478 lines
13 KiB
Bash
478 lines
13 KiB
Bash
#!/usr/bin/env bash
|
||
# dic — 命令列字典查詢工具（bash 版）
|
||
# 用法：
|
||
# dic 香蕉 兩本字典都查
|
||
# dic -r 香蕉 只查重編修訂本
|
||
# dic -c 香蕉 只查簡編本
|
||
# dic -a 香蕉 顯示所有匹配（包含部分匹配）
|
||
# dic -l 香蕉 只列出匹配的字詞名清單
|
||
# dic --no-color 香蕉 不染色
|
||
# dic -h 顯示說明
|
||
|
||
set -u
|
||
|
||
# ========================================
|
||
# 設定區：改成你的 CSV 檔案完整路徑
|
||
DICT_NAME_1="重編修訂本"
|
||
DICT_PATH_1="$HOME/.local/share/dic/dict_revised.csv"
|
||
DICT_NAME_2="簡編本"
|
||
DICT_PATH_2="$HOME/.local/share/dic/dict_concise.csv"
|
||
# ========================================
|
||
|
||
PROG_NAME="dic"
|
||
|
||
show_help() {
|
||
cat <<'EOF'
|
||
dic — 命令列字典查詢工具
|
||
用法：
|
||
dic 香蕉 兩本字典都查
|
||
dic -r 香蕉 只查重編修訂本
|
||
dic -c 香蕉 只查簡編本
|
||
dic -a 香蕉 顯示所有匹配（包含部分匹配）
|
||
dic -l 香蕉 只列出匹配的字詞名清單
|
||
dic --no-color 香蕉 不染色
|
||
dic -h 顯示說明
|
||
EOF
|
||
}
|
||
|
||
# --- 顏色設定 ---
|
||
COLOR_ENABLED=1
|
||
|
||
setup_color() {
|
||
local force_off="$1"
|
||
if [[ "$force_off" == "1" ]] || [[ ! -t 1 ]]; then
|
||
COLOR_ENABLED=0
|
||
fi
|
||
}
|
||
|
||
c_wrap() {
|
||
# $1 = 顏色 code, $2 = 文字
|
||
if [[ "$COLOR_ENABLED" == "0" ]]; then
|
||
printf '%s' "$2"
|
||
else
|
||
printf '\033[%sm%s\033[0m' "$1" "$2"
|
||
fi
|
||
}
|
||
|
||
c_bold() { c_wrap "1" "$1"; }
|
||
c_dim() { c_wrap "2" "$1"; }
|
||
c_title() { c_wrap "1;36" "$1"; } # 粗體青色
|
||
c_zhuyin() { c_wrap "33" "$1"; } # 黃色
|
||
c_book() { c_wrap "1;35" "$1"; } # 粗體紫色（字典名）
|
||
c_hint() { c_wrap "2;37" "$1"; } # 淡灰色提示
|
||
# 詞性 [名][動] 等，要在釋義裡 inline 染色，由 awk 處理
|
||
|
||
# --- CSV 解析 ---
|
||
# 完整處理 RFC 4180 風格 CSV：
|
||
# - 雙引號內可以有逗號、換行
|
||
# - 連續兩個雙引號 "" 代表一個字面雙引號
|
||
# 輸出格式：每筆紀錄一行，欄位之間用 \x1f (US, unit separator) 分隔，
|
||
# 欄位內的換行保留為字面 \n（兩個字元），方便後面再還原。
|
||
#
|
||
# 第一行（header）也會被輸出，呼叫端負責跳過。
|
||
|
||
parse_csv() {
|
||
# $1 = 檔案路徑
|
||
awk -v FS="" '
|
||
BEGIN {
|
||
in_quote = 0
|
||
field = ""
|
||
nfields = 0
|
||
# 用陣列存當前 record 的所有欄位
|
||
delete fields
|
||
}
|
||
{
|
||
# awk 一行一行讀進來；如果上一行還在引號裡，就把換行補回去
|
||
if (in_quote) {
|
||
field = field "\\n" # 字面 \n 兩字元，避免破壞分隔
|
||
}
|
||
line = $0
|
||
n = length(line)
|
||
for (i = 1; i <= n; i++) {
|
||
ch = substr(line, i, 1)
|
||
if (in_quote) {
|
||
if (ch == "\"") {
|
||
# 看下一個字元，判斷是 escaped quote 還是收尾
|
||
next_ch = (i < n) ? substr(line, i+1, 1) : ""
|
||
if (next_ch == "\"") {
|
||
field = field "\""
|
||
i++
|
||
} else {
|
||
in_quote = 0
|
||
}
|
||
} else {
|
||
field = field ch
|
||
}
|
||
} else {
|
||
if (ch == "\"") {
|
||
in_quote = 1
|
||
} else if (ch == ",") {
|
||
fields[nfields++] = field
|
||
field = ""
|
||
} else {
|
||
field = field ch
|
||
}
|
||
}
|
||
}
|
||
# 行尾：如果不在引號內，代表一筆 record 結束
|
||
if (!in_quote) {
|
||
fields[nfields++] = field
|
||
# 印出這筆 record，欄位用 \x1f 分隔
|
||
out = ""
|
||
for (k = 0; k < nfields; k++) {
|
||
if (k > 0) out = out "\x1f"
|
||
out = out fields[k]
|
||
}
|
||
print out
|
||
# 重置
|
||
field = ""
|
||
nfields = 0
|
||
delete fields
|
||
}
|
||
}
|
||
' "$1"
|
||
}
|
||
|
||
# 找出 header 中「字詞名」「注音一式」「釋義」的欄位編號
|
||
# 輸出三個數字，用空格分隔
|
||
find_columns() {
|
||
local file="$1"
|
||
parse_csv "$file" | head -n 1 | awk -F$'\x1f' '
|
||
{
|
||
idx_name = -1; idx_zhuyin = -1; idx_def = -1
|
||
for (i = 1; i <= NF; i++) {
|
||
if ($i == "字詞名") idx_name = i
|
||
if ($i == "注音一式") idx_zhuyin = i
|
||
if ($i == "釋義") idx_def = i
|
||
}
|
||
printf "%d %d %d\n", idx_name, idx_zhuyin, idx_def
|
||
}
|
||
'
|
||
}
|
||
|
||
# 載入字典並依條件篩選
|
||
# $1 = 檔案路徑
|
||
# $2 = query
|
||
# $3 = mode: "exact" 或 "all"
|
||
# 輸出每筆結果：name \x1f zhuyin \x1f definition（其中 definition 內的換行還是字面 \n）
|
||
search_dict() {
|
||
local file="$1"
|
||
local query="$2"
|
||
local mode="$3"
|
||
|
||
if [[ ! -f "$file" ]]; then
|
||
# 警告寫到 stderr
|
||
if [[ "$COLOR_ENABLED" == "1" ]]; then
|
||
printf '\033[2m警告：\033[0m找不到字典檔 %s\n' "$file" >&2
|
||
else
|
||
printf '警告：找不到字典檔 %s\n' "$file" >&2
|
||
fi
|
||
return
|
||
fi
|
||
|
||
local cols
|
||
cols=$(find_columns "$file")
|
||
local idx_name idx_zhuyin idx_def
|
||
read -r idx_name idx_zhuyin idx_def <<< "$cols"
|
||
|
||
if [[ "$idx_name" == "-1" ]]; then
|
||
return
|
||
fi
|
||
|
||
parse_csv "$file" | awk -F$'\x1f' \
|
||
-v idx_name="$idx_name" \
|
||
-v idx_zhuyin="$idx_zhuyin" \
|
||
-v idx_def="$idx_def" \
|
||
-v query="$query" \
|
||
-v mode="$mode" '
|
||
NR == 1 { next } # 跳過 header
|
||
{
|
||
name = $idx_name
|
||
if (name == "") next
|
||
|
||
if (mode == "exact") {
|
||
if (name != query) next
|
||
sort_key = "0\t0\t" length(name) "\t" name
|
||
} else {
|
||
if (index(name, query) == 0) next
|
||
exact = (name == query) ? 0 : 1
|
||
starts = (index(name, query) == 1) ? 0 : 1
|
||
sort_key = exact "\t" starts "\t" length(name) "\t" name
|
||
}
|
||
|
||
zhuyin = (idx_zhuyin > 0) ? $idx_zhuyin : ""
|
||
def = (idx_def > 0) ? $idx_def : ""
|
||
|
||
# 用 sort_key 開頭，方便外面用 sort 排，後面再砍掉
|
||
# 欄位：sort_key \x1f name \x1f zhuyin \x1f definition
|
||
printf "%s\x1f%s\x1f%s\x1f%s\n", sort_key, name, zhuyin, def
|
||
}
|
||
' | LC_ALL=C sort -t $'\x1f' -k1,1 | awk -F$'\x1f' '
|
||
{
|
||
# 砍掉第一欄 sort_key
|
||
out = ""
|
||
for (i = 2; i <= NF; i++) {
|
||
if (i > 2) out = out "\x1f"
|
||
out = out $i
|
||
}
|
||
print out
|
||
}
|
||
'
|
||
}
|
||
|
||
# 算字典裡「包含 query 但不完全相符」的筆數
|
||
count_partial() {
|
||
local file="$1"
|
||
local query="$2"
|
||
|
||
if [[ ! -f "$file" ]]; then
|
||
echo 0
|
||
return
|
||
fi
|
||
|
||
local cols
|
||
cols=$(find_columns "$file")
|
||
local idx_name _z _d
|
||
read -r idx_name _z _d <<< "$cols"
|
||
|
||
if [[ "$idx_name" == "-1" ]]; then
|
||
echo 0
|
||
return
|
||
fi
|
||
|
||
parse_csv "$file" | awk -F$'\x1f' \
|
||
-v idx_name="$idx_name" \
|
||
-v query="$query" '
|
||
NR == 1 { next }
|
||
{
|
||
name = $idx_name
|
||
if (name == "") next
|
||
if (index(name, query) > 0 && name != query) c++
|
||
}
|
||
END { print c+0 }
|
||
'
|
||
}
|
||
|
||
# 把釋義裡的 [名][動][形] 之類染色
|
||
# 從 stdin 讀，印到 stdout
|
||
colorize_definition() {
|
||
if [[ "$COLOR_ENABLED" == "0" ]]; then
|
||
cat
|
||
return
|
||
fi
|
||
# 詞性標記：[ 後面 1~4 個非 []\n 字元，接 ]
|
||
# 用 sed 的 ERE
|
||
sed -E $'s/(\\[[^][\\n]{1,4}\\])/\033[1;32m\\1\033[0m/g'
|
||
}
|
||
|
||
# 印一筆字典資料
|
||
# $1 = book_name
|
||
# $2 = name
|
||
# $3 = zhuyin
|
||
# $4 = definition（其中換行為字面 \n 兩字元）
|
||
print_entry() {
|
||
local book="$1"
|
||
local name="$2"
|
||
local zhuyin="$3"
|
||
local def="$4"
|
||
|
||
printf ' %s\n' "$(c_book "▎$book")"
|
||
printf ' %s %s\n' "$(c_title "$name")" "$(c_zhuyin "$zhuyin")"
|
||
|
||
if [[ -n "$def" ]]; then
|
||
# 把字面 \n 還原成真的換行，每行縮排 4 格，然後染色詞性標記
|
||
# 並且把行尾的空白砍掉（對應 Python 的 .rstrip()）
|
||
printf '%s' "$def" \
|
||
| awk 'BEGIN{RS="\\\\n"} { sub(/[ \t\r]+$/, ""); print }' \
|
||
| sed -E '$ { /^$/d; }' \
|
||
| colorize_definition \
|
||
| sed 's/^/ /'
|
||
fi
|
||
printf '\n'
|
||
}
|
||
|
||
# --- 主程式 ---
|
||
|
||
# 解析參數
|
||
QUERY=""
|
||
OPT_REVISED=0
|
||
OPT_CONCISED=0
|
||
OPT_ALL=0
|
||
OPT_LIST=0
|
||
OPT_NO_COLOR=0
|
||
OPT_HELP=0
|
||
|
||
# 自己處理參數（不用 getopt，因為要支援 --long 且不想引外部依賴）
|
||
while [[ $# -gt 0 ]]; do
|
||
case "$1" in
|
||
-h|--help)
|
||
OPT_HELP=1; shift ;;
|
||
-r|--revised)
|
||
OPT_REVISED=1; shift ;;
|
||
-c|--concised)
|
||
OPT_CONCISED=1; shift ;;
|
||
-a|--all)
|
||
OPT_ALL=1; shift ;;
|
||
-l|--list)
|
||
OPT_LIST=1; shift ;;
|
||
--no-color)
|
||
OPT_NO_COLOR=1; shift ;;
|
||
--)
|
||
shift
|
||
if [[ $# -gt 0 ]]; then QUERY="$1"; shift; fi
|
||
;;
|
||
-*)
|
||
# 支援 -rc 這種合併短參數
|
||
arg="${1#-}"
|
||
if [[ "$arg" =~ ^[rcalh]+$ ]]; then
|
||
for (( i=0; i<${#arg}; i++ )); do
|
||
ch="${arg:$i:1}"
|
||
case "$ch" in
|
||
r) OPT_REVISED=1 ;;
|
||
c) OPT_CONCISED=1 ;;
|
||
a) OPT_ALL=1 ;;
|
||
l) OPT_LIST=1 ;;
|
||
h) OPT_HELP=1 ;;
|
||
esac
|
||
done
|
||
shift
|
||
else
|
||
printf 'dic: 未知選項 %s\n' "$1" >&2
|
||
exit 2
|
||
fi
|
||
;;
|
||
*)
|
||
if [[ -z "$QUERY" ]]; then
|
||
QUERY="$1"
|
||
fi
|
||
shift ;;
|
||
esac
|
||
done
|
||
|
||
if [[ "$OPT_HELP" == "1" ]] || [[ -z "$QUERY" ]]; then
|
||
show_help
|
||
exit 0
|
||
fi
|
||
|
||
setup_color "$OPT_NO_COLOR"
|
||
|
||
# 決定要查哪幾本
|
||
declare -a CHOSEN_NAMES CHOSEN_PATHS
|
||
if [[ "$OPT_REVISED" == "1" && "$OPT_CONCISED" == "1" ]]; then
|
||
CHOSEN_NAMES=("$DICT_NAME_1" "$DICT_NAME_2")
|
||
CHOSEN_PATHS=("$DICT_PATH_1" "$DICT_PATH_2")
|
||
elif [[ "$OPT_REVISED" == "1" ]]; then
|
||
CHOSEN_NAMES=("$DICT_NAME_1")
|
||
CHOSEN_PATHS=("$DICT_PATH_1")
|
||
elif [[ "$OPT_CONCISED" == "1" ]]; then
|
||
CHOSEN_NAMES=("$DICT_NAME_2")
|
||
CHOSEN_PATHS=("$DICT_PATH_2")
|
||
else
|
||
CHOSEN_NAMES=("$DICT_NAME_1" "$DICT_NAME_2")
|
||
CHOSEN_PATHS=("$DICT_PATH_1" "$DICT_PATH_2")
|
||
fi
|
||
|
||
# -l 列表模式 → 自動切到 all
|
||
if [[ "$OPT_ALL" == "1" || "$OPT_LIST" == "1" ]]; then
|
||
MODE="all"
|
||
else
|
||
MODE="exact"
|
||
fi
|
||
|
||
# 收集每本字典的結果。
|
||
# 因為 bash 沒有結構化資料，把每本字典的結果暫存到一個 tmp 檔。
|
||
TMP_DIR=$(mktemp -d)
|
||
trap 'rm -rf "$TMP_DIR"' EXIT
|
||
|
||
TOTAL_RESULTS=0
|
||
TOTAL_PARTIAL=0
|
||
NUM_BOOKS=${#CHOSEN_NAMES[@]}
|
||
|
||
# partial counts per book（給 exact 模式用，提示「另有 N 筆包含」）
|
||
declare -a BOOK_RESULT_COUNTS BOOK_PARTIAL_COUNTS BOOK_RESULT_FILES
|
||
|
||
for (( bi=0; bi<NUM_BOOKS; bi++ )); do
|
||
name="${CHOSEN_NAMES[$bi]}"
|
||
path="${CHOSEN_PATHS[$bi]}"
|
||
out_file="$TMP_DIR/book_$bi"
|
||
|
||
search_dict "$path" "$QUERY" "$MODE" > "$out_file"
|
||
count=$(wc -l < "$out_file" | tr -d ' ')
|
||
BOOK_RESULT_COUNTS[$bi]="$count"
|
||
BOOK_RESULT_FILES[$bi]="$out_file"
|
||
TOTAL_RESULTS=$(( TOTAL_RESULTS + count ))
|
||
|
||
if [[ "$MODE" == "exact" ]]; then
|
||
partial=$(count_partial "$path" "$QUERY")
|
||
else
|
||
partial=0
|
||
fi
|
||
BOOK_PARTIAL_COUNTS[$bi]="$partial"
|
||
TOTAL_PARTIAL=$(( TOTAL_PARTIAL + partial ))
|
||
done
|
||
|
||
# --- 列表模式：只印字詞名 ---
|
||
if [[ "$OPT_LIST" == "1" ]]; then
|
||
for (( bi=0; bi<NUM_BOOKS; bi++ )); do
|
||
count="${BOOK_RESULT_COUNTS[$bi]}"
|
||
[[ "$count" == "0" ]] && continue
|
||
name="${CHOSEN_NAMES[$bi]}"
|
||
file="${BOOK_RESULT_FILES[$bi]}"
|
||
printf '%s（%d 筆）\n' "$(c_book "$name")" "$count"
|
||
while IFS=$'\x1f' read -r n zh _; do
|
||
printf ' %s %s\n' "$(c_title "$n")" "$(c_dim "$zh")"
|
||
done < "$file"
|
||
printf '\n'
|
||
done
|
||
if [[ "$TOTAL_RESULTS" == "0" ]]; then
|
||
printf '找不到「%s」\n' "$QUERY"
|
||
exit 1
|
||
fi
|
||
exit 0
|
||
fi
|
||
|
||
# --- 一般輸出 ---
|
||
if [[ "$TOTAL_RESULTS" == "0" ]]; then
|
||
printf '找不到「%s」\n' "$(c_bold "$QUERY")"
|
||
if [[ "$TOTAL_PARTIAL" -gt 0 && "$MODE" == "exact" ]]; then
|
||
printf '%s\n' "$(c_hint "但有 $TOTAL_PARTIAL 筆字詞包含「$QUERY」，加 -a 查看全部")"
|
||
fi
|
||
exit 1
|
||
fi
|
||
|
||
# 標題
|
||
if [[ "$MODE" == "all" ]]; then
|
||
printf '%s%s %s\n\n' \
|
||
"$(c_dim "查詢：")" \
|
||
"$(c_bold "$QUERY")" \
|
||
"$(c_dim "（共 $TOTAL_RESULTS 筆匹配）")"
|
||
else
|
||
printf '%s%s\n\n' "$(c_dim "查詢：")" "$(c_bold "$QUERY")"
|
||
fi
|
||
|
||
# 各本字典的結果
|
||
for (( bi=0; bi<NUM_BOOKS; bi++ )); do
|
||
name="${CHOSEN_NAMES[$bi]}"
|
||
count="${BOOK_RESULT_COUNTS[$bi]}"
|
||
partial="${BOOK_PARTIAL_COUNTS[$bi]}"
|
||
file="${BOOK_RESULT_FILES[$bi]}"
|
||
|
||
if [[ "$count" == "0" ]]; then
|
||
if [[ "$MODE" == "exact" && "$partial" -gt 0 ]]; then
|
||
printf ' %s\n' "$(c_book "▎$name")"
|
||
printf ' %s\n\n' "$(c_hint "沒有完全相符，但有 $partial 筆包含此字")"
|
||
fi
|
||
continue
|
||
fi
|
||
while IFS=$'\x1f' read -r n zh def; do
|
||
print_entry "$name" "$n" "$zh" "$def"
|
||
done < "$file"
|
||
done
|
||
|
||
# 結尾提示
|
||
if [[ "$MODE" == "exact" && "$TOTAL_PARTIAL" -gt 0 ]]; then
|
||
printf '%s\n' "$(c_hint "另有 $TOTAL_PARTIAL 筆字詞包含「$QUERY」，加 -a 查看全部，或 -l 只看清單")"
|
||
fi
|
||
|
||
exit 0
|