git-deltafy-script

   1 #!/bin/bash
   2
   3 # Example script to deltafy an entire GIT repository based on the commit list.
   4 # The most recent version of a file is the reference and previous versions
   5 # are made delta against the best earlier version available. And so on for
   6 # successive versions going back in time.  This way the increasing delta
   7 # overhead is pushed towards older versions of any given file.
   8 #
   9 # The -d argument allows to provide a limit on the delta chain depth.
  10 # If 0 is passed then everything is undeltafied.  Limiting the delta
  11 # depth is meaningful for subsequent access performance to old revisions.
  12 # A value of 16 might be a good compromize between performance and good
  13 # space saving.  Current default is unbounded.
  14 #
  15 # The --max-behind=30 argument is passed to git-mkdelta so to keep
  16 # combinations and memory usage bounded a bit.  If you have lots of memory
  17 # and CPU power you may remove it (or set to 0) to let git-mkdelta find the
  18 # best delta match regardless of the number of revisions for a given file.
  19 # You can also make the value smaller to make it faster and less
  20 # memory hungry.  A value of 5 ought to still give pretty good results.
  21 # When set to 0 or ommitted then look behind is unbounded.  Note that
  22 # git-mkdelta might die with a segmentation fault in that case if it
  23 # runs out of memory.  Note that the GIT repository will still be consistent
  24 # even if git-mkdelta dies unexpectedly.
  25
  26 set -e
  27
  28 depth=
  29 [ "$1" == "-d" ] && depth="--max-depth=$2" && shift 2
  30
  31 function process_list() {
  32         if [ "$list" ]; then
  33                 echo "Processing $curr_file"
  34                 echo "$head $list" | xargs git-mkdelta $depth --max-behind=30 -v
  35         fi
  36 }
  37
  38 curr_file=""
  39
  40 git-rev-list HEAD |
  41 git-diff-tree -r -t --stdin |
  42 awk '/^:/ { if ($5 == "M" || $5 == "N") print $4, $6;
  43             if ($5 == "M") print $3, $6 }' |
  44 LC_ALL=C sort -s -k 2 | uniq |
  45 while read sha1 file; do
  46         if [ "$file" == "$curr_file" ]; then
  47                 list="$list $sha1"
  48         else
  49                 process_list
  50                 curr_file="$file"
  51                 list=""
  52                 head="$sha1"
  53         fi
  54 done
  55 process_list
  56
  57 curr_file="root directory"
  58 head=""
  59 list="$(
  60         git-rev-list HEAD |
  61         while read commit; do
  62                 git-cat-file commit $commit |
  63                 sed -n 's/tree //p;Q'
  64         done
  65         )"
  66 process_list
  67