January 2000 - Zope-Dev

[Petition] Kludge for Splitter.c (long)
by LEE, Kwan Soo 17 Jan '00

17 Jan '00

Hi, I've been try to (mis)understand why ZCatalog does not support Full Text Search for Non-Latin character text. Being Ignorant of C or C++(even regex), I did with pure hodgepodge trial and error approach, sigh ... Then I come to the Splitter.py module(at the end of message) which seemed to work (with obvious limitations) on Zope 2.0.1 Windows binary installation. It did Full Text Search for Korean texts. But IT FAILED with Zope 2.1.2 Linux binary Installation. So now my 2 petitions(please, correct me if the petitions based on my day-dream). 1. How about a dumb kludge Splitter.c which treats the characters in the user-specifiable/configuable list as white space and all other characters upto char(255) as meaningfull character and splits the text. In Korean, the current approach based on 'stem' words and 'stop' words will simple not work. For we have quite different writing convention. I guess many other (small) languages have simillar problems. Still, Full Text Search capabilities are so valuable to live without it. Furthermore, what if a Zope site contains documents in many languages? I guess the approach based on _ONE_ locale will not work greatly. Does one need several personalities of Splitter? Before the "Full I18N/Localization Support"(I'm not sure what that mean ...) of Python & ZOPE, a (maybe unsupported or community supported) kludge Splitter module with adequate warning may relieve the lives of lots of none-English/European Language Zopistas. 2. Can any one eplain(or give the clue of) the difference of SearchIndex/ZCatalog i Zope 2.0.x and 2.1.x? Especially the role of subindex in TextIndex.py and UnTextIndex.py? My Splitter.py gets errors whenever subindex is related. TIA LEE, Kwan Soo ##### #Splitter.py ##### import string t_t=['',]*256 for i in range(256): if i < 48 or 57 < i < 64: t_t[i] = ' ' else: t_t[i]= chr(i) tt=string.join(t_t,'') class Splitter: """w/o stop word list support""" def __init__(self, isrc, *kw): self.isrc = string.translate(isrc, tt) tempsrc = string.split(isrc, ' ') xx=[] for x in tempsrc: if x: xx.append(x) self.src = xx def __getslice__(self, a, b): return self.src[a:b] def __getitem__(self, a): return self.src[a] def __len__(self): return len(self.src) def indexes(self, a): res =[] for i in range(len(self)): if self[i]==a: res.append(i) return res def pos(self, a): i = int(a/2) x = string.find(self.isrc, self[i]) return (x, x+len(self[i]) )

2 1

ZMySQLDA & MySQLdb question need help.
by Victor.Zhai＠ogilvy.com 17 Jan '00

17 Jan '00

Hi,all My platform is : Redhat 6.1 . In my project, I need to use the access the MYSQL database. First: I download the source of mysql: mysql-2.22.27.tar.gz for http://www.tcx.se. Then I compiled it and make it can run like: Shell> safe-mysqld &. Second: I try the ZMySQLDA. But when I expended the ZMySQLDA to the folder of /lib/python/Producs/, I tried to compile it.But then error occured: It seems that "&" cannot support by my compiler. Line 384: tf = &(mysql_fetch_field_direct(res, i)); ... And MySQLmodule.so cannot be produced!. And when I use MySQLdb expend in my Products directory. When I want to build the MySQLdb it appeared like this: Traceback (innermost last): File "build.py", line 14, in ? import MySQLdb File "MySQLdb.py", line 19, in ? from _mysql import * ImportError: No module named _mysql Please help me with this Thanks a lot!. Vitor zhai. "WorldSecure Server <ogilvy.com>" made the following annotations on 01/16/00 20:57:22 ------------------------------------------------------------------------------ Privileged/Confidential Information may be contained in this message. If you are not the addressee indicated in this message (or responsible for delivery of the message to such person), you may not copy or deliver this message to anyone. In such case, you should destroy this message and kindly notify the sender by reply email. Please advise immediately if you or your employer does not consent to email for messages of this kind. Opinions, conclusions and other information in this message that do not relate to the official business of the Ogilvy Group shall be understood as neither given nor endorsed by it. ==============================================================================

1 0

ZDP News
by Tom Deprez 16 Jan '00

16 Jan '00

Hi all, Weekly ZDP News at http://zdp.zope.org:8080/ZDP/news/ Regards, Tom

1 0

URL quoting, '/' problems with ZCatalog
by Itamar Shtull-Trauring 16 Jan '00

16 Jan '00

(This is going into the Collector too.) I've been working extensively with the ZCatalog in the past few weeks, using it to link to cataloged object using REQUEST.resolve_url(BASE0 + catalog.getpath(data_record_id_)). Along the way I've encountered a number of problems that have come up here on the list, mainly with people complaing about "Uncatalog of absent id". I'm running Zope 2.1.0 and 2.1.1. 1. Find Items to ZCatalog vs. IndexAware: Finf and IndexAware index object in the ZCatalog. The problem is that they do it using completely different formats. Find creates ZCatalog entries with a / in the beginning, and not url_quoted. IndexAware creates ZCatalog entries without a / in the beginning, and url_quoted. For example, the following object /A Folder/an object will be cataloged two ways using Find: /A Folder/an object using IndexAware: A%20Folder/an%20object 2. REQUEST.resolve_url will only accept non url-quoted URLs, so if you want to use the result of a ZCatalog search on IndexAware objects you'll need to unurlquote them. ZCatalog.getobject uses REQUEST.resolve_url, so I assume it will also break on IndexAware objects. You have to pass a full URL for some reason, too - http://zope:8080/index_html, instead of /index_html. The problem here is that Zope does not define what format an URL should be in. I propose that the standard format be unurl_quoted (spaces, not %20) with a '/' in the beginning, but the other way around is okay as long as there is a STANDARD. Is it too late to fix REQUEST.resolve_url, or is there some good reason that I need to prepend BASE0 every time I call it? -- Itamar S.T. itamars(a)ibm.net

1 0

patch for FTP editing of SQL Methods
by Brian Takashi Hooper 16 Jan '00

16 Jan '00

Hi there - I don't know if this will interest anyone, but I added rudimentary FTP support to SQL Methods (so that people like me who edit through emacs or some other FTP interface can change stuff that way). Of course there is no provision for changing arguments to the SQL Method and what have you, but exactly how to do that right is kinda unclear at this point... well, better than nothing? Here's the patch: --Brian Hooper diff -u ZSQLMethods.orig/SQL.py ZSQLMethods/SQL.py --- ZSQLMethods.orig/SQL.py Sun Jan 16 22:20:28 2000 +++ ZSQLMethods/SQL.py Sun Jan 16 22:05:23 2000 @@ -189,3 +189,27 @@ manage_main=HTMLFile('edit', globals()) + __ac_permissions__=( + ('FTP access', ('manage_FTPstat','manage_FTPget','manage_FTPlist')), + ('Change Database Methods', + ('manage_edit','manage_advanced', 'manage_testForm','manage_test', + 'PUT')), + ) + + def manage_FTPget(self): + "Get source for FTP download" + return self.src + + def PUT(self, REQUEST, RESPONSE): + """Handle HTTP PUT requests.""" + self.dav__init(REQUEST, RESPONSE) + body=REQUEST.get('BODY', '') + # self._validateProxy(REQUEST) is this OK? + try: + self.src=body + self.template=t=Shared.DC.ZRDB.DA.SQL(body) + t.cook() + RESPONSE.setStatus(204) + except ParseError: + RESPONSE.setStatus(426) + return RESPONSE

1 0

[ANNOUNCE] ZopeMozilla CVS, and a first setup for ZopeStudio
by Martijn Pieters 15 Jan '00

15 Jan '00

Hello all, A new CVS module has been set up for code relating to the Zope Mozilla project. It can be browsed with cvsweb: http://cvs.zope.org/cgi-bin/cvsweb.cgi/ZopeMozilla/ If you want to check out files from this module, read the Public CVS access instructions: http://classic.zope.org:8080/Community/CVS_public_access and substitute ZopeMozilla for Zope2. We plan to give write access to the CVS to certain people in the Zope-community in the very near future. There are different sub-projects within the Zope Mozilla project, each with their own sub directory in the ZopeMozilla module. Next to code related to these projects, documentation will also be checked into these directories. I hope to get check-in messages emailed to either Zope-Mozilla or Zope-Checkins, when Ken Manheimer returns next week. I have checked a first setup for ZopeStudio in. It consists of two related pieces; one Zope Product, that generates the data needed in the interface, and one Mozilla chrome package, that defines the interface. Each live in their own subdir again. To look at what I've done, you can go one of two ways: 1. Check out the whole ZopeStudio package into a new directory named ZopeStudio: cvs -z7 checkout -d ZopeStudio ZopeMozilla/ZopeStudio And then link or copy ZopeProduct/ZopeStudio into your Zope lib/python/Products subdirectory, and do the same with MozillaChrome/zopestudio, into your Mozilla bin/chrome directory. 2. Check out the separate packages out into your Zope lib/python/Products and Mozilla bin/chrome directories. In your Zope install lib/python/Products: cvs -z7 checkout -d ZopeStudio ZopeMozilla/ZopeStudio/ZopeProduct/ZopeStudio and in your Mozilla install bin/chrome: cvs -z7 checkout -d zopestudio ZopeMozilla/ZopeStudio/MozillaChrome/zopestudio After either one of these steps, you restart your Zope server, and start up Mozilla with the command: mozilla -chrome chrome://zopestudio/content When the interface has loaded, enter the URL of your Zope server and press Enter. Have fun! Requirements: Zope 2.x (tested on Zope 2.1.2 but should work on older versions) Mozilla M12 (tested on a recent nightly build). Thanks go to Shalabh Chaturverdi for taking my simple tree and extending it into ZIDE, and for identifying implementation difficulties. Please do play around and extend on it! -- Martijn Pieters, Software Engineer | Digital Creations http://www.digicool.com | Creators of Zope http://www.zope.org | mailto:mj@digicool.com ICQ: 4532236 | PGP: http://wwwkeys.nl.pgp.net:11371/pks/lookup?op=get&search=0xA8A32149 -------------------------------------------

1 0

fastcgi mangles images
by John Edstrom 14 Jan '00

14 Jan '00

I'm trying to set up a Zope site to be accessible through an Apache server via fastcgi. The suggested procedures in the how-tos seems to work. Except! the pages displayed on the browser are badly distorted. The effect is variable, but I get things like missing background tiles, blocks of black color, i.e. general weirdness. The page looks fine when displayed through ZServer directly. I've checked the HTML & text and what gets through is fine. I use a jpeg as a background tile and some images here and there. If I take out all images from the document then the page looks fine through fastcgi. If I include just the background tile image from the fastcgi/ZServer, the whole page is displayed weirdly. If I take the exact same image from the apache server the page looks good through fastcgi. I'm thinking that the images are getting munged somehow and confusing the browser. I'm using a DEC UNIX machine (I don't have a problem on intel Linux). I wonder if there is a problem somewhere with how FCGIServer.py handles images (and possibly other blocks of binary data). Looking in FCGIServer.py I see things like: self.contentLength = (vals[4] << 8) + vals[5] I don't know python internals at all. Might this sort of quick calculation by shifts generate erroneous transfers on 64-bit machines? -- John Edstrom

1 0

RFI: OpenLink ODBC drivers
by Christopher Petrilli 14 Jan '00

14 Jan '00

Gang, Does anyone have any experience with these drivers positive or negative? If you do, we'd love to hear about it privately! email me directly, I'd really appreciate it. Chris -- | Christopher Petrilli Python Powered Digital Creations, Inc. | petrilli(a)digicool.com http://www.digicool.com

1 0

RE:Catalog bug ZClass constructor
by Jason Spisak 14 Jan '00

14 Jan '00

Brian, Here is the code I have for the ZClass is was building when I found this out. Does any of this look like "custom code" that would account for a leading slash? <dtml-comment> Set a REQUEST object called 'id' as a unique number by representing the current time as an integer, and then convert that to a string. </dtml-comment> <dtml-call "REQUEST.set('id', _.str(_.int(ZopeTime())))"> <dtml-comment> Create an instance of the CompanyClass with the 'id' we just set in the REQUEST. </dtml-comment> <dtml-with "CompanyClass.createInObjectManager(REQUEST['id'], REQUEST)"> <dtml-comment> Pack all of our for values and anything else in the REQUEST into the proerties of the CompanyClass instance we just created. </dtml-comment> <dtml-call "propertysheets.Company_properties.manage_editProperties(REQUEST)"> <dtml-comment> Because the 'CreateInObjectManager' we called above indexes the object before we called 'manage_editProperties', all of the properties of our new Company are blank as far as the Catalog is concerned. Not Good! So we tell the Ctalog to take another look, now that they have values. </dtml-comment> <dtml-call reindex_object> </dtml-with> <dtml-comment> Okay, now that we've created a Company, let's be nice and send the user into it. We do this by redirecting the browser to a page inside our new object. The URL2 tells us to chop off the last 'two' elements of the current URL and the the '+' adds on the id of the object we just created. So they end up inside the new Company! </dtml-comment> <dtml-call "RESPONSE.redirect( URL2 + '/' + REQUEST['id'] )"> This is very basic stuff, so I was suprized to see a problem. I am calling something in a nold way that doesn't need to be? Please let me know. -- Jason Spisak webmaster(a)hiretechs.com

1 0

RE: [Zope-dev] RE: [Zope] Very important for CatalogWare zopists
by Michel Pelletier 14 Jan '00

14 Jan '00

> -----Original Message----- > From: Brian Lloyd [mailto:Brian@digicool.com] > While I won't totally rule out the possibility of something > in Zope, the fact that we don't see this behavior seems to > suggest that there might be something going awry in your > particular ZClasses. In particular, I'd look for some > difference in the way that it registers itself with the > catalog at the time that it is created versus when it is > "found" by the catalog. The difference could be a difference > in either behavior, state or environment (i.e. is the > acquisition context, etc. the same in both cases when the > "absolute_url" method gets called to generate the url for > the catalog? Is there any custom code around this logic?). BTW, absolute_url did change to support virtual hosting in 2.x. This could be causing a problem in your ZClass constructors. Jason, try inspecting the return value of absolute_url in your constructors, you may have to prepend a / -Michel

1 0