changeset 11:d4a9c31d0046 draft

planemo upload for repository https://github.com/statonlab/docker-GRReport/tree/master/my_tools/rmarkdown_fastq_dump commit affb362adeb095389646b0c51738d24c2de3838d-dirty

--- a/fastq_dump_pe.Rmd	Wed Sep 27 16:17:40 2017 -0400
+++ b/fastq_dump_pe.Rmd	Wed Sep 27 16:43:52 2017 -0400
@@ -0,0 +1,48 @@
+---
+title: 'Fastq-dump: download and extract paired end reads into FASTQ/FASTA file'
+output:
+    html_document:
+      number_sections: true
+      toc: true
+      theme: cosmo
+      highlight: tango
+---
+
+```{r setup, include=FALSE, warning=FALSE, message=FALSE}
+knitr::opts_chunk$set(
+  echo = ECHO
+)
+```
+
+# Command line arguments
+
+```{r 'command line arguments'}
+str(opt)
+```
+
+# Download and extract reads
+
+```{r 'download and extract reads'}
+# create a directory to store read files
+dir.create('read_files_directory')
+# download and extract reads
+sra_accessions = strsplit(gsub(',', ' ', 'SRA_ACCESSION'), ' ')[[1]]
+sra_accessions = sra_accessions[sra_accessions != '']
+# loop through SRA accessions to download and extract reads.
+for(id in sra_accessions) {
+  if('FORMAT' == 'fasta') {
+    command = paste0('fastq-dump --fasta ', '-O read_files_directory ')
+  } else {
+    command = paste0('fastq-dump ', '-O read_files_directory ')
+  }
+  
+  if('SPLIT_FILES' == TRUE) {
+    command = paste0(command, '--split_files ', id)
+  } else {
+    command = paste0(command, id)
+  }
+  print(command)
+  system(command = command)
+}
+```
+

--- a/fastq_dump_pe.xml	Wed Sep 27 16:17:40 2017 -0400
+++ b/fastq_dump_pe.xml	Wed Sep 27 16:43:52 2017 -0400
@@ -12,46 +12,43 @@
         collection.
     </description>
     <stdio>
-        <regex match="Execution halted"
-               source="both"
-               level="fatal"
-               description="Execution halted." />
-        <regex match="Error in"
-               source="both"
-               level="fatal"
-               description="An undefined error occured, please check your intput carefully and contact your administrator." />
-        <regex match="Fatal error"
-               source="both"
-               level="fatal"
-               description="An undefined error occured, please check your intput carefully and contact your administrator." />
+        <!--All stderr are redirected to a file. "XXX" is used to match with nothing-->
+        <regex match="XXX"
+               source="stderr"
+               level="warning"
+               description="Check the warnings_and_errors.txt file for more details."/>
     </stdio>
     <command>
         <![CDATA[
-            Rscript '${__tool_directory__}/fastq_dump_se_render.R'
-                -i $sra_accession
+            Rscript '${__tool_directory__}/fastq_dump_pe_render.R'
+                -i '$sra_accession'
                 -e $echo
                 -f $format
+                -p $split_files
 
                 -r $report
                 -d $report.files_path
+                -s $sink_message
 
-                -t '${__tool_directory__}/fastq_dump_se.Rmd'
+                -t '${__tool_directory__}/fastq_dump_pe.Rmd'
         ]]>
     </command>
     <inputs>
         <param type="text" name="sra_accession" label="SRR/DRR/ERR accessions" optional="false"
                help="A list of SRR/DRR/ERR accessions separated by comma or space. e.g. SRR6077558,ERR343809"/>
-        <param type="boolean" name="format" truevalue="fastq" falsevalue="fasta" checked="false" label="output files in fastq (true) or fasta (false)?"/>
+        <param type="boolean" name="format" truevalue="fastq" falsevalue="fasta" checked="false"
+               label="output files in fastq (true) or fasta (false)?"/>
         <param type="boolean" name="split_files" truevalue="TRUE" falsevalue="FALSE" checked="false" label="Split files"
-               help="Dump each read into separate file.Files
-                                   will receive suffix corresponding to read
-                                   number"/>
+               help="Dump each read into separate file. Files will receive suffix corresponding to read number"/>
 
-        <param type="boolean" name="echo" truevalue="TRUE" falsevalue="FALSE" checked="false" label="Display analysis code in report?" />
+        <param type="boolean" name="echo" truevalue="TRUE" falsevalue="FALSE" checked="false"
+               label="Display analysis code in report?"/>
     </inputs>
     <outputs>
-        <collection type="list" name="list_collection">
-            <discover_datasets pattern="__name_and_ext__" directory="read_files_directory" />
+        <data format="html" name="report" label="Fastq-dump report" />
+        <collection type="list" name="list_collection" label="Fastq-dump (paired end reads)">
+            <discover_datasets pattern="(?P&lt;identifier_0&gt;[^_]+)_(?P&lt;identifier_1&gt;[^_]+)\.(?P&lt;ext&gt;[^\._]+)?" directory="read_files_directory"/>
         </collection>
+        <data format="txt" name="sink_message" label="Warnings and Errors" from_work_dir="warnings_and_errors.txt" />
     </outputs>
 </tool>
\ No newline at end of file

--- a/fastq_dump_pe_render.R	Wed Sep 27 16:17:40 2017 -0400
+++ b/fastq_dump_pe_render.R	Wed Sep 27 16:43:52 2017 -0400
@@ -0,0 +1,100 @@
+##======= Handle arguments from command line ========
+# setup R error handline to go to stderr
+options(show.error.messages=FALSE,
+        error=function(){
+          cat(geterrmessage(), file=stderr())
+          quit("no", 1, F)
+        })
+
+# we need that to not crash galaxy with an UTF8 error on German LC settings.
+loc = Sys.setlocale("LC_MESSAGES", "en_US.UTF-8")
+
+# suppress warning
+options(warn = -1)
+
+options(stringsAsFactors=FALSE, useFancyQuotes=FALSE)
+args = commandArgs(trailingOnly=TRUE)
+
+suppressPackageStartupMessages({
+  library(getopt)
+  library(tools)
+})
+
+#/////////////////////// SINK WARNINGS AND ERRORS TO A FILE FOR DEBUGGING ///////////
+zz = file('warnings_and_errors.txt')
+sink(zz)
+sink(zz, type = 'message')
+
+# column 1: the long flag name
+# column 2: the short flag alias. A SINGLE character string
+# column 3: argument mask
+#           0: no argument
+#           1: argument required
+#           2: argument is optional
+# column 4: date type to which the flag's argument shall be cast.
+#           possible values: logical, integer, double, complex, character.
+##------- 1. input data ---------------------
+spec_list=list()
+spec_list$SRA_ACCESSION = c('sra_accession', 'i', '1', 'character')
+spec_list$FORMAT = c('format', 'f', '1', 'character')
+spec_list$ECHO = c('echo', 'e', '1', 'character')
+spec_list$SPLIT_FILES = c('split_files', 'p', '1', 'character')
+##--------2. output report and outputs --------------
+spec_list$REPORT_HTML = c('report_html', 'r', '1', 'character')
+spec_list$OUTPUT_DIR = c('output_dir', 'd', '1', 'character')
+spec_list$SINK_OUTPUT = c('sink_output', 's', '1', 'character')
+##--------3. Rmd templates in the tool directory ----------
+spec_list$FASTQ_DUMP_SE_RMD = c('fastq_dump_pe_rmd', 't', '1', 'character')
+
+spec = t(as.data.frame(spec_list))
+opt = getopt(spec)
+
+#------ Load libraries ---------
+library(rmarkdown)
+library(htmltools)
+library(dplyr)
+
+#----- 1. create the report directory ------------------------
+system(paste0('mkdir -p ', opt$output_dir))
+
+#----- 2. generate Rmd files with Rmd templates --------------
+#   a. templates without placeholder variables:
+#         copy templates from tool directory to the working directory.
+#   b. templates with placeholder variables:
+#         substitute variables with user input values and place them in the working directory.
+
+#----- 01 fastq_dump_pe.Rmd -----------------------
+readLines(opt$fastq_dump_pe_rmd) %>%
+  (function(x) {
+    gsub('SRA_ACCESSION', opt$sra_accession, x)
+  }) %>%
+  (function(x) {
+    gsub('FORMAT', opt$format, x)
+  }) %>%
+  (function(x) {
+    gsub('ECHO', opt$echo, x)
+  }) %>%
+  (function(x) {
+    gsub('SPLIT_FILES', opt$split_files, x)
+  }) %>%
+  (function(x) {
+    gsub('OUTPUT_DIR', opt$output_dir, x)
+  }) %>%
+  (function(x) {
+    fileConn = file('fastq_dump_pe.Rmd')
+    writeLines(x, con=fileConn)
+    close(fileConn)
+  })
+
+#------ 3. render all Rmd files --------
+render('fastq_dump_pe.Rmd', output_file = opt$report_html)
+
+
+#-------4. manipulate outputs -----------------------------
+
+
+
+
+
+sink()
+#/////////// END OF SINK OUTPUT ///////////////////////////
\ No newline at end of file

--- a/fastq_dump_se.Rmd	Wed Sep 27 16:17:40 2017 -0400
+++ b/fastq_dump_se.Rmd	Wed Sep 27 16:43:52 2017 -0400
@@ -36,7 +36,7 @@
     command = paste0('fastq-dump ', '-O read_files_directory ', id)
   }
   print(command)
-  #system(command = command)
+  system(command = command)
 }
 ```
 

--- a/fastq_dump_se.xml	Wed Sep 27 16:17:40 2017 -0400
+++ b/fastq_dump_se.xml	Wed Sep 27 16:43:52 2017 -0400
@@ -12,6 +12,7 @@
         collection.
     </description>
     <stdio>
+        <!--All stderr are redirected to a file. "XXX" is used to match with nothing-->
         <regex match="XXX"
                source="stderr"
                level="warning"
@@ -38,8 +39,8 @@
         <param type="boolean" name="echo" truevalue="TRUE" falsevalue="FALSE" checked="false" label="Display analysis code in report?" />
     </inputs>
     <outputs>
-        <data format="html" name="report" label="Fastq-dump (single end reads)" />
-        <collection type="list" name="list_collection">
+        <data format="html" name="report" label="Fastq-dump report" />
+        <collection type="list" name="list_collection" label="Fastq-dump (single end reads)">
             <discover_datasets pattern="__name_and_ext__" directory="read_files_directory" />
         </collection>
         <data format="txt" name="sink_message" label="Warnings and Errors" from_work_dir="warnings_and_errors.txt" />