pacnpal/greptweet
1
0
Fork 0
mirror of https://github.com/thewesker/greptweet.git synced 2025-12-22 21:11:06 -05:00
Code Issues Packages Projects Releases Wiki Activity

Switch create.cgi to output HTML instead of text to allow for better feedback

Browse Source
This commit is contained in:
Kai Hendry
2011-11-22 00:22:53 +08:00
parent 64c5b9c836
commit 1c680703c6
4 changed files with 43 additions and 21 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
README.markdown
View File

@@ -12,7 +12,13 @@
* Won't work on protected accounts (duh!)
* No @mentions or DMs from other accounts
# Twitter can be flaky
## Fetching already!
Closing a tab whilst creating an account,
<http://greptweet/create.cgi?id=example>, can cause issues. Need to study
<http://mywiki.wooledge.org/ProcessManagement>.
## Twitter can be flaky
Twitter does not allow the possibility of retrieving more than 3200 tweets.
However twitter generally stalls before coming close to this limit. Please
@@ -23,8 +29,6 @@ to any already existing tweets.
I did file <https://dev.twitter.com/discussions/3414>, which later seemed to be fixed.
# Shell script feedback on the Web is unsolved :(
# Shell script feedback (progressive loading) on the Web is solved by outputing more than one should
<http://stackoverflow.com/questions/3547488>
Could do with some help here to make shell scripts show nicely, like `tail` in a console.

48
create.cgi
View File

@@ -1,11 +1,5 @@
#!/bin/bash -e
# vim: set ts=4 sw=4
cat <<END
Cache-Control: no-cache
Content-Type: text/plain
END
exec 2>&1
oldpwd=$PWD
@@ -29,16 +23,32 @@ then
test "${parm[3]}" && old=1
fi
# Trying to workaround:
# http://stackoverflow.com/questions/3547488/showing-a-long-running-shell-process-with-apache
# Maybe needs a bigger buffer ???
figlet $id
cat <<END
Cache-Control: no-cache
Content-Type: text/html
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8" />
<title>Fetching tweets of $id</title>
<link rel="stylesheet" type="text/css" href="/style.css">
</head>
<body>
<h1>Greptweet is running a long operation to fetch upto 3200 tweets from $id</h1>
<p>Please be patient. If you close this page prematurely you can limit the tweets <a href="https://github.com/kaihendry/Greptweet/blob/master/fetch-tweets.sh">fetch-tweets.sh</a> gets and trigger a locking bug.</p>
<pre>
END
hash figlet 2>/dev/null && figlet $id
if test -d u/$id
then
echo Directory $id already exists
echo Visit http://greptweet.com/u/$id
echo Attempting an update
cd u/$id
@@ -48,18 +58,18 @@ then
if ! test -f lock
then
touch lock
touch lock # Bug here if tab is closed before it's finished
../../fetch-tweets.sh $id $old
rm lock
rm lock # We need to also clear to lock if fetch-tweets was killed by Apache
else
echo Fetching already!
fi
else
if curl -I http://api.twitter.com/1/users/lookup.xml?screen_name=${id} | grep -q "Status: 404 Not Found"
if curl -sI http://api.twitter.com/1/users/lookup.xml?screen_name=${id} | grep -q "Status: 404 Not Found"
then
echo $id does not exist on twitter.com
echo "$id does not exist on twitter.com :("
exit
fi
@@ -85,3 +95,11 @@ fi
test -s $oldpwd/u/$id/$id.txt || rm -rf $oldpwd/u/$id
cd $oldpwd; ./users.sh > users.shtml
cat <<END
</pre>
<h1>Visit <a href="http://$HTTP_HOST/u/$id">http://$HTTP_HOST/u/$id</a></h1>
</body>
</html>
END

2
fetch-tweets.sh
View File

@@ -63,7 +63,7 @@ mv $temp2 $temp
if test $(xmlstarlet sel -t -v "count(//statuses/status)" $temp) -eq 0
then
head $temp
head $temp | perl -MHTML::Entities -pe 'encode_entities($_)'
if test "$2" && test "$since"
then
echo No old tweets ${since}

2
grep.php
View File

@@ -1,4 +1,4 @@
<?
<?php
if (empty($_GET['q'])) { die(); }
$QUERY=urldecode(escapeshellarg(urlencode($_GET['q'])));
//$fp = fopen('debug.log', 'a');