diff options
| -rwxr-xr-x | src/extract-elm.sh | 2 | ||||
| -rwxr-xr-x | src/extract-generic.sh | 55 | ||||
| -rwxr-xr-x | src/extract-nix.sh | 2 | ||||
| -rwxr-xr-x | src/extract-sh.sh | 2 | ||||
| -rwxr-xr-x | src/extract.sh | 13 |
5 files changed, 55 insertions, 19 deletions
diff --git a/src/extract-elm.sh b/src/extract-elm.sh index 9fe1d4c..7ab3e17 100755 --- a/src/extract-elm.sh +++ b/src/extract-elm.sh @@ -3,7 +3,7 @@ set -efu TREE_GREPPER_LANGUAGE=elm \ -TREE_GREPPER_QUERY='([(line_comment) (block_comment)]+)' \ +TREE_GREPPER_QUERY='([(line_comment) (block_comment)])' \ LINE_COMMENT_START='--' \ BLOCK_COMMENT_START='{-|\?' \ BLOCK_COMMENT_END='-}' \ diff --git a/src/extract-generic.sh b/src/extract-generic.sh index cadbb30..f4aef22 100755 --- a/src/extract-generic.sh +++ b/src/extract-generic.sh @@ -4,12 +4,57 @@ set -efu input_file=${1-/dev/stdin} +# note: merge sibling comments into matched comment blocks, as tree-sitter does +# not allow for matching sibling nodes +function merge_matches() { + matches=$(jq -c .matches[]) + new_matches= + last_match= + new_match= + while read -r cur_match; do + if test -z "$new_match"; then + new_match=$cur_match + fi + if ! test -z "$last_match"; then + last_end_row=$(echo "$last_match" | jq -r .end.row) + cur_start_row=$(echo "$cur_match" | jq -r .start.row) + if test "$(($last_end_row + 1))" = "$cur_start_row"; then + new_text=$(echo "$new_match" | jq -r .text) + cur_text=$(echo "$cur_match" | jq -r .text) + text=$(printf "%s\n%s" "$new_text" "$cur_text") + end=$(echo "$cur_match" | jq -c .end) + new_match=$( + echo "$new_match" | + # XXX "$end" seems to be forbidden as variable name + jq -c \ + --argjson end_ "$end" \ + --arg text "$text" \ + '. + { + "end": $end_, + "text": $text + }' + ) + else + new_matches=$(printf "%s\n%s" "$new_matches" "$new_match") + new_match=$cur_match + fi + fi + last_match=$cur_match + done <<< "$matches" + if ! test -z "$new_match"; then + new_matches=$(printf "%s\n%s" "$new_matches" "$new_match") + fi + new_matches=$(echo "$new_matches" | jq -cs .) + echo "$line" | jq -c --argjson matches "$new_matches" '. + { "matches": $matches }' +} + tree-grepper \ --query "$TREE_GREPPER_LANGUAGE" "$TREE_GREPPER_QUERY" \ --format json $input_file | - jq 'sort_by(.file)' | - jq '.[]' --indent 0 | + jq -c '.[]' | while read -r line; do + echo "$line" | merge_matches + done | while read -r line; do file=$(echo "$line" | jq -r .file) file_type=$(echo "$line" | jq -r .file_type) items=$(echo "$line" | @@ -32,19 +77,19 @@ tree-grepper \ text=$(echo "$item" | jq .match.text -r | - sed 's/^'"$LINE_COMMENT_START"' *TODO *//' | + sed 's/^'"$LINE_COMMENT_START"'\( *TODO *\| \?\)//g' | { if test -z "${BLOCK_COMMENT_START-}"; then cat else - sed 's/^'"$BLOCK_COMMENT_START"' *TODO *//' + sed 's/^'"$BLOCK_COMMENT_START"'\( *TODO *\| \?\)//' fi } | { if test -z "${BLOCK_COMMENT_END-}"; then cat else - sed 's/ *'"$BLOCK_COMMENT_END"'$//' + sed 's/ *'"$BLOCK_COMMENT_END"' *$//' fi } ) diff --git a/src/extract-nix.sh b/src/extract-nix.sh index 335cbba..575131d 100755 --- a/src/extract-nix.sh +++ b/src/extract-nix.sh @@ -3,7 +3,7 @@ set -efu TREE_GREPPER_LANGUAGE=nix \ -TREE_GREPPER_QUERY='((comment)+)' \ +TREE_GREPPER_QUERY='(comment)' \ LINE_COMMENT_START='#' \ BLOCK_COMMENT_START='\/\*' \ BLOCK_COMMENT_END='\*\/' \ diff --git a/src/extract-sh.sh b/src/extract-sh.sh index c31df74..a8bb1ff 100755 --- a/src/extract-sh.sh +++ b/src/extract-sh.sh @@ -3,6 +3,6 @@ set -efu TREE_GREPPER_LANGUAGE=sh \ -TREE_GREPPER_QUERY='((comment)+)' \ +TREE_GREPPER_QUERY='(comment)' \ LINE_COMMENT_START='#' \ "$(dirname "$0")"/extract-generic.sh "$@" diff --git a/src/extract.sh b/src/extract.sh index 6422694..07346fc 100755 --- a/src/extract.sh +++ b/src/extract.sh @@ -80,17 +80,8 @@ git ls-files --cached --exclude-standard --other | head -n 1) created_at=$(git show $first_commit --no-patch --format=%ad) - heading=$(echo "$text" | sed '/^$/Q' | sed 's/.*TODO //') - # TODO body is not extracted in shell scripts - # - # @assigned aforemny - # @assigned kirchner@posteo.de - # @scheduled 2023-10-03 - body=$(echo "$text" | tail -n +$(($(echo "$heading" | wc -l) + 2)) | \ - awk -F '[^ ]' ' - NR == 1 {n = length($1)} - {sub("^ {1,"n"}", ""); print}' - ) + heading=$(echo "$text" | sed -n '0,/^$/p') + body=$(echo "$text" | tail -n +$(($(echo "$heading" | wc -l) + 2))) echo "$item" | jq -c \ --arg body "$body" \ |