From 8b7af72b22d4604e3d73e3ddb8fcac9ac4c1449c Mon Sep 17 00:00:00 2001
From: Lukas Krupcik <lukas.krupcik@vsb.cz>
Date: Mon, 18 Jul 2016 16:17:31 +0200
Subject: [PATCH] update filtering
---
html_md.sh | 16 +++++++---------
1 file changed, 7 insertions(+), 9 deletions(-)
diff --git a/html_md.sh b/html_md.sh
index 33bfe6c8a..929568cbe 100755
--- a/html_md.sh
+++ b/html_md.sh
@@ -26,7 +26,8 @@ if [ "$1" = "-w" ]; then
wget -bqc --mirror --convert-links --adjust-extension --page-requisites --no-parent https://docs.it4i.cz;
pid=$(pgrep wget);
- FILE="docs.it4i.cz/@@search?sort_on=sortable_title&.html"
+ # loop for all download html files
+ FILE="docs.it4i.cz/@@search?sort_on=sortable_title&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&&.html"
# test exists file -> yes - kill process wget
@@ -56,6 +57,8 @@ if [ "$1" = "-c" ]; then
rm "$i";
echo "$i deleted";
done
+
+ rm -rf info;
# create folder info
mkdir info;
@@ -115,11 +118,11 @@ if [ "$1" = "-c" ]; then
while read y;
do
# search and delete according with filter_auto
- cat "${i%.*}.md" | sed -e 's/{'"$y"'}//g' | sed -e 's/\\//g' | sed -e 's/: //g' | sed -e 's/<\/div>//g' | sed '/^<div/d' | sed '/^$/d' > "${i%.*}TMP.md";
+ cat "${i%.*}.md" | sed -e 's/{'"$y"'}//g' | sed -e 's/\\//g' | sed -e 's/^Q//g' | sed -e 's/: //g' | sed -e 's/<\/div>//g' | sed '/^<div/d' | awk -v RS='\n\n\n\n\n' 1 > "${i%.*}TMP.md";
cat "${i%.*}TMP.md" > "${i%.*}.md";
done
- echo "\t\tother filter..."
+ #echo "\t\tother filter..."
cat filter_other |
while read a;
do
@@ -127,7 +130,7 @@ if [ "$1" = "-c" ]; then
cat "${i%.*}.md" | sed -e 's/'"$a"'//g' > "${i%.*}TMP.md";
cat "${i%.*}TMP.md" > "${i%.*}.md";
done
-
+
# delete temporary files
rm "${i%.*}TMP.md";
@@ -135,8 +138,3 @@ if [ "$1" = "-c" ]; then
rm filter_autoTMP
rm filter_auto
fi
-
-
-
-
-
--
GitLab