Maintenance Commands

  -prepare    [release|report] Compare daily update to archive
  -ignore     Ignore contents of object in -prepare comparisons
  -damaged    Report UIDs containing damaged embedded HTML tags
  -missing    Print list of missing identifiers
  -unique     File of UIDs for skipping all but last version

Miscellaneous

  -head       Print before everything else
  -tail       Print after everything else
  -hd         Print before each record
  -tl         Print after each record

Update Candidate Report

  cd "$MASTER/Pubmed"
  gunzip -c *.xml.gz | xtract -strict -compress -format flush |
  rchive -prepare report -ignore DateRevised -archive "$MASTER/Archive" \
    -index MedlineCitation/PMID -pattern PubmedArticle

Unnecessary Update Removal

  cd "$MASTER/Pubmed"
  gunzip -c *.xml.gz | xtract -strict -compress -format flush |
  rchive -prepare release -ignore DateRevised -archive "$MASTER/Archive" -index MedlineCitation/PMID \
    -head "<PubmedArticleSet>" -tail "</PubmedArticleSet>" -pattern PubmedArticle |
  xtract -format indent -xml '<?xml version="1.0" encoding="UTF-8" ?>' \
    -doctype '<!DOCTYPE PubmedArticleSet SYSTEM "http://dtd.nlm.nih.gov/ncbi/pubmed/out/pubmed_180101.dtd">' |
  gzip > newupdate.xml.gz

Get Archive UID List

  pm-uids "$MASTER/Archive" > complete.uid

Reconstruct List of Versioned PMIDs

  cd "$MASTER/Pubmed"
  rm -f "$MASTER/Archive/versioned.uid"
  gunzip -c *.xml.gz |
  xtract -strict -pattern PubmedArticle -if MedlineCitation/PMID@Version -gt 1 \
    -element MedlineCitation/PMID > "$MASTER/Archive/versioned.uid"

Reconstruct Release Files

  split -a 3 -l 30000 release.uid uids-
  n=1
  for x in uids-???
  do
    xmlfile=$(printf "pubmed18n%04d.xml.gz" "$n")
    n=$((n+1))
    echo "$xmlfile"
    cat "$x" |
    rchive -fetch "$MASTER/Archive" -head "<PubmedArticleSet>" -tail "</PubmedArticleSet>" |
    xtract -strict -format indent -xml '<?xml version="1.0" encoding="UTF-8" ?>' \
      -doctype '<!DOCTYPE PubmedArticleSet SYSTEM "http://dtd.nlm.nih.gov/ncbi/pubmed/out/pubmed_180101.dtd">' |
    gzip > "$xmlfile"
  done
  rm -rf uids-???

Damaged Embedded HTML Tag Search

  for fl in *.xml.gz
  do
    echo "$fl"
    gunzip -c "$fl" | rchive -mixed -damaged -index MedlineCitation/PMID^Version -pattern PubmedArticle
  done

  grep -v pubmed18n | grep AMPER | cut -f 1,6

Reconstruct Term List Keys

  rm -f "$MASTER/Postings/sections.txt"
  find "$MASTER/Postings" -name "*.mst" |
  sed -e 's,.*/\(.*\)\.mst,\1,' |
  sort | uniq > "$MASTER/Postings/sections.txt"

Generate Term List Paths

  find "$MASTER/Postings" -name "*.trm" |
  sed -e 's,\(.*/\)\(.*\.trm\),\1 \2,' |
  sort -k 2 | uniq | tr -d ' '