forks/archivebox
1
0
Fork 0
Code Releases Activity

encoding bugfixes

Browse source
This commit is contained in:
Nick Sweeting 2017-10-18 17:47:19 -05:00
parent eb47155a12
commit d74394fa45
1 changed files with 6 additions and 6 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

12
archive_methods.py
View file

@ -103,7 +103,7 @@ def fetch_wget(out_dir, link, requisites=FETCH_WGET_REQUISITES, timeout=TIMEOUT)
output = html_appended_url(link) output = html_appended_url(link)
if result.returncode > 0: if result.returncode > 0:
print(' got wget response code {}:'.format(result.returncode)) print(' got wget response code {}:'.format(result.returncode))
print('\n'.join(' ' + line for line in result.stderr.decode().rsplit('\n', 10)[-10:] if line.strip())) print('\n'.join(' ' + line for line in (result.stderr or result.stdout).decode().rsplit('\n', 10)[-10:] if line.strip()))
# raise Exception('Failed to wget download') # raise Exception('Failed to wget download')
chmod_file(link['domain'], cwd=out_dir) chmod_file(link['domain'], cwd=out_dir)
except Exception as e: except Exception as e:
@ -135,10 +135,10 @@ def fetch_pdf(out_dir, link, timeout=TIMEOUT):
] ]
end = progress(timeout, prefix=' ') end = progress(timeout, prefix=' ')
try: try:
result = run(CMD, stdout=DEVNULL, stderr=PIPE, cwd=out_dir, timeout=timeout + 1) # output.pdf result = run(CMD, stdout=PIPE, stderr=PIPE, cwd=out_dir, timeout=timeout + 1) # output.pdf
end() end()
if result.returncode: if result.returncode:
print(' ', result.stderr.decode()) print(' ', (result.stderr or result.stdout).decode())
raise Exception('Failed to print PDF') raise Exception('Failed to print PDF')
chmod_file('output.pdf', cwd=out_dir) chmod_file('output.pdf', cwd=out_dir)
output = 'output.pdf' output = 'output.pdf'
@ -172,10 +172,10 @@ def fetch_screenshot(out_dir, link, timeout=TIMEOUT, resolution=RESOLUTION):
] ]
end = progress(timeout, prefix=' ') end = progress(timeout, prefix=' ')
try: try:
result = run(CMD, stdout=DEVNULL, stderr=DEVNULL, cwd=out_dir, timeout=timeout + 1) # sreenshot.png result = run(CMD, stdout=PIPE, stderr=PIPE, cwd=out_dir, timeout=timeout + 1) # sreenshot.png
end() end()
if result.returncode: if result.returncode:
print(' ', result.stderr.decode()) print(' ', (result.stderr or result.stdout).decode())
raise Exception('Failed to take screenshot') raise Exception('Failed to take screenshot')
chmod_file('screenshot.png', cwd=out_dir) chmod_file('screenshot.png', cwd=out_dir)
output = 'screenshot.png' output = 'screenshot.png'
@ -212,7 +212,7 @@ def archive_dot_org(out_dir, link, timeout=TIMEOUT):
# Parse archive.org response headers # Parse archive.org response headers
headers = result.stdout.splitlines() headers = result.stdout.splitlines()
content_location = [h for h in headers if b'Content-Location: ' in h] content_location = [h for h in headers if b'Content-Location: ' in h]
errors = [h for h in headers if b'X-Archive-Wayback-Runtime-Error: ' in h] errors = [h for h in headers if h and b'X-Archive-Wayback-Runtime-Error: ' in h]
if content_location: if content_location:
archive_path = content_location[0].split(b'Content-Location: ', 1)[-1].decode('utf-8') archive_path = content_location[0].split(b'Content-Location: ', 1)[-1].decode('utf-8')