all the other xml marshalling formats as well. Here's the patch for ZPublisher/Converters.py: *** Converters.py Sat Jan 23 13:29:38 1999 --- Converters.py Sat Jan 23 13:19:58 1999 *************** *** 159,164 **** --- 159,168 ---- if type(v) is not ListType: v=(v,) return tuple(v) + def field2xmlmarshal(v): + import xml.marshal + return xml.marshal.loads(v) + type_converters = { 'float': field2float, 'int': field2int, *************** *** 171,175 **** --- 175,180 ---- 'tokens': field2tokens, 'lines': field2lines, 'text': field2text, + 'xmlmarshal': field2xmlmarshal, } The only thing that's required is to add the xml sig's xml package to some place Zope can get at, like lib/python1.5 (hopefully the xml package will be in Zope by default in the next release) and of course get the xml-marshal module (which I believe was posted to the xml sig a while ago.) Voila--now you can enter xml into form elements and have Zope parse it, simply by naming your form elements with the 'xmlmarshal' type converter, for example, 'myInput:xmlmarshal'. Here's a Document that allows you to experiment: <h1>XML Marshal Demo</h1> <form action="" method="post"> <textarea name="spam:xmlmarshal" cols="60" rows="5"> </textarea><br> <input type="submit"> </form>  spam=  This document calls itself and displays the marshalled contents of the textarea. So if you type this into the textarea: <marshal><tuple><integer>12</integer><string>sdsdf</string></tuple></marshal> You'll see this: spam=(12, 'sdsdf') So, I guess you probably won't be entering a lot of this stuff by hand ;-) But maybe it would be a nice format for cookies, or Zope properties, or for programs that call Zope methods (oh, I guess that's what xml-rpc is for...) etc. -Amos

1 0