Hi Jitu, For jobs with multiple outputs, MCF keeps track of which uploads succeed and which fail, and thus which uploads need to be redone and which don't. MCF will keep retrying a given document until all the uploads succeed.
It sounds like you are trying to build the same thing outside of MCF. That doesn't make much sense. Karl On Wed, Aug 27, 2014 at 3:18 PM, Jitu <[email protected]> wrote: > Hi Karl, > > My requirement is i have to crawl files from google drive and output to 2 > different servers. one is my own cloud server to store files and other to > store my file metadata information. Since both are 2 different servers its > difficult to handle if one succeeds and other fails and retries which would > eventually re-upload the same file again. The mandatory condition is if > upload of file fails then i should not upload metadata information and > metadata should not be sent to server before file is uploaded successfully > to my own cloud. > > secondly while uploading files to my own cloud i would like to maintain > the relative path of the files even in my own cloud but not the absolute > path like " > http://88.123.123.123/personal/administrator/Shared%20Documents/TestNewFolder/images%283%29"; > but like this "/Shared%20Documents/TestNewFolder/images%283%29". > > Thanks, > Jitu > > > On Wed, Aug 27, 2014 at 4:53 PM, Karl Wright <[email protected]> wrote: > >> Hi Jitu, >> >> The only data that an output connector receives from a repository >> connector is encapsulated entirely in the RepositoryDocument object. >> However, that object is flexible enough so that you can pass pretty near >> anything down from one connector to another, in the form of metadata. >> >> On a broader note, it is clear to me that you are attempting to use >> ManifoldCF in a manner that is far removed from what it was designed to >> do. So it seems to me that you have two choices: >> (1) Post your ENTIRE problem statement, so we can recommend an approach >> that might work for you, or >> (2) Hack away at the code yourself, and don't expect much help from us >> for your endeavor. >> >> Thanks, >> Karl >> >> >> >> >> On Wed, Aug 27, 2014 at 6:34 AM, Jitu <[email protected]> wrote: >> >>> Hi, >>> >>> is there a way to find the relative path of the files crawled by input >>> connector in output connector. for example " >>> http://88.123.123.123/personal/administrator/Shared%20Documents/TestNewFolder/images%283%29"; >>> is the absolute/complete path but >>> "/Shared%20Documents/TestNewFolder/images%283%29" is relative path from >>> where the documents are crawled from. So i need this info in output >>> connector how can i get it. Please help? >>> >>> >>> The information about relative path is getting stored in "repohistory" >>> table during fetch activity but during ingest activity i am getting >>> complete URI. is there a way to get relative path. >>> >>> Thanks, >>> Jitu >>> >>> >>> >> >
