Hi all, I finally resolved the issue I'll mention it here to help anyone else who has the same issue. In "Edit" mode on the workflow you add the "Workflow control->Inputs" to use the api. Final question.... Everything gets created and stored in the database, but the history column on the left of the browser doesn't get updated. Can user see if the task has completed successfully and all files uploaded from the browser? Thanks Neil From: Burdett, Neil (ICT Centre, Herston - RBWH) Sent: Thursday, 9 May 2013 2:56 PM To: galaxy-dev@lists.bx.psu.edu Subject: Getting example_watch_folder.py to work... Hi, I'm still struggling to get the example_watch_folder.py to work. So any help much appreciated. I've created a simple workflow, which essentially opens a text file and then writes out the data to a html file. My xml file: <tool id="CopyTool" name="Copy Tool"> <description>Test</description> <command interpreter="perl">$__root_dir__/tools/copy/copy.pl --input_image "$inputImage" --output_html "$output" </command> <inputs> <param format="txt" name="inputImage" type="data" label="Input Image" /> </inputs> <outputs> <data format="html" name="output" label="${tool.name} on #echo os.path.basename( str ( $inputImage.name ) )#" /> </outputs> <help> Copy Tool </help> </tool> And copy.pl: #!/usr/bin/perl use strict; use warnings; use Getopt::Long; my $input_image ; my $output_html ; # Get options from command line (i.e. galaxy) my $result = GetOptions ( "input_image=s" => \$input_image, "output_html=s" => \$output_html ) or die "Bad options"; print "input_image is: $input_image\n"; open FILE,"$input_image" or die $!; my @lines = <FILE>; close FILE; my $numOfLines = scalar @lines; # Create the output HTML, with links to the files and the square gif. # open HTML, ">", $output_html or die "Failed to create output HTML file '$output_html': $! "; print HTML<<EOF; <html> <head> <style> iframe { border: 0px; background: #eeeeee ; } </style> </head> <body> <h1>Copy Files tool</h1> <h2>Generated $numOfLines files</h2> EOF # Put direct links to each output file foreach my $sub_filename ( @lines ) { print HTML "Direct link to the $sub_filename file.<br/>\n"; } print HTML "<br/><br/><br/>\n"; print HTML "</body></html>\n" ; close HTML ; So pretty basic. I run: ./example_watch_folder.py cce1b01926646d548f6ddc32ff01aa2e http://140.253.78.234/galaxy/api/ /home/galaxy/data_input /home/galaxy/data_output "My API Import" f2db41e1fa331b3e and get the following output: {'outputs': ['a799d38679e985db'], 'history': '33b43b4e7093c91f'} I can see the sample.txt file I placed in the /home/galaxy/data_input has been put into the database: /home/galaxy/galaxy-dist/database/files/000/dataset_8.dat and: http://140.253.78.234/galaxy/api/libraries/f2db41e1fa331b3e/contents/1cd8e2f... shows the file has been uploaded: { "data_type": "txt", "date_uploaded": "2013-05-09T04:41:10.579602", "file_name": "/home/galaxy/galaxy-dist/database/files/000/dataset_8.dat", "file_size": 76, "genome_build": "?", "id": "1cd8e2f6b131e891", "ldda_id": "1cd8e2f6b131e891", "message": "", "metadata_data_lines": 8, "metadata_dbkey": "?", "misc_blurb": "8 lines", "misc_info": "uploaded txt file", "model_class": "LibraryDataset", "name": "sam2.txt", "template_data": {}, "uploaded_by": "test@test.com<mailto:test@test.com>", "uuid": null } However, looking in the histories: http://140.253.78.234/galaxy/api/histories/33b43b4e7093c91f/contents/a799d38... The output file is zero as the input file can not be found as shown below in bold ... { "accessible": true, "api_type": "file", "data_type": "html", "deleted": false, "display_apps": [], "display_types": [], "download_url": "/galaxy/api/histories/33b43b4e7093c91f/contents/a799d38679e985db/display", "file_ext": "html", "file_name": "/home/galaxy/galaxy-dist/database/files/000/dataset_9.dat", "file_size": 0, "genome_build": "?", "hid": 1, "history_id": "33b43b4e7093c91f", "id": "a799d38679e985db", "metadata_data_lines": null, "metadata_dbkey": "?", "misc_blurb": "error", "misc_info": "Thu May 9 14:42:10 2013input_image is: None\nNo such file or directory at /home/galaxy/galaxy-dist/tools/copy/copy.pl line 27.\n", "model_class": "HistoryDatasetAssociation", "name": "Copy Tool on None", "peek": null, "purged": false, "state": "error", "uuid": null, "visible": true, "visualizations": [] } Does anyone have any idea why the input_file is not passed in/obtained in the perl script? as the script obviously copies it to the database so part of the script is working? the file sam2.txt looks like this Thanks again, sorry for swamping the list with this issue Neil