FME Form

Hi all!I’m looking for some possible help with my FME workflow (dwg to geodatabase). I have a DWG reader followed by several geometry transformers, and the geometry is working well. Where I’m stuck is the attribute schema.I have a CSV/Excel file that defines my target schema. it’s structured row-by-row. each row represents a single attribute, not a feature. All the attribute fields are under the same "attribute_value" column. So, each feature class is represented by many rows. For example, electrical pole has 8 rows—one row each for pole_id, owner, project#, etc.—all sharing the same source layer, output feature class, and dataset. Another feature class (vegetation) follows the same pattern but with fewer rows because it has fewer attribute fields.I need the geodatabase to have each feature class only contain the attributes listed for it in the CSV, and nothing extra. I understand this requires a dynamic writer, but I haven’t had success yet. I currently have a SchemaMapper set up just

Hi, I am writing data to Esri ArcSDE in FME Form and have tested using both CoordinateSystemSetter and Reprojector to define the coordinate system. However, when I open the feature class in ArcGIS Pro I get a message that the coordinate system is missing. Has anyone had similar experiences?

In the answer to this questionEsri ArcGIS Feature Service - Reads Dates as Epoch Milliseconds | Communitya URL is given to the FME hub which doesn’t match the address I’m used to seeing for FME Hub, crucially the info on this package is all quite different too This is what i’m familiar withhub.safe.com/publishers/safe/packages/esri-agolEsri ArcGIS Connector - FME Hub and this is the link providedhub.fme.co/publishers/safe/packages/esri-agolEsri ArcGIS Connector - FME Hub

Greetings everyone,I have a large number of 2D polygons for which I want to determine the Z value. I also have a DTM (digital terrain model).Now I have observed the following: If I take only one polygon and, using the FeatureReader, read in only exactly those triangles of the DTM that are touched by the polygon in 2D, and then determine the Z values of the polygon using the SurfaceDraper (always carrying forward only z_max and using that as the Z value via the 3D_Forcer), I get different Z values than when I first create a buffer around the polygon and then pass the buffer to the FeatureReader and include the triangles around the polygon in the SurfaceDraper.The triangles from the DTM do not touch each other.There are deviations of several meters, depending on how strongly the terrain height varies around the polygon.It seems as if the SurfaceDraper somehow takes the surrounding terrain into account, even if it does not touch the polygon.It cannot be due to the tolerance; I have alread

Oh no, I put my question into the wrong Category (Community Q+A). I cant move the thread.Sorry for the double posting, but it seemed, that in the Q+A category is nobody who can move my thread (or?).So I post this one here with the link to my wrong settet Post.  

We have multiple workspaces that write to ArcGIS Online with multipart geometries. The old ARCGISONLINEFEATURES writer works and correctly writes the geometries. The new (2025) Esri ArcGIS Feature Service Writer returns an error rejecting the features and the following warning:ArcGISFeatureServiceWriter: INSERT error for a feature in Project Stats by State. The error code from the server was '1000' and the message was 'Adding/updating geometries that result in SQL GeometryCollections is currently not supported (SQL Server performance).' Using a Deaggregator seems to fix this in most cases. However, we want to retain the multipart geometries. In other cases, I see a similar warning and the following error after using a Deaggregator:ArcGISFeatureServiceWriter: INSERT error for a feature in Terracon Offices Closest Areas. The error code from the server was '1000' and the message was 'Data is Null. This method or property cannot be called on Null values.' I’ve tried a NullAttributeMapper a

Hello,I am testing out the Esri ArcGIS Feature Service writer that is in the Esri ArcGIS Online Connector package (Esri ArcGIS Connector v3.13.2). I’m using Form 2024.2.2.0 (20250114 - Build 24820 - WIN64). Screenshots are edited to remove confidential information. I have noticed that the “Target WKID” cannot be left blank even if you have selected “Feature Service Handling: Use Existing” If I am understanding the documentation correctly, this parameter is only needed if the writer is creating a Feature Service. However, as the default is set to WKID 102100, if you don’t change this to the correct WKID, it will write the data out to the wrong location. I initially made this mistake before I realized I had to input my data’s WKID. For the old Esri ArcGIS Portal Feature Service Writer, you were able to leave this parameter blank: I think this parameter being a required parameter is unnecessary when Feature Service Handling is set to “Use Existing” and I like the way it was in the older w

Hi everyone I have been having some issues with the new ArcGIS feature service writer. When using this writer in combination with a change detector, it doesn't seem to update or insert any of my changed records and i have been getting weird behaviour all over the place. Let me preface this by saying that everything works as expected on the old portal feature service writer, it's when switching to the new feature service writer that weird things start to happen.  Under the advanced parameters I have enabled the following fields to adress the first problem I encountered (namely that the writer no longer seems to care that global id is a read only field and skips it for writing):  Afterwards, I ran into the next set of errors:All global id's for the update actions are filled and added as an attribute. I don't really get what i am doing wrong, but furthermore I really don't understand why the behaviour seems to be so different in the new arcgis feature servicewriter. Why do i have to jump

Hi,I’d like to read a folder of documents and a set of points (say from a csv) and then join these together (filename of documents might match an attribute on the points), then write them all to an ArcGIS Online feature service with the documents as attachments.  I would ideally like to do this in one workspace and not have to hardcode anything.I don’t think this is possible without using one workspace to write the CSV to feature service, and then another to read it and add attachments using the ArcGISAttachmentCOnnector, as that requires the feature service URL, which may not be known at the time of creating the feature service.Thanks,Marc

hi everyone! some of my line features (pipes) have text features floating above (top left) of the line, describing what type of pipe it is. I want to add that text as an attribute field to that feature. but also not all of the lines have one and I want it to only match if it's describing that feature (not grabbing a random text field nearby). I'll also note that everytime I've tried to use neighborfinder, the autocad_text_string AND fme_text_string values of lines show numbers (ex. 356.77, 623.44) instead of the descriptive text it should have shown. I've been trying this for hours and would appreciate any help given. thanks in advance!my workflow is: DWG Reader →  Attribute exposer →  Feature joiner1 CSV Reader →  Duplicate filter →  Feature joiner1 Feature joiner1 →  Deaggregator → 2D forcer → Geometry filter → point/text & lines Point/Text → Tester (if fme_type = text) → Attributemanager (create new attribute field: final_textstring, set to @value(autocad_text_string) & dele

I’m struggling with dynamic worflow when i want to change the target MS SQL DB schema (table qualifier). When i want to migrate data between databases with the same DB schema, it’s fine. But when i want to change the schema i’m running into errors. So far I could manage to migrate data from custom schema (e.g. “ldm”) to default “dbo” schema manipulating “fme_feature_type” and fme_feature_type_name” (also not sure about the differnce of these 2 attributes) in Attribute Manager assigning just Table name and omitting the schema qualifier.When i want to use another target schema than “dbo” (e.g. “test”) i always run into error:Microsoft SQL Server Spatial Writer: A feature has arrived with 'fme_feature_type' of test.OPSiteEK, but the feature type is not recognized. Halting translation1) Does anyone know how to reliably change the target schema for dynamic worflow?2) I have also noticed that if I want drop and recreate tables with no data, i have to send at least 1 data feature, because if

Hi,In the headers section, it seems it’s mandatory to define the header parameter immediately.Is there any way to push a Json or a FME attribute to define a dynamic header to use?My issue is, the client API Rest has many methods with different header parameter. I’m trying to setup a custom transformer that will be able to handle dynamically the header to set. But it doesn’t seem possible to do it right now?Any workaround to achieve it?Regards,Jean  

When using the new ESRI ArcGis Feature Service I get this error when trying to write fme_aggregate geometry to AGOL: Python Exception <FMEException>: FMEException: 1: <class 'fmeobjects.FMEAggregate'> cannot be converted to esriGeometryPolylineWith the old writer I did not have this problem. Is there some solution for this? Deaggregating is not the solution i neede. The multi part geometry is only contaning lines. I have tried GeometryCoercer, but without effect. 

Has anyone in the UK who is using the Ordnance Survey Maps API managed to get them added into FME for use as a background map.  I have tried and get nothing to display….Any advice welcome...

Hi FME Community,I’m working on a project where I need to create a 3D cone shape in FME 2025 and publish it to ArcGIS Online 3D viewer. The cone should:Start at ground level (let’s say Cape Canaveral X, Y, Z coordinates). Have a base radius of 500 m. Extend in a cone shape getting wider upwards at a 7° angle. Reach approximately 50 km into space (so it’s a very tall, tapered cone).I’ve tried using combinations of Creator, Taper Line Bufferer, Rotator, and SurfaceBuilder, but I’m struggling to generate the correct geometry and export it in a format that ArcGIS Online accepts (e.g., Collada, OBJ, or multipatch).Challenges I’m facing:Generating the circular base and apex points accurately. Applying the 7° tilt without breaking the geometry. Ensuring the output works in ArcGIS Online’s 3D environment.Has anyone successfully built something similar in FME?Which transformers or workflow would you recommend? Any tips for handling such large Z values (50 km) and coordinate systems? Variable in

Does anyone have experience working with the MCAP file format in FME? MCAP is an open file format designed to store large amounts of time-stamped data, for example from robots, sensor recordings, and other systems that generate a high volume of messages.https://mcap.dev, https://foxglove.dev/blog/introducing-the-mcap-file-formatIn my case, I am working with MCAP files from a sensor box that determines the position of rail vehicles with track-level accuracy and records the track sections where a track-laying machine has been operating.I am aware that there is currently no dedicated Reader, Writer, or Transformer for MCAP in FME, nor is there anything available on the FME Hub. At the moment, I am at a very early stage of understanding and analyzing this data. There is, however, a Python library available for working with MCAP files: https://mcap.dev/docs/python/.

Hi, I have a workspace in which I read features from different postgis tables and attach their schema with the FeatureReader.Then I want to write dynamically the features in a geopackage.When writing, the gpkg writer tries to use the fields defined as primary key in the schema, but sometimes I have several or the primary key is not an int as requested by the gpkg format.How can I easily delete the information on the primary key fields in the list attribute{} created by the FeatureReader ? I assume I have to delete if in the subattributes fme_data_type and native_data_type in the list.Furthermore, some features read contains id column, how to handle this to not interfere with the gpkg writer ? In the past there was an option in the writer to specify the name of the gkpg primary key.Maybe the creation of a gpkg_id field defined as Primary key is enough ? Thanks for any advice !P.S. : I use FME Form 2025.

I have a csv spreadsheet of arcgis server feaureservice urls with the featureclass, name attribute , and projection attribute. This list is long and has multiple servers and I want to pass this list into my workspace and have the FeatureReader transformer read the url and featureservice with a spatial filter for my area of interest (AOI) and then clip out the data I need inside the AOI and only keep the attribute name field in the result, run some additional analysis, then write out the clipped results to a GDB, then move to the next record url in the spreadsheet.. My plan is to reproject the AOI to each projection listed in the csv and then pass the AOI as the creator to the FeatureReader transformer to only read based on the spatial filter. This works great if I do separate feature readers for each URL but that makes for a gigantic workspace and isn’t very dynamic as the spreadsheet and AOI changes frequently. Is there a way to pass a csv list to the FeatureReader in this way and acc

Hi All,I’m using FME 2021 32Bit on GESW 4.1 and unable to read the tables. Screenshots showing starting of TICS server & modules which I used in GESW along with error from FME are attached for reference. Can someone point out to the location which I’m missing anything such that tables can be read… Thank you

I am encountering an error in FME Workbench when closing the GENERIC Reader. During execution, the translation fails with a transcoding error related to a binary string, and the program terminates unexpectedly. The error message indicates that FME is unable to transcode a binary string containing non-UTF characters (apparently related to a shapefile) into another encoding. As a result, the translation process stops with a “Translation FAILED” message. This issue occurs consistently and prevents successful completion of the workspace. Any guidance on how to handle encoding or binary data in this scenario would be appreciated. Menssage Log:NullAttributeMapperFactory (NullAttributeMapperFactory): Transcoding the binary string '1e\xA8\x83\x16\x00\xE8\xBB\xE9\x96\x1DfH\xC0 -\xB12\x1A9:\xC0\x00\x00\x00\x00\x00\x001\x01\x00\x00\x00\x09\x00\x00\x00shapefile' to another encoding can not be doneOGCGEOPACKAGE writer: Committing transaction of 1451 features to feature types 'AREA_IMOVEL_1'...Closi

N/A

I have a workspace where the DateTimeCalculator is failing for all records. The rejection code is always INVALID_INPUT. And yet, when I view the data in Inspector the two input fields I’m using both list as Date under Data Type. The dates look fine fine (e.g. - 20251216). It fails for any mode - Add or Subtract, Calculate Interval, etc. Any ideas why it’s failing and what I need to do to fix it? (FME Form 2025.1.0.0)

Hi,I need to use many hosted GDB from the web. They aren’t in the same root folder, they all have sub folders.Quebec Hydrographic Network GeobaseIs there a way I could use Multiple URL instead of having to use 15 FeatureReader, one for each GDB?Or a way to download everything from the root folder ?Thank you.

Hi.I just had a job on our FME server 2022.2.2 go haywire.When reading from one or more database tables, it suddenly spews out a _lot_ of error messages, repeating the same two lines from line 319 to line 865997 (on page 1734):Microsoft SQL Server Spatial Reader: An internal error has occurred. Please visit http://www.safe.com/supportMicrosoft SQL Server Spatial Reader: Failed to parse datetime, attribute set as nullWhat gives ? Should I be concerned ?I've occasionally encountered huge log files before, but haven't had any idea why. Next time I'll see if it's the same issue.The job took several minutes (as opposed to normally a handful of seconds) to finish, which is not good.Cheers.

I’m getting an error when reading an Oracle database on the Datetime. Error converting from OCI datetime to C character string : `ORA-01805: possible error in date/time operation' These are my current installs