fix extracting body for shell issues

The underlying problem is that tree-sitter does not allow for matching
sibling nodes. So any issues comprised of line comments (vs. block
comments) suffered from this issue, independent of the language (shell).

Thus, we manually merge such sibling matches after running tree-grepper.

1 files changed, 50 insertions, 5 deletions

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
           }
         )