mirrors/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2025-05-15 15:44:26 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

show warning when killing archivebox during index writing

Browse source
This commit is contained in:
Nick Sweeting 2020-08-18 04:38:29 -04:00
parent 92de20af15
commit f371032b71
1 changed files with 24 additions and 14 deletions
Download patch file Download diff file Expand all files Collapse all files

38
archivebox/index/__init__.py
View file

@ -236,16 +236,23 @@ def write_main_index(links: List[Link], out_dir: str=OUTPUT_DIR, finished: bool=
log_indexing_process_started(len(links)) log_indexing_process_started(len(links))
with timed_index_update(os.path.join(out_dir, SQL_INDEX_FILENAME)): try:
write_sql_main_index(links, out_dir=out_dir) with timed_index_update(os.path.join(out_dir, SQL_INDEX_FILENAME)):
os.chmod(os.path.join(out_dir, SQL_INDEX_FILENAME), int(OUTPUT_PERMISSIONS, base=8)) # set here because we don't write it with atomic writes write_sql_main_index(links, out_dir=out_dir)
os.chmod(os.path.join(out_dir, SQL_INDEX_FILENAME), int(OUTPUT_PERMISSIONS, base=8)) # set here because we don't write it with atomic writes
with timed_index_update(os.path.join(out_dir, JSON_INDEX_FILENAME)):
write_json_main_index(links, out_dir=out_dir)
with timed_index_update(os.path.join(out_dir, JSON_INDEX_FILENAME)): with timed_index_update(os.path.join(out_dir, HTML_INDEX_FILENAME)):
write_json_main_index(links, out_dir=out_dir) write_html_main_index(links, out_dir=out_dir, finished=finished)
except (KeyboardInterrupt, SystemExit):
with timed_index_update(os.path.join(out_dir, HTML_INDEX_FILENAME)): stderr('[!] Warning: Still writing index to disk...', color='lightyellow')
write_html_main_index(links, out_dir=out_dir, finished=finished) stderr(' Run archivebox init to fix any inconsisntencies from an ungraceful exit.')
with timed_index_update(os.path.join(out_dir, SQL_INDEX_FILENAME)):
write_sql_main_index(links, out_dir=out_dir)
os.chmod(os.path.join(out_dir, SQL_INDEX_FILENAME), int(OUTPUT_PERMISSIONS, base=8)) # set here because we don't write it with atomic writes
raise SystemExit(0)
log_indexing_process_finished() log_indexing_process_finished()
@ -255,13 +262,16 @@ def load_main_index(out_dir: str=OUTPUT_DIR, warn: bool=True) -> List[Link]:
"""parse and load existing index with any new links from import_path merged in""" """parse and load existing index with any new links from import_path merged in"""
all_links: List[Link] = [] all_links: List[Link] = []
all_links = list(parse_json_main_index(out_dir)) try:
links_from_sql = list(parse_sql_main_index(out_dir)) all_links = list(parse_json_main_index(out_dir))
links_from_sql = list(parse_sql_main_index(out_dir))
if warn and not set(l.url for l in all_links) == set(l.url for l in links_from_sql): if warn and not set(l.url for l in all_links) == set(l.url for l in links_from_sql):
stderr('{red}[!] Warning: SQL index does not match JSON index!{reset}'.format(**ANSI)) stderr('{red}[!] Warning: SQL index does not match JSON index!{reset}'.format(**ANSI))
stderr(' To repair the index and re-import any orphaned links run:') stderr(' To repair the index and re-import any orphaned links run:')
stderr(' archivebox init') stderr(' archivebox init')
except (KeyboardInterrupt, SystemExit):
raise SystemExit(0)
return all_links return all_links