mirror of
https://github.com/thewesker/greptweet.git
synced 2025-12-24 05:51:06 -05:00
expand urls
This commit is contained in:
@@ -37,8 +37,10 @@ echo Trying to get $(($twitter_total - $saved))
|
|||||||
temp=$(mktemp)
|
temp=$(mktemp)
|
||||||
temp2=$(mktemp)
|
temp2=$(mktemp)
|
||||||
|
|
||||||
echo "curl -s \"${api}screen_name=${1}&count=200&page=${page}${since}&include_rts=true&trim_user=1\""
|
url="${api}screen_name=${1}&count=200&page=${page}${since}&include_rts=true&trim_user=1&include_entities=1"
|
||||||
curl -si "${api}screen_name=${1}&count=200&page=${page}${since}&include_rts=true&trim_user=1" > $temp
|
|
||||||
|
echo "curl -s \"$url\""
|
||||||
|
curl -si "$url" > $temp
|
||||||
echo $?
|
echo $?
|
||||||
|
|
||||||
{
|
{
|
||||||
@@ -46,9 +48,9 @@ echo $?
|
|||||||
do
|
do
|
||||||
if test "$REPLY" = $'\r'
|
if test "$REPLY" = $'\r'
|
||||||
then
|
then
|
||||||
break
|
break
|
||||||
else
|
else
|
||||||
echo "$REPLY" >&2 # print header to stderr
|
echo "$REPLY" >&2 # print header to stderr
|
||||||
fi
|
fi
|
||||||
done
|
done
|
||||||
cat; } < $temp > $temp2
|
cat; } < $temp > $temp2
|
||||||
@@ -59,24 +61,34 @@ grep -iE 'rate|status' # show the interesting twitter rate limits
|
|||||||
mv $temp2 $temp
|
mv $temp2 $temp
|
||||||
|
|
||||||
if test $(xmlstarlet sel -t -v "count(//statuses/status)" $temp) -eq 0
|
if test $(xmlstarlet sel -t -v "count(//statuses/status)" $temp) -eq 0
|
||||||
then
|
then
|
||||||
|
|
||||||
head $temp
|
head $temp
|
||||||
if test "$2" && test "$since"
|
if test "$2" && test "$since"
|
||||||
then
|
then
|
||||||
echo No old tweets ${since}
|
echo No old tweets ${since}
|
||||||
elif test "$since"
|
elif test "$since"
|
||||||
then
|
then
|
||||||
echo No new tweets ${since}
|
echo No new tweets ${since}
|
||||||
else
|
else
|
||||||
echo "Twitter is returning empty responses on page ${page} :("
|
echo "Twitter is returning empty responses on page ${page} :("
|
||||||
fi
|
fi
|
||||||
rm -f $temp $temp2
|
rm -f $temp $temp2
|
||||||
exit
|
exit
|
||||||
|
|
||||||
fi
|
fi
|
||||||
|
|
||||||
xmlstarlet sel -t -m "//statuses/status" -v "id" -o "|" -v "created_at" -o "|" -v "normalize-space(text)" -n $temp > $temp2
|
xmlstarlet sel -t -m "statuses/status" -n -o "text " -v "text" -m "entities/urls/url" -i "expanded_url != ''" -n -o "url " -v "url" -o " " -v "expanded_url" $temp | {
|
||||||
|
while read -r first rest
|
||||||
|
do
|
||||||
|
case $first in
|
||||||
|
"text") echo $text; text=$rest ;;
|
||||||
|
"url") set -- $(echo $rest); text=$(echo $text | sed s,$1,$2,g) ;;
|
||||||
|
esac
|
||||||
|
done
|
||||||
|
echo $text
|
||||||
|
} > $temp2
|
||||||
|
|
||||||
cat $temp2 | perl -MHTML::Entities -pe 'decode_entities($_)' > $temp
|
cat $temp2 | perl -MHTML::Entities -pe 'decode_entities($_)' > $temp
|
||||||
cat $temp | sed '/^$/d' > $temp2
|
cat $temp | sed '/^$/d' > $temp2
|
||||||
|
|
||||||
|
|||||||
Reference in New Issue
Block a user