diff options
Diffstat (limited to 'third_party/python/pyasn1/TODO.rst')
| -rw-r--r-- | third_party/python/pyasn1/TODO.rst | 92 |
1 files changed, 92 insertions, 0 deletions
diff --git a/third_party/python/pyasn1/TODO.rst b/third_party/python/pyasn1/TODO.rst new file mode 100644 index 0000000000..5c79ee7cdf --- /dev/null +++ b/third_party/python/pyasn1/TODO.rst @@ -0,0 +1,92 @@ + +Things to be done +================= + +Big things to tackle, anyone interested is welcome to fork pyasn1, work on +it and come up with a PR! + +New codecs +---------- + +* PER +* OER +* XER +* LWER +* JSON (alinged with existing experimental schemas) + +Lazy codecs +----------- + +Implement a thin layer over base types to cache pieces +of substrate being decoded till the very moment of ASN.1 +object access in the parse tree. + +Codecs generator interface +-------------------------- + +For indefinite length or chunked encoding mode, make codecs +iterable producing/consuming substrate/objects. + +ASN.1 schema compiler +--------------------- + +Ideally, the compiler should parse modern schema files and be +designed to emit code for arbitrary languages (including SQL). + +Base types +---------- + +Implement X.680 constructs, including information schema. + +Examples +-------- + +Add examples, including advanced/obscure use cases. + +Documentation +------------- + +Document more API, add notes and example snippets. + +More fresh modules +------------------ + +Compile and ship more Pythonized ASN.1 modules for +various ASN.1-based protocols (e.g. Kerberos etc). +Refresh outdated modules in pyasn1-packages. + +Minor, housekeeping things +-------------------------- + +* more PEP8'ing at places +* consider simplifying repr(), otherwise it tend to be too hard to grasp +* Specialize ASN.1 character and useful types + +* ber.decoder: + + * suspend codec on underrun error ? + * present subtypes ? + * component presence check wont work at innertypeconst + * type vs value, defaultValue + +* ber.encoder: + + * Asn1Item.clone() / shallowcopy issue + * large length encoder? + * lookup type by tag first to allow custom codecs for non-base types + +* type.useful: + + * may need to implement prettyIn/Out + +* type.char: + + * may need to implement constraints + +* type.namedtypes + + * type vs tagset name convention + +* how untagged TagSet should be initialized? + +* type and codecs for Real needs refactoring |