Поддержка множества pageId в rag/download.sh
This commit is contained in:
@@ -1,11 +1,12 @@
|
|||||||
#!/bin/bash
|
#!/bin/bash
|
||||||
|
# shellcheck disable=SC1091,SC2181
|
||||||
|
|
||||||
command -v curl >/dev/null 2>&1 || { echo >&2 "Ошибка: curl не установлен"; exit 1; }
|
command -v curl >/dev/null 2>&1 || { echo >&2 "Ошибка: curl не установлен"; exit 1; }
|
||||||
command -v jq >/dev/null 2>&1 || { echo >&2 "Ошибка: jq не установлен"; exit 1; }
|
command -v jq >/dev/null 2>&1 || { echo >&2 "Ошибка: jq не установлен"; exit 1; }
|
||||||
|
|
||||||
if [ $# -lt 1 ]; then
|
if [ $# -lt 1 ]; then
|
||||||
echo >&2 "Ошибка: не указан ID страницы для загрузки"
|
echo >&2 "Ошибка: не указан ID страницы для загрузки"
|
||||||
echo "Использование: $0 <pageId>"
|
echo "Использование: $0 <pageId1> [<pageId2>...]"
|
||||||
exit 1
|
exit 1
|
||||||
fi
|
fi
|
||||||
|
|
||||||
@@ -16,45 +17,46 @@ source ../.env
|
|||||||
[ -z "$CONF_USERNAME" ] && { echo >&2 "Ошибка: CONF_USERNAME не указан в файле .env"; exit 1; }
|
[ -z "$CONF_USERNAME" ] && { echo >&2 "Ошибка: CONF_USERNAME не указан в файле .env"; exit 1; }
|
||||||
[ -z "$CONF_PASSWORD" ] && { echo >&2 "Ошибка: CONF_PASSWORD не указан в файле .env"; exit 1; }
|
[ -z "$CONF_PASSWORD" ] && { echo >&2 "Ошибка: CONF_PASSWORD не указан в файле .env"; exit 1; }
|
||||||
|
|
||||||
PAGE_ID="$1"
|
|
||||||
API_ENDPOINT="${CONF_URL}/rest/api/content/${PAGE_ID}?expand=body.view,children.page"
|
|
||||||
|
|
||||||
OUTPUT_PATH="./input_html"
|
OUTPUT_PATH="./input_html"
|
||||||
[ ! -d "$OUTPUT_PATH" ] && mkdir -p "$OUTPUT_PATH"
|
[ ! -d "$OUTPUT_PATH" ] && mkdir -p "$OUTPUT_PATH"
|
||||||
|
|
||||||
|
for PAGE_ID in "$@"; do
|
||||||
|
API_ENDPOINT="${CONF_URL}/rest/api/content/${PAGE_ID}?expand=body.view,children.page"
|
||||||
echo
|
echo
|
||||||
echo "Загрузка: $API_ENDPOINT"
|
echo "Загрузка: $API_ENDPOINT"
|
||||||
|
|
||||||
response=$(curl -s -u "$CONF_USERNAME:$CONF_PASSWORD" -H "Accept: application/json" "${API_ENDPOINT}")
|
RESPONSE=$(curl -s -u "$CONF_USERNAME:$CONF_PASSWORD" -H "Accept: application/json" "${API_ENDPOINT}")
|
||||||
if [ $? -ne 0 ]; then
|
if [ $? -ne 0 ]; then
|
||||||
echo "$response"
|
echo "$RESPONSE"
|
||||||
exit 1
|
exit 1
|
||||||
fi
|
fi
|
||||||
|
|
||||||
error_message=$(echo "$response" | jq -r '.message' 2>/dev/null)
|
ERROR_MESSAGE=$(echo "$RESPONSE" | jq -r '.message' 2>/dev/null)
|
||||||
if [ -n "$error_message" ] && [ "$error_message" != "null" ]; then
|
if [ -n "$ERROR_MESSAGE" ] && [ "$ERROR_MESSAGE" != "null" ]; then
|
||||||
echo "$response"
|
echo "$RESPONSE"
|
||||||
exit 1
|
exit 1
|
||||||
fi
|
fi
|
||||||
|
|
||||||
title=$(echo "$response" | jq -r .title)
|
TITLE=$(echo "$RESPONSE" | jq -r .title)
|
||||||
content=$(echo "$response" | jq -r .body.view.value)
|
CONTENT=$(echo "$RESPONSE" | jq -r .body.view.value)
|
||||||
|
|
||||||
if [ -z "$content" ]; then
|
if [ -z "$CONTENT" ]; then
|
||||||
echo "Пустая страница, пропущено"
|
echo "Пустая страница, пропущено"
|
||||||
exit
|
exit
|
||||||
fi
|
fi
|
||||||
|
|
||||||
path="$output_path/${title//\//_}.html"
|
HTML_FILEPATH="$OUTPUT_PATH/${TITLE//\//_}.html"
|
||||||
content=${content//href=\"\//href=\"$CONF_URL}
|
CONTENT=${CONTENT//href=\"\//href=\"$CONF_URL}
|
||||||
content=${content//src=\"\//src=\"$CONF_URL}
|
CONTENT=${CONTENT//src=\"\//src=\"$CONF_URL}
|
||||||
|
|
||||||
url="$CONF_URL/pages/viewpage.action?pageId=$PAGE_ID"
|
URL="$CONF_URL/pages/viewpage.action?pageId=$PAGE_ID"
|
||||||
echo -e "@@$url@@\n<br><html><body>Исходная страница: <a href=$url>$url</a><br><br><h1>$title</h1>$content</body></html>" > "$path"
|
echo -e "@@$URL@@\n<br><html><body>Исходная страница: <a href=$URL>$URL</a><br><br><h1>$TITLE</h1>$CONTENT</body></html>" > "$HTML_FILEPATH"
|
||||||
echo "Сохранено: $output_path/$title.html"
|
echo "Сохранено: $OUTPUT_PATH/$TITLE.html"
|
||||||
|
|
||||||
|
CHILD_IDS=$(echo "$RESPONSE" | jq -r '.children.page.results[]?.id' 2>/dev/null)
|
||||||
|
for CHILD_ID in $CHILD_IDS; do
|
||||||
|
echo "Переход к дочерней странице: $PAGE_ID => $CHILD_ID"
|
||||||
|
$0 "$CHILD_ID"
|
||||||
|
done
|
||||||
|
|
||||||
child_ids=$(echo "$response" | jq -r '.children.page.results[]?.id' 2>/dev/null)
|
|
||||||
for child_id in $child_ids; do
|
|
||||||
echo "Переход к дочерней странице: $child_id"
|
|
||||||
$0 "$child_id"
|
|
||||||
done
|
done
|
||||||
|
|||||||
Reference in New Issue
Block a user