-*- mode: outline; coding: utf-8 -*-

* About.

    http://mercurial.selenic.com/wiki/ProjectsUsingMercurial

                Some Projects that Use Mercurial

* User config.

Put to your ~/.hgrc:

  [ui]

  ; Editor for editing commit message.

  editor = gvim

  ; Who commit.

  username = Oleksandr Gavenko <gavenkoa@gmail.com>

  ; Use internal merge algorithm, which mark conflict like <<<<<< ====== >>>>>>.

  ; Save previous file version in '*.orig' file, after merge must be marked as

  ; resolved by running 'hg resolve -m <file>'.

  merge = internal:merge

  [web]

  ; Default encoding for file hosted by 'hg serv'.

  encoding = utf-8

** Useful extension.

Put to your ~/.hgrc:

  [extensions]

  ; To allow fetch command.

  fetch =

  ; To allow Mercurial Queues.

  hgext.mq =

** Multiline log message for log command.

By default 'hg log' show only first line of log message. To see all message run:

  $ hg log -v

or put into ~/.hgrc:

  [defaults]

  log = -v

** Follow history ever when file copied.

By default 'hg log' show only history after last file copy. To see log message

before copying run:

  $ hg log -f

or put into ~/.hgrc:

  [defaults]

  log = -f

* Publishing repo.

With static HTTP hosting you can copy via rsync, ftp, scp, etc., so long as all the files beneath

.hg are copied. Also since 1.1 pull protocol can detect static HTTP hosting:

  $ hg clone http://example.com/project

  http://mercurial.selenic.com/wiki/hgserve

  http://mercurial.selenic.com/wiki/HgWebDirStepByStep

  http://mercurial.selenic.com/wiki/StaticHTTP

** hgweb.config.

Set allowed project by specifying paths to they (keys are URL, values are fs paths):

  [paths]

  myproject = /home/user/hg/myproject

  otherproject = /home/user/hg/otherproject

You can use single wildcard '*' to search current subdirs or double wildcard '**' to search subdirs

recursively:

  [paths]

  myproject = /home/user/hg/my/*

  otherproject = /home/user/hg/other/**

Alternatively you can set a collection of repos (keys and values are both filesystem paths, keys

should be prefixes of the values and are "subtracted" from the values in order to generate the URL

paths to each repository):

  [collections]

  /home/user/hg = /home/user/hg

  /home/another/hg = /home/another/hg

Allow archive downloads:

  [web]

  allow_archive = gz, zip, bz2

Make web page look nice:

  [web]

  # Use 'cd /lib/python2.x/site-packages/mercurial/templates; find . -type d' to see available

  # styles. Some interesting: gitweb, coal, monoblue.

  style = gitweb

Set another settings:

  [web]

  encoding = UTF-8

  maxchanges = 100

  maxfiles = 100

In each $proj/.hg/hgrc put:

  [web]

  contact = ADMIN <admin@example.com>

  description = <p style="color: red;">$proj</b> allow make a <a href="http://example.com">BIG Thing.</a>

  # Do not use name, in this case you see dir name where project lcated.

  # name = $proj

To allow push in 'hg serv':

  [web]

  allow_push = *

  push_ssl = false

  http://mercurial.selenic.com/wiki/PublishingRepositories

                Publishing Mercurial Repositories

** init.d script.

  #!/bin/sh

  CMD=/usr/bin/hg

  PORT=7878

  SRC=/srv/hg

  CONGIG=/srv/hg/hgweb.config

  PIDFILE=/var/run/hg.pid

  case "$1" in

  start)

    echo "Mecurial Server service starting."

    (cd "$SRC"; $CMD serve -d -p $PORT --pid-file "$PIDFILE")

    ;;

  stop)

    if [ -f "$PIDFILE" ]; then

      PID=`cat "$PIDFILE"`

      if [ "$PID" -gt 1 ]; then

        kill -TERM $PID

        echo "Stopping the Mercurial service PID=$PID."

      else

        echo Bad PID for Mercurial -- \"$PID\".

        echo You may remove \"$PIDFILE\" manually.

      fi

    else

      echo No PID file recorded for mercurial.

    fi

    ;;

  *)

    echo "$0 {start|stop}"

    exit 1

    ;;

  esac

  http://mercurial.selenic.com/wiki/hgserve

* Manage patches with MQ.

First enable MQ, add following to your ~/.hgrc:

  [extensions]

  hgext.mq =

Second get unpatched sources and put it to hg repository:

  $ tar zxf proj-x.y.z.tar.gz

  $ mv proj-x.y.z proj

  $ cd proj

  $ hg init

  $ hg ci -m "Added x.y.z version of proj."

Init MQ and take name for first patch:

  $ hg qinit

  $ hg qnew first.patch

Next make changes by editing source and save it to patch:

  $ $EDITOR file.c

  ...

  $ hg diff

  ...

  $ hg qrefresh

  $ hg diff      # <-- have zero diff

You can make second patch by applying existing one:

  $ hg qnew second.patch

  $ patch -p1 <bugfix.patch

  $ hg qrefresh

You can take list of patches (from old to new) and revert or apply patches by

qpop/qpush command:

  $ hg qseries   # <-- what patches have

  first.patch

  second.patch

  $ hg qapplied  # <-- what patches applied

  first.patch

  second.patch

  $ hg qpop

  $ hg qseries

  first.patch

  second.patch

  $ hg qapplied

  first.patch

You can revert or apply all patches by single command:

  $ hg qpop -a

  $ hg qpush -a

You can delete patch from patch list (before that you need de-apply patch):

  $ hg qpop -a

  $ hg qdelete first.patch

To add new version of source and fix patches for it first de apply patches,

then pull new changes and try apply patches on top of new sources:

  $ hg qpop -a

  $ rm *       # .hg dir not deleted because its name start with dot

  $ cd ..

  $ tar zxf proj-a.b.c.tar.gz

  $ cp -R proj-a.b.c/* proj

  $ cd proj

  $ hg addremove -s 70

  $ hg ci -m "Added a.b.c version of proj."

  $ hg qpush -a

To apply series of already done patches use:

  $ ls /path/to/bugfixes/*.patch | xargs hg qimport

* Proxy.

  $ hg clone --config http_proxy.host=$host:$port \

    --config http_proxy.user=$user --config http_proxy.passwd=$password  $addr

* Free Mercurial hosting.

  http://mercurial.selenic.com/wiki/MercurialHosting

                Free Hosting of Mercurial Repositories