diff options
| author | Yaohan Chen <yaohan.chen@gmail.com> | 2013-10-18 21:18:02 -0400 |
|---|---|---|
| committer | Eric Anderson <ejona86@gmail.com> | 2013-10-19 18:09:37 -0700 |
| commit | c8512ab578621c40d87ddb53946e4976e78b2e4a (patch) | |
| tree | bde1570e22c02d03d8369f0869bfcc4893aa06c8 /hib-dlagent | |
| parent | 85ea8628f420d0de3d679b0530436e9d991eb912 (diff) | |
| download | hib-dlagent-c8512ab578621c40d87ddb53946e4976e78b2e4a.tar.gz hib-dlagent-c8512ab578621c40d87ddb53946e4976e78b2e4a.zip | |
Fix URL extraction
The listing page source code has changed so that href= is on its own
line. The previous discover_url() picks this up and prints/uses download
URLs in the form of "href='https...". This change fixes the problem.
Diffstat (limited to 'hib-dlagent')
| -rwxr-xr-x | hib-dlagent | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/hib-dlagent b/hib-dlagent index e55ca61..f191b7c 100755 --- a/hib-dlagent +++ b/hib-dlagent @@ -21,7 +21,7 @@ login() { discover_url() { local LISTING_PAGE="$1" - curl -s --cookie "$COOKIE_JAR" "$LISTING_PAGE" | grep "/$FILE?" | \ + curl -s --cookie "$COOKIE_JAR" "$LISTING_PAGE" | grep "/$FILE?" | grep 'data-web=' | \ sed -e "s/.* data-web='\([^']*\)'.*/\1/" | head -n 1 } |