Exporting Container Field to File

[Jonathan Perel]

As  you can't use the Export Field Contents step when running a script through FileMaker Server, I threw together the following ScriptMaster script. Any suggestions or improvements welcome. 

 

Write Container to File ( filePath ; containerField )

import org.apache.commons.io.*;

InputStream input;
FileOutputStream out;

try{
	input = fmpro.getContainerStream(containerField);
}
catch(e){
 return false;
}

if ( input == null ) {
 return false;
}

try{
	output= new FileOutputStream( filePath );
}
catch(e){
 return false;
}

IOUtils.copy (input, output );

//int read = 0;
//byte[] bytes = new byte[1024];
//while ((read = input.read(bytes)) != -1) {
// output.write(bytes, 0, read);
//}

return true;

[Jonathan Perel]

Oh... you'll need Apache.org's Common-IO library for this to work.

 

http://commons.apache.org/proper/commons-io/

 

[Leo Di Croce]

Hi Jonathan,

 

Thank you posting this. I think it's exactly what I'm looking for, but I'm having trouble implementing.

 

The function returns 0, and the containers fail to "export".

 

I'm trying to write two container fields to a server's temporary folder.

 

 

The variable that defines the folder:

 $PATHfolder = /private/var/folders/g6/3h4y328x1q5gc59q4hr_4xjr0000gn/T/tempKMZ_9A1902/

The variables that define the file paths:

$PATHxml = /private/var/folders/g6/3h4y328x1q5gc59q4hr_4xjr0000gn/T/tempKMZ_9A1902/sample.xml
$PATHpng = /private/var/folders/g6/3h4y328x1q5gc59q4hr_4xjr0000gn/T/tempKMZ_9A1902/19-tag.png

Does that look right?

 

Or could it be an error in my ScriptMaster module? I pasted your code, downloaded Apache Commons IO, and added five jars:

commons-io-2.4-javadoc.jar

commons-io-2.4-sources.jar

commons-io-2.4-test-sources.jar

commons-io-2.4-tests.jar

commons-io-2.4.jar

 

Not sure how to attach a screenshot, but it would show five Commons-Io selected in the Jars tab of the "Edit Script" layout.

 

Thanks again for the module. Any further help is greatly appreciated.

=L=

_____

Leo Di Croce

 

 

Feb 19, 2014 - Found a solution here.

[Jonathan Perel]

Hey there,

 

1. The only jar you should need is "Commons-Io-2.4.jar". You shouldn't need any of the other jars.

2. There are two input variables.

containerField - is the NAME of the container field. It should NOT be the value of container field itself.

filePath - is the POSIX file path destination for the file in the container field.

 

 

Hope that helps,

JP

[Ibrahim Bittar]

Hi Jonathan

 

Can you post a sample file. I'm having a hart time trying to make it work.

 

Sorry, I'm new to Java and Groovy.

 

Regards

 

Ibrahim

[Jonathan Perel]

Here's the sample file. The call to the function is:

WriteContainertoFile( "/Users/Shared/" & WriteContainerField::ContainerName ; "WriteContainerField::ContainerField" )

Also attached is a screenshot of the script in ScriptMaster and the JAR config in ScriptMaster. Its a pretty simple script to implement.

WriteContainerField.fmp12.zip

[john renfrew]

Jonathan

 

simplified code...

import org.apache.commons.io.*
 
if ( !input ) {
	return false
}

try{
	input = fmpro.getContainerStream(containerField)
	output= new FileOutputStream( filePath )
	IOUtils.copy (input, output )
}
catch(e){
	return false
}
 
return true

[Bailey Kessing]

Hmmmm. When I run this I get a PDF file that is not complete or damaged. See the two PDFs below. 

About Downloads (Original).pdf

About Downloads (Writen from Container).pdf

[Bailey Kessing]

I am having problems with this…any one else?

[john renfrew]

Just tested it and am getting the same issue as you...

It's stopping at about 299k

 

Tried it with about 3 or 4 other methods, but still the same result, indicating choking on the input side...

 

And then i tried it with other PDF files (of which I have many for test, up to 50MB)

And it all works as expected.

I suspect that source PDF is in some way not lovely on the inside...

 

Opened it with Acrobat Pro, re-saved it. All lovely. It's the PDF not the code.

 

 My code however is bad..

 

try

import org.apache.commons.io.IOUtils

try{
	input = fmpro.getContainerStream(containerField)
	if ( !input ) {
		return false
	}
	output= new FileOutputStream( filePath )
	IOUtils.copy(input, output)
        //for files over 2GB
        //IOUtils.copyLarge(input, output)
}
catch(e){
	return false
	//return e
}
 
return true

[Bailey Kessing]

Thank you so much. Can I ask about the large file option…I am a geneticist and have VERY large files sometimes. Is there some "cost" to implementing the Large file option all the time?

 

Also, I have an issue with the java code reading a file into a container with large files…is there a similar solution for that? And also reading the content of a large file! Sorry to bother you guys but I'm a scientist not great programmer and these 3 things will make my life complete!! 

[john renfrew]

add another parameter which indicates if the size is large..

 

for the other issues you might need to be using Buffered Stream methods if the files are large. posting code will help us to help you here..

import org.apache.commons.io.IOUtils

try{
	input = fmpro.getContainerStream(containerField)
	if ( !input ) {
		return false
	} //end if
	output= new FileOutputStream( filePath )
	if (sizeParam){
		//for files over 2GB
		IOUtils.copyLarge(input, output)
	} else {
		IOUtils.copy(input, output)
	} //end if
}
catch(e){
	//turn on to see what errors are
	//return e
	return false
} //end try
 
return true

[Bailey Kessing]

It's the standard "Read File Contents" module in Scriptmaster:

 

if( pathToFile == null ) throw new Exception("You must supply a pathToFile parameter");

return new InputStreamReader( new FileInputStream( pathToFile ), "utf-8" );

 

And the standard "Get File As Container" module:

 

File f = new File( pathToFile );

fileExists = f.exists();

return f;

 

I have run out of memory with both I believe with large files

[Bailey Kessing]

Also, your code from Posted 13 February 2015 - 03:03 PM works but your latest one doesn't.  I guess you didn't finish it with the size parameter part?

[TheRoth]

Have you try with Mail-it ?

 

Emai_ExportFile( fileObject; filePath {; saveMode; protectFile } )

/*
Exports file object from container field to disk [i.e. an attachment that has
been retrieved from a message and stored in FileMaker]
fileObject - File object to be exported
filePath - Path [including file name and extension] to which the file will
be saved. If path does not include file name the default file name
from object will be used.
saveMode - "strict" [default], "rename" or "overwrite"
protectFile - Used to write-protect the exported [temp] file if it should only
be used for viewing purposes by the user. "modify" [default] does
not write-protect the file. "read-only" protects it from modification.
saveMode="strict" this function fails if there is an existing file.
If saveMode="rename" plug-in adds a counter to the file name if
there is a conflict with an existing file. If saveMode="overwrite"
plug-in tries to overwrite existing file with the same name [if any].

*/

[martinspanjaard]

The free BaseElements plug-in has a similar function:

BE_ExportFieldContents ( field ; outputPath )

Exports the contents of the container field to disk at the outputPath specified. Container fields should be inserted via "Insert File". This provides a similar functionality to the Export Field Contents script step, but is also available via FileMaker Server.

 

[jrr316]

I am playing around with this, but found that it is not compatible with the Get(DocumentsPath) or Get (TemporaryPath) commands in Filemaker. Anyone have any luck with that?

 

[bcooney]

I would guess that BE expects OS paths, not FM paths.

https://www.briandunning.com/cf/902

[john renfrew]

https://baseelementsplugin.zendesk.com/hc/en-us/articles/205350417-Notes-about-File-Paths

<<The paths used by the plugin are Operating System paths, not FileMaker paths.......>> and same is true for ScriptMaster functions.

[shmert]

Here's another ScriptMaster module which doesn't require Apache IOUtils and supports FileMaker-style paths. It uses some internal ScriptMaster utilities instead, I'll add it to the ScriptMaster list of modules. Returns number of bytes transferred. The 4096 is the buffer size. 

Parameters:

• containerName
• path

Script:

import com.prosc.io.IOUtils;
import com.prosc.fmkit.PluginUtils;

inputStream = fmpro.getContainerStream(containerName)
if (inputStream == null) throw new IOException(containerName + " is empty");

try {
	File f = PluginUtils.fileForFMPath(path);
	FileOutputStream out = new FileOutputStream(f);
	IOUtils.writeInputToOutput(inputStream, out, 4096);
} finally {
	out.close();
	inputStream.close();
}

 

 

[Thorsten Kaltenborn]

Hi shmert,

On 17. August 2017 at 8:25 PM, shmert said:

Here's another ScriptMaster module which doesn't require Apache IOUtils and supports FileMaker-style paths. It uses some internal ScriptMaster utilities instead, I'll add it to the ScriptMaster list of modules. Returns number of bytes transferred. The 4096 is the buffer size. 

can you explain me, which extensions or java files I have to install and where to put them? If I try your script, it returns just "error".

Regards,

Thorsten