mirrors/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2025-05-12 22:25:44 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

add filesizes and stray files in snapshot dir to snapshot_live ui

Browse source
This commit is contained in:
Nick Sweeting 2024-06-03 02:31:35 -07:00
parent c570674798
commit 1cd62ecc61
No known key found for this signature in database
2 changed files with 48 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

38
archivebox/core/views.py
View file

@ -85,9 +85,47 @@ class SnapshotView(View):
'name': result.extractor,
'path': embed_path,
'ts': ts_to_date_str(result.end_ts),
'size': abs_path.stat().st_size or '?',
}
archiveresults[result.extractor] = result_info
existing_files = {result['path'] for result in archiveresults.values()}
min_size_threshold = 128 # bytes
allowed_extensions = {
'txt',
'html',
'htm',
'png',
'jpg',
'jpeg',
'gif',
'webp'
'svg',
'webm',
'mp4',
'mp3',
'pdf',
'md',
}
# iterate through all the files in the snapshot dir and add the biggest ones to the result list
for result_file in Path(snapshot.link_dir).glob('*/*/*'):
extension = result_file.suffix.lstrip('.').lower()
if result_file.is_dir() or result_file.name.startswith('.') or extension not in allowed_extensions:
continue
if result_file.name in existing_files:
continue
file_size = result_file.stat().st_size or 0
if file_size > min_size_threshold:
archiveresults[result_file.name] = {
'name': result_file.stem,
'path': result_file.relative_to(snapshot.link_dir),
'ts': ts_to_date_str(result_file.stat().st_mtime or 0),
'size': file_size,
}
preferred_types = ('singlefile', 'wget', 'screenshot', 'dom', 'media', 'pdf', 'readability', 'mercury')
all_types = preferred_types + tuple(result_type for result_type in archiveresults.keys() if result_type not in preferred_types)

11
archivebox/templates/core/snapshot_live.html
View file

@ -395,12 +395,14 @@
</div>
<div class="header-bottom container-fluid">
<div class="row header-bottom-frames">
{% for result in archiveresults %}
<div class="col-lg-2">
<div class="card {% if forloop.first %}selected-card{% endif %}">
<div class="card-body">
<a href="{{result.path}}" target="preview" title="./{{result.path}} (downloaded {{result.ts}})">
<h4>{{result.name}}</h4>
<h4>{{result.name}} <small>({{result.size|filesizeformat}})</small></h4>
<!-- <p class="card-text" ><code>./{{result.path|truncatechars:30}}</code></p> -->
</a>
<!--<a href="{{result.path}}" target="preview"><h4 class="card-title">{{result.name}}</h4></a>-->
@ -409,6 +411,8 @@
</div>
</div>
{% endfor %}
<div class="col-lg-2">
<div class="card">
<div class="card-body">
@ -423,8 +427,13 @@
</div>
</div>
</header>
<iframe id="main-frame" sandbox="allow-same-origin allow-top-navigation-by-user-activation allow-scripts allow-forms" class="full-page-iframe" src="{{best_result.path}}" name="preview"></iframe>
<script src="{% static 'jquery.min.js' %}" type="text/javascript"></script>
<script>