I use the ZipExtractor to do this. Can you explain why this does not work for you?

As an alternative a Reader / FeatureReader can read directly from zip.

See Reading from Archive Files.

Thank you for your answer !

I'll try to explain my problem more clearly !

I have an archive wich I want to download from a ftp (so I use "FTPCaller")

The content of the archive wich I require is located on the 4th sub folder of the archive so I want to explore this archive using FME to only get the SHP I need.

Here are some screenshots of my script and transformers...

When I launch the script with those transformers activated I get those errors:

When I run the script with the FTPCaller activating (using a previously downloaded of the .7z) I get no error but 0 feature output...

I don't understand where my mistake is and I've tried to resolve this problem for the entire day so I hope you have a solution !

Regards,

Andy METAYER

This probably has to do with FeatureCaching? Do you see the downloaded file in the explorer?

I do ! And as you can see in the 4th screen, feature catching is activated!

I tried to run the script after disabling FeatureCatching but there is no change... Would this be possible for you to try to make the same script as I did (creator > ftpcaller > zipextractor) with this ftp link "ftp://RPG_ext:quoojaicaiqu6ahD@ftp3.ign.fr/RPG_2-0__SHP_LAMB93_R11_2020-01-01.7z" (Origin: https://geoservices.ign.fr/rpg) so you can tell me if problems comes from me or something else ?

Thank you in advance

I do not understand why but I am able to reproduce your issue with this archive.

ERROR |Specified zip file or destination root folder is invalid.

But when using a FeatureReader Esri Shapefile, dataset 

@Value(_downloaded_file_path)t**\*.shp]

it does read the data. Sample attached.

I still have one more question ! How did you set the Output ports per feature type ? I tried to reproduce your script to understand it but can't manage to set the output ports.

When I try to access "Feature Types to Read" it opens a dialog box saying that I need to select the dataset

And when I just copy the features types from your script (ILOTS_ANONYMES PARCELLES_GRAPHIQUES) and try to validate another dialog box is opened

I don't understand, If I launch your script everything works perfectly but when I try to reproduce it it doesn't work at all wich is a big problem as long as I'll have to do this kind of scripts again in the future...

If I try to avoid the FTPCaller by using a downloaded .7z as dataset it works perfectly so maybe the problems come from the caller ?

If you have an idea please let me know,

Regards

I think that it doesn't work with the attribute paths, I don't get why it works on yours but it's the only clue I have..

For the Generating Output Ports part, I manually unzipped the shapes and selected those as source dataset.

But you have to do that every time you edit the FeatureReader. Personally, I prefer to use the Generic Output Port, split the features based on fme_feature_type and expose attributes using an AttributeExposer in the workspace later. But I mostly work on server workspaces trying to handle any dataset a user throws at it.

Oh ok I get it ! Thank your for your answer, I'll try t use this method !