Mercurial > repos > mingchen0919 > rmarkdown_collection_builder
changeset 6:a0c8b2b25774 draft
planemo upload for repository https://github.com/statonlab/docker-GRReport/tree/master/my_tools/rmarkdown_collection_builder commit 88ca36a41aa577ba888cee39cf81b176bf7e68db
| author | mingchen0919 |
|---|---|
| date | Tue, 26 Sep 2017 13:33:22 -0400 |
| parents | 66022ba24ac7 |
| children | abf9cbd7578f |
| files | collection_list.Rmd collection_list.xml collection_list_paired.Rmd collection_list_paired.xml collection_list_paired_render.R collection_paired.Rmd collection_paired.xml collection_paired_render.R |
| diffstat | 8 files changed, 181 insertions(+), 13 deletions(-) [+] |
line wrap: on
line diff
--- a/collection_list.Rmd Tue Sep 26 00:16:09 2017 -0400 +++ b/collection_list.Rmd Tue Sep 26 13:33:22 2017 -0400 @@ -1,5 +1,5 @@ --- -title: 'Collection builder (collection type: list)' +title: 'Build collection: a list of datasets' output: html_document: number_sections: true
--- a/collection_list.xml Tue Sep 26 00:16:09 2017 -0400 +++ b/collection_list.xml Tue Sep 26 13:33:22 2017 -0400 @@ -1,4 +1,4 @@ -<tool id="rmarkdown_collection_list" name="Collection type: list" version="1.0.0"> +<tool id="rmarkdown_collection_list" name="Collection type (list)" version="1.0.0"> <requirements> <requirement type="package" version="1.15.0.6-0">pandoc</requirement> <requirement type="package" version="1.20.0">r-getopt</requirement> @@ -7,7 +7,7 @@ <requirement type="package" version="0.5.0">r-dplyr</requirement> </requirements> <description> - Create a 'list' dataset collection from a set of files in Galaxy history. + Create a list of datasets collection from files in history. </description> <stdio> <regex match="Execution halted"
--- a/collection_list_paired.Rmd Tue Sep 26 00:16:09 2017 -0400 +++ b/collection_list_paired.Rmd Tue Sep 26 13:33:22 2017 -0400 @@ -1,5 +1,5 @@ --- -title: 'Collection builder (collection type: list:paired)' +title: 'Build collection: a list of paired datasets' output: html_document: number_sections: true @@ -41,7 +41,7 @@ original_files = list.files('./forward_files_directory') # create new files names from original file names new_files = gsub('\\.dat', '_forward\\.FORMAT', original_files) -new_files = paste0('input_', 1:length(original_files), '_forward.FORMAT') +new_files = paste0('input', 1:length(original_files), '_forward.FORMAT') new_files # rename files file.copy(paste0('./forward_files_directory/', original_files), @@ -61,10 +61,9 @@ original_files = list.files('./reverse_files_directory') # create new files names from original file names new_files = gsub('\\.dat', '_reverse\\.FORMAT', original_files) -new_files = paste0('input_', 1:length(original_files), '_reverse.FORMAT') +new_files = paste0('input', 1:length(original_files), '_reverse.FORMAT') new_files # rename files file.copy(paste0('./reverse_files_directory/', original_files), paste0('./files_directory/', new_files)) -``` - +``` \ No newline at end of file
--- a/collection_list_paired.xml Tue Sep 26 00:16:09 2017 -0400 +++ b/collection_list_paired.xml Tue Sep 26 13:33:22 2017 -0400 @@ -1,4 +1,4 @@ -<tool id="rmarkdown_collection_list_paired" name="Collection type: list:paired" version="1.0.0"> +<tool id="rmarkdown_collection_list_paired" name="Collection type (list:paired)" version="1.0.0"> <requirements> <requirement type="package" version="1.15.0.6-0">pandoc</requirement> <requirement type="package" version="1.20.0">r-getopt</requirement> @@ -7,7 +7,7 @@ <requirement type="package" version="0.5.0">r-dplyr</requirement> </requirements> <description> - Create a 'list:paired' dataset collection from a set of files in Galaxy history. + Create a list of paired dataset collection from files in history. </description> <stdio> <regex match="Execution halted" @@ -40,14 +40,14 @@ <inputs> <param type="data" name="forward_file_list" multiple="true" label="forward inputs"/> <param type="data" name="reverse_file_list" multiple="true" label="reverse inputs"/> - <param type="text" name="format" label="the format of input files (e.g., fastq, fastqsanger, fasta)"/> + <param type="text" name="format" label="the format (extension) of input files (e.g., fastq, fastqsanger, fasta)"/> <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="Collection builder (list:paired collection)" /> <collection type="list:paired" name="list_collection"> - <discover_datasets pattern="__name_and_ext__" directory="files_directory" visible="true"/> + <discover_datasets pattern="(?P<identifier_0>[^_]+)_(?P<identifier_1>[^_]+)\.(?P<ext>[^\._]+)?" directory="files_directory"/> </collection> </outputs> </tool> \ No newline at end of file
--- a/collection_list_paired_render.R Tue Sep 26 00:16:09 2017 -0400 +++ b/collection_list_paired_render.R Tue Sep 26 13:33:22 2017 -0400 @@ -82,4 +82,4 @@ render('list_paired_collection.Rmd', output_file = opt$list_paired_collection_html) -#-------4. manipulate outputs ----------------------------- +#-------4. manipulate outputs ----------------------------- \ No newline at end of file
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/collection_paired.Rmd Tue Sep 26 13:33:22 2017 -0400 @@ -0,0 +1,29 @@ +--- +title: 'Build collection: a paired datasets' +output: + html_document: + number_sections: true + toc: true + theme: cosmo + highlight: tango +--- + +```{r setup, include=FALSE, warning=FALSE, message=FALSE} +knitr::opts_chunk$set( + echo = ECHO +) +``` + +# Command line arguments + +```{r 'command line arguments'} +str(opt) +``` + +# Rename files + +```{r} +file.copy("FORWARD_INPUT", './OUTPUT_DIR/forward.FORMAT') +file.copy("REVERSE_INPUT", './OUTPUT_DIR/reverse.FORMAT') +``` +
--- a/collection_paired.xml Tue Sep 26 00:16:09 2017 -0400 +++ b/collection_paired.xml Tue Sep 26 13:33:22 2017 -0400 @@ -0,0 +1,52 @@ +<tool id="rmarkdown_collection_paired" name="Collection type (paired)" version="1.0.0"> + <requirements> + <requirement type="package" version="1.15.0.6-0">pandoc</requirement> + <requirement type="package" version="1.20.0">r-getopt</requirement> + <requirement type="package" version="1.2">r-rmarkdown</requirement> + <requirement type="package" version="0.3.5">r-htmltools</requirement> + <requirement type="package" version="0.5.0">r-dplyr</requirement> + </requirements> + <description> + Create a paired dataset collection from files in history. + </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." /> + </stdio> + <command> + <![CDATA[ + Rscript '${__tool_directory__}/collection_paired_render.R' + -L $forward_input + -R $reverse_input + -e $echo + -f $format + + -r $report + -d $report.files_path + + -t '${__tool_directory__}/collection_paired.Rmd' + ]]> + </command> + <inputs> + <param type="data" name="forward_input" multiple="false" label="forward input"/> + <param type="data" name="reverse_input" multiple="false" label="reverse input"/> + <param type="text" name="format" label="the format (extension) of input files (e.g., fastq, fastqsanger, fasta)"/> + + <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="Collection builder (paired collection)" /> + <collection type="paired" name="list_collection"> + </collection> + </outputs> +</tool> \ No newline at end of file
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/collection_paired_render.R Tue Sep 26 13:33:22 2017 -0400 @@ -0,0 +1,88 @@ +##======= 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) +}) + +# 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$FORWARD_INPUT = c('forward_input', 'L', '1', 'character') +spec_list$REVERSE_INPUT = c('reverse_input', 'R', '1', 'character') +spec_list$ECHO = c('echo', 'e', '1', 'character') +spec_list$FORMAT = c('format', 'f', '1', 'character') +##--------2. output report and outputs -------------- +spec_list$OUTPUT_HTML = c('paired_collection_html', 'r', '1', 'character') +spec_list$OUTPUT_DIR = c('paired_collection_dir', 'd', '1', 'character') +##--------3. Rmd templates in the tool directory ---------- +spec_list$PAIRED_COLLECTION_RMD = c('paired_collection_rmd', 't', '1', 'character') + +spec = t(as.data.frame(spec_list)) +opt = getopt(spec) +##====== End of arguments handling ========== + +#------ Load libraries --------- +library(rmarkdown) +library(htmltools) +library(dplyr) + +#----- 1. create the report directory ------------------------ +system(paste0('mkdir -p ', opt$paired_collection_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 paired_collection.Rmd ----------------------- +readLines(opt$paired_collection_rmd) %>% + (function(x) { + gsub('ECHO', opt$echo, x) + }) %>% + (function(x) { + gsub('FORMAT', opt$format, x) + }) %>% + (function(x) { + gsub('FORWARD_INPUT', opt$forward_input, x) + }) %>% + (function(x) { + gsub('REVERSE_INPUT', opt$reverse_input, x) + }) %>% + (function(x) { + gsub('OUTPUT_DIR', opt$paired_collection_dir, x) + }) %>% + (function(x) { + fileConn = file('paired_collection.Rmd') + writeLines(x, con=fileConn) + close(fileConn) + }) + +#------ 3. render all Rmd files -------- +render('paired_collection.Rmd', output_file = opt$paired_collection_html) + + +#-------4. manipulate outputs -----------------------------