#!/bin/sh
#$ -S /bin/sh

# このシェルは以下のコマンドで投入
# sh chipatlas/sh/listForClassify.sh  # <<=== コマンド (DDBJ)

projectDir=`echo $0| sed 's[/sh/listForClassify.sh[['`
rm -rf tmpDirForFastq

if [ "$1" = "" ]; then
  qstat| awk '$3 == "TimeCourse" && $4 == "okishinya" {print "qdel " $1}'| sh
  rm -rf tmpDirForListForClassify
  rm -rf ncbi
  mkdir tmpDirForListForClassify
  
  for Genome in `ls $projectDir/results`; do
    tagDir=$projectDir/results/$Genome/tag
    rm -rf $tagDir
    mkdir $tagDir
    N=`echo $projectDir/results/$Genome/summary/*RX*.txt $projectDir/results/$Genome/bmap/summary/*RX*.txt| tr ' ' '\n'| tee $projectDir/results/$Genome/summary/summary_list.txt| wc -l| awk '{printf 1 + int($1/1000)}'`
    for num in `seq $N`; do
      qsub -o /dev/null -e /dev/null $projectDir/sh/listForClassify.sh $projectDir $Genome $num
    done
  done
  

  for Genome in `ls $projectDir/results`; do
    for Type in ag ct; do # ct: cell type, ag: antigen
      qsub -N classifyList -o /dev/null -e /dev/null $projectDir/sh/listForClassify.sh tsv $Genome $Type $projectDir
    done
  done

  bin/wait4qsub classifyLi

  for org in `cat $projectDir/sh/preferences.txt| awk '$1 == "Genome"'| cut -f2| tr ' ' '\n'| cut -d '=' -f1`; do
    for Type in ag ct; do
      for Genome in `echo $org| tr ',' ' '`; do
        cat $projectDir/classification/$Type"_"Statistics.$Genome.tab
      done| awk -F '\t' '!a[$4]++' > $projectDir/classification/Statistics.tmp
      for Genome in `echo $org| tr ',' ' '`; do
        cat $projectDir/classification/Statistics.tmp > $projectDir/classification/$Type"_"Statistics.$Genome.tab
      done
      rm $projectDir/classification/Statistics.tmp
    done
  done
  exit

elif [ "$1" = "tsv" ]; then
# ct_Statistics.$Genome.tab を作成。これをもとに Mac で分類作業をする。
  bin/wait4qsub listForCla
  
  Genome=$2
  Type=$3
  projectDir=$4
  TSV=$projectDir/classification/$Type"_"Statistics-$Genome-tab.tsv
  rm -rf $projectDir/results/$Genome/summary/summary_list.txt
  
  if [ $Type = "ag" ]; then
    echo tmpDirForListForClassify/$Genome.tag*txt| xargs cat| sort > $projectDir/results/$Genome/tag/allTags.txt
  fi
  
  echo tmpDirForListForClassify/$Genome.tag*txt| xargs cat| awk -F '\t' -v TYPE=$Type '{
    if (TYPE == "ag") print $1 "\t" $2 "\t" $5  # $1=SRX, $2=SRA, $3=フル抗体名
    if (TYPE == "ct") print $1 "\t" $2 "\t" $7  # $1=SRX, $2=SRA, $3=フル細胞名
  }'| sed 's/ /__SPACENOKAWARI__/g'| sort -k2| sort -k3| uniq -c1| awk '{
    Cell[$4] = $4
    numSrx[$4] = numSrx[$4] + $1
    SRX[$4] = $2
    SRA[$4] = SRA[$4]$3
  } END {
    for (key in Cell) print gsub("RA", "", SRA[key]) "\t" numSrx[key] "\thttps://www.ncbi.nlm.nih.gov/sra?term=" SRX[key] "\t" Cell[key] "\t" Cell[key]
  }' | sort -nr -k1| sort -nr -k2| sed 's/__SPACENOKAWARI__/ /g'| awk -F '\t' -v TSV=$TSV '
  BEGIN {
    while ((getline < TSV) > 0) arr[$4] = $5
  } {
    if (length(arr[$4]) > 0) $5 = arr[$4]  # 旧 tsv と照合する。
    printf "%s\t%s\t%s\t%s\t%s\n", $1, $2, $3, $4, $5
  }'| sort -t $'\t' -k5| awk '
  BEGIN {
    print "SRA\tSRX\tMGI\told\tnew\tjudge"
  } {print}' > $projectDir/classification/$Type"_"Statistics.$Genome.tab
                                    # $1 SRA の数, $2 SRX の数, $3 SRX の例, $4 細胞組織名, $5 細胞組織名
  exit
fi

####################################################################################################################################
#                                                             以下、qsub モード
####################################################################################################################################
projectDir=$1
Genome=$2
num=$3

smallestQ=`ls $projectDir/results/$Genome/| grep Bed[0-9]| tr -d Bed|sort -n|head -n1` # 例 05
tagDir=$projectDir/results/$Genome/tag

# 英数字以外で用いられる文字
# µ́Ž ³¼»­¾¥ɛ–—⁰�∆≥fi＇Ⅱ‑!"#%&'()*+,-./:;=?@\^_`{|}~€†‡ˆŠŒ‘’“”•–—™š›œž，．´¨‐…‘’“”（）
# −±×÷°′″℃¢£△→¬¶ΒΔΦΩαβγδκλμσφАМСб¸¯˚¡¦º©®™¤ÆØæıøßÁÀÄÂĄÅÃÈÊÌÎÑÖÕŔŠáàäâăąåãčçéèêęíïñóöôřúüûý
# IGV で表示できないものは URL エンコーディングに変換
#  % > %25
#  + > %2B
#  ; > %3B
#  = > %3D
#  " > %22


# FastQ サイズがゼロでないBed を入力
metaDataFull=`ls $projectDir/lib/metadata/NCBI_SRA_Metadata_Full_20*.metadata.tab| tail -n1`
cat $projectDir/results/$Genome/summary/summary_list.txt| awk -v num=$num 'NR > (num-1)*1000 && NR <= num*1000'| xargs cat | awk -v metaDataFull=$metaDataFull '{
  if ($3 > 0) x[$1]++
} END {
  while ((getline < metaDataFull) > 0) {
    if (x[$1] > 0) {
      fn = "'$projectDir'/results/'$Genome'/metadata/" $1 ".meta.txt"
      print $0 > fn
      close(fn)
      print $0
    }
  }
  close(metaDataFull)
}'| sed 's/-/___HYPHEN___/g'\
| sed 's/[^0-9a-zA-Z\t"#%&\(\)'\''*+,-./:;=?\^_{}~" ]//g'\
|tr -d '"'| sed 's/___HYPHEN___/-/g'| awk -F '\t' -v Dir=$tagDir -v PD=$projectDir ' #"
                                                                  # 削除される記号は @ † ‡ § ¶ ダブルクオート パイプ など。 この時点でスペースはそのまま。
BEGIN {
  agAttributes = PD"/sh/ag_attributes.txt"
  ctAttributes = PD"/sh/ct_attributes.txt"
  while ((getline < agAttributes) > 0) {
    nrAb++
    x[$1] = nrAb
  }
  while ((getline < ctAttributes) > 0) {
    nrCl++
    y[$1] = nrCl
  }
} {
  for (i=18; i<=NF; i++) {
    split($i, atribute, "=")
    if (x[atribute[1]] > 0) AB[x[atribute[1]]] = atribute[2]   # 抗体名の取得
    if (y[atribute[1]] > 0) CL[y[atribute[1]]] = atribute[2]   # 細胞組織名の取得
  }
  # 抗体名の決定
  arr["Ab"] = "ND"
  if ($4 == "DNase-Hypersensitivity") {
    arr["Ab"] = "DNase-seq"
  } else if ($4 == "Bisulfite-Seq") {
    arr["Ab"] = "Bisulfite-seq"
  } else if ($4 == "ATAC-seq") {
    arr["Ab"] = "ATAC-seq"
  } else {
    for (i=1; i<=nrAb; i++) {
      if (AB[i] != "") {
        a++
        allAb = allAb "|" AB[i]
        if (a == 1) arr["Ab"] = AB[i]
      }
    }
  }
  # 細胞組織名の決定
  arr["Cell"]="ND"
  for (i=1; i<=nrCl; i++) {
    if (CL[i] != "") {
      c++
      allCell = allCell "|" CL[i]
      if (c == 1) arr["Cell"] = CL[i]
    }
  }
  # タイトルの決定
  if ($3 != "xxx") arr["Tytle"] = $3
  else             arr["Tytle"] = "ND"
  # タイトルから GSMxxx を除く
  if (substr(arr["Tytle"], 1, 3) == "GSM") {
    arr["Tytle"]=substr(arr["Tytle"], match(arr["Tytle"], /_/)+1)
  }
  # 特殊符合を除く
  for( key in arr) {
    gsub("&lt;", "<", arr[key])
    gsub("&gt;", ">", arr[key])
    gsub("&amp;", "and", arr[key])
  }
  # 長いタイトルを短縮
  fullTtle = arr["Tytle"]
  if (length(arr["Tytle"]) > 50) longTytle=substr(arr["Tytle"], 1, 47)"..."
  else                           longTytle=arr["Tytle"]
  # 長い抗体、細胞、タイトルを短縮
  for( key in arr) {
    if (length(arr[key]) > 25) arr[key]=substr(arr[key], 1, 22)"..."
  }
  # 全属性の整理
  if (sub("\\|", "", allAb) == 0) allAb = "NoAb"
  if (arr["Ab"] == "DNase-seq") allAb = "DNase-seq"
  if (arr["Ab"] == "ATAC-seq") allAb = "ATAC-seq"
  if (arr["Ab"] == "Bisulfite-seq") allAb = "Bisulfite-seq"
  if (sub("\\|", "", allCell) == 0) allCell = "NoCell"
  # 出力
  outFn = Dir"/"$1".tag.txt"
  tagStr = $1 "\t" $12 "\t" fullTtle "\t" longTytle "\t" allAb "\t" arr["Ab"] "\t" allCell "\t" arr["Cell"] "\t" $4
  print tagStr > outFn
  print tagStr
  delete atribute
  delete AB
  delete CL
  delete arr
  allAb = ""
  allCell = ""
  a = 0
  c = 0
  close(outFn)
  # 例: SRX021071	H9_Oct4_technical_replicate_2	H9_Oct4_technical_replicate_2	Oct4	Oct4	Embyonic_stem_cells	Embyonic_stem_cells	ChIP-Seq
    # $1 = SRX
    # $2 = SRA
    # $3 = フルタイトル (GSM 除く)
    # $4 = 短縮タイトル (50 字以内)
    # $5 = 抗原抗体フル (パイプ 区切り)
    # $6 = 短縮 抗原抗体 (どれかひとつ; 25字以内)
    # $7 = 細胞組織フル (パイプ 区切り)
    # $8 = 短縮 細胞組織 (どれかひとつ; 25字以内)
    # $9 = ChIP-Seq or DNase-Hypersensitivity
}' > tmpDirForListForClassify/$Genome.tag.$num.txt
exit