Mirrors/ArchiveBox
2
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox synced 2024-11-26 06:00:22 +00:00
Code Issues Projects Releases Packages Wiki Activity

show which format file was parsed as

Browse source
This commit is contained in:
Nick Sweeting 2019-02-04 18:45:18 -08:00
parent ae0c20dc76
commit c32a385e8f
2 changed files with 5 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

5
archivebox/archive.py
View file

@ -59,7 +59,7 @@ def merge_links(archive_path=OUTPUT_DIR, import_path=None, only_new=False):
all_links = []
if import_path:
# parse and validate the import file
raw_links = parse_links(import_path)
raw_links, parser_name = parse_links(import_path)
all_links = validate_links(raw_links)
# merge existing links in archive_path and new links
@ -70,11 +70,12 @@ def merge_links(archive_path=OUTPUT_DIR, import_path=None, only_new=False):
num_new_links = len(all_links) - len(existing_links)
if num_new_links and not only_new:
print('[{green}+{reset}] [{}] Adding {} new links from {} to {}/index.json'.format(
print('[{green}+{reset}] [{}] Adding {} new links from {} to {}/index.json (parsed as {})'.format(
datetime.now().strftime('%Y-%m-%d %H:%M:%S'),
num_new_links,
pretty_path(import_path),
pretty_path(archive_path),
parser_name,
**ANSI,
))
# else:

4
archivebox/parse.py
View file

@ -53,7 +53,7 @@ def parse_links(path):
links = []
with open(path, 'r', encoding='utf-8') as file:
for parser_func in get_parsers(file).values():
for parser_name, parser_func in get_parsers(file).items():
# otherwise try all parsers until one works
try:
links += list(parser_func(file))
@ -63,7 +63,7 @@ def parse_links(path):
# parser not supported on this file
pass
return links
return links, parser_name
def parse_pocket_export(html_file):