mirrors/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2025-05-13 22:54:27 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

colorize output and add progress bar

Browse source
This commit is contained in:
Nick Sweeting 2017-07-05 16:33:51 -05:00
parent 172c4ad5b8
commit a167d2a1f4
5 changed files with 226 additions and 89 deletions
Download patch file Download diff file Expand all files Collapse all files

16
parse.py
View file

@ -64,8 +64,8 @@ def parse_pocket_export(html_file):
without_scheme = fixed_url.replace('http://', '').replace('https://', '')
info = {
'url': fixed_url,
'domain': without_scheme.split('/')[0], # without pathname
'base_url': without_scheme.split('?')[0], # without query args
'domain': without_scheme.split('/', 1)[0], # without pathname
'base_url': without_scheme.split('?', 1)[0], # without query args
'time': datetime.fromtimestamp(int(match.group(2))).strftime('%Y-%m-%d %H:%M'),
'timestamp': match.group(2),
'tags': match.group(3),
@ -84,10 +84,10 @@ def parse_json_export(json_file):
erg = line
info = {
'url': erg['href'],
'domain': erg['href'].replace('http://', '').replace('https://', '').split('/')[0],
'base_url': erg['href'].replace('https://', '').replace('http://', '').split('?')[0],
'time': datetime.fromtimestamp(int(time.mktime(time.strptime(erg['time'].split(',')[0], '%Y-%m-%dT%H:%M:%SZ')))),
'timestamp': str(int(time.mktime(time.strptime(erg['time'].split(',')[0], '%Y-%m-%dT%H:%M:%SZ')))),
'domain': erg['href'].replace('http://', '').replace('https://', '').split('/', 1)[0],
'base_url': erg['href'].replace('https://', '').replace('http://', '').split('?', 1)[0],
'time': datetime.fromtimestamp(int(time.mktime(time.strptime(erg['time'].split(',', 1)[0], '%Y-%m-%dT%H:%M:%SZ')))),
'timestamp': str(int(time.mktime(time.strptime(erg['time'].split(',', 1)[0], '%Y-%m-%dT%H:%M:%SZ')))),
'tags': erg['tags'],
'title': erg['description'].replace(' — Readability', ''),
}
@ -108,8 +108,8 @@ def parse_bookmarks_export(html_file):
info = {
'url': url,
'domain': url.replace('http://', '').replace('https://', '').split('/')[0],
'base_url': url.replace('https://', '').replace('http://', '').split('?')[0],
'domain': url.replace('http://', '').replace('https://', '').split('/', 1)[0],
'base_url': url.replace('https://', '').replace('http://', '').split('?', 1)[0],
'time': dt,
'timestamp': secs,
'tags': "",