Clear where non-American English appears

Previously, the hook could notify us when non-American English words
were used in a commit, but it did not output the exact words that needed
correction. Additionally, the warning line number was always set to a
constant value of 1.

This commit enhances the functionality by providing the exact line
number and the corresponding word that triggered the warning. This
allows users to quickly and clearly identify which word is causing the
issue.

Implementation Details:

The function 'get_match_position' is designed to find the first
occurrence of a target string in a multi-line text. It takes the text,
target string, starting line, and starting column as input, and returns
the line and column where the target is found.

The 'get_all_match_positions' function iterates through multiple target
words, using 'get_match_position' to locate each word’s first
occurrence. The function maintains a cursor ('start_line' and
'start_col'), ensuring that searches for subsequent words continue from
the last found position. The output format is 'target: line,column"',
making it easier to pinpoint problematic words.

Change-Id: If6cbb943ac5ee5b450486686f32ad55ed1d4f234

Author: Dennis40816

scripts/commit-msg.hook

@@ -124,6 +124,54 @@ read_commit_message() {
   done < $COMMIT_MSG_FILE
 }
 
+### Get the position (line and column) of the first occurrence of a target string.
+# Usage: get_match_position "$text" "$target" [start_line] [start_col]
+# Parameters:
+#   text       - multiline text to search in.
+#   target     - string to search for.
+#   start_line - (optional) starting line number (default: 1).
+#   start_col  - (optional) starting column number (default: 1).
+get_match_position() {
+  local text="$1" target="$2"
+  local start_line="${3:-1}" start_col="${4:-1}"
+  awk -v t="$target" -v sl="$start_line" -v sc="$start_col" '{
+    if (NR < sl) next;
+    if (NR == sl) {
+      pos = index(substr($0, sc), t);
+      if (pos) { print NR, pos+sc-1; exit }
+    } else {
+      pos = index($0, t);
+      if (pos) { print NR, pos; exit }
+    }
+  }' <<< "$text"
+}
+
+### For each target (one per line) in a multiline string,
+### search the text and call add_warning with the target's first match.
+# Parameters:
+#   $1 - multiline text to search in.
+#   $2 - targets (each line is a target string).
+get_all_match_positions() {
+  local text="$1"
+  local targets="$2"
+  local start_line=1
+  local start_col=1
+  local target result line col
+
+  while IFS= read -r target; do
+    [ -z "$target" ] && continue
+    result=$(get_match_position "$text" "$target" "$start_line" "$start_col")
+
+    [ -z "$result" ] && continue
+    read line col <<< "$result"
+    echo "$target: $line,$col"
+
+    # Update global progress for subsequent searches.
+    start_line="$line"
+    start_col=$((col + 1))
+  done <<< "$targets"
+}
+
 #
 # Validate the contents of the commmit msg agains the good commit guidelines.
 #
@@ -350,6 +398,8 @@ done
 
   FULL_COMMIT_MSG=$(sed '/^#/d;/^[[:space:]]*$/d;/^[[:space:]]*Change-Id:/d' "$COMMIT_MSG_FILE" | \
                     sed -E "s@${URL_REGEX#^}@@g")
+  FULL_COMMIT_MSG_WITH_EMPTY=$(sed '/^#/d;/^[[:space:]]*Change-Id:/d' "$COMMIT_MSG_FILE" | \
+                    sed -E "s@${URL_REGEX#^}@@g")
   # Extended list of abusive words (case-insensitive).
   # Adjust the list as needed.
   ABUSIVE_WORDS_REGEX='\b(fuck|fucking|dick|shit|bitch|asshole|cunt|motherfucker|damn|crap|dumbass|piss)\b'
@@ -372,11 +422,24 @@ done
   MSG_FOR_SPELLCHECK=$(echo "$FULL_COMMIT_MSG" | sed -E \
     -e "s/(['\"][^'\"]*['\"])//g" \
     -e "s/\bcommit[[:space:]]+[0-9a-fA-F]{7,40}\b/commit/g")
+  MSG_SPELLCHECK_FOR_LINE_FINDING=$(echo "$FULL_COMMIT_MSG_WITH_EMPTY" | sed -E \
+    -e "s/(['\"][^'\"]*['\"])//g" \
+    -e "s/\bcommit[[:space:]]+[0-9a-fA-F]{7,40}\b/commit/g")
 
   # Use aspell to list misspelled words according to American English, ignoring quoted text.
   MISSPELLED_WORDS=$(echo "$MSG_FOR_SPELLCHECK" | $ASPELL --lang=en --list --home-dir=scripts --personal=aspell-pws)
   if [ -n "$MISSPELLED_WORDS" ]; then
-    add_warning 1 "Avoid using non-American English words"
+    results=$(get_all_match_positions "$MSG_SPELLCHECK_FOR_LINE_FINDING" "$MISSPELLED_WORDS")
+
+    while IFS= read -r result; do
+      # Expected format: "target: line,column"
+      local target=$(echo "$result" | cut -d: -f1)
+      local pos=$(echo "$result" | cut -d: -f2 | tr -d ' ')
+      local line=$(echo "$pos" | cut -d, -f1)
+
+      add_warning "$line" "Avoid using non-American English words: $target"
+    done <<< "$results"
+
   fi
 }