view TODO @ 827:a61728b58dc0

Fix array overflow bug in bdiff I ran into a bug while importing a large repository into mercurial. The diff algorithm does not allocate a big enough array of hunks for some test cases. This results in memory corruption, and possibly, as in my case, a seg fault. You should be able to reproduce this problem with any case of more than a few lines that follows this pattern: a b = = 1 1 2 2 3 4 3 5 . 4 . . 5 . . . I.e., "a" has blank lines on every other line that have been removed in "b". In this case, the number of matching hunks is equal to the number of lines in "b". This is more than ((an + bn)/4 + 2). I'm not sure what motivates this formula, but when I changed it to the smaller of an or bn (+ 1), it works. [comment added by mpm]
author "Wallace, Eric S" <eric.s.wallace@intel.com>
date Thu, 04 Aug 2005 13:25:59 -0800
parents d2c52a502473
children f1555f48f884 c2e77581bc84
line wrap: on
line source

General:
- Better documentation
- More regression tests
- More specific try/except.
- less code duplication, more code in the right places
- python 2.2 support
- export to git
- Code cleanup: apply http://python.org/peps/pep-0008.html

Core:
- difflib creating/removing files (fixed except dates: should be epoch)
- directory foo.d or foo.i with existing file foo (use some quoting?)
- get various options from hgrc (e.g. history always -v, tip always -q)
- hg over https:// and rsync://
- hooks for new changesets getting pulled/imported etc.
- make showing removed files (in history etc.) faster.
- hgmerge error: merge should abort nicely and running it again should work
- if hardlinking fails, pull should be used
- .hgignore should use new patterns

Commands:
- hg add <directory> should work (currently only: hg add -I <dir>)
- hg status <filename>: file rev, changeset rev, changed, added,
  deleted, sha-1
- select to pull a subset of the heads
- commands.py: number of args too much magic (e.g. in import_())
- optionally only show merges (two parents)
- automatic pull fallback to old-http://
- hg pull http://foo tries hostname
  'foo?cmd=changegroup&roots=0000000000000000000000000000000000000000'
- pass options to ssh (debug/verbose/remote hg command etc.)
- create a commented .hg/hgrc on init/clone
- hg pull default in a subdir doesn't work, if it is a relative path
- hg clone should store corrected relative paths, so moving a directory
  containing related repositories works again
- mention synonyms (co, ci, ...) in help
- show global options in help
- if everyone knows 'hg clone': hg init [DIR]
- if everyone knows 'hg update -m': remove -t

Web:
- optionally only show merges (two parents)
- one hgweb with many repos (another script)
- hgweb tip link too long (URL?cmd=changelog;rev=)
- hgweb: shorter links (e.g. cs=... instead of cmd=changeset;node=...?)
- hgweb: deliver static files (e.g. favicon, stylesheets)
- hgweb personalization: timezone (display/change), display of
  features, number of entries per page
- some web servers think hgweb.cgi.[di] is a CGI script with old-http://
  (use quoting (see foo.d in Core) or document server configurations?)
- link children in hgweb
- allow verbose mode
- hide trivial parent (like in show_changeset)