Transformers

Sorting happens mostly by one attribute across a set of features. I would like to sort multiple attribute values (all of type datetime) across one feature and pick out the most recent date (of that particular feature) and assign it to a new attribute. I can imagine a lot of complex Testers, but it seems so obvious to use sorting of some kind. Any suggestions?

Is there a way to control the size of labels within AutoCAD blocks during DWG rendering? The text looks smaller and is misaligned when rendered by FME (and therefore the inspector and, more importantly, the PNG writer), but it's the right size and aligned properly in AutoCAD.Bad in FME (for Linux):vs. Good in AutoCAD: 

The DateTimeParse function is driving me mad (in FME 2018.1)!What is the problem with this...?Failed to evaluate expression '@DateTimeParse(3-Jun-2020,%Y%m%d,repair)'. Result is set to nullI have also tried @DateTimeFormat and @DateTimeCast and things like putting quotes around the input etc. Nothing works.I want to use the function because I want to convert a value of "3 June 2020" to 2020-06-03 to write into a date field in SQL Server. I have already use the string functions to extract the substrings for day (3), 3 character month (Jun) and year (2020). I could go the long winded way and write these to separate attributes then use the DateTimeConverter transformer, but why should I?

mes données de départ  Tableau 1  numéroheure 1 11302124031850414005153069307155081530  ce que je veux obtenir Tableau 2numéroheure1 11:30212:403 18:50414:00515:30609:30715:5083: 30 J'ai essayer avec le transformer StringRemplacer @Left (@Value (heure), 2): @ Right (@Value (heure), 2)et j'obtiens :j'aimerais que quand y'a 3 nombre, il ajoute les " : " derrière le premier nombre ex: 930 devient 9:30 merci.

Hi there, I would like to use a Conditional Region Growing algorithm from the PCL library (See tutorial) inside a workbench. I make some point cloud preprocessing using vector features and then, at some point, I would like to introduce this algorithm and go further with my processing. How can I make use of that? I have tried with Python bindings using PythonCaller but it is not so helpful in this case.Is there a chance that this solution will work? :-> Preprocessing... -> SystemCaller(call .exe file triggering C++ algorithm) -> grab results (still within FME) -> process further Would you like to share some other possible solution?The biggest advantage of FME is that you can construct pipeline from A to Z and you do not have to interrupt it. I can, of course, export results after preprocessing and then that get back to FME but it is not what I would like to have at the end.

Bonjour,J'ai deux tables qui ont des noms de colonnes identiques et j'aimerais ajouter l'une en dessous de l'autre et pas à la suite. merci.exemple: tableau premier: nom prénomville paiemax romainParisFrance  deuxième tableau: nom prénom villepaielucievernLondres Royaume-Uni J'aimerais avoir ce résultat: nom  prénomvillepays maxromainparisFrance lucievernlondresUK

Hi, I have date and time value in a column which look like '19960108031100'. Now I want to split into two columns i.e. date and time. Could anyone guide on how to do that? RegardsYasir

Hi, On FME Cloud I tried to execute the statement "FME PYTHON" and a python script. I am using a SystemCaller, which works on my Windows FME Desktop. The Python script is from a third party that should be run over the outcome of a workspace, and I don't intend to adapt it. However, the workspace did not run on FME Cloud. After some experimenting I learned that - since Cloud runs on Ubuntu - this still works, however, the path to the FME executable should be /opt/fme-engine-2019/fme. My question: Is this a correct way to run a Python script? Or is there a way to just run Python script from the PythonCaller?

Hi guys,I have a line file (1 line) and a point file (Multiple pairs of start/end points). I'm trying to clip multiple line segments from a single line based on a start and end points with matching ID attribute. I have been trying PointOnLineOverlayer but it adds the attributes from only the first point to every line. When I generate a list based on the ID the list includes every point ID on every line, not just the points that the new line intersects.  Is there another way to achieve this? Any help would be greatly aprreciated. Cheers, Dave.

Hi I am new to using FME.  I have imported an IFC file in a reader.  I would like to take the ifcSpace geometry (an Aggregate Solid) and break it down to its constituent faces.  Any ideas please?  Thanks  Gareth

Hi im trying to find out if its possible to apply a photoshop curve to a large raster data set. the file format of the curve is .acv, which cant be read into FME. I Was wondering if it were possible to convert it into something that can be read in. regards 

Good Day,Still learning and I need help on this. THANKS A LOT IN ADVANCEHAVE THIS DMDLOCD6_1_2020D1_13_20D1_20_2012039GDL0.1980.2010.28412039MEX1.1811.2011.6912039MTY0.8950.911.28112039TIJ1.2761.2390.297 First Row is HeadersNEED THIS 12039GDLD6_1_20200.19812039GDLD1_13_200.20112039GDLD1_20_200.28412039MEXD6_1_20201.18112039MEXD1_13_201.20112039MEXD1_20_201.6912039MTYD6_1_20200.89512039MTYD1_13_200.9112039MTYD1_20_201.28112039TIJD6_1_20201.27612039TIJD1_13_201.23912039TIJD1_20_200.297

Hi, I have different DXF/DWG file and want to calculate the BoundingBox of the geometries for every file.The BoundingBox should be a seperate layer in the output file.What is the best way to solve my problem? Thanks and best regards

I think I have the easiest question ever but I've wasted too much time on it and it just won't work.I just want to create a new attribute for each record with the result of the sum of 3 other values. The result of nieuw + renovatie + stroomopwaarts. It looks very easy but i have already tried everything and nothing works. What am I missing? 

Hi, How can I take a point that has overlaps at a distance of 150 ft from the pointonpointoverlayer and take those overlapping points to form a circle?  I used the https://knowledge.safe.com/questions/25031/offset-points-on-a-circle.html reference and I was able to build a circle from eriks but I go from 798 records to over 10,000. 

Not really a question but a warning...Reading data directly from inside a ZIP file is great, but there are a few issues that can catch you out. Firstly, you need enough space in the FME_TEMP folder to extract the entire contents of the ZIP file, which can be annoying if you only want one file. Read and vote for my idea on this here:https://knowledge.safe.com/idea/100199/reading-from-zip-files-extract-only-file-specified.htmlSecondly, I've discovered that if you read data from inside a ZIP file using the File and Pathnames Reader, the metadata about the files will be based on the extracted data in FME_TEMP, not the original ZIP file location or properties of the files inside (such as the date created). So in the screenshot below, the path is the FME_TEMP folder, not the original source folder. And the created, modified and accessed dates are all today, because that's when I ran the Workspace and therefore when the data was extracted from the ZIP to the FME_TEMP folder.

I've attached my xml and xfMap file. I read 'Consectetur' and 'Purus' as featuretypes. I want to add the attributes from 'Lorem' (version) and 'Ipsum' (date) to my featuretypes. Is this possible with xfMap? I already tried some things with references but had no result.Another problem I encounter is that I cannot read geometry by using 'Location' in my Feature Content Map. I don't know what I'm doing wrong. I've used the geometry part from the TOP10NL Reader.

Hi all,I am currently loading data into a SQL Server database from geodatabases. When I load in the data in via arccatalog, i successfully see the data.  Now when I do this via FME, i seem to get a different icon and then the spatial index has to be recalculated. The data is visible but this icon appears at every point. Does this mean the spatial index has not been created? Is there a easy automated way, where the spatial index gets created when loading into a enterprise geodatabase SQL server?

 HI, I have a dwg file with a lot of layers which can be seen in Civil 3D: Not every Layer have a block entitiy in it yetI am trying to rename layername in FME, but I am 1. Very unsure on how to set up my reader, since there are many options and parameters to choose form when reading a dwg file, I can choose to open per Layername or geometry etc, this is the settings I have now which just shows up as "<All>" and chose Autocad dwg/dwf as reader I DO NOT whant to change specific entities layername, which is the setting now, the problem is that some layers do not have entities so only the number of entities are read and I can choose "autocad_layer" and see that value but only for 3 layers.  I want to read in ALL layernames, and use a attributemanager and say if layername="T-KST" rename to "Roads" for example and be able to do that for all my layer names (and possible the description of the layer) which can be seen to the right in the first ima

I am new to FME. I think this is a simple quetion for you guys. I have a SQL table with two fields Longitude and Latitude and other attributes in SQL Server database. I need to it to Feature class into same database. how to do that? how to setup Longitude and Latitude fields as geometry fields in the reader?

I have a polyline where each point has its own heightfor these polylines I need to recreate the geometry (a pipe is round with a diameter of attributes, a pipe is rectangular with different ratios of width and height from attributes)but difficulties appearedthe pipe has fillets at the endand pipes of square section are either built on a plane or along non-joined segmentsFriends I ask for help on how to build a solid geometry, taking into account different heights, and make the end of the pipe flat

Part of my FMW gets an area, creates a grid of points within, converts them to blocks via the dwg styler and then clips it from a negative buffer of itself. This is an attempt to create a symbolic fill for an area.Blocks still expand beyond the Area as seen below (1st screenshot is the current result, 2nd screenshot is desired output). Is there a way to clip the blocks to the area? I would also like to randomise the fill pattern rather than use a strict grid, so if anyone has an idea on how to do this I would be very interested.    

Hello, everyone. I have points (attributes: X, Y, ID) and I want to merge points that are under 50 m distance and new point should have attributes ( new_ID, X, Y, all id of the points from which the new point was created) and if there is a duplicate point, also making one with attributes ( new_ID, X, Y, all id of the points from which the new point was created) Thanks a lot.

Hi there, I have Incident_no in the csv which I don't want to include into the calculation. Data is coming from the database. Is there any way that I exclude the records that contain incident number present in CSV? I tried with Feature joiner but it didn't give an option of not equal.

Using the logic from this post and this post, I wanted to rename my log file using the Pyhton shutdown script.For some reason, os.rename works only with python 2.7 and not python 3.x.I tried single, double, and triple quotes; used raw, and I can't seem to find a good way to correct os.rename. Used FME_macro and just a string for the source file. Nothing worked so far.# Import modulesimport fme, datetime, time, os# Get current date and timedateToday = datetime.date.today().strftime('%Y/%m/%d')timeNow = datetime.datetime.now().time().strftime('%H.%M.%S')logger = open(FME_LogFileName,'a')logger.write('Date: {}\n'.format(dateToday))logger.write('Time: {}\n'.format(timeNow))logger.write('ProcessID: {}\n'.format(fme.processID))logger.write('Status: {}\n'.format(fme.status))logger.write('Workspace duration was {} seconds\n'.format(fme.elapsedRunTime))logger.write('Total feature read {}\n'.format(fme.totalFeaturesRead))logger.write('Total features written {}\n'.format(fme.totalFeaturesWritten)