<!-- Author: Davide Albanese - Fondazione Edmud Mach, 2013 -->

<tool id="check_id_map" name="Check ID Map" version="1.6.0-2">
  <description>
    Checks user's metadata mapping file for required data, valid
    format
  </description>
  <requirements>
    <requirement type="binary">check_id_map.py</requirement>
  </requirements>
  <command>
    check_id_map.py
    
    -m $mapping_fp
    
    #if str($char_replace):
    -c $char_replace
    #end if
    
    #if $not_barcoded:
    -b
    #end if
    
    #if $variable_len_barcodes:
    -B
    #end if
    
    #if $disable_primer_check:
    -p
    #end if
    
    #if str($added_demultiplex_field):
    -j $added_demultiplex_field
    #end if
    ;
    rm `basename $mapping_fp .txt`'.html'
    ;
    rm overlib.js
    ;
    mv `basename $mapping_fp .txt`'.log' $out_log
    ;
    mv `basename $mapping_fp .txt`'_corrected.txt' $out_txt
  </command>
  <inputs>
    <param name="mapping_fp" label="Metadata mapping file" optional="False" type="data" format="tabular"/>
  
    <param name="char_replace" value="_" label="Character used to replace invalid characters found in the mapping file. Must be a valid character (alphanumeric, period, or underscore)" optional="False" type="text"/>
    
    <param name="not_barcoded" label="Set if barcodes are not present. BarcodeSequence header still required" selected="False" type="boolean"/>
    
    <param name="variable_len_barcodes" label="Set if variable length barcodes are present to suppress warnings about barcodes of unequal length"  selected="False" type="boolean"/>
    
    <param name="disable_primer_check" label="Set to disable checks for primers. LinkerPrimerSequence header still required" selected="False" type="boolean"/>
    
    <param name="added_demultiplex_field" label="Use to add a field to use in the mapping file as additional demultiplexing (can be used with or without barcodes).  All combinations of barcodes/primers and the these fields must be unique. The fields must contain values that can be parsed from the fasta labels such as 'plate=R_2008_12_09'.  In this case, 'plate' would be the column header and 'R_2008_12_09' would be the field data (minus quotes) in the mapping file.  To use the run prefix from the fasta label, such as '>FLP3FBN01ELBSX', where 'FLP3FBN01' is generated from the run ID, use '-j run_prefix' and set the run prefix to be used as the data under the column header 'run_prefix'" optional="False" type="text" value=""/>
  </inputs>
  <outputs>
    <data format="txt" name="out_log" label="${tool.name} on ${mapping_fp.name}"/>
    <data format="tabular" name="out_txt" label="Corrected ${mapping_fp.name}"/>
  </outputs>
  <help>
    Check ID Map checks::
    
    1. The BarcodeSequence, LinkerPrimerSequences, and ReversePrimer fields 
       have valid IUPAC DNA characters, and BarcodeSequence characters
       are non-degenerate (error)

    2. The SampleID, BarcodeSequence, LinkerPrimerSequence, and Description
       headers are present (error)

    3. There are not duplicate header fields (error)
    
    4. There are not duplicate barcodes (error)

    5. Barcodes are of the same length.  Suppressed when
       variable_len_barcode flag is passed (warning)

    6. The headers do not contain invalid characters (alphanumeric and 
       underscore only) (warning)

    7. The data fields do not contain invalid characters (alphanumeric, 
       underscore, space, and +-%./:,; characters) (warning)

    8. SampleID fields are MIENS compliant (only alphanumeric
       and . characters). (warning)

    9. There are no duplicates when the primer and variable length 
       barcodes are appended (error)

    10. There are no duplicates when barcodes and added demultiplex 
       fields (-j option) are combined (error)

    11. Data fields are not found beyond the Description column (warning)

    Details about the metadata mapping file format can be found here:
    http://www.qiime.org/documentation/file_formats.html#metadata-mapping-files

    Errors and warnings are saved to a log file.  Errors can be caused
    by problems with the headers, invalid characters in barcodes or
    primers, or by duplications in SampleIDs or barcodes.
    
    Warnings can arise from invalid characters and variable length
    barcodes that are not specified with the --variable_len_barcode.
    Warnings will contain a reference to the cell (row,column) that
    the warning arose from.

    In addition to the log file, a 'corrected_mapping' file will be
    created.  Any invalid characters will be replaced with '.'
    characters in the SampleID fields (to enforce MIENS compliance)
    and text in other data fields will be replaced with the character
    specified by the -c parameter, which is an underscore '_' by
    default.

    If pooled primers are used, separate with a comma. For instance,
    a pooled set of three 27f primers (used to increase taxonomic
    coverage) could be specified in the LinkerPrimerSequence fields as
    such:
    AGGGTTCGATTCTGGCTCAG,AGAGTTTGATCCTGGCTTAG,AGAATTTGATCTTGGTTCAG
  </help>
</tool>