hgweb: all protocol functions have become generators Using the write() callable supplied by the start_response() call is frowned upon by the WSGI spec, returning an iterable over the content chunks is the recommended way. Be aware, though: returning many small chunks will slow down responses, because the server has to flush each chunk separately.

templater: make a template a string-only iterator

streamclone yields chunks instead of accepting a file-like object

hgweb: protocol functions take repo instead of web This makes it much easier for implementers of hgweb alternatives to simply call into protocol functions after setting up a repo and a request.

hgweb: move capabilities calculation back into hgweb.protocol

hgweb: centralize permission checks for protocol commands Consistently enforces authorization checks set up in hgrc up front, so that the actual commands don't have to worry about them and implementers of hgweb alternatives can easily implement their own permission checks.

tests: add some tests for web.allowpull configurations

hgweb: protocol requests are processed immediately This makes separation of interface/webcommands and protocol easier and will make http repos slightly faster because there's less overhead.

merge with mpm

coal: include current node in log, graph menu links

hgweb: move shortcut expansion to request instantiation

fix iso date help

context: simplify parents code

workingctx: add __contains__ method

status: various cleanups - rename lookup to cmp for formatting reasons - reduce nesting level of fixup code - sort fixup results - group comparison cases in a separate clause - make list initialization simpler - delay flag lookup for working directory - sort removed list in comparison - avoid sorting already sorted lists

status: use contexts

hgk: fix parent breakage

dirstate: simplify/optimize path checking - add fast _finddirs function - remove recursion from incpath/decpath - split changepath into addpath/droppath - change relax arg to check - move incpathcheck logic into addpath - move incpath into addpath - move decpath into droppath - inline code in self._dirs creation

acl: refactoring - undo class structure - make buildmatch return None for no function - use contexts properly - simplify check loop

manifest: make checkforbidden take a list

context: add walk method

context: add __int__ and hex methods

util: add sort helper

remove: work directly off status This allows us to use a single directory walk and to trivially ignore unknown files. The resulting code is also easier to follow.