Trim leading zeros from chunk size, Varnish Cache problem #73
Conversation
| 'varnish-type-response-extra-line' => [ | ||
| ["006\r\nWi\r\nki\r\n005\r\npedia\r\n00e\r\n in\r\n\r\nchunks.\r\n0\r\n\r\n"], | ||
| "Wi\r\nkipedia in\r\n\r\nchunks." | ||
| ] |
There was a problem hiding this comment.
While the ltrim will definitely handle multiple leading 0's, could you add examples with more and less leading 0's then two. Just to ensure we don't accidentally refactor it specifically to two in the future. (You could even add a test adding a random number of 0's.)
clue
left a comment
clue
left a comment
There was a problem hiding this comment.
Nice catch! 👍 This actually appears to be valid according to the RFC: https://tools.ietf.org/html/rfc7230#section-4.1
May I ask you to test this also for the end chunk?
Also ping @legionth for the reference, who added a somewhat similar decoder for our Http component 👍
|
I added tests for end |
|
Thanks for your quick response! 👍 It's actually tempting to confuse ABNF syntax with PCRE modifiers, the RFC linked above uses an ABNF for this that is defined as |
|
Sorry about that, I didn't know about this ABNF. I took the code from @legionth and moved tests to valid ones. |
clue
left a comment
clue
left a comment
There was a problem hiding this comment.
Nice, thanks your quick response! 👍
|
Awesome and thank you 👍 . I'll tag a new release soon |
4 participants
Varnish Cache returns data with chunk size lead by 2 zeros
00. Examples:00einstead ofe001instead of1001234instead of1234Leading zeros have no value so we can trim them. I'm not 100% sure that there are always 2 zeros, but as far as I observed, there was 2 (to be sure I trim them all). Last zero is passed normally, so
\r\n0\r\n\r\n.Tests added.