Improve error reporting

[julianharty]

AFAIK this project currently uses zimwritefs. This exited when I was processing a ZIM file in the Docker container (started using the command line in the project's README.md). As the following extract from the command-line Docker console zimwritefs complains about opening a particular html file (which seems to exist, at least in the source ZIP file) and then terminates after reporting an error.

0|index    | Visiting directory ./pool/7-angry-crabs-conspire/content/ridesfiles/bikepix
0|index    | stdout: Visiting directory ./pool/7-angry-crabs-conspire/content/ridesfiles/ridesfiles
0|index    | Visiting directory ./pool/7-angry-crabs-conspire/content/ridesfiles/cubclub
0|index    | stdout: A:13000; CA:2103; UA:10897; FA:10899; IA:2092; C:390; CC:14; UC:376
0|index    | stdout: Visiting directory ./pool/7-angry-crabs-conspire/content/ridesfiles/greenwood
0|index    | stderr: zimwriterfs: unable to open file at path: ./pool/7-angry-crabs-conspire/content/ridesfiles/castle-combe-grand-national-motorcycle-race-2015.html
0|index    | terminate called after throwing an instance of 'int'
0|index    | child process exited with code null
0|index    | Deleted pool 7-angry-crabs-conspire
0|index    | Someone is downloading 7-angry-crabs-conspire

The user sees a generic error page, so is no wiser about what the problem might be. It would be great to improve the error reporting, perhaps when issue #16 is implemented?

Hmm, we can't find the Zim file you're looking for.
This may be because we delete them to save space. 
If someone sent you this link, ask them for the Zim file itself or the zip file they used. 
Alternatively, you could take your frustration to Twitter with the button below. 

[ISNIT0]

Yes, see this: https://github.com/openzim/zip2zim/issues/24