signal-finder: add script

[bin.git] / tweet-grab

diff --git a/tweet-grab b/tweet-grab
index 8e94dc8daaf50f10058c2a26367f96f9bf6c42ee..b6e7fc091da43fc04eecbbb932acbe6b8b93b2c5 100755 (executable)
--- a/tweet-grab
+++ b/tweet-grab
@@ -1,7 +1,13 @@
 #!/usr/bin/env sh
+# Grabs tweets and writes them to a file. Avoids using the twitter API by
+# scrapping nitter.net (or any other nitter instance) which can easily be
+# scraped, unlike the intentionally crippled twitter.com. See the output of
+# this script at https://ishaan.bhojwani.org
+
 FILE=/root/tweets/index.html
 LOGFILE=/root/tweets/cron.log
-> $FILE
-curl https://nitter.net/keshinko | grep tweet-content | head -n 1 | sed -e 's/<[^>]*>//g' -e 's/^[ \t]*\(.*$\)/\1/' >> $FILE
-echo "$(tail -n 1000 $LOGFILE)" > $LOGFILE
-date -R >> $LOGFILE
+INSTANCE=https://nitter.fdn.fr
+
+curl $INSTANCE/keshinko | grep tweet-content | head -n 1 \
+  | sed -e 's/<[^>]*>//g' -e 's/^[ \t]*\(.*$\)/\1/' > $FILE
+date -R > $LOGFILE