lib.yang: support for pattern restrictions #1263
Conversation
…..]" This reverts commit 958c37d.
General categories are only implemented for the ASCII range, the predicates return false for all codepoints outside of ASCII. For lib.xsd_regexp that means it can’t match category escapes on non-ASCII input, i.e. it can’t fully handle Unicode input. It can however apply all valid regular expressions to ASCII input. Caveat: users of category escapes must restrict their input to ASCII. In order to fix this, we would need (access to) a full Unicode database in which we could look up codepoints and their attributes.
This reverts commit 76acfd3.
This reverts commit ea0978a.
…I-only This reverts commit c3ef659.
lib.maxpc: add backtracking combinators
|
Right, this can’t work right now since IP addresses are converted to numbers before the validators see them… I didn’t notice this before because I have disabled that behavior on my downstream branch, my bad! I guess I can guard the pattern validator on the invariant that the value to be validated is a string. |
This extra guard is necessary because some intrinsic lib.yang extensions will parse/validate some types natively (e.g. IP addresses using lib.protocol.ipvX) circumventing the YANG validation stack.
|
That being fixed, the next error appears to be a lwaftr test assuming ranges are single intervals: https://gist.github.com/SnabbBot/47076ad6d12879c3ce4d9ba4bf8ccbd7#file-log-L816 Edit: Actually they just use a different format. |
When hacking the range validator I had misunderstood the format used to express YANG range restrictions (I thought it didn’t feature multiple intervals), and changed it along the way. I think the new format is a bit simpler though (more regular) and requires less special cases, so I decided to keep it.
eugeneia
force-pushed
the
yang-restrictions-upstream
branch
from
8b27755 to
e577721
Compare
|
All good to go! |
|
Given that wingo-next has a lot of changes to some of the same stuff and that it's ready to go modulo some unrelated perf regression, I propose holding off here; unless you want to change the base branch to wingo-next of course. WDYT? |
|
I am in no particular rush to get this merged, so I guess feel free to pick it up whenever convenient. |
eugeneia
force-pushed
the
yang-restrictions-upstream
branch
2 times, most recently
from
5579f1e to
4396c52
Compare
eugeneia
force-pushed
the
yang-restrictions-upstream
branch
from
4396c52 to
03daccb
Compare
eugeneia
changed the title
|
Merged the current master. Note that besides the pattern validator this PR also...
|
|
Actually, I found a glaring omission in the Regex implementation (duh) over the weekend, so please put merging this on hold until I come up with a fix. :-)
|
Removes the broken match.range and replaces match.all with a simple implementation by means of match.plus. In lib.xsd_regexp implement backtracking match.range combinator by means of match.all. Simplify variable argument variants of match.plus and match.alternate along the way.
|
Ok should be better now! :-) |
Adds support for pattern and to lib.yang. The XSD Regex implementation is of questionable design (at least it was fun to hack), and has some restrictions:
Generally speaking, its not advisable to use it for sanitizing non-ASCII input. It’s good enough to parse things like the IP address patterns we use though.
I could add full unicode support (I have a branch that interprets the strings as UTF-8), it’s just that I have decided not to, because I wasn’t sure if sticking a whole Unicode databse into the
snabbbinary was worth it. If anyone actually needs Unicode support in Yang patterns, I will happily add it back in!Cc @wingo