From 732eae532c8e8632db95ab80e0dde5071e744386 Mon Sep 17 00:00:00 2001 From: sigoden Date: Wed, 10 Jul 2024 18:53:32 +0800 Subject: feat: adjust the way of returning data to LLM (#69) --- tools/get_web_page.sh | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'tools/get_web_page.sh') diff --git a/tools/get_web_page.sh b/tools/get_web_page.sh index 119dff7..7a516bb 100755 --- a/tools/get_web_page.sh +++ b/tools/get_web_page.sh @@ -9,7 +9,8 @@ main() { # span and div tags are dropped from the HTML https://pandoc.org/MANUAL.html#raw-htmltex and sed removes any inline SVG images in image tags from the Markdown content. curl -fsSL "$argc_url" | \ pandoc -f html-native_divs-native_spans -t gfm-raw_html | \ - sed -E 's/!\[.*?\]\((data:image\/svg\+xml[^)]+)\)//g' + sed -E 's/!\[.*?\]\((data:image\/svg\+xml[^)]+)\)//g' \ + >> "$LLM_OUTPUT" } eval "$(argc --argc-eval "$0" "$@")" -- cgit v1.2.3