[java] 2007-03-30 11:47:59,366 (main) [
DatabaseUtil.java:2710:ERROR] Could not create foreign key indices for entity [TrackingCodeOrderReturn]: SQL Exception while executing the following: [java] CREATE INDEX TKNG_CODODR_ODRTN ON OFBIZ.TRACKING_CODE_ORDER_RETURN (ORDER_ID) [java] Error was: java.sql.SQLException: Index 'TKNG_CODODR_ODRTN' already exists in Schema 'OFBIZ'. [java] 2007-03-30 11:47:59,382 (main) [ DatabaseUtil.java:2710:ERROR] Could not create foreign key indices for entity [TrackingCodeOrderReturn]: SQL Exception while executing the following: [java] CREATE INDEX TKNG_CODODR_TKCD ON OFBIZ.TRACKING_CODE_ORDER_RETURN (TRACKING_CODE_ID) [java] Error was: java.sql.SQLException: Index 'TKNG_CODODR_TKCD' already exists in Schema 'OFBIZ'. [java] 2007-03-30 11:47:59,398 (main) [ DatabaseUtil.java:2710:ERROR] Could not create foreign key indices for entity [TrackingCodeOrderReturn]: SQL Exception while executing the following: [java] CREATE INDEX TKNG_CODODR_TYP ON OFBIZ.TRACKING_CODE_ORDER_RETURN (TRACKING_CODE_TYPE_ID) [java] Error was: java.sql.SQLException: Index 'TKNG_CODODR_TYP' already exists in Schema 'OFBIZ'. |
I'll look into this.
Regards Anil On 3/30/07, Si Chen <[hidden email]> wrote: > > [java] 2007-03-30 11:47:59,366 (main) [ > DatabaseUtil.java:2710:ERROR] Could not create foreign key indices for > entity [TrackingCodeOrderReturn]: SQL Exception while executing the > following: > [java] CREATE INDEX TKNG_CODODR_ODRTN ON > OFBIZ.TRACKING_CODE_ORDER_RETURN (ORDER_ID) > [java] Error was: java.sql.SQLException: Index 'TKNG_CODODR_ODRTN' > already exists in Schema 'OFBIZ'. > [java] 2007-03-30 11:47:59,382 (main) [ > DatabaseUtil.java:2710:ERROR] Could not create foreign key indices for > entity [TrackingCodeOrderReturn]: SQL Exception while executing the > following: > [java] CREATE INDEX TKNG_CODODR_TKCD ON > OFBIZ.TRACKING_CODE_ORDER_RETURN (TRACKING_CODE_ID) > [java] Error was: java.sql.SQLException: Index 'TKNG_CODODR_TKCD' > already exists in Schema 'OFBIZ'. > [java] 2007-03-30 11:47:59,398 (main) [ > DatabaseUtil.java:2710:ERROR] Could not create foreign key indices for > entity [TrackingCodeOrderReturn]: SQL Exception while executing the > following: > [java] CREATE INDEX TKNG_CODODR_TYP ON > OFBIZ.TRACKING_CODE_ORDER_RETURN (TRACKING_CODE_TYPE_ID) > [java] Error was: java.sql.SQLException: Index 'TKNG_CODODR_TYP' > already exists in Schema 'OFBIZ'. > |
In reply to this post by Si Chen-2
Hi list,
I want to use a catalog named in Chinese. The creation is OK. But when click the link of /catalog/control/EditProdCatalog?prodCatalogId=ChineseCatalogName on Catalog Main page, there's an error said "Could not Find Product Catalog with Id". I think it's because ProductCatalogId is not escaped/unescaped. I searched the code and could not find escape/unescape method. If there are no such methods, I'll add them. Did I miss something? TIA, Shi Yusen/Beijing Langhua Ltd. |
HTTP in general is NOT capable of UTF-8 arguments, and you should always stick with ISO-8859(-1) characters for that. For OFBiz this means no *Id fields should have more general, ie UTF-8 or whatever, characters. There are quite a few good discussions on this in the OFBiz mailing list archives, and hundreds of general discussions and information pages about it around the internet and even in various RFCs. -David On Mar 31, 2007, at 4:48 AM, Shi Yusen wrote: > Hi list, > > I want to use a catalog named in Chinese. The creation is OK. But when > click the link > of /catalog/control/EditProdCatalog? > prodCatalogId=ChineseCatalogName on > Catalog Main page, there's an error said "Could not Find Product > Catalog > with Id". I think it's because ProductCatalogId is not > escaped/unescaped. I searched the code and could not find > escape/unescape method. > > If there are no such methods, I'll add them. > > Did I miss something? > > TIA, > > Shi Yusen/Beijing Langhua Ltd. > > smime.p7s (3K) Download Attachment |
Hi David,
> HTTP in general is NOT capable of UTF-8 arguments, and you should > always stick with ISO-8859(-1) characters for that. Yes, that's true. > For OFBiz this > means no *Id fields should have more general, ie UTF-8 or whatever, > characters. OK, I'll keep this as a rule of OFBiz. Thanks, Shi Yusen/Beijing Langhua Ltd. |
BTW, if this is necessary there are things we can do in certain parts of the system, like using the 0x style encoding for UTF-8 characters and then decoding the parameters as they come in. For most pages in OFBiz this could even be done in generic places in the framework rather than per-page or something, but it hasn't been done yet. -David On Mar 31, 2007, at 9:45 AM, Shi Yusen wrote: > Hi David, > >> HTTP in general is NOT capable of UTF-8 arguments, and you should >> always stick with ISO-8859(-1) characters for that. > Yes, that's true. > >> For OFBiz this >> means no *Id fields should have more general, ie UTF-8 or whatever, >> characters. > OK, I'll keep this as a rule of OFBiz. > > Thanks, > > Shi Yusen/Beijing Langhua Ltd. > > smime.p7s (3K) Download Attachment |
> BTW, if this is necessary there are things we can do in certain parts > of the system, like using the 0x style encoding for UTF-8 characters > and then decoding the parameters as they come in. Yes, this will work. Some functions such as search may have to contain unicode characters in url parameters. BTW, % is also a useful pattern. > For most pages in > OFBiz this could even be done in generic places in the framework > rather than per-page or something, but it hasn't been done yet. I'll try to study this later when I have a good understanding on the widget. Currently some of the getting-parameter impletements are scattered in beanshell files. Do you mean this program style will be changed in the future? Kind Regards, Shi Yusen/Beijing Langhua Ltd. |
In reply to this post by Shi Yusen
unsubscribe
--------------------------------- What kind of emailer are you? Find out today - get a free analysis of your email personality. Take the quiz at the Yahoo! Mail Championship. |
In reply to this post by David E Jones
David,
I'm curious why Product.productId isn't a meaningless neutral integer as prescribed by the Data Model book. Chris Howe described some issues caused by using the Product.productId under some circumstances, and they seems closely related to this. HTTP should support characters 0-9. Is there any language in the world that doesn't use 0-9 but something else instead? Jonathon David E. Jones wrote: > > HTTP in general is NOT capable of UTF-8 arguments, and you should always > stick with ISO-8859(-1) characters for that. For OFBiz this means no *Id > fields should have more general, ie UTF-8 or whatever, characters. > > There are quite a few good discussions on this in the OFBiz mailing list > archives, and hundreds of general discussions and information pages > about it around the internet and even in various RFCs. > > -David > > > On Mar 31, 2007, at 4:48 AM, Shi Yusen wrote: > >> Hi list, >> >> I want to use a catalog named in Chinese. The creation is OK. But when >> click the link >> of /catalog/control/EditProdCatalog?prodCatalogId=ChineseCatalogName on >> Catalog Main page, there's an error said "Could not Find Product Catalog >> with Id". I think it's because ProductCatalogId is not >> escaped/unescaped. I searched the code and could not find >> escape/unescape method. >> >> If there are no such methods, I'll add them. >> >> Did I miss something? >> >> TIA, >> >> Shi Yusen/Beijing Langhua Ltd. >> >> > |
Hi Jonathon,
> I'm curious why Product.productId isn't a meaningless neutral integer as prescribed by the Data > Model book. Chris Howe described some issues caused by using the Product.productId under some > circumstances, and they seems closely related to this. I think ProductId is not discussed in this topic. There's an implicit rule: as OFBiz is a productivity application, NO EFFICIENCY of OFBiz should be lost when making any improvements :). Take it easy. Regards, Shi Yusen/Beijing Langhua Ltd. |
Hi Shi Yusen,
You mentioned "prodCatalogId". That goes for all other fields of type id-ne, including Product.productId? I do agree that improvements need to be made incrementally, so we don't have to wait say a decade before we can start using the "corrected" OFBiz. As I told Chris, I'd rather not have to do/witness the painfully arduous task of creating say OFBiz version "ERD Corrected". It works now for me, and that's all I need (at least until I need to put chinese characters into URLs for lookups, but I'd use POST usually). I'm just curious why it was done that way, that's all. By the way, I was just thinking of importing data, and I realized I wouldn't know the primary keys when entering data for import if those keys were auto-incremented integer ids. Is that why all id fields are strings instead of integers? Maybe I'll dig into the data reader (that imports data from xml files). What happens if my data for import omits primary key values? Jonathon Shi Yusen wrote: > Hi Jonathon, > >> I'm curious why Product.productId isn't a meaningless neutral integer as prescribed by the Data >> Model book. Chris Howe described some issues caused by using the Product.productId under some >> circumstances, and they seems closely related to this. > I think ProductId is not discussed in this topic. There's an implicit rule: as OFBiz is a productivity application, NO EFFICIENCY of OFBiz should be lost when making any improvements :). Take it easy. > > Regards, > > Shi Yusen/Beijing Langhua Ltd. > > |
It is nice to have a primary key be descriptive, as it allows less
programming thought and verification to go into an application. However that should be weighed against the dangers and limitations that convention may put on you. I don't think prodCatalogId or prodCategoryId, for that matter, have as many potential pitfalls as productId does, at least not in the limited ways they have been used thus far. Also, any convention used for a prodCatalogId or prodCategoryId can be discarded and not cause as large of a business interruption as changing a convention on productId might, I would imagine. Tossing my 2 cents in since my name was invoked. :-) --- Jonathon -- Improov <[hidden email]> wrote: > Hi Shi Yusen, > > You mentioned "prodCatalogId". That goes for all other fields of type > id-ne, including > Product.productId? > > I do agree that improvements need to be made incrementally, so we > don't have to wait say a decade > before we can start using the "corrected" OFBiz. As I told Chris, I'd > rather not have to > do/witness the painfully arduous task of creating say OFBiz version > "ERD Corrected". It works now > for me, and that's all I need (at least until I need to put chinese > characters into URLs for > lookups, but I'd use POST usually). > > I'm just curious why it was done that way, that's all. > > By the way, I was just thinking of importing data, and I realized I > wouldn't know the primary keys > when entering data for import if those keys were auto-incremented > integer ids. Is that why all > id fields are strings instead of integers? > > Maybe I'll dig into the data reader (that imports data from xml > files). What happens if my data > for import omits primary key values? > > Jonathon > > Shi Yusen wrote: > > Hi Jonathon, > > > >> I'm curious why Product.productId isn't a meaningless neutral > integer as prescribed by the Data > >> Model book. Chris Howe described some issues caused by using the > Product.productId under some > >> circumstances, and they seems closely related to this. > > I think ProductId is not discussed in this topic. There's an > implicit rule: as OFBiz is a productivity application, NO EFFICIENCY > of OFBiz should be lost when making any improvements :). Take it > easy. > > > > Regards, > > > > Shi Yusen/Beijing Langhua Ltd. > > > > > > |
Free forum by Nabble | Edit this page |