feat: WARC link points to the first warc result in target path

This commit is contained in:
Cristian 2020-11-12 10:57:31 -05:00
parent b237e412df
commit e594e6a75a
2 changed files with 10 additions and 4 deletions

View file

@ -1,6 +1,7 @@
from django.utils.html import format_html from django.utils.html import format_html
from core.models import Snapshot, EXTRACTORS from core.models import Snapshot, EXTRACTORS
from pathlib import Path
def get_icons(snapshot: Snapshot) -> str: def get_icons(snapshot: Snapshot) -> str:
@ -34,9 +35,10 @@ def get_icons(snapshot: Snapshot) -> str:
exists, extractor, icons.get(extractor, "?")) exists, extractor, icons.get(extractor, "?"))
if extractor == "wget": if extractor == "wget":
# warc isn't technically it's own extractor, so we have to add it after wget # warc isn't technically it's own extractor, so we have to add it after wget
exists = list((Path(path) / canon["warc_path"]).glob("*.warc.gz"))
output += output_template.format(path, canon["warc_path"], if exists:
exists, "warc", icons.get("warc", "?")) output += output_template.format(exists[0], "",
True, "warc", icons.get("warc", "?"))
except Exception as e: except Exception as e:
print(e) print(e)

View file

@ -223,6 +223,10 @@
.title-col a { .title-col a {
color: black; color: black;
} }
.exists-False {
display: none;
}
</style> </style>
<link rel="stylesheet" href="{% static 'bootstrap.min.css' %}"> <link rel="stylesheet" href="{% static 'bootstrap.min.css' %}">
<link rel="stylesheet" href="{% static 'jquery.dataTables.min.css' %}" /> <link rel="stylesheet" href="{% static 'jquery.dataTables.min.css' %}" />
@ -283,4 +287,4 @@
</footer> </footer>
</body> </body>
</html> </html>