Debian Codegraph @

share/analysis-scripts/concat-csv.sh

raw
1 #!/bin/bash -eu
2
3 set -o pipefail
4
5 header_column=""
6 delete_first=""
7 sort=""
8 files=""
9 show_usage=""
10
11 while [[ $# > 0 ]]
12 do
13 case $1 in
14 --add-header-column)
15 header_column="yes"
16 ;;
17
18 --delete-first)
19 delete_first="$2"
20 shift
21 ;;
22
23 --sort)
24 sort="yes"
25 ;;
26
27 -h|--help)
28 show_usage="yes"
29 ;;
30
31 *)
32 files="$files $1"
33 ;;
34 esac
35 shift
36 done
37
38 target=""
39 sources=""
40 read target sources <<<$files
41
42 if [ -z "$sources" -o -n "$show_usage" ]
43 then
44 echo "Usage: $0 TARGET SOURCE1 ..."
45 echo "Concat and/or transform CSV files."
46 echo ""
47 echo "The following arguments can be given:"
48 echo " --add-header-column, add a header column consisting of the base name of"
49 echo " the CSV file"
50 echo " --delete-first N delete the first N columns"
51 echo " --sort sort the CSV file"
52 echo " -h, --help prints this help and quits"
53 exit 1
54 fi
55
56
57 # Read header line
58 read first rest <<<$sources
59 header=$(head --quiet --lines 1 $first)
60 if [ -n "$header_column" ]
61 then
62 header="project $header"
63 fi
64 if [ -n "$delete_first" ]
65 then
66 header=$(cut -f1-$delete_first --complement <<<"$header")
67 fi
68 # Read sources
69 {
70 for f in $sources
71 do
72 if [ -n "$header_column" ]
73 then
74 base=$(basename "$f")
75 project="${base%%.*}"
76 sed -e '1d' -e "s/^/$project\t/" "$f";
77 else
78 sed -e '1d' "$f";
79 fi
80 done
81 } |
82 # Remove first columns
83 {
84 if [ -n "$delete_first" ]
85 then
86 cut -f1-$delete_first --complement
87 else
88 cat
89 fi
90 } |
91 # Remove blanks
92 sed '/^$/d' |
93 # Sort the file
94 {
95 echo "$header"
96 if [ -n $sort ]
97 then
98 sort -u
99 else
100 cat
101 fi
102 } > "$target"
103