Transformers

We are using FME Desktop 2022.0 Beta and have created many workbenches for nightly data processing. I want to run these in varying sequences using python but cannot get fmeobjects to load. We are running Python 3.7.11 64-bit (arcgispro-py3:conda) installed with ArcGIS Pro on a windows machine. The following code fails on import fmeobjects: ImportError: DLL load failed: The specified module could not be found.import syssys.path.append(r"C:\Program Files\FME\fmeobjects\python37") import fmeobjectsI see many posts pertaining to this issue but none of the recommend approaches work. I have also tried adding the path above to my PYTHONPATH with the same results. Input?  

ContextI am building a custom transformer. I need to transform input features into points to process them and then build back the features geometry before to output them. Output features are expected to have the same geometry type as input ones. ProblemI thought I could just reconstruct the geometry of the features using the LineBuilder transformer, but it creates polygons with closed lines, even when the input feature is a line (fme_line), not a polygon. QuestionHow can I reset the features geometry to their original ones before sending the features to the output port? The GeometryCoercer could do the job if I could set the output geometry at runtime (e.g. using an attribute with original fme_geometry value).

I have imported data, from a shapefile, into an existing object in Smallworld before but we recently acquired a smaller Telco and will be importing their data into ours. I have a reader (shapefile) transforming into an existing object in our Smallworld writer.  In the type_ field of the shapefile there is data that matches what the spec_id should be in the SW object, and the spec_id field is an enumerated (integer-string) list. I have changed the data on the shapefile side to match what the spec should be, both integer wise and string wise. I'm still getting the Translator Error: Value "x" not in enumerator set.  I've tried to run just a basic import without setting the spec_id field from the shapefile and I get specification is mandatory error in the magik prompt box on the SW FME session. I'm sure this is a user based error, but have tried this on multiple objects and still running into the spec_id issue.  Thanks for any help.   

I am using a featurewriter to overwrite 2 feature classes in a file geodatabase. The attribute definition panel shows the correct GlobalID and guid types. However, the result is string fields. Any ideas?

This has been asked before, at least here and here.  The problem is best described in the picture, with the starting point on top and desired outcome below. The data is huge, so manual attribute creation is not an option, the attributes should be generated like "Year@Value(number)" and "Explanation@Value(number)"I have tried the method used in these, with AttributeCreator with every option I can imagine, but to no avail. I've used it in conjunction with AttributeExposer, but without any good outcome.  Dear all, is there a solution to this?

Hello, here is my question. I have two table. The first table is like a glossary. The second table is the input data.​Table 1 (glossary)​Number Fruit1-1 banana;apple <= (meaning can be either banana or apple)1-2 orange​Table 2 (Input data)Number Fruit1-1 banana1-1 apple​So my goal is to see if table 2 match with table 1. I have a input data (table 2). Both are number 1-1 and the fruit name also match with table 1. Hence, both the result should be passed.  However, I am not sure how to do the matching. As the table 1 fruit name is "banana;apple", separating by ";". I would like to have some guidance. Thank you very much!​

Hello friends,Im a new user of FME. I have a dgn file with hundreds of points with or without tags. i need to convert this dgn to shp with unique custom ids. Is it any procedure or transformers to do this in fme. I mean each point have unique id numbers I tried with many videos and tutorials didn't solve my issue thanks

Hi, I'm looking to update a value of an existing JSON structure once it has already been constructed.I feel I'm using JSONUpdater for its correct purpose, but might be messing up the syntax. Attached is how it looks for me:Does this appear to be the correct syntax for trying to modifying a nested key/value pair?I also wonder if using the attribute "ProgramProfile" for the JSON Text field would be a better approach, as it does show up in the list of selectable attributes. I have not had luck with this approach either though. JSON sample being used:{   "UserInfoList" : [      {         "IntegrationUserID" : "123456789",         "UserProfile" : {            "FirstName" : "Johnny"         },         "ProgramProfile" : {            "UserCategoryName" : "PLACEHOLDER"         },         "AddressProfileList" : [            {               "CountryName" : "Japan"            }         ]      }   ]} 

Hi, I'm fairly new to FME. I'm curently trying to set up a workbench that reads data from multiple sources and merges it. However, I have to be able to work repeatly within a specific extent which is provided with a shapefile (see pic below). For now, I just wan to manually input a shapefile with one feature ( hopefully I manage to find a way to loop the procedure with many features but that might be for another question). Right now, I can't seem to be able to pass the extent coordinate (which I get from a BoundsExtractor) to the search envelop of a FeatureReader. This post (https://community.safe.com/s/question/0D54Q000080hXiFSAU/assign-bounding-extents-to-reader-search-envelope) talks about using spatialFilter after the Feature Reader but that means I have to read ALL the data and then filter it (which I don't find very optimal).  Has anyone have an idea on how I can go around this?Kind regards,Anne-Marie  

Hello all,Since updating to FME 2020 (FME(R) 2020.0.1.0 (20200407 - Build 20218 - linux-x64)) the columns filter in the data inspector table view does not work most of the time.When clicking on the "Columns" button then selecting some columns to display, on clicking "ok" the columns displayed in the table view are filtered to what appears to be a random set of columns. eg. After selecting the following in the column selector:  Clicking ok results in the following columns being displayed in the table view: As can be seen, they are not the same columns. This happens most of the time, though it does occasionally work correctly. I do not know what is different when it does work. This happens when Data Inspector is opened standalone and also in the Table View in the Visual Preview in Workbench.  This is also occurring in the Windows version of 2020 (FME(R) 2020.0.1.0 (20200407 - Build 20218 - WIN64)). It did not occur when we were using 2019.Does anyone know what is caus

Hello, I'm using a WMS Reader to convert the wms to mbtiles. The workbench is called repeatedly for all my polygons (Bounding Box) with "FMEServerJobSubmitter" from a second workbench. The first conversions run succesfully but after a few iterations the process stops with the error:    HTTP/FTP transfer error: 'SSL connect error' Please ensure that your network connection is properly set up |No proxy settings have been entered. If you require a proxy to access external URLs, please ensure the appropriate information has been entered   The error occurs only in FME Server, in FME Desktop 2017.1 I do not have this problem.

I'm using the EsriReprojector with source RD_New to destination GCS_WGS_1984 and geographic Transformation Amersfoort_To_WGS_1984_NTv2. Later I need my features to go back to RD New so I use the EsriReprojector again with the source and destination switched and the geographic transformation direction set to Reverse. But this gives me a different coordinate system: _Netherlands-RDNew-2008_0 instead of Netherlands-RDNew-2008 (what it used to be). Why is this a problem? Everything is projected with an offset of 100m.  When using geographic transformation Amersfoort_To_WGS_1984_2008_MB in reverse it seems to be fixed. Yet it still gives me the new coordinate system _Netherlands-RDNew-2008_0 Why, how, and will this give me problems when overwriting layers with Netherlands-RDNew-2008?  Any dutchies with the same question? 

Hello friends, I’m new to FME. I have two dgn files for roads. One file for the central line of roads with all tag details. The other one is a new file, filled roads (create as region in microstation) based on the center line file. I would like to combine these two files and the output should be only one shp file. Not required the center line on the final shp. All tags should be as attributes in shp. Only the tags to copy from centerlinefile to new file which is falling on the top of the filled shapes. (like spatial join in GIS) Using Products : Microstaion SS4, FME Trial Thank you Chithra

Hello FME-Users,I have a problem with my CityGML Data. I want the final results to look like clean OBJ Files. When I load them into e.g. 3Ds Max they should have no errors. I also do not want faces to be seen in my 3D software. Instead I would like to see proper polygons (if quads are possible, I want to see them; they can be made out of triangles, but these triangles should be hidden).I found this image on Wikipedia showing the different possible Elements of a mesh:   My current FME workflow looks like this:  I have buildings made out of 3 elements: Groundsurface, Wallsurface and Roofsurface. I would like to bring all of these together as a mesh (with polygons). I used the GeometryCoercer to convert the rawdata into polygons. The MeshBuilder is used to create a mesh. Here I think my polygons were converted into surfaces... Below I also provide you with my settings from the Deaggregator and the Coerser.  In 3Ds Max the buildings look like that:this looks really faulty... The files are

I have a workbench where I use a python shutdown script. This script is printing some results to the translation log window.I would like to the see the text that the python shutdown script is printing to the translation log window in a different color than black. I used some suggestions from the internet but they are not working as the translation log is not a "real" python terminal.I tried a few things without installing any library to python though. As I will be sharing the workbench with colleagues who will run the wb locally, I would prefer if it's not necessary to install a library.Another question is if it is possible to print emojis to the translation window using a python shutdown script? I know it is possible in python e.g. print("\\U0001f600") but this print statement is not working when printing it to the translation log window. Thanks.

I'm trying to build a workspace that will allow users to select a file, identify X and Y attribute names using published parameters and have it be translated AND transformed. I found the Easy Translator demo, which covers the translated part of what I'm trying to do (and is broken - I submitted an issue on GitHub). Now I'm getting stuck on the transformation part as the attribute names are read in as a feature, and the actual attribute names are coming through as col0 ... col5, etc.I need to get to a point where I can either identify what column number that the attribute names are that they listed in the published parameters, or ideally rename the col0 attribute to the actual name that it is.

As you see, value is null so it gets ORA-01449(null).So, How can I setting value collumn ?

When reading in a zip file with three text files, it results in garbled data. Any ways to fix this?

 FME 2020: How to get all the layers automatically selected when the dwg file is selected as a source file in the reader? I found out that each time a dwg file is selected (as per the screenshot below), the “select all” option is not checked by default.  How this option can automatically be checked as any dwg is selected as a source?                   

add variable in json reader from user parameter like this;

Hello FME users. I can't solve the problem of passing all the points along the shortest path. I can't make a cycle with sorting. Please help me.

I am noticing that the number of features and feature types between the Merged Feature and Individual Feature mode are quite different. My expectation was that when you select "Read All Elements", it would...read all elements, regardless if merged or individual. I compared two different files and both had more features and feature types in Merged mode. I am not specifying a view and reading geometry as non-spatial. One had 5 more FeatureTypes, the other had 13 more! Why?  Thanks,@larue​    

for an example:why is the result not is 1 but 4

coordinates of given type extract coordinates and type

I have a column in a table with names and I want to find the ones with alphanumeric string and replace it with a null value. Apparently this is erroneous data in the table.  Sample Data in the column NAME:GREEN-BROWN TREES --> Leave as isFORT/STRONG -> Leave as isPEACE RIVER --> Leave as is3T7H8HF54FD --> This is the data I want to find and replace with Null Greatly appreciated. Thank you.