Mercurial > repos > devteam > windowsplitter
changeset 3:1011f028f09d draft default tip
"planemo upload for repository https://github.com/galaxyproject/tools-devteam/tree/master/tools/windowsplitter commit 6960b37b09831a1b83f18775677b83a3cc5c2cd0"
| author | devteam |
|---|---|
| date | Wed, 03 Jun 2020 13:44:45 +0000 |
| parents | 1fb384fb950e |
| children | |
| files | windowSplitter.py windowSplitter.xml |
| diffstat | 2 files changed, 99 insertions(+), 84 deletions(-) [+] |
line wrap: on
line diff
--- a/windowSplitter.py Tue Oct 13 12:36:17 2015 -0400 +++ b/windowSplitter.py Wed Jun 03 13:44:45 2020 +0000 @@ -1,39 +1,46 @@ #!/usr/bin/env python - """ Split into windows. usage: %prog input size out_file -l, --cols=N,N,N,N: Columns for chrom, start, end, strand in file """ +from __future__ import print_function import sys -from galaxy import eggs -import pkg_resources -pkg_resources.require( "bx-python" ) from bx.cookbook import doc_optparse -from galaxy.tools.util.galaxyops import * + + +# Default chrom, start, end, strand cols for a bed file +BED_DEFAULT_COLS = 0, 1, 2, 5 + -def stop_err( msg ): - sys.stderr.write( msg ) - sys.exit() +def parse_cols_arg(cols): + """Parse a columns command line argument into a four-tuple""" + if cols: + # Handle case where no strand column included - in this case, cols + # looks something like 1,2,3, + if cols.endswith(','): + cols += '0' + col_list = [int(x) - 1 for x in cols.split(",")] + return col_list + else: + return BED_DEFAULT_COLS def main(): # Parsing Command Line here - options, args = doc_optparse.parse( __doc__ ) - + options, args = doc_optparse.parse(__doc__) + try: - chr_col_1, start_col_1, end_col_1, strand_col_1 = parse_cols_arg( options.cols ) + chr_col_1, start_col_1, end_col_1, strand_col_1 = parse_cols_arg(options.cols) inp_file, winsize, out_file, makesliding, offset = args winsize = int(winsize) offset = int(offset) makesliding = int(makesliding) - except: - stop_err( "Data issue, click the pencil icon in the history item to correct the metadata attributes of the input dataset." ) - - fo = open(out_file,'w') + except Exception: + sys.exit("Data issue, click the pencil icon in the history item to correct the metadata attributes of the input dataset.") skipped_lines = 0 first_invalid_line = 0 @@ -41,43 +48,42 @@ if offset == 0: makesliding = 0 - for i, line in enumerate( file( inp_file ) ): - line = line.strip() - if line and line[0:1] != "#": - try: - elems = line.split('\t') - start = int(elems[start_col_1]) - end = int(elems[end_col_1]) - if makesliding == 0: - numwin = (end - start)/winsize - else: - numwin = (end - start)/offset - if numwin > 0: - for win in range(numwin): - elems_1 = elems - elems_1[start_col_1] = str(start) - elems_1[end_col_1] = str(start + winsize) - fo.write( "%s\n" % '\t'.join( elems_1 ) ) - if makesliding == 0: - start = start + winsize - else: - start = start + offset - if start+winsize > end: - break - except: - skipped_lines += 1 - if not invalid_line: - first_invalid_line = i + 1 - invalid_line = line - - fo.close() + with open(out_file, 'w') as fo, open(inp_file) as fi: + for i, line in enumerate(fi): + line = line.strip() + if line and line[0:1] != "#": + try: + elems = line.split('\t') + start = int(elems[start_col_1]) + end = int(elems[end_col_1]) + if makesliding == 0: + numwin = (end - start) // winsize + else: + numwin = (end - start) // offset + if numwin > 0: + for _ in range(numwin): + elems_1 = elems + elems_1[start_col_1] = str(start) + elems_1[end_col_1] = str(start + winsize) + fo.write("%s\n" % '\t'.join(elems_1)) + if makesliding == 0: + start = start + winsize + else: + start = start + offset + if start + winsize > end: + break + except Exception: + skipped_lines += 1 + if not invalid_line: + first_invalid_line = i + 1 + invalid_line = line if makesliding == 1: - print 'Window size=%d, Sliding=Yes, Offset=%d' % ( winsize, offset ) + print('Window size=%d, Sliding=Yes, Offset=%d' % (winsize, offset)) else: - print 'Window size=%d, Sliding=No' % (winsize) + print('Window size=%d, Sliding=No' % (winsize)) if skipped_lines > 0: - print 'Skipped %d invalid lines starting with #%d: "%s"' % ( skipped_lines, first_invalid_line, invalid_line ) + print('Skipped %d invalid lines starting with #%d: "%s"' % (skipped_lines, first_invalid_line, invalid_line)) if __name__ == "__main__":
--- a/windowSplitter.xml Tue Oct 13 12:36:17 2015 -0400 +++ b/windowSplitter.xml Wed Jun 03 13:44:45 2020 +0000 @@ -1,36 +1,47 @@ -<tool id="winSplitter" name="Make windows" version="1.0.0"> - <description></description> - <command interpreter="python">windowSplitter.py $input $size $out_file1 ${wintype.choice} ${wintype.offset} -l ${input.metadata.chromCol},${input.metadata.startCol},${input.metadata.endCol},${input.metadata.strandCol}</command> - <inputs> - <!--<param label="Genome" name="dbkey" type="genomebuild"/>--> - <param format="interval" name="input" type="data" label="Select data"/> - <param name="size" type="integer" value="500" label="Window size"/> - <conditional name="wintype"> - <param name="choice" type="select" label="Make sliding windows?"> - <option value="0" selected="true">No</option> - <option value="1">Yes</option> - </param> - <when value="0"> - <param name="offset" type="hidden" value="0" /> - </when> - <when value="1"> - <param name="offset" type="integer" value="10" label="Offset size"/> - </when> - </conditional> - </inputs> - <outputs> - <data format="interval" name="out_file1" metadata_source="input"/> - </outputs> - <tests> - <test> - <param name="input" value="4.bed"/> - <param name="size" value="5000"/> - <param name="choice" value="1"/> - <param name="offset" value="4000"/> - <output name="out_file1" file="4_windows.bed"/> - </test> - </tests> - <help> +<tool id="winSplitter" name="Make windows" version="1.0.1"> + <description></description> + <requirements> + <requirement type="package" version="0.8.8">bx-python</requirement> + </requirements> + <command> + python $__tool_directory__/windowSplitter.py + $input + $size + $out_file1 + ${wintype.choice} + ${wintype.offset} + -l ${input.metadata.chromCol},${input.metadata.startCol},${input.metadata.endCol},${input.metadata.strandCol} + </command> + <inputs> + <!--<param label="Genome" name="dbkey" type="genomebuild"/>--> + <param format="interval" name="input" type="data" label="Select data"/> + <param name="size" type="integer" value="500" label="Window size"/> + <conditional name="wintype"> + <param name="choice" type="select" label="Make sliding windows?"> + <option value="0" selected="true">No</option> + <option value="1">Yes</option> + </param> + <when value="0"> + <param name="offset" type="hidden" value="0" /> + </when> + <when value="1"> + <param name="offset" type="integer" value="10" label="Offset size"/> + </when> + </conditional> + </inputs> + <outputs> + <data format="interval" name="out_file1" metadata_source="input"/> + </outputs> + <tests> + <test> + <param name="input" value="4.bed"/> + <param name="size" value="5000"/> + <param name="choice" value="1"/> + <param name="offset" value="4000"/> + <output name="out_file1" file="4_windows.bed"/> + </test> + </tests> + <help> .. class:: infomark @@ -98,7 +109,5 @@ chr22 3000 4000 NM_174568 0 + chr22 3500 4500 NM_174568 0 + - </help> - - + </help> </tool>