I'm trying to create a BATCH file to run my workbench.fmw in witch I have tow inputs files : 8 pipelines shapefiles and 96 communal areas shapefiles . The workbench allows cutting pipeline relative to the areas by using "Spatial Realtor" Transformer. I tried to create my script DOS in a very complicated way so I come to you for help and suggestions and if someone had ever created a similar BATCH ?
I thank you in advance. OAD
As long as reading your description, I don't think you need to do such a complicated batch processing.
To cut pipelines by the communal areas, you can use the Clipper transformer. If you need to create multiple destination datasets separately for each input pipeline file or communal area file, the "fanout" functionality performs it effectively.
If you explain the reason that you need to do batch processing, we might be able to find a more effective solution.
Takashi
The workbench is not very complcated and it does what I wanted. I must not cut pipelines by the areas I only need to lacate spatially my pipelines over the corresponding area.
I need to do batch processing in order to run this workbench several times with diffrent inputs automaticlly (by double clicking on the .bat).
I hope I'm more explicit.
OAD
If you are looking for a quick way to specify different source files for each run-time, how about creating a workspace with a WorkspaceRunner to launch the main workspace?
The WorkspaceRunner provides an interface on which you can select any source files like a general Open File dialog box. I think it's easier to use and more flexible than a batch file.
Takashi
Takashi
I didn't find appropriate examples in the FMEpedia, so I created simple examples. Download them here (3MB zipped file including test data).
"runner1.fmw" has two WorkspaceRunners; each of them runs "main.fmw". This maybe is an example of the simplest batch processing.
"runner2.fmw" has just one WorkspaceRunner, but it reads input / output feature type name pairs from an external csv table and runs "main.fmw" for every pair. I often use this pattern.
There are many patterns to use the WorkspaceRunner depending on the actual situation, I cannot show all of them. Please customize these examples to create your own best solution.
Hope this helps.
Takashi
I tried to customize it and by defining my own Published Parameters, INPUTS/OUTPUTS paths but I must be forgotten a little detail. This is my example I hope you help me to find out what I missed.https://docs.google.com/file/d/0BwhE3DdmajjtME0zOGZ0QWNqNFU/edit?usp=sharing https://docs.google.com/file/d/0BwhE3Ddmajjtc3hkT21aRHVxRVU/edit?usp=sharing
note: \\ (back slash) is displayed as ? symbol on the Japanese Windows. I think the main workspace will work expectedly if you specify the parameter values (colored boxes) appropriately. You can specify a parameter value with one of: 1) setting a fixed value manually (select file / directory, write directly) 2) specifying an attribute of input feature 3) linking to a workspace parameter etc.
Takashi
3) linking to a user parameter
The first (left hand side) WorkspaceRunner seems to fail. No feature came from its SUCCEEDED port, so the second one didn't work.
The image shows the second (right) WorkspaceRunner parameters, but in fact there could be problems in the first one.
How did you specify parameters of the first WorkspaceRunner?
Takashi
The problem was with the LOG_FILE parameter and its value. I didn't correct it but I removed it for that part of my project.
However it would be so useful for the second part to create a published parameter to generate my Logs in the right directory. I need you advice please.
https://docs.google.com/file/d/0BwhE3DdmajjtdVdhclZSMXFjYkk/edit?usp=sharing
Possibly I found the reason for the failure.
If LOG_FILE (user parameter) is linked with LOG_FILE (workspace parameter), you should specify a file path (not a directory path) to the parameter.
e.g. C:\\foo\\bar\\LOGS\\DIFF_MPC\\filename.log
Takashi
It worked very well.
Thank you very much for your recommendations very useful.
Best Regards
OAD
Your suggestions are very useful for me, thank you very much.