Authoring

Custom transformer weirdnesses

Hi there, I thought this seemed simple but i've become a bit stuck:  I have x features (many millions but im running a testset of 20k) and they all have an Parcel number. Each feature has an associated Category (one of eight) that is made up of 1 to 3 letters.  Now, each parcel number does not necessarily occur just once; there could be up to 9 instances of each parcel number and each instance may, or may not, have the same Category code.   I wanted to test the Category attribute - grouped by Parcel - to see if there was a complete set of Category duplicates ('Yes') or if the Category differed at all ('No') for a given parcel.   (Bonus qu: In some cases where there are 6 instances of a Parcel and there are 3 different categories, it is clear to see that the 3 different categories have been duplicated by mistake - because there are associated identical areas - does anyone know how to clean those out??)  thanks all

Hi,       I have a bunch of datasets passing through a counter. The datasets are grouped by area (region) so that each region gets its own count.  This works and I get a nice count of datasets per region  as they are sorted by date.  1 being the most recent dataset.  The problem is.  There is another attribute of [Catagory] and I need it so that every time [Catagory] = Update, the counter does something to freeze the count on the next record to go through.  So the [Catagory=] Update record gets counted like normal, but the next one to go through after it does not increase in count, as follows:  Dataset 1 - [Catagory] = [], Count = 1 Dataset 2 - [Catagory] = [], Count = 2 Dataset 3 - [Catagory] = [], Count = 3 Dataset 4 - [Catagory] = [Update], Count = 4 Dataset 5 - [Catagory] = [], Count = 4 Dataset 6 - [Catagory] = [], Count = 5 Dataset 7 - [Catagory] = [Update], Count = 6 Dataset 8 - [Catagory] = [], Count = 6  And so on...       My mind is already twisting with the logic I have used

Hello,  When having a seemly elaborate workbench, would run it using a specific data. Now when this same workbench is to be used for several different data... traces of the attributes of the earlier data exists. Anyway to refresh the workbench such that these traces may be removed from each transformer ?  Thank You.  

Hi Folks, Conceptually simple problem:  I have a collection of features which include two attributes each:  Unique_ID--- value 1  10 1  9 2  8 1  7 2  6 2  5 2  4 1  3 2  2 1  1  What I want to do is keep only one feature for each Unique_ID, and that feature should be the one with the highest Value. So in the above example, I'd only want to keep: 1  10 2  8  I can sort the values to get the above order of features (Sorter), but can't think of how to get FME to discard all features but the one with the largets value.  (I considered using a variable setter/getter with the Sampler, but the Sampler can't sample based off variables/attributes, only parameters.)  Any suggestions? Thanks!

FMEers,  I have a heaping helping of parcel data that I would like to query or filter by extent BEFORE FME reads every feature and builds its own index (which is what I perceive it is doing because it loads all million plus features before anything gets piped to anywhere)   I've set the min,max,x,y values in the Oracle Spatial reader, but it seems to completely ignore it.   I loaded the ALL_SDO_GEOM_METADATA table with the appropriate DIM_INFO and created spatial indexes.   FME still reads every feature.   I have Oracle Locator, but not Oracle Spatial Object.   Does the reader require the Spatial Object product to do filtering?  BUT--  When I use a subquery with SDO_ANYINTERACT in the select statement for a feature_type, THEN I get the filtering I want.   Trouble is, there are more than just parcels, so I need to add a subquery to every other feature in the workspace, querying to get the same MBR a gazillion times:  select * from my_geometry a join my_data b on a.ID=b.ID where b.da

Workbench finishes in 12.9 seconds in Desktop but never ends in Server,  For the same data source, the workbench finishes in 12.9 seconds in Desktop but when the same workbench is run from the FME Serve with same data source it never ends            What might be the issue here?   Thank you  Best  Jamal

“Feature Types to Read” parameter issue in the published workbench,  I wondering why there is no option in the published workbench to browse to select the required features. In the Desktop, if the 3 dots icon is clicked, then a new window (Select Feature Types) is open to allow selecting the required types as shown in the screenshot below    How the “Select Feature Types” can be open while working in the published workbench such that one can selected the required types?  Thank you  Best  Jamal

Is there any way to get the FeatureReader to return Geodatabase Metadata? On the Geodatabase readers there is a "Feature Read Mode" parameter that allows you to choose to get Features or Metadata, but how do you specify that with the FeatureReader?

In this article you show how to create a JSON template from attributes,  http://docs.safe.com/fme/html/FME_Transformers/Default.htm#Transformers/jsontemplater.htm  But I have some values that are in Published Parameters. How would I get those?  I don't see an XQUERY request for that:  http://docs.safe.com/fme/html/FME_Transformers/Default.htm#XQuery/XQuery_functions.htm

Hi, as mentioned earlier, I am currently working with FME Base edition, which has certain limitations. I want to ask if somebody happen to know if parallel processing is available in FME Base edition. In the dialog, the parallel processing level may be set in range from "No parallelism" to "Extreme". However, for options other than "No parallelism" I sometimes get an error message saying  that TransporterFactory is not licensed. Does it have something to do with Base edition or am i doing wrong something else? The available documentation doesn't say anything about parellelism in FME Base. Thanks,  Jaro

Hi,  We are 2 guys and want to create some work bench to handle dozen of different data from dozen of different DB (btw the purpose is to create some XML and populate a single common DB with all this data ).  The thing is we also want some "end user" to be able to write some work bench.  Actually we have made ~30 custom transformer (CT) already (all are linked)  Here is our question: What is the best way to specify (or not) CT parameters ?  1_ copy feature attribute in CT parameters and add a AttributeCreator from parameters to feature attribute. We thinks it's a bit ugly, but the clearer for the end user: specify what we expect.  2_assume the feature have an attribute called "foo" and handle it that way ... and let the user deal with errors if his feature are missing attributes.  3_?? we don't know how to say "we want a feature with 'foo' and 'bar'".

Hi All,  I have made a fun script. It makes ford circles with most math done by using listfunctions sorting etc.. So realy a minimal of actual math involved.  It works fine, except when one zooms in the some of the circles dont realy touch exactly as they should. Anyone got a tip?    Still cant upload workbenches here. Drop box is out of the question for me, can't install it due to rigorous network security. I heard someone nmae another one, wich does'nt require software install. Forgot the name tho.  Im currently trying to do the circles in circle version, but i hav'nt yet figured out how to get complexnumber math working using fme.  (I made for instance a 3th-root custom transformer using aproximation (invovlves looping) rather then complexnumbers. Wich was needed to get that geeky heart formula working...but also needs complexnumber calculation..i got half a heart... )

Hi everybody I'm pretty new on this community, so please forgive all my mistakes. I have problem with DB connection. I'm using win xp, FME 2014. I have access to DB, but when i'm  trying to connect with db i recived message as you see on image. All settings in FME and ODBC should be correct. On the other machine (win7) where settings are the same , connection issue doesn't exist. Could you give me some advises where the problem is? Best regards Maciek 

Has anybody tried to update specific attributes to an existing table?  I want to update a few attributes without changing the structure of my table.  Ex: I have an existing Hydrants feature class (GeoMedia SQL database) and I add 5 new hydrants to my workspace.  I update the Neighbourhood and Mapsheet attributes they are located in using the PointOnAreaOverlayer.  The only problem is i'm reading and writing to the same Feature .....this in return doubles my amount of records or overwirtes (drops) my table completely...    Bottom line is, can FME Read AND Write to the same feature class without changing the output name / structure of the table?     

Hello, I have an ESRI Geodatabase Reader using a WHERE clause (Parameters / Advanced / .. that works good. Now I like to import Feature Types to perform the same query.  Unforunately the Where Clause needs the feature class name in the sub-query:  (ATTR1, ATTRDATE) IN (SELECT ATTR1, MAX(ATTRDATE) FROM SDE.FCNAME GROUP BY ATTR1)  So my question is: is there a way to set the where clause on the Feature Type level instead of the Reader level? Or is there a way to replace the feature class name with a variable/link that points to the appropriate Feature Type?  I appreciate any hint klaus

Hi there,    Simply renaming a value is quite easy using the AttributeValueMapper, however, I'm completely stumped as to how I should go about this when the source value are dynamic.    To get into details:    What I need is very basic. For example, If I have a value like "ROAD" I want to rename that value (and all other present values) to something like "ROAD_X". So I simply want to add the suffix, "_X" to the existing value.    This would be simple if my source was not dynamic, in the sense that it simply points to a folder, and whatever's in there gets run through the workspace.    So, is it possible to add "_X" or something similair to an unknown value with ValueMapper, or any other transformer?    Thanks in advance,  

Hi there,  Slightly odd question that I think I know the anwser to but thought I'd just see if anyone out there has any bright ideas.  We have reader file that undergoes some transformations through a workspace, but what the customer wants to be able to do at the end of the workbench is bascially rename the file based on a new filename that has been generated through the process (but not modify the file in any way).  Is there anyway to do this?    I'm guessing not and already using AttributeFileWriter to write the name of the file out to a simple text file where I can wrap the workbench call up in a batch file where I can write some simple code to do the renaming there.  But if anyone can suggest a better way I'd love to hear it!  Cheers  Dan

I have just downloaded FME 2014, when I edit the reader feature type properties and de-select the exposed user attributes, they still come through in the workbench.

In regards to Linked Looping Custom Transformers, is there a way for Variables to persist through iterations? Currently the Global Variables persist per feature even though FME seems to process a single iteration at a time rather than perform all iterations for a single feature before going to the next feature.  What i'm seeing is: Iteration 1: 1st feature goes in, global variable set to a. Needs to be looped. 2nd feature goes in, variable appended to a, b. Needs to be looped. 3rd feature goes in, variable appended to a, b, c. Needs to be looped  Iteration 2:  1st feature goes through loop, yet variable's value is a. 2nd feature goes through loop, its variable value is b.  This is likely following the flow of features yet the log file is extremely confusing as it shows the processing via iterations rather than tracing a single feature through all iterations.   

Has the option to update a reader changed in FME2014? Is it now part of a different component on the reader right-click context menu?  The 'Edit source FileFormat location'  option doesn't seem to re-read the file, just change source path.  Any ideas?  Thanks, Simon

Hi,  i just noticed that if i have more than one writer in my workspace and choose only the output of one for the DataDownload- or DataStreaming-Service only this output gets written to FME_SERVER_DEST_DIR. The other writers write their output to the default, in the workspace file configured path. How do I avoid conflicting file access with concurrent user session for such a workspace? I need a unique filepath for a temporary file in my workspace which should not be included in the download/streaming output. I could use a private python parameter which generates a filename (e.g. with a timestamp) or i could perhaps utilize the UUIDGenerator to achive unique filenames... Or I could use the FME_SERVER_DEST_DIR parameter and extract the unique part of it to use it e.g. in %TEMP%.  But perhaps FME Server has some variable/function i could use directly?

I've run into some incompatibilities between the Desktop edition I'm developing on (2013 SP4 build 13547) and the server I'm deploying to (2013 SP3 build 13528). I was hoping someone could point me to a desktop copy of build 13528.  Thanks!

I have connection details to connect a Sybase DB an I try to figure out how to connect using the ODBC 3.x reader. I have no knowledge about ODBC. so my question is, what do I have to type in the Database field of the ODBC 3.x Parameter window. I have a username and pwd, and my connection details are similar to the following  Default DB: db_prod Server Name: MONUM99 DBMS: Sybase ASE  Any hint is appreciated 

Hi,  Iam attempting to write data using dynaminc schema in a workspace, it fails with the following warning:  Unexpected execution path. Please contact Safe Software A fatal error has occurred. Check the logfile above for details ... Last line repeated 2 times ...  I am writing to ESRI geodatabase feature classes. When i disable the writers and run the workspace, everything works as it should do. I have used dynamic schema before succesfully and have tried many different variations of parameter setting but to no avail.  Any help much appreciated!  Thanks,  Chris.