monkmachine commented on PR #558:
URL: https://github.com/apache/tika/pull/558#issuecomment-1130493082
> No, that probably won't work. Sorry. If you send me some examples, I can
try some things.
Can't send you examples unfortunately :(
I did manage to process over 1000's files today w
monkmachine commented on PR #558:
URL: https://github.com/apache/tika/pull/558#issuecomment-1130489653
> Can you guarantee that reading per line will be ok on this json-disaster?
If so, that's the way to go.
>
> The other thing is that you'll want to specify the encoding on your
read
monkmachine commented on PR #558:
URL: https://github.com/apache/tika/pull/558#issuecomment-1130489224
> Can you tell if they're writing utf8? Are there any ascii accented data
items or non-ascii characters that you can use to figure out what they're
default encoding is?
If you can h
monkmachine commented on PR #558:
URL: https://github.com/apache/tika/pull/558#issuecomment-1130486579
> No, that probably won't work. Sorry. If you send me some examples, I can
try some things.
Yeah we'd be ok if Jackson allowed "nan" as well as "NaN" as we could use
JsonReadFeature
monkmachine commented on PR #558:
URL: https://github.com/apache/tika/pull/558#issuecomment-1130393422
If I use buffer reader I get the correct output but it's slower: 3s vs 10s
(it's quite a large file)
```public void jsonConvert() throws FileNotFoundException, IOException {
monkmachine commented on PR #558:
URL: https://github.com/apache/tika/pull/558#issuecomment-1130370795
If I read byte by byte (i.e. byte[] bytes = new byte[1];) I get the correct
result:
