patch: fix eolmode=auto with new files

If target file does not exist or has no eol, current code was normalizing eols
to LF. Preserve patch file eols instead.

doc/hgrc.5.txt

@@ -1,958 +1,959 b''
 ======
  hgrc
 ======
 
 ---------------------------------
 configuration files for Mercurial
 ---------------------------------
 
 :Author:         Bryan O'Sullivan <[e-mail]>
 :Organization:   Mercurial
 :Manual section: 5
 :Manual group:   Mercurial Manual
 
 .. contents::
    :backlinks: top
    :class: htmlonly
 
 
 Synopsis
 --------
 
 The Mercurial system uses a set of configuration files to control
 aspects of its behavior.
 
 Files
 -----
 
 Mercurial reads configuration data from several files, if they exist.
 The names of these files depend on the system on which Mercurial is
 installed. ``*.rc`` files from a single directory are read in
 alphabetical order, later ones overriding earlier ones. Where multiple
 paths are given below, settings from earlier paths override later
 ones.
 
 | (Unix, Windows) ``<repo>/.hg/hgrc``
 
     Per-repository configuration options that only apply in a
     particular repository. This file is not version-controlled, and
     will not get transferred during a "clone" operation. Options in
     this file override options in all other configuration files. On
     Unix, most of this file will be ignored if it doesn't belong to a
     trusted user or to a trusted group. See the documentation for the
     trusted_ section below for more details.
 
 | (Unix) ``$HOME/.hgrc``
 | (Windows) ``%USERPROFILE%\.hgrc``
 | (Windows) ``%USERPROFILE%\Mercurial.ini``
 | (Windows) ``%HOME%\.hgrc``
 | (Windows) ``%HOME%\Mercurial.ini``
 
     Per-user configuration file(s), for the user running Mercurial. On
     Windows 9x, ``%HOME%`` is replaced by ``%APPDATA%``. Options in these
     files apply to all Mercurial commands executed by this user in any
     directory. Options in these files override per-system and per-installation
     options.
 
 | (Unix) ``/etc/mercurial/hgrc``
 | (Unix) ``/etc/mercurial/hgrc.d/*.rc``
 
     Per-system configuration files, for the system on which Mercurial
     is running. Options in these files apply to all Mercurial commands
     executed by any user in any directory. Options in these files
     override per-installation options.
 
 | (Unix) ``<install-root>/etc/mercurial/hgrc``
 | (Unix) ``<install-root>/etc/mercurial/hgrc.d/*.rc``
 
     Per-installation configuration files, searched for in the
     directory where Mercurial is installed. ``<install-root>`` is the
     parent directory of the **hg** executable (or symlink) being run. For
     example, if installed in ``/shared/tools/bin/hg``, Mercurial will look
     in ``/shared/tools/etc/mercurial/hgrc``. Options in these files apply
     to all Mercurial commands executed by any user in any directory.
 
 | (Windows) ``C:\Mercurial\Mercurial.ini``
 | (Windows) ``HKEY_LOCAL_MACHINE\SOFTWARE\Mercurial``
 | (Windows) ``<install-dir>\Mercurial.ini``
 
     Per-installation/system configuration files, for the system on
     which Mercurial is running. Options in these files apply to all
     Mercurial commands executed by any user in any directory. Registry
     keys contain PATH-like strings, every part of which must reference
     a ``Mercurial.ini`` file or be a directory where ``*.rc`` files will
     be read.
 
 Syntax
 ------
 
 A configuration file consists of sections, led by a ``[section]`` header
 and followed by ``name = value`` entries::
 
     [spam]
     eggs=ham
     green=
        eggs
 
 Each line contains one entry. If the lines that follow are indented,
 they are treated as continuations of that entry. Leading whitespace is
 removed from values. Empty lines are skipped. Lines beginning with
 ``#`` or ``;`` are ignored and may be used to provide comments.
 
 A line of the form ``%include file`` will include ``file`` into the
 current configuration file. The inclusion is recursive, which means
 that included files can include other files. Filenames are relative to
 the configuration file in which the ``%include`` directive is found.
 
 A line with ``%unset name`` will remove ``name`` from the current
 section, if it has been set previously.
 
 
 Sections
 --------
 
 This section describes the different sections that may appear in a
 Mercurial "hgrc" file, the purpose of each section, its possible keys,
 and their possible values.
 
 ``alias``
 """""""""
 Defines command aliases.
 Aliases allow you to define your own commands in terms of other
 commands (or aliases), optionally including arguments.
 
 Alias definitions consist of lines of the form::
 
     <alias> = <command> [<argument]...
 
 For example, this definition::
 
     latest = log --limit 5
 
 creates a new command ``latest`` that shows only the five most recent
 changesets. You can define subsequent aliases using earlier ones::
 
     stable5 = latest -b stable
 
 .. note:: It is possible to create aliases with the same names as
    existing commands, which will then override the original
    definitions. This is almost always a bad idea!
 
 
 ``auth``
 """"""""
 Authentication credentials for HTTP authentication. Each line has
 the following format::
 
     <name>.<argument> = <value>
 
 where ``<name>`` is used to group arguments into authentication
 entries. Example::
 
     foo.prefix = hg.intevation.org/mercurial
     foo.username = foo
     foo.password = bar
     foo.schemes = http https
 
     bar.prefix = secure.example.org
     bar.key = path/to/file.key
     bar.cert = path/to/file.cert
     bar.schemes = https
 
 Supported arguments:
 
 ``prefix``
     Either ``*`` or a URI prefix with or without the scheme part.
     The authentication entry with the longest matching prefix is used
     (where ``*`` matches everything and counts as a match of length
     1). If the prefix doesn't include a scheme, the match is performed
     against the URI with its scheme stripped as well, and the schemes
     argument, q.v., is then subsequently consulted.
 ``username``
     Optional. Username to authenticate with. If not given, and the
     remote site requires basic or digest authentication, the user
     will be prompted for it.
 ``password``
     Optional. Password to authenticate with. If not given, and the
     remote site requires basic or digest authentication, the user
     will be prompted for it.
 ``key``
     Optional. PEM encoded client certificate key file.
 ``cert``
     Optional. PEM encoded client certificate chain file.
 ``schemes``
     Optional. Space separated list of URI schemes to use this
     authentication entry with. Only used if the prefix doesn't include
     a scheme. Supported schemes are http and https. They will match
     static-http and static-https respectively, as well.
     Default: https.
 
 If no suitable authentication entry is found, the user is prompted
 for credentials as usual if required by the remote.
 
 
 ``decode/encode``
 """""""""""""""""
 Filters for transforming files on checkout/checkin. This would
 typically be used for newline processing or other
 localization/canonicalization of files.
 
 Filters consist of a filter pattern followed by a filter command.
 Filter patterns are globs by default, rooted at the repository root.
 For example, to match any file ending in ``.txt`` in the root
 directory only, use the pattern ``*.txt``. To match any file ending
 in ``.c`` anywhere in the repository, use the pattern ``**.c``.
 
 The filter command can start with a specifier, either ``pipe:`` or
 ``tempfile:``. If no specifier is given, ``pipe:`` is used by default.
 
 A ``pipe:`` command must accept data on stdin and return the transformed
 data on stdout.
 
 Pipe example::
 
   [encode]
   # uncompress gzip files on checkin to improve delta compression
   # note: not necessarily a good idea, just an example
   *.gz = pipe: gunzip
 
   [decode]
   # recompress gzip files when writing them to the working dir (we
   # can safely omit "pipe:", because it's the default)
   *.gz = gzip
 
 A ``tempfile:`` command is a template. The string ``INFILE`` is replaced
 with the name of a temporary file that contains the data to be
 filtered by the command. The string ``OUTFILE`` is replaced with the name
 of an empty temporary file, where the filtered data must be written by
 the command.
 
 .. note:: The tempfile mechanism is recommended for Windows systems,
    where the standard shell I/O redirection operators often have
    strange effects and may corrupt the contents of your files.
 
 The most common usage is for LF <-> CRLF translation on Windows. For
 this, use the "smart" converters which check for binary files::
 
   [extensions]
   hgext.win32text =
   [encode]
   ** = cleverencode:
   [decode]
   ** = cleverdecode:
 
 or if you only want to translate certain files::
 
   [extensions]
   hgext.win32text =
   [encode]
   **.txt = dumbencode:
   [decode]
   **.txt = dumbdecode:
 
 
 ``defaults``
 """"""""""""
 
 (defaults are deprecated. Don't use them. Use aliases instead)
 
 Use the ``[defaults]`` section to define command defaults, i.e. the
 default options/arguments to pass to the specified commands.
 
 The following example makes ``hg log`` run in verbose mode, and ``hg
 status`` show only the modified files, by default::
 
   [defaults]
   log = -v
   status = -m
 
 The actual commands, instead of their aliases, must be used when
 defining command defaults. The command defaults will also be applied
 to the aliases of the commands defined.
 
 
 ``diff``
 """"""""
 
 Settings used when displaying diffs. They are all Boolean and
 defaults to False.
 
 ``git``
     Use git extended diff format.
 ``nodates``
     Don't include dates in diff headers.
 ``showfunc``
     Show which function each change is in.
 ``ignorews``
     Ignore white space when comparing lines.
 ``ignorewsamount``
     Ignore changes in the amount of white space.
 ``ignoreblanklines``
     Ignore changes whose lines are all blank.
 
 ``email``
 """""""""
 Settings for extensions that send email messages.
 
 ``from``
     Optional. Email address to use in "From" header and SMTP envelope
     of outgoing messages.
 ``to``
     Optional. Comma-separated list of recipients' email addresses.
 ``cc``
     Optional. Comma-separated list of carbon copy recipients'
     email addresses.
 ``bcc``
     Optional. Comma-separated list of blind carbon copy recipients'
     email addresses. Cannot be set interactively.
 ``method``
     Optional. Method to use to send email messages. If value is ``smtp``
     (default), use SMTP (see the SMTP_ section for configuration).
     Otherwise, use as name of program to run that acts like sendmail
     (takes ``-f`` option for sender, list of recipients on command line,
     message on stdin). Normally, setting this to ``sendmail`` or
     ``/usr/sbin/sendmail`` is enough to use sendmail to send messages.
 ``charsets``
     Optional. Comma-separated list of character sets considered
     convenient for recipients. Addresses, headers, and parts not
     containing patches of outgoing messages will be encoded in the
     first character set to which conversion from local encoding
     (``$HGENCODING``, ``ui.fallbackencoding``) succeeds. If correct
     conversion fails, the text in question is sent as is. Defaults to
     empty (explicit) list.
 
     Order of outgoing email character sets:
 
     1. ``us-ascii``: always first, regardless of settings
     2. ``email.charsets``: in order given by user
     3. ``ui.fallbackencoding``: if not in email.charsets
     4. ``$HGENCODING``: if not in email.charsets
     5. ``utf-8``: always last, regardless of settings
 
 Email example::
 
   [email]
   from = Joseph User <[e-mail]>
   method = /usr/sbin/sendmail
   # charsets for western Europeans
   # us-ascii, utf-8 omitted, as they are tried first and last
   charsets = iso-8859-1, iso-8859-15, windows-1252
 
 
 ``extensions``
 """"""""""""""
 
 Mercurial has an extension mechanism for adding new features. To
 enable an extension, create an entry for it in this section.
 
 If you know that the extension is already in Python's search path,
 you can give the name of the module, followed by ``=``, with nothing
 after the ``=``.
 
 Otherwise, give a name that you choose, followed by ``=``, followed by
 the path to the ``.py`` file (including the file name extension) that
 defines the extension.
 
 To explicitly disable an extension that is enabled in an hgrc of
 broader scope, prepend its path with ``!``, as in
 ``hgext.foo = !/ext/path`` or ``hgext.foo = !``  when path is not
 supplied.
 
 Example for ``~/.hgrc``::
 
   [extensions]
   # (the mq extension will get loaded from Mercurial's path)
   hgext.mq =
   # (this extension will get loaded from the file specified)
   myfeature = ~/.hgext/myfeature.py
 
 
 ``format``
 """"""""""
 
 ``usestore``
     Enable or disable the "store" repository format which improves
     compatibility with systems that fold case or otherwise mangle
     filenames. Enabled by default. Disabling this option will allow
     you to store longer filenames in some situations at the expense of
     compatibility and ensures that the on-disk format of newly created
     repositories will be compatible with Mercurial before version 0.9.4.
 
 ``usefncache``
     Enable or disable the "fncache" repository format which enhances
     the "store" repository format (which has to be enabled to use
     fncache) to allow longer filenames and avoids using Windows
     reserved names, e.g. "nul". Enabled by default. Disabling this
     option ensures that the on-disk format of newly created
     repositories will be compatible with Mercurial before version 1.1.
 
 ``merge-patterns``
 """"""""""""""""""
 
 This section specifies merge tools to associate with particular file
 patterns. Tools matched here will take precedence over the default
 merge tool. Patterns are globs by default, rooted at the repository
 root.
 
 Example::
 
   [merge-patterns]
   **.c = kdiff3
   **.jpg = myimgmerge
 
 ``merge-tools``
 """""""""""""""
 
 This section configures external merge tools to use for file-level
 merges.
 
 Example ``~/.hgrc``::
 
   [merge-tools]
   # Override stock tool location
   kdiff3.executable = ~/bin/kdiff3
   # Specify command line
   kdiff3.args = $base $local $other -o $output
   # Give higher priority
   kdiff3.priority = 1
 
   # Define new tool
   myHtmlTool.args = -m $local $other $base $output
   myHtmlTool.regkey = Software\FooSoftware\HtmlMerge
   myHtmlTool.priority = 1
 
 Supported arguments:
 
 ``priority``
   The priority in which to evaluate this tool.
   Default: 0.
 ``executable``
   Either just the name of the executable or its pathname.
   Default: the tool name.
 ``args``
   The arguments to pass to the tool executable. You can refer to the
   files being merged as well as the output file through these
   variables: ``$base``, ``$local``, ``$other``, ``$output``.
   Default: ``$local $base $other``
 ``premerge``
   Attempt to run internal non-interactive 3-way merge tool before
   launching external tool.
   Default: True
 ``binary``
   This tool can merge binary files. Defaults to False, unless tool
   was selected by file pattern match.
 ``symlink``
   This tool can merge symlinks. Defaults to False, even if tool was
   selected by file pattern match.
 ``checkconflicts``
   Check whether there are conflicts even though the tool reported
   success.
   Default: False
 ``checkchanged``
   Check whether outputs were written even though the tool reported
   success.
   Default: False
 ``fixeol``
   Attempt to fix up EOL changes caused by the merge tool.
   Default: False
 ``gui``
   This tool requires a graphical interface to run. Default: False
 ``regkey``
   Windows registry key which describes install location of this
   tool. Mercurial will search for this key first under
   ``HKEY_CURRENT_USER`` and then under ``HKEY_LOCAL_MACHINE``.
   Default: None
 ``regname``
   Name of value to read from specified registry key. Defaults to the
   unnamed (default) value.
 ``regappend``
   String to append to the value read from the registry, typically
   the executable name of the tool.
   Default: None
 
 
 ``hooks``
 """""""""
 Commands or Python functions that get automatically executed by
 various actions such as starting or finishing a commit. Multiple
 hooks can be run for the same action by appending a suffix to the
 action. Overriding a site-wide hook can be done by changing its
 value or setting it to an empty string.
 
 Example ``.hg/hgrc``::
 
   [hooks]
   # update working directory after adding changesets
   changegroup.update = hg update
   # do not use the site-wide hook
   incoming =
   incoming.email = /my/email/hook
   incoming.autobuild = /my/build/hook
 
 Most hooks are run with environment variables set that give useful
 additional information. For each hook below, the environment
 variables it is passed are listed with names of the form ``$HG_foo``.
 
 ``changegroup``
   Run after a changegroup has been added via push, pull or unbundle.
   ID of the first new changeset is in ``$HG_NODE``. URL from which
   changes came is in ``$HG_URL``.
 ``commit``
   Run after a changeset has been created in the local repository. ID
   of the newly created changeset is in ``$HG_NODE``. Parent changeset
   IDs are in ``$HG_PARENT1`` and ``$HG_PARENT2``.
 ``incoming``
   Run after a changeset has been pulled, pushed, or unbundled into
   the local repository. The ID of the newly arrived changeset is in
   ``$HG_NODE``. URL that was source of changes came is in ``$HG_URL``.
 ``outgoing``
   Run after sending changes from local repository to another. ID of
   first changeset sent is in ``$HG_NODE``. Source of operation is in
   ``$HG_SOURCE``; see "preoutgoing" hook for description.
 ``post-<command>``
   Run after successful invocations of the associated command. The
   contents of the command line are passed as ``$HG_ARGS`` and the result
   code in ``$HG_RESULT``. Hook failure is ignored.
 ``pre-<command>``
   Run before executing the associated command. The contents of the
   command line are passed as ``$HG_ARGS``. If the hook returns failure,
   the command doesn't execute and Mercurial returns the failure
   code.
 ``prechangegroup``
   Run before a changegroup is added via push, pull or unbundle. Exit
   status 0 allows the changegroup to proceed. Non-zero status will
   cause the push, pull or unbundle to fail. URL from which changes
   will come is in ``$HG_URL``.
 ``precommit``
   Run before starting a local commit. Exit status 0 allows the
   commit to proceed. Non-zero status will cause the commit to fail.
   Parent changeset IDs are in ``$HG_PARENT1`` and ``$HG_PARENT2``.
 ``preoutgoing``
   Run before collecting changes to send from the local repository to
   another. Non-zero status will cause failure. This lets you prevent
   pull over HTTP or SSH. Also prevents against local pull, push
   (outbound) or bundle commands, but not effective, since you can
   just copy files instead then. Source of operation is in
   ``$HG_SOURCE``. If "serve", operation is happening on behalf of remote
   SSH or HTTP repository. If "push", "pull" or "bundle", operation
   is happening on behalf of repository on same system.
 ``pretag``
   Run before creating a tag. Exit status 0 allows the tag to be
   created. Non-zero status will cause the tag to fail. ID of
   changeset to tag is in ``$HG_NODE``. Name of tag is in ``$HG_TAG``. Tag is
   local if ``$HG_LOCAL=1``, in repository if ``$HG_LOCAL=0``.
 ``pretxnchangegroup``
   Run after a changegroup has been added via push, pull or unbundle,
   but before the transaction has been committed. Changegroup is
   visible to hook program. This lets you validate incoming changes
   before accepting them. Passed the ID of the first new changeset in
   ``$HG_NODE``. Exit status 0 allows the transaction to commit. Non-zero
   status will cause the transaction to be rolled back and the push,
   pull or unbundle will fail. URL that was source of changes is in
   ``$HG_URL``.
 ``pretxncommit``
   Run after a changeset has been created but the transaction not yet
   committed. Changeset is visible to hook program. This lets you
   validate commit message and changes. Exit status 0 allows the
   commit to proceed. Non-zero status will cause the transaction to
   be rolled back. ID of changeset is in ``$HG_NODE``. Parent changeset
   IDs are in ``$HG_PARENT1`` and ``$HG_PARENT2``.
 ``preupdate``
   Run before updating the working directory. Exit status 0 allows
   the update to proceed. Non-zero status will prevent the update.
   Changeset ID of first new parent is in ``$HG_PARENT1``. If merge, ID
   of second new parent is in ``$HG_PARENT2``.
 ``tag``
   Run after a tag is created. ID of tagged changeset is in ``$HG_NODE``.
   Name of tag is in ``$HG_TAG``. Tag is local if ``$HG_LOCAL=1``, in
   repository if ``$HG_LOCAL=0``.
 ``update``
   Run after updating the working directory. Changeset ID of first
   new parent is in ``$HG_PARENT1``. If merge, ID of second new parent is
   in ``$HG_PARENT2``. If the update succeeded, ``$HG_ERROR=0``. If the
   update failed (e.g. because conflicts not resolved), ``$HG_ERROR=1``.
 
 .. note:: It is generally better to use standard hooks rather than the
    generic pre- and post- command hooks as they are guaranteed to be
    called in the appropriate contexts for influencing transactions.
    Also, hooks like "commit" will be called in all contexts that
    generate a commit (e.g. tag) and not just the commit command.
 
 .. note:: Environment variables with empty values may not be passed to
    hooks on platforms such as Windows. As an example, ``$HG_PARENT2``
    will have an empty value under Unix-like platforms for non-merge
    changesets, while it will not be available at all under Windows.
 
 The syntax for Python hooks is as follows::
 
   hookname = python:modulename.submodule.callable
   hookname = python:/path/to/python/module.py:callable
 
 Python hooks are run within the Mercurial process. Each hook is
 called with at least three keyword arguments: a ui object (keyword
 ``ui``), a repository object (keyword ``repo``), and a ``hooktype``
 keyword that tells what kind of hook is used. Arguments listed as
 environment variables above are passed as keyword arguments, with no
 ``HG_`` prefix, and names in lower case.
 
 If a Python hook returns a "true" value or raises an exception, this
 is treated as a failure.
 
 
 ``http_proxy``
 """"""""""""""
 Used to access web-based Mercurial repositories through a HTTP
 proxy.
 
 ``host``
     Host name and (optional) port of the proxy server, for example
     "myproxy:8000".
 ``no``
     Optional. Comma-separated list of host names that should bypass
     the proxy.
 ``passwd``
     Optional. Password to authenticate with at the proxy server.
 ``user``
     Optional. User name to authenticate with at the proxy server.
 
 ``smtp``
 """"""""
 Configuration for extensions that need to send email messages.
 
 ``host``
     Host name of mail server, e.g. "mail.example.com".
 ``port``
     Optional. Port to connect to on mail server. Default: 25.
 ``tls``
     Optional. Whether to connect to mail server using TLS. True or
     False. Default: False.
 ``username``
     Optional. User name to authenticate to SMTP server with. If
     username is specified, password must also be specified.
     Default: none.
 ``password``
     Optional. Password to authenticate to SMTP server with. If
     username is specified, password must also be specified.
     Default: none.
 ``local_hostname``
     Optional. It's the hostname that the sender can use to identify
     itself to the MTA.
 
 
 ``patch``
 """""""""
 Settings used when applying patches, for instance through the 'import'
 command or with Mercurial Queues extension.
 
 ``eol``
     When set to 'strict' patch content and patched files end of lines
-    are preserved. When set to ``lf`` or ``crlf``, both files end of lines
-    are ignored when patching and the result line endings are
+    are preserved. When set to ``lf`` or ``crlf``, both files end of
+    lines are ignored when patching and the result line endings are
     normalized to either LF (Unix) or CRLF (Windows). When set to
     ``auto``, end of lines are again ignored while patching but line
     endings in patched files are normalized to their original setting
-    on a per-file basis.
+    on a per-file basis. If target file does not exist or has no end
+    of line, patch line endings are preserved.
     Default: strict.
 
 
 ``paths``
 """""""""
 Assigns symbolic names to repositories. The left side is the
 symbolic name, and the right gives the directory or URL that is the
 location of the repository. Default paths can be declared by setting
 the following entries.
 
 ``default``
     Directory or URL to use when pulling if no source is specified.
     Default is set to repository from which the current repository was
     cloned.
 ``default-push``
     Optional. Directory or URL to use when pushing if no destination
     is specified.
 
 
 ``profiling``
 """""""""""""
 Specifies profiling format and file output. In this section
 description, 'profiling data' stands for the raw data collected
 during profiling, while 'profiling report' stands for a statistical
 text report generated from the profiling data. The profiling is done
 using lsprof.
 
 ``format``
     Profiling format.
     Default: text.
 
     ``text``
       Generate a profiling report. When saving to a file, it should be
       noted that only the report is saved, and the profiling data is
       not kept.
     ``kcachegrind``
       Format profiling data for kcachegrind use: when saving to a
       file, the generated file can directly be loaded into
       kcachegrind.
 ``output``
     File path where profiling data or report should be saved. If the
     file exists, it is replaced. Default: None, data is printed on
     stderr
 
 ``server``
 """"""""""
 Controls generic server settings.
 
 ``uncompressed``
     Whether to allow clients to clone a repository using the
     uncompressed streaming protocol. This transfers about 40% more
     data than a regular clone, but uses less memory and CPU on both
     server and client. Over a LAN (100 Mbps or better) or a very fast
     WAN, an uncompressed streaming clone is a lot faster (~10x) than a
     regular clone. Over most WAN connections (anything slower than
     about 6 Mbps), uncompressed streaming is slower, because of the
     extra data transfer overhead. Default is False.
 
 
 ``trusted``
 """""""""""
 For security reasons, Mercurial will not use the settings in the
 ``.hg/hgrc`` file from a repository if it doesn't belong to a trusted
 user or to a trusted group. The main exception is the web interface,
 which automatically uses some safe settings, since it's common to
 serve repositories from different users.
 
 This section specifies what users and groups are trusted. The
 current user is always trusted. To trust everybody, list a user or a
 group with name ``*``.
 
 ``users``
   Comma-separated list of trusted users.
 ``groups``
   Comma-separated list of trusted groups.
 
 
 ``ui``
 """"""
 
 User interface controls.
 
 ``archivemeta``
     Whether to include the .hg_archival.txt file containing meta data
     (hashes for the repository base and for tip) in archives created
     by the hg archive command or downloaded via hgweb.
     Default is True.
 ``askusername``
     Whether to prompt for a username when committing. If True, and
     neither ``$HGUSER`` nor ``$EMAIL`` has been specified, then the user will
     be prompted to enter a username. If no username is entered, the
     default ``USER@HOST`` is used instead.
     Default is False.
 ``debug``
     Print debugging information. True or False. Default is False.
 ``editor``
     The editor to use during a commit. Default is ``$EDITOR`` or ``vi``.
 ``fallbackencoding``
     Encoding to try if it's not possible to decode the changelog using
     UTF-8. Default is ISO-8859-1.
 ``ignore``
     A file to read per-user ignore patterns from. This file should be
     in the same format as a repository-wide .hgignore file. This
     option supports hook syntax, so if you want to specify multiple
     ignore files, you can do so by setting something like
     ``ignore.other = ~/.hgignore2``. For details of the ignore file
     format, see the |hgignore(5)|_ man page.
 ``interactive``
     Allow to prompt the user. True or False. Default is True.
 ``logtemplate``
     Template string for commands that print changesets.
 ``merge``
     The conflict resolution program to use during a manual merge.
     There are some internal tools available:
 
     ``internal:local``
         keep the local version
     ``internal:other``
         use the other version
     ``internal:merge``
         use the internal non-interactive merge tool
     ``internal:fail``
         fail to merge
 
 For more information on configuring merge tools see the
 merge-tools_ section.
 
 ``patch``
     command to use to apply patches. Look for ``gpatch`` or ``patch`` in
     PATH if unset.
 ``quiet``
     Reduce the amount of output printed. True or False. Default is False.
 ``remotecmd``
     remote command to use for clone/push/pull operations. Default is ``hg``.
 ``report_untrusted``
     Warn if a ``.hg/hgrc`` file is ignored due to not being owned by a
     trusted user or group. True or False. Default is True.
 ``slash``
     Display paths using a slash (``/``) as the path separator. This
     only makes a difference on systems where the default path
     separator is not the slash character (e.g. Windows uses the
     backslash character (``\``)).
     Default is False.
 ``ssh``
     command to use for SSH connections. Default is ``ssh``.
 ``strict``
     Require exact command names, instead of allowing unambiguous
     abbreviations. True or False. Default is False.
 ``style``
     Name of style to use for command output.
 ``timeout``
     The timeout used when a lock is held (in seconds), a negative value
     means no timeout. Default is 600.
 ``traceback``
     Mercurial always prints a traceback when an unknown exception
     occurs. Setting this to True will make Mercurial print a traceback
     on all exceptions, even those recognized by Mercurial (such as
     IOError or MemoryError). Default is False.
 ``username``
     The committer of a changeset created when running "commit".
     Typically a person's name and email address, e.g. ``Fred Widget
     <[e-mail]>``. Default is ``$EMAIL`` or ``username@hostname``. If
     the username in hgrc is empty, it has to be specified manually or
     in a different hgrc file (e.g. ``$HOME/.hgrc``, if the admin set
     ``username =``  in the system hgrc).
 ``verbose``
     Increase the amount of output printed. True or False. Default is False.
 
 
 ``web``
 """""""
 Web interface configuration.
 
 ``accesslog``
     Where to output the access log. Default is stdout.
 ``address``
     Interface address to bind to. Default is all.
 ``allow_archive``
     List of archive format (bz2, gz, zip) allowed for downloading.
     Default is empty.
 ``allowbz2``
     (DEPRECATED) Whether to allow .tar.bz2 downloading of repository
     revisions.
     Default is False.
 ``allowgz``
     (DEPRECATED) Whether to allow .tar.gz downloading of repository
     revisions.
     Default is False.
 ``allowpull``
     Whether to allow pulling from the repository. Default is True.
 ``allow_push``
     Whether to allow pushing to the repository. If empty or not set,
     push is not allowed. If the special value ``*``, any remote user can
     push, including unauthenticated users. Otherwise, the remote user
     must have been authenticated, and the authenticated user name must
     be present in this list (separated by whitespace or ``,``). The
     contents of the allow_push list are examined after the deny_push
     list.
 ``allow_read``
     If the user has not already been denied repository access due to
     the contents of deny_read, this list determines whether to grant
     repository access to the user. If this list is not empty, and the
     user is unauthenticated or not present in the list (separated by
     whitespace or ``,``), then access is denied for the user. If the
     list is empty or not set, then access is permitted to all users by
     default. Setting allow_read to the special value ``*`` is equivalent
     to it not being set (i.e. access is permitted to all users). The
     contents of the allow_read list are examined after the deny_read
     list.
 ``allowzip``
     (DEPRECATED) Whether to allow .zip downloading of repository
     revisions. Default is False. This feature creates temporary files.
 ``baseurl``
     Base URL to use when publishing URLs in other locations, so
     third-party tools like email notification hooks can construct
     URLs. Example: ``http://hgserver/repos/``.
 ``contact``
     Name or email address of the person in charge of the repository.
     Defaults to ui.username or ``$EMAIL`` or "unknown" if unset or empty.
 ``deny_push``
     Whether to deny pushing to the repository. If empty or not set,
     push is not denied. If the special value ``*``, all remote users are
     denied push. Otherwise, unauthenticated users are all denied, and
     any authenticated user name present in this list (separated by
     whitespace or ``,``) is also denied. The contents of the deny_push
     list are examined before the allow_push list.
 ``deny_read``
     Whether to deny reading/viewing of the repository. If this list is
     not empty, unauthenticated users are all denied, and any
     authenticated user name present in this list (separated by
     whitespace or ``,``) is also denied access to the repository. If set
     to the special value ``*``, all remote users are denied access
     (rarely needed ;). If deny_read is empty or not set, the
     determination of repository access depends on the presence and
     content of the allow_read list (see description). If both
     deny_read and allow_read are empty or not set, then access is
     permitted to all users by default. If the repository is being
     served via hgwebdir, denied users will not be able to see it in
     the list of repositories. The contents of the deny_read list have
     priority over (are examined before) the contents of the allow_read
     list.
 ``descend``
     hgwebdir indexes will not descend into subdirectories. Only repositories
     directly in the current path will be shown (other repositories are still
     available from the index corresponding to their containing path).
 ``description``
     Textual description of the repository's purpose or contents.
     Default is "unknown".
 ``encoding``
     Character encoding name.
     Example: "UTF-8"
 ``errorlog``
     Where to output the error log. Default is stderr.
 ``hidden``
     Whether to hide the repository in the hgwebdir index.
     Default is False.
 ``ipv6``
     Whether to use IPv6. Default is False.
 ``name``
     Repository name to use in the web interface. Default is current
     working directory.
 ``maxchanges``
     Maximum number of changes to list on the changelog. Default is 10.
 ``maxfiles``
     Maximum number of files to list per changeset. Default is 10.
 ``port``
     Port to listen on. Default is 8000.
 ``prefix``
     Prefix path to serve from. Default is '' (server root).
 ``push_ssl``
     Whether to require that inbound pushes be transported over SSL to
     prevent password sniffing. Default is True.
 ``staticurl``
     Base URL to use for static files. If unset, static files (e.g. the
     hgicon.png favicon) will be served by the CGI script itself. Use
     this setting to serve them directly with the HTTP server.
     Example: ``http://hgserver/static/``.
 ``stripes``
     How many lines a "zebra stripe" should span in multiline output.
     Default is 1; set to 0 to disable.
 ``style``
     Which template map style to use.
 ``templates``
     Where to find the HTML templates. Default is install path.
 
 
 Author
 ------
 Bryan O'Sullivan <[e-mail]>.
 
 Mercurial was written by Matt Mackall <[e-mail]>.
 
 See Also
 --------
 |hg(1)|_, |hgignore(5)|_
 
 Copying
 -------
 This manual page is copyright 2005 Bryan O'Sullivan.
 Mercurial is copyright 2005-2009 Matt Mackall.
 Free use of this software is granted under the terms of the GNU General
 Public License version 2.
 
 .. include:: common.txt

mercurial/patch.py

@@ -1,1447 +1,1485 b''
 # patch.py - patch file parsing routines
 #
 # Copyright 2006 Brendan Cully <[e-mail]>
 # Copyright 2007 Chris Mason <[e-mail]>
 #
 # This software may be used and distributed according to the terms of the
 # GNU General Public License version 2, incorporated herein by reference.
 
 from i18n import _
 from node import hex, nullid, short
 import base85, cmdutil, mdiff, util, diffhelpers, copies
 import cStringIO, email.Parser, os, re
 import sys, tempfile, zlib
 
 gitre = re.compile('diff --git a/(.*) b/(.*)')
 
 class PatchError(Exception):
     pass
 
 class NoHunks(PatchError):
     pass
 
 # helper functions
 
 def copyfile(src, dst, basedir):
     abssrc, absdst = [util.canonpath(basedir, basedir, x) for x in [src, dst]]
     if os.path.exists(absdst):
         raise util.Abort(_("cannot create %s: destination already exists") %
                          dst)
 
     dstdir = os.path.dirname(absdst)
     if dstdir and not os.path.isdir(dstdir):
         try:
             os.makedirs(dstdir)
         except IOError:
             raise util.Abort(
                 _("cannot create %s: unable to create destination directory")
                 % dst)
 
     util.copyfile(abssrc, absdst)
 
 # public functions
 
 def extract(ui, fileobj):
     '''extract patch from data read from fileobj.
 
     patch can be a normal patch or contained in an email message.
 
     return tuple (filename, message, user, date, node, p1, p2).
     Any item in the returned tuple can be None. If filename is None,
     fileobj did not contain a patch. Caller must unlink filename when done.'''
 
     # attempt to detect the start of a patch
     # (this heuristic is borrowed from quilt)
     diffre = re.compile(r'^(?:Index:[ \t]|diff[ \t]|RCS file: |'
                         r'retrieving revision [0-9]+(\.[0-9]+)*$|'
                         r'(---|\*\*\*)[ \t])', re.MULTILINE)
 
     fd, tmpname = tempfile.mkstemp(prefix='hg-patch-')
     tmpfp = os.fdopen(fd, 'w')
     try:
         msg = email.Parser.Parser().parse(fileobj)
 
         subject = msg['Subject']
         user = msg['From']
         if not subject and not user:
             # Not an email, restore parsed headers if any
             subject = '\n'.join(': '.join(h) for h in msg.items()) + '\n'
 
         gitsendmail = 'git-send-email' in msg.get('X-Mailer', '')
         # should try to parse msg['Date']
         date = None
         nodeid = None
         branch = None
         parents = []
 
         if subject:
             if subject.startswith('[PATCH'):
                 pend = subject.find(']')
                 if pend >= 0:
                     subject = subject[pend+1:].lstrip()
             subject = subject.replace('\n\t', ' ')
             ui.debug('Subject: %s\n' % subject)
         if user:
             ui.debug('From: %s\n' % user)
         diffs_seen = 0
         ok_types = ('text/plain', 'text/x-diff', 'text/x-patch')
         message = ''
         for part in msg.walk():
             content_type = part.get_content_type()
             ui.debug('Content-Type: %s\n' % content_type)
             if content_type not in ok_types:
                 continue
             payload = part.get_payload(decode=True)
             m = diffre.search(payload)
             if m:
                 hgpatch = False
                 ignoretext = False
 
                 ui.debug('found patch at byte %d\n' % m.start(0))
                 diffs_seen += 1
                 cfp = cStringIO.StringIO()
                 for line in payload[:m.start(0)].splitlines():
                     if line.startswith('# HG changeset patch'):
                         ui.debug('patch generated by hg export\n')
                         hgpatch = True
                         # drop earlier commit message content
                         cfp.seek(0)
                         cfp.truncate()
                         subject = None
                     elif hgpatch:
                         if line.startswith('# User '):
                             user = line[7:]
                             ui.debug('From: %s\n' % user)
                         elif line.startswith("# Date "):
                             date = line[7:]
                         elif line.startswith("# Branch "):
                             branch = line[9:]
                         elif line.startswith("# Node ID "):
                             nodeid = line[10:]
                         elif line.startswith("# Parent "):
                             parents.append(line[10:])
                     elif line == '---' and gitsendmail:
                         ignoretext = True
                     if not line.startswith('# ') and not ignoretext:
                         cfp.write(line)
                         cfp.write('\n')
                 message = cfp.getvalue()
                 if tmpfp:
                     tmpfp.write(payload)
                     if not payload.endswith('\n'):
                         tmpfp.write('\n')
             elif not diffs_seen and message and content_type == 'text/plain':
                 message += '\n' + payload
     except:
         tmpfp.close()
         os.unlink(tmpname)
         raise
 
     if subject and not message.startswith(subject):
         message = '%s\n%s' % (subject, message)
     tmpfp.close()
     if not diffs_seen:
         os.unlink(tmpname)
         return None, message, user, date, branch, None, None, None
     p1 = parents and parents.pop(0) or None
     p2 = parents and parents.pop(0) or None
     return tmpname, message, user, date, branch, nodeid, p1, p2
 
 GP_PATCH  = 1 << 0  # we have to run patch
 GP_FILTER = 1 << 1  # there's some copy/rename operation
 GP_BINARY = 1 << 2  # there's a binary patch
 
 class patchmeta(object):
     """Patched file metadata
 
     'op' is the performed operation within ADD, DELETE, RENAME, MODIFY
     or COPY.  'path' is patched file path. 'oldpath' is set to the
     origin file when 'op' is either COPY or RENAME, None otherwise. If
     file mode is changed, 'mode' is a tuple (islink, isexec) where
     'islink' is True if the file is a symlink and 'isexec' is True if
     the file is executable. Otherwise, 'mode' is None.
     """
     def __init__(self, path):
         self.path = path
         self.oldpath = None
         self.mode = None
         self.op = 'MODIFY'
         self.lineno = 0
         self.binary = False
 
     def setmode(self, mode):
         islink = mode & 020000
         isexec = mode & 0100
         self.mode = (islink, isexec)
 
 def readgitpatch(lr):
     """extract git-style metadata about patches from <patchname>"""
 
     # Filter patch for git information
     gp = None
     gitpatches = []
     # Can have a git patch with only metadata, causing patch to complain
     dopatch = 0
 
     lineno = 0
     for line in lr:
         lineno += 1
         line = line.rstrip(' \r\n')
         if line.startswith('diff --git'):
             m = gitre.match(line)
             if m:
                 if gp:
                     gitpatches.append(gp)
                 dst = m.group(2)
                 gp = patchmeta(dst)
                 gp.lineno = lineno
         elif gp:
             if line.startswith('--- '):
                 if gp.op in ('COPY', 'RENAME'):
                     dopatch |= GP_FILTER
                 gitpatches.append(gp)
                 gp = None
                 dopatch |= GP_PATCH
                 continue
             if line.startswith('rename from '):
                 gp.op = 'RENAME'
                 gp.oldpath = line[12:]
             elif line.startswith('rename to '):
                 gp.path = line[10:]
             elif line.startswith('copy from '):
                 gp.op = 'COPY'
                 gp.oldpath = line[10:]
             elif line.startswith('copy to '):
                 gp.path = line[8:]
             elif line.startswith('deleted file'):
                 gp.op = 'DELETE'
                 # is the deleted file a symlink?
                 gp.setmode(int(line[-6:], 8))
             elif line.startswith('new file mode '):
                 gp.op = 'ADD'
                 gp.setmode(int(line[-6:], 8))
             elif line.startswith('new mode '):
                 gp.setmode(int(line[-6:], 8))
             elif line.startswith('GIT binary patch'):
                 dopatch |= GP_BINARY
                 gp.binary = True
     if gp:
         gitpatches.append(gp)
 
     if not gitpatches:
         dopatch = GP_PATCH
 
     return (dopatch, gitpatches)
 
 class linereader(object):
     # simple class to allow pushing lines back into the input stream
     def __init__(self, fp, textmode=False):
         self.fp = fp
         self.buf = []
         self.textmode = textmode
         self.eol = None
 
     def push(self, line):
         if line is not None:
             self.buf.append(line)
 
     def readline(self):
         if self.buf:
             l = self.buf[0]
             del self.buf[0]
             return l
         l = self.fp.readline()
         if not self.eol:
             if l.endswith('\r\n'):
                 self.eol = '\r\n'
             elif l.endswith('\n'):
                 self.eol = '\n'
         if self.textmode and l.endswith('\r\n'):
             l = l[:-2] + '\n'
         return l
 
     def __iter__(self):
         while 1:
             l = self.readline()
             if not l:
                 break
             yield l
 
 # @@ -start,len +start,len @@ or @@ -start +start @@ if len is 1
 unidesc = re.compile('@@ -(\d+)(,(\d+))? \+(\d+)(,(\d+))? @@')
 contextdesc = re.compile('(---|\*\*\*) (\d+)(,(\d+))? (---|\*\*\*)')
 eolmodes = ['strict', 'crlf', 'lf', 'auto']
 
 class patchfile(object):
     def __init__(self, ui, fname, opener, missing=False, eolmode='strict'):
         self.fname = fname
         self.eolmode = eolmode
         self.eol = None
         self.opener = opener
         self.ui = ui
         self.lines = []
         self.exists = False
         self.missing = missing
         if not missing:
             try:
                 self.lines = self.readlines(fname)
                 self.exists = True
             except IOError:
                 pass
         else:
             self.ui.warn(_("unable to find '%s' for patching\n") % self.fname)
 
         self.hash = {}
         self.dirty = 0
         self.offset = 0
         self.rej = []
         self.fileprinted = False
         self.printfile(False)
         self.hunks = 0
 
     def readlines(self, fname):
         if os.path.islink(fname):
             return [os.readlink(fname)]
         fp = self.opener(fname, 'r')
         try:
             lr = linereader(fp, self.eolmode != 'strict')
             lines = list(lr)
             self.eol = lr.eol
             return lines
         finally:
             fp.close()
 
     def writelines(self, fname, lines):
         # Ensure supplied data ends in fname, being a regular file or
         # a symlink. updatedir() will -too magically- take care of
         # setting it to the proper type afterwards.
         islink = os.path.islink(fname)
         if islink:
             fp = cStringIO.StringIO()
         else:
             fp = self.opener(fname, 'w')
         try:
-            if self.eolmode == 'auto' and self.eol:
+            if self.eolmode == 'auto':
                 eol = self.eol
             elif self.eolmode == 'crlf':
                 eol = '\r\n'
             else:
                 eol = '\n'
 
-            if self.eolmode != 'strict' and eol != '\n':
+            if self.eolmode != 'strict' and eol and eol != '\n':
                 for l in lines:
                     if l and l[-1] == '\n':
                         l = l[:-1] + eol
                     fp.write(l)
             else:
                 fp.writelines(lines)
             if islink:
                 self.opener.symlink(fp.getvalue(), fname)
         finally:
             fp.close()
 
     def unlink(self, fname):
         os.unlink(fname)
 
     def printfile(self, warn):
         if self.fileprinted:
             return
         if warn or self.ui.verbose:
             self.fileprinted = True
         s = _("patching file %s\n") % self.fname
         if warn:
             self.ui.warn(s)
         else:
             self.ui.note(s)
 
 
     def findlines(self, l, linenum):
         # looks through the hash and finds candidate lines.  The
         # result is a list of line numbers sorted based on distance
         # from linenum
 
         cand = self.hash.get(l, [])
         if len(cand) > 1:
             # resort our list of potentials forward then back.
             cand.sort(key=lambda x: abs(x - linenum))
         return cand
 
     def hashlines(self):
         self.hash = {}
         for x, s in enumerate(self.lines):
             self.hash.setdefault(s, []).append(x)
 
     def write_rej(self):
         # our rejects are a little different from patch(1).  This always
         # creates rejects in the same form as the original patch.  A file
         # header is inserted so that you can run the reject through patch again
         # without having to type the filename.
 
         if not self.rej:
             return
 
         fname = self.fname + ".rej"
         self.ui.warn(
             _("%d out of %d hunks FAILED -- saving rejects to file %s\n") %
             (len(self.rej), self.hunks, fname))
 
         def rejlines():
             base = os.path.basename(self.fname)
             yield "--- %s\n+++ %s\n" % (base, base)
             for x in self.rej:
                 for l in x.hunk:
                     yield l
                     if l[-1] != '\n':
                         yield "\n\ No newline at end of file\n"
 
         self.writelines(fname, rejlines())
 
     def write(self, dest=None):
         if not self.dirty:
             return
         if not dest:
             dest = self.fname
         self.writelines(dest, self.lines)
 
     def close(self):
         self.write()
         self.write_rej()
 
     def apply(self, h):
         if not h.complete():
             raise PatchError(_("bad hunk #%d %s (%d %d %d %d)") %
                             (h.number, h.desc, len(h.a), h.lena, len(h.b),
                             h.lenb))
 
         self.hunks += 1
 
         if self.missing:
             self.rej.append(h)
             return -1
 
         if self.exists and h.createfile():
             self.ui.warn(_("file %s already exists\n") % self.fname)
             self.rej.append(h)
             return -1
 
         if isinstance(h, binhunk):
             if h.rmfile():
                 self.unlink(self.fname)
             else:
                 self.lines[:] = h.new()
                 self.offset += len(h.new())
                 self.dirty = 1
             return 0
 
+        horig = h
+        if self.eolmode == 'auto' and self.eol:
+            # If eolmode == 'auto' and target file exists and has line
+            # endings we have to normalize input data before patching.
+            # Otherwise, patchfile operates in 'strict' mode. If
+            # eolmode is set to 'crlf' or 'lf', input hunk is already
+            # normalized to avoid data copy.
+            h = h.getnormalized()
+
         # fast case first, no offsets, no fuzz
         old = h.old()
         # patch starts counting at 1 unless we are adding the file
         if h.starta == 0:
             start = 0
         else:
             start = h.starta + self.offset - 1
         orig_start = start
         if diffhelpers.testhunk(old, self.lines, start) == 0:
             if h.rmfile():
                 self.unlink(self.fname)
             else:
                 self.lines[start : start + h.lena] = h.new()
                 self.offset += h.lenb - h.lena
                 self.dirty = 1
             return 0
 
         # ok, we couldn't match the hunk.  Lets look for offsets and fuzz it
         self.hashlines()
         if h.hunk[-1][0] != ' ':
             # if the hunk tried to put something at the bottom of the file
             # override the start line and use eof here
             search_start = len(self.lines)
         else:
             search_start = orig_start
 
         for fuzzlen in xrange(3):
             for toponly in [ True, False ]:
                 old = h.old(fuzzlen, toponly)
 
                 cand = self.findlines(old[0][1:], search_start)
                 for l in cand:
                     if diffhelpers.testhunk(old, self.lines, l) == 0:
                         newlines = h.new(fuzzlen, toponly)
                         self.lines[l : l + len(old)] = newlines
                         self.offset += len(newlines) - len(old)
                         self.dirty = 1
                         if fuzzlen:
                             fuzzstr = "with fuzz %d " % fuzzlen
                             f = self.ui.warn
                             self.printfile(True)
                         else:
                             fuzzstr = ""
                             f = self.ui.note
                         offset = l - orig_start - fuzzlen
                         if offset == 1:
                             msg = _("Hunk #%d succeeded at %d %s"
                                     "(offset %d line).\n")
                         else:
                             msg = _("Hunk #%d succeeded at %d %s"
                                     "(offset %d lines).\n")
                         f(msg % (h.number, l+1, fuzzstr, offset))
                         return fuzzlen
         self.printfile(True)
         self.ui.warn(_("Hunk #%d FAILED at %d\n") % (h.number, orig_start))
-        self.rej.append(h)
+        self.rej.append(horig)
         return -1
 
 class hunk(object):
     def __init__(self, desc, num, lr, context, create=False, remove=False):
         self.number = num
         self.desc = desc
         self.hunk = [ desc ]
         self.a = []
         self.b = []
         self.starta = self.lena = None
         self.startb = self.lenb = None
-        if context:
-            self.read_context_hunk(lr)
-        else:
-            self.read_unified_hunk(lr)
+        if lr is not None:
+            if context:
+                self.read_context_hunk(lr)
+            else:
+                self.read_unified_hunk(lr)
         self.create = create
         self.remove = remove and not create
 
+    def getnormalized(self):
+        """Return a copy with line endings normalized to LF."""
+
+        def normalize(lines):
+            nlines = []
+            for line in lines:
+                if line.endswith('\r\n'):
+                    line = line[:-2] + '\n'
+                nlines.append(line)
+            return nlines
+
+        # Dummy object, it is rebuilt manually
+        nh = hunk(self.desc, self.number, None, None, False, False)
+        nh.number = self.number
+        nh.desc = self.desc
+        nh.a = normalize(self.a)
+        nh.b = normalize(self.b)
+        nh.starta = self.starta
+        nh.startb = self.startb
+        nh.lena = self.lena
+        nh.lenb = self.lenb
+        nh.create = self.create
+        nh.remove = self.remove
+        return nh
+
     def read_unified_hunk(self, lr):
         m = unidesc.match(self.desc)
         if not m:
             raise PatchError(_("bad hunk #%d") % self.number)
         self.starta, foo, self.lena, self.startb, foo2, self.lenb = m.groups()
         if self.lena is None:
             self.lena = 1
         else:
             self.lena = int(self.lena)
         if self.lenb is None:
             self.lenb = 1
         else:
             self.lenb = int(self.lenb)
         self.starta = int(self.starta)
         self.startb = int(self.startb)
         diffhelpers.addlines(lr, self.hunk, self.lena, self.lenb, self.a, self.b)
         # if we hit eof before finishing out the hunk, the last line will
         # be zero length.  Lets try to fix it up.
         while len(self.hunk[-1]) == 0:
             del self.hunk[-1]
             del self.a[-1]
             del self.b[-1]
             self.lena -= 1
             self.lenb -= 1
 
     def read_context_hunk(self, lr):
         self.desc = lr.readline()
         m = contextdesc.match(self.desc)
         if not m:
             raise PatchError(_("bad hunk #%d") % self.number)
         foo, self.starta, foo2, aend, foo3 = m.groups()
         self.starta = int(self.starta)
         if aend is None:
             aend = self.starta
         self.lena = int(aend) - self.starta
         if self.starta:
             self.lena += 1
         for x in xrange(self.lena):
             l = lr.readline()
             if l.startswith('---'):
                 lr.push(l)
                 break
             s = l[2:]
             if l.startswith('- ') or l.startswith('! '):
                 u = '-' + s
             elif l.startswith('  '):
                 u = ' ' + s
             else:
                 raise PatchError(_("bad hunk #%d old text line %d") %
                                  (self.number, x))
             self.a.append(u)
             self.hunk.append(u)
 
         l = lr.readline()
         if l.startswith('\ '):
             s = self.a[-1][:-1]
             self.a[-1] = s
             self.hunk[-1] = s
             l = lr.readline()
         m = contextdesc.match(l)
         if not m:
             raise PatchError(_("bad hunk #%d") % self.number)
         foo, self.startb, foo2, bend, foo3 = m.groups()
         self.startb = int(self.startb)
         if bend is None:
             bend = self.startb
         self.lenb = int(bend) - self.startb
         if self.startb:
             self.lenb += 1
         hunki = 1
         for x in xrange(self.lenb):
             l = lr.readline()
             if l.startswith('\ '):
                 s = self.b[-1][:-1]
                 self.b[-1] = s
                 self.hunk[hunki-1] = s
                 continue
             if not l:
                 lr.push(l)
                 break
             s = l[2:]
             if l.startswith('+ ') or l.startswith('! '):
                 u = '+' + s
             elif l.startswith('  '):
                 u = ' ' + s
             elif len(self.b) == 0:
                 # this can happen when the hunk does not add any lines
                 lr.push(l)
                 break
             else:
                 raise PatchError(_("bad hunk #%d old text line %d") %
                                  (self.number, x))
             self.b.append(s)
             while True:
                 if hunki >= len(self.hunk):
                     h = ""
                 else:
                     h = self.hunk[hunki]
                 hunki += 1
                 if h == u:
                     break
                 elif h.startswith('-'):
                     continue
                 else:
                     self.hunk.insert(hunki-1, u)
                     break
 
         if not self.a:
             # this happens when lines were only added to the hunk
             for x in self.hunk:
                 if x.startswith('-') or x.startswith(' '):
                     self.a.append(x)
         if not self.b:
             # this happens when lines were only deleted from the hunk
             for x in self.hunk:
                 if x.startswith('+') or x.startswith(' '):
                     self.b.append(x[1:])
         # @@ -start,len +start,len @@
         self.desc = "@@ -%d,%d +%d,%d @@\n" % (self.starta, self.lena,
                                              self.startb, self.lenb)
         self.hunk[0] = self.desc
 
     def fix_newline(self):
         diffhelpers.fix_newline(self.hunk, self.a, self.b)
 
     def complete(self):
         return len(self.a) == self.lena and len(self.b) == self.lenb
 
     def createfile(self):
         return self.starta == 0 and self.lena == 0 and self.create
 
     def rmfile(self):
         return self.startb == 0 and self.lenb == 0 and self.remove
 
     def fuzzit(self, l, fuzz, toponly):
         # this removes context lines from the top and bottom of list 'l'.  It
         # checks the hunk to make sure only context lines are removed, and then
         # returns a new shortened list of lines.
         fuzz = min(fuzz, len(l)-1)
         if fuzz:
             top = 0
             bot = 0
             hlen = len(self.hunk)
             for x in xrange(hlen-1):
                 # the hunk starts with the @@ line, so use x+1
                 if self.hunk[x+1][0] == ' ':
                     top += 1
                 else:
                     break
             if not toponly:
                 for x in xrange(hlen-1):
                     if self.hunk[hlen-bot-1][0] == ' ':
                         bot += 1
                     else:
                         break
 
             # top and bot now count context in the hunk
             # adjust them if either one is short
             context = max(top, bot, 3)
             if bot < context:
                 bot = max(0, fuzz - (context - bot))
             else:
                 bot = min(fuzz, bot)
             if top < context:
                 top = max(0, fuzz - (context - top))
             else:
                 top = min(fuzz, top)
 
             return l[top:len(l)-bot]
         return l
 
     def old(self, fuzz=0, toponly=False):
         return self.fuzzit(self.a, fuzz, toponly)
 
     def newctrl(self):
         res = []
         for x in self.hunk:
             c = x[0]
             if c == ' ' or c == '+':
                 res.append(x)
         return res
 
     def new(self, fuzz=0, toponly=False):
         return self.fuzzit(self.b, fuzz, toponly)
 
 class binhunk:
     'A binary patch file. Only understands literals so far.'
     def __init__(self, gitpatch):
         self.gitpatch = gitpatch
         self.text = None
         self.hunk = ['GIT binary patch\n']
 
     def createfile(self):
         return self.gitpatch.op in ('ADD', 'RENAME', 'COPY')
 
     def rmfile(self):
         return self.gitpatch.op == 'DELETE'
 
     def complete(self):
         return self.text is not None
 
     def new(self):
         return [self.text]
 
     def extract(self, lr):
         line = lr.readline()
         self.hunk.append(line)
         while line and not line.startswith('literal '):
             line = lr.readline()
             self.hunk.append(line)
         if not line:
             raise PatchError(_('could not extract binary patch'))
         size = int(line[8:].rstrip())
         dec = []
         line = lr.readline()
         self.hunk.append(line)
         while len(line) > 1:
             l = line[0]
             if l <= 'Z' and l >= 'A':
                 l = ord(l) - ord('A') + 1
             else:
                 l = ord(l) - ord('a') + 27
             dec.append(base85.b85decode(line[1:-1])[:l])
             line = lr.readline()
             self.hunk.append(line)
         text = zlib.decompress(''.join(dec))
         if len(text) != size:
             raise PatchError(_('binary patch is %d bytes, not %d') %
                              len(text), size)
         self.text = text
 
 def parsefilename(str):
     # --- filename \t|space stuff
     s = str[4:].rstrip('\r\n')
     i = s.find('\t')
     if i < 0:
         i = s.find(' ')
         if i < 0:
             return s
     return s[:i]
 
 def selectfile(afile_orig, bfile_orig, hunk, strip):
     def pathstrip(path, count=1):
         pathlen = len(path)
         i = 0
         if count == 0:
             return '', path.rstrip()
         while count > 0:
             i = path.find('/', i)
             if i == -1:
                 raise PatchError(_("unable to strip away %d dirs from %s") %
                                  (count, path))
             i += 1
             # consume '//' in the path
             while i < pathlen - 1 and path[i] == '/':
                 i += 1
             count -= 1
         return path[:i].lstrip(), path[i:].rstrip()
 
     nulla = afile_orig == "/dev/null"
     nullb = bfile_orig == "/dev/null"
     abase, afile = pathstrip(afile_orig, strip)
     gooda = not nulla and util.lexists(afile)
     bbase, bfile = pathstrip(bfile_orig, strip)
     if afile == bfile:
         goodb = gooda
     else:
         goodb = not nullb and os.path.exists(bfile)
     createfunc = hunk.createfile
     missing = not goodb and not gooda and not createfunc()
 
     # some diff programs apparently produce create patches where the
     # afile is not /dev/null, but rather the same name as the bfile
     if missing and afile == bfile:
         # this isn't very pretty
         hunk.create = True
         if createfunc():
             missing = False
         else:
             hunk.create = False
 
     # If afile is "a/b/foo" and bfile is "a/b/foo.orig" we assume the
     # diff is between a file and its backup. In this case, the original
     # file should be patched (see original mpatch code).
     isbackup = (abase == bbase and bfile.startswith(afile))
     fname = None
     if not missing:
         if gooda and goodb:
             fname = isbackup and afile or bfile
         elif gooda:
             fname = afile
 
     if not fname:
         if not nullb:
             fname = isbackup and afile or bfile
         elif not nulla:
             fname = afile
         else:
             raise PatchError(_("undefined source and destination files"))
 
     return fname, missing
 
 def scangitpatch(lr, firstline):
     """
     Git patches can emit:
     - rename a to b
     - change b
     - copy a to c
     - change c
 
     We cannot apply this sequence as-is, the renamed 'a' could not be
     found for it would have been renamed already. And we cannot copy
     from 'b' instead because 'b' would have been changed already. So
     we scan the git patch for copy and rename commands so we can
     perform the copies ahead of time.
     """
     pos = 0
     try:
         pos = lr.fp.tell()
         fp = lr.fp
     except IOError:
         fp = cStringIO.StringIO(lr.fp.read())
     gitlr = linereader(fp, lr.textmode)
     gitlr.push(firstline)
     (dopatch, gitpatches) = readgitpatch(gitlr)
     fp.seek(pos)
     return dopatch, gitpatches
 
 def iterhunks(ui, fp, sourcefile=None, textmode=False):
     """Read a patch and yield the following events:
     - ("file", afile, bfile, firsthunk): select a new target file.
     - ("hunk", hunk): a new hunk is ready to be applied, follows a
     "file" event.
     - ("git", gitchanges): current diff is in git format, gitchanges
     maps filenames to gitpatch records. Unique event.
 
     If textmode is True, input line-endings are normalized to LF.
     """
     changed = {}
     current_hunk = None
     afile = ""
     bfile = ""
     state = None
     hunknum = 0
     emitfile = False
     git = False
 
     # our states
     BFILE = 1
     context = None
     lr = linereader(fp, textmode)
     dopatch = True
     # gitworkdone is True if a git operation (copy, rename, ...) was
     # performed already for the current file. Useful when the file
     # section may have no hunk.
     gitworkdone = False
 
     while True:
         newfile = False
         x = lr.readline()
         if not x:
             break
         if current_hunk:
             if x.startswith('\ '):
                 current_hunk.fix_newline()
             yield 'hunk', current_hunk
             current_hunk = None
             gitworkdone = False
         if ((sourcefile or state == BFILE) and ((not context and x[0] == '@') or
             ((context is not False) and x.startswith('***************')))):
             try:
                 if context is None and x.startswith('***************'):
                     context = True
                 gpatch = changed.get(bfile)
                 create = afile == '/dev/null' or gpatch and gpatch.op == 'ADD'
                 remove = bfile == '/dev/null' or gpatch and gpatch.op == 'DELETE'
                 current_hunk = hunk(x, hunknum + 1, lr, context, create, remove)
             except PatchError, err:
                 ui.debug(err)
                 current_hunk = None
                 continue
             hunknum += 1
             if emitfile:
                 emitfile = False
                 yield 'file', (afile, bfile, current_hunk)
         elif state == BFILE and x.startswith('GIT binary patch'):
             current_hunk = binhunk(changed[bfile])
             hunknum += 1
             if emitfile:
                 emitfile = False
                 yield 'file', ('a/' + afile, 'b/' + bfile, current_hunk)
             current_hunk.extract(lr)
         elif x.startswith('diff --git'):
             # check for git diff, scanning the whole patch file if needed
             m = gitre.match(x)
             if m:
                 afile, bfile = m.group(1, 2)
                 if not git:
                     git = True
                     dopatch, gitpatches = scangitpatch(lr, x)
                     yield 'git', gitpatches
                     for gp in gitpatches:
                         changed[gp.path] = gp
                 # else error?
                 # copy/rename + modify should modify target, not source
                 gp = changed.get(bfile)
                 if gp and gp.op in ('COPY', 'DELETE', 'RENAME', 'ADD'):
                     afile = bfile
                     gitworkdone = True
             newfile = True
         elif x.startswith('---'):
             # check for a unified diff
             l2 = lr.readline()
             if not l2.startswith('+++'):
                 lr.push(l2)
                 continue
             newfile = True
             context = False
             afile = parsefilename(x)
             bfile = parsefilename(l2)
         elif x.startswith('***'):
             # check for a context diff
             l2 = lr.readline()
             if not l2.startswith('---'):
                 lr.push(l2)
                 continue
             l3 = lr.readline()
             lr.push(l3)
             if not l3.startswith("***************"):
                 lr.push(l2)
                 continue
             newfile = True
             context = True
             afile = parsefilename(x)
             bfile = parsefilename(l2)
 
         if newfile:
             emitfile = True
             state = BFILE
             hunknum = 0
     if current_hunk:
         if current_hunk.complete():
             yield 'hunk', current_hunk
         else:
             raise PatchError(_("malformed patch %s %s") % (afile,
                              current_hunk.desc))
 
     if hunknum == 0 and dopatch and not gitworkdone:
         raise NoHunks
 
 def applydiff(ui, fp, changed, strip=1, sourcefile=None, eolmode='strict'):
     """
     Reads a patch from fp and tries to apply it.
 
     The dict 'changed' is filled in with all of the filenames changed
     by the patch. Returns 0 for a clean patch, -1 if any rejects were
     found and 1 if there was any fuzz.
 
     If 'eolmode' is 'strict', the patch content and patched file are
     read in binary mode. Otherwise, line endings are ignored when
     patching then normalized according to 'eolmode'.
     """
     rejects = 0
     err = 0
     current_file = None
     gitpatches = None
     opener = util.opener(os.getcwd())
-    textmode = eolmode != 'strict'
+    # In 'auto' mode, we must preserve original eols if target file
+    # eols are undefined. Otherwise, hunk data will be normalized
+    # later.
+    textmode = eolmode not in ('strict', 'auto')
 
     def closefile():
         if not current_file:
             return 0
         current_file.close()
         return len(current_file.rej)
 
     for state, values in iterhunks(ui, fp, sourcefile, textmode):
         if state == 'hunk':
             if not current_file:
                 continue
             current_hunk = values
             ret = current_file.apply(current_hunk)
             if ret >= 0:
                 changed.setdefault(current_file.fname, None)
                 if ret > 0:
                     err = 1
         elif state == 'file':
             rejects += closefile()
             afile, bfile, first_hunk = values
             try:
                 if sourcefile:
                     current_file = patchfile(ui, sourcefile, opener, eolmode=eolmode)
                 else:
                     current_file, missing = selectfile(afile, bfile, first_hunk,
                                             strip)
                     current_file = patchfile(ui, current_file, opener, missing, eolmode)
             except PatchError, err:
                 ui.warn(str(err) + '\n')
                 current_file, current_hunk = None, None
                 rejects += 1
                 continue
         elif state == 'git':
             gitpatches = values
             cwd = os.getcwd()
             for gp in gitpatches:
                 if gp.op in ('COPY', 'RENAME'):
                     copyfile(gp.oldpath, gp.path, cwd)
                 changed[gp.path] = gp
         else:
             raise util.Abort(_('unsupported parser state: %s') % state)
 
     rejects += closefile()
 
     if rejects:
         return -1
     return err
 
 def diffopts(ui, opts=None, untrusted=False):
     def get(key, name=None, getter=ui.configbool):
         return ((opts and opts.get(key)) or
                 getter('diff', name or key, None, untrusted=untrusted))
     return mdiff.diffopts(
         text=opts and opts.get('text'),
         git=get('git'),
         nodates=get('nodates'),
         showfunc=get('show_function', 'showfunc'),
         ignorews=get('ignore_all_space', 'ignorews'),
         ignorewsamount=get('ignore_space_change', 'ignorewsamount'),
         ignoreblanklines=get('ignore_blank_lines', 'ignoreblanklines'),
         context=get('unified', getter=ui.config))
 
 def updatedir(ui, repo, patches, similarity=0):
     '''Update dirstate after patch application according to metadata'''
     if not patches:
         return
     copies = []
     removes = set()
     cfiles = patches.keys()
     cwd = repo.getcwd()
     if cwd:
         cfiles = [util.pathto(repo.root, cwd, f) for f in patches.keys()]
     for f in patches:
         gp = patches[f]
         if not gp:
             continue
         if gp.op == 'RENAME':
             copies.append((gp.oldpath, gp.path))
             removes.add(gp.oldpath)
         elif gp.op == 'COPY':
             copies.append((gp.oldpath, gp.path))
         elif gp.op == 'DELETE':
             removes.add(gp.path)
     for src, dst in copies:
         repo.copy(src, dst)
     if (not similarity) and removes:
         repo.remove(sorted(removes), True)
     for f in patches:
         gp = patches[f]
         if gp and gp.mode:
             islink, isexec = gp.mode
             dst = repo.wjoin(gp.path)
             # patch won't create empty files
             if gp.op == 'ADD' and not os.path.exists(dst):
                 flags = (isexec and 'x' or '') + (islink and 'l' or '')
                 repo.wwrite(gp.path, '', flags)
             elif gp.op != 'DELETE':
                 util.set_flags(dst, islink, isexec)
     cmdutil.addremove(repo, cfiles, similarity=similarity)
     files = patches.keys()
     files.extend([r for r in removes if r not in files])
     return sorted(files)
 
 def externalpatch(patcher, args, patchname, ui, strip, cwd, files):
     """use <patcher> to apply <patchname> to the working directory.
     returns whether patch was applied with fuzz factor."""
 
     fuzz = False
     if cwd:
         args.append('-d %s' % util.shellquote(cwd))
     fp = util.popen('%s %s -p%d < %s' % (patcher, ' '.join(args), strip,
                                        util.shellquote(patchname)))
 
     for line in fp:
         line = line.rstrip()
         ui.note(line + '\n')
         if line.startswith('patching file '):
             pf = util.parse_patch_output(line)
             printed_file = False
             files.setdefault(pf, None)
         elif line.find('with fuzz') >= 0:
             fuzz = True
             if not printed_file:
                 ui.warn(pf + '\n')
                 printed_file = True
             ui.warn(line + '\n')
         elif line.find('saving rejects to file') >= 0:
             ui.warn(line + '\n')
         elif line.find('FAILED') >= 0:
             if not printed_file:
                 ui.warn(pf + '\n')
                 printed_file = True
             ui.warn(line + '\n')
     code = fp.close()
     if code:
         raise PatchError(_("patch command failed: %s") %
                          util.explain_exit(code)[0])
     return fuzz
 
 def internalpatch(patchobj, ui, strip, cwd, files=None, eolmode='strict'):
     """use builtin patch to apply <patchobj> to the working directory.
     returns whether patch was applied with fuzz factor."""
 
     if files is None:
         files = {}
     if eolmode is None:
         eolmode = ui.config('patch', 'eol', 'strict')
     if eolmode.lower() not in eolmodes:
         raise util.Abort(_('Unsupported line endings type: %s') % eolmode)
     eolmode = eolmode.lower()
 
     try:
         fp = open(patchobj, 'rb')
     except TypeError:
         fp = patchobj
     if cwd:
         curdir = os.getcwd()
         os.chdir(cwd)
     try:
         ret = applydiff(ui, fp, files, strip=strip, eolmode=eolmode)
     finally:
         if cwd:
             os.chdir(curdir)
     if ret < 0:
         raise PatchError
     return ret > 0
 
 def patch(patchname, ui, strip=1, cwd=None, files=None, eolmode='strict'):
     """Apply <patchname> to the working directory.
 
     'eolmode' specifies how end of lines should be handled. It can be:
     - 'strict': inputs are read in binary mode, EOLs are preserved
     - 'crlf': EOLs are ignored when patching and reset to CRLF
     - 'lf': EOLs are ignored when patching and reset to LF
     - None: get it from user settings, default to 'strict'
     'eolmode' is ignored when using an external patcher program.
 
     Returns whether patch was applied with fuzz factor.
     """
     patcher = ui.config('ui', 'patch')
     args = []
     if files is None:
         files = {}
     try:
         if patcher:
             return externalpatch(patcher, args, patchname, ui, strip, cwd,
                                  files)
         else:
             try:
                 return internalpatch(patchname, ui, strip, cwd, files, eolmode)
             except NoHunks:
                 patcher = util.find_exe('gpatch') or util.find_exe('patch') or 'patch'
                 ui.debug('no valid hunks found; trying with %r instead\n' %
                          patcher)
                 if util.needbinarypatch():
                     args.append('--binary')
                 return externalpatch(patcher, args, patchname, ui, strip, cwd,
                                      files)
     except PatchError, err:
         s = str(err)
         if s:
             raise util.Abort(s)
         else:
             raise util.Abort(_('patch failed to apply'))
 
 def b85diff(to, tn):
     '''print base85-encoded binary diff'''
     def gitindex(text):
         if not text:
             return '0' * 40
         l = len(text)
         s = util.sha1('blob %d\0' % l)
         s.update(text)
         return s.hexdigest()
 
     def fmtline(line):
         l = len(line)
         if l <= 26:
             l = chr(ord('A') + l - 1)
         else:
             l = chr(l - 26 + ord('a') - 1)
         return '%c%s\n' % (l, base85.b85encode(line, True))
 
     def chunk(text, csize=52):
         l = len(text)
         i = 0
         while i < l:
             yield text[i:i+csize]
             i += csize
 
     tohash = gitindex(to)
     tnhash = gitindex(tn)
     if tohash == tnhash:
         return ""
 
     # TODO: deltas
     ret = ['index %s..%s\nGIT binary patch\nliteral %s\n' %
            (tohash, tnhash, len(tn))]
     for l in chunk(zlib.compress(tn)):
         ret.append(fmtline(l))
     ret.append('\n')
     return ''.join(ret)
 
 def _addmodehdr(header, omode, nmode):
     if omode != nmode:
         header.append('old mode %s\n' % omode)
         header.append('new mode %s\n' % nmode)
 
 def diff(repo, node1=None, node2=None, match=None, changes=None, opts=None):
     '''yields diff of changes to files between two nodes, or node and
     working directory.
 
     if node1 is None, use first dirstate parent instead.
     if node2 is None, compare node1 with working directory.'''
 
     if opts is None:
         opts = mdiff.defaultopts
 
     if not node1 and not node2:
         node1 = repo.dirstate.parents()[0]
 
     def lrugetfilectx():
         cache = {}
         order = []
         def getfilectx(f, ctx):
             fctx = ctx.filectx(f, filelog=cache.get(f))
             if f not in cache:
                 if len(cache) > 20:
                     del cache[order.pop(0)]
                 cache[f] = fctx.filelog()
             else:
                 order.remove(f)
             order.append(f)
             return fctx
         return getfilectx
     getfilectx = lrugetfilectx()
 
     ctx1 = repo[node1]
     ctx2 = repo[node2]
 
     if not changes:
         changes = repo.status(ctx1, ctx2, match=match)
     modified, added, removed = changes[:3]
 
     if not modified and not added and not removed:
         return
 
     date1 = util.datestr(ctx1.date())
     man1 = ctx1.manifest()
 
     if repo.ui.quiet:
         r = None
     else:
         hexfunc = repo.ui.debugflag and hex or short
         r = [hexfunc(node) for node in [node1, node2] if node]
 
     if opts.git:
         copy, diverge = copies.copies(repo, ctx1, ctx2, repo[nullid])
         copy = copy.copy()
         for k, v in copy.items():
             copy[v] = k
 
     gone = set()
     gitmode = {'l': '120000', 'x': '100755', '': '100644'}
 
     for f in sorted(modified + added + removed):
         to = None
         tn = None
         dodiff = True
         header = []
         if f in man1:
             to = getfilectx(f, ctx1).data()
         if f not in removed:
             tn = getfilectx(f, ctx2).data()
         a, b = f, f
         if opts.git:
             if f in added:
                 mode = gitmode[ctx2.flags(f)]
                 if f in copy:
                     a = copy[f]
                     omode = gitmode[man1.flags(a)]
                     _addmodehdr(header, omode, mode)
                     if a in removed and a not in gone:
                         op = 'rename'
                         gone.add(a)
                     else:
                         op = 'copy'
                     header.append('%s from %s\n' % (op, a))
                     header.append('%s to %s\n' % (op, f))
                     to = getfilectx(a, ctx1).data()
                 else:
                     header.append('new file mode %s\n' % mode)
                 if util.binary(tn):
                     dodiff = 'binary'
             elif f in removed:
                 # have we already reported a copy above?
                 if f in copy and copy[f] in added and copy[copy[f]] == f:
                     dodiff = False
                 else:
                     header.append('deleted file mode %s\n' %
                                   gitmode[man1.flags(f)])
             else:
                 omode = gitmode[man1.flags(f)]
                 nmode = gitmode[ctx2.flags(f)]
                 _addmodehdr(header, omode, nmode)
                 if util.binary(to) or util.binary(tn):
                     dodiff = 'binary'
             r = None
             header.insert(0, mdiff.diffline(r, a, b, opts))
         if dodiff:
             if dodiff == 'binary':
                 text = b85diff(to, tn)
             else:
                 text = mdiff.unidiff(to, date1,
                                     # ctx2 date may be dynamic
                                     tn, util.datestr(ctx2.date()),
                                     a, b, r, opts=opts)
             if header and (text or len(header) > 1):
                 yield ''.join(header)
             if text:
                 yield text
 
 def export(repo, revs, template='hg-%h.patch', fp=None, switch_parent=False,
            opts=None):
     '''export changesets as hg patches.'''
 
     total = len(revs)
     revwidth = max([len(str(rev)) for rev in revs])
 
     def single(rev, seqno, fp):
         ctx = repo[rev]
         node = ctx.node()
         parents = [p.node() for p in ctx.parents() if p]
         branch = ctx.branch()
         if switch_parent:
             parents.reverse()
         prev = (parents and parents[0]) or nullid
 
         if not fp:
             fp = cmdutil.make_file(repo, template, node, total=total,
                                    seqno=seqno, revwidth=revwidth,
                                    mode='ab')
         if fp != sys.stdout and hasattr(fp, 'name'):
             repo.ui.note("%s\n" % fp.name)
 
         fp.write("# HG changeset patch\n")
         fp.write("# User %s\n" % ctx.user())
         fp.write("# Date %d %d\n" % ctx.date())
         if branch and (branch != 'default'):
             fp.write("# Branch %s\n" % branch)
         fp.write("# Node ID %s\n" % hex(node))
         fp.write("# Parent  %s\n" % hex(prev))
         if len(parents) > 1:
             fp.write("# Parent  %s\n" % hex(parents[1]))
         fp.write(ctx.description().rstrip())
         fp.write("\n\n")
 
         for chunk in diff(repo, prev, node, opts=opts):
             fp.write(chunk)
 
     for seqno, rev in enumerate(revs):
         single(rev, seqno+1, fp)
 
 def diffstatdata(lines):
     filename, adds, removes = None, 0, 0
     for line in lines:
         if line.startswith('diff'):
             if filename:
                 isbinary = adds == 0 and removes == 0
                 yield (filename, adds, removes, isbinary)
             # set numbers to 0 anyway when starting new file
             adds, removes = 0, 0
             if line.startswith('diff --git'):
                 filename = gitre.search(line).group(1)
             else:
                 # format: "diff -r ... -r ... filename"
                 filename = line.split(None, 5)[-1]
         elif line.startswith('+') and not line.startswith('+++'):
             adds += 1
         elif line.startswith('-') and not line.startswith('---'):
             removes += 1
     if filename:
         isbinary = adds == 0 and removes == 0
         yield (filename, adds, removes, isbinary)
 
 def diffstat(lines, width=80, git=False):
     output = []
     stats = list(diffstatdata(lines))
 
     maxtotal, maxname = 0, 0
     totaladds, totalremoves = 0, 0
     hasbinary = False
     for filename, adds, removes, isbinary in stats:
         totaladds += adds
         totalremoves += removes
         maxname = max(maxname, len(filename))
         maxtotal = max(maxtotal, adds+removes)
         if isbinary:
             hasbinary = True
 
     countwidth = len(str(maxtotal))
     if hasbinary and countwidth < 3:
         countwidth = 3
     graphwidth = width - countwidth - maxname - 6
     if graphwidth < 10:
         graphwidth = 10
 
     def scale(i):
         if maxtotal <= graphwidth:
             return i
         # If diffstat runs out of room it doesn't print anything,
         # which isn't very useful, so always print at least one + or -
         # if there were at least some changes.
         return max(i * graphwidth // maxtotal, int(bool(i)))
 
     for filename, adds, removes, isbinary in stats:
         if git and isbinary:
             count = 'Bin'
         else:
             count = adds + removes
         pluses = '+' * scale(adds)
         minuses = '-' * scale(removes)
         output.append(' %-*s |  %*s %s%s\n' % (maxname, filename, countwidth,
                                                count, pluses, minuses))
 
     if stats:
         output.append(_(' %d files changed, %d insertions(+), %d deletions(-)\n')
                       % (len(stats), totaladds, totalremoves))
 
     return ''.join(output)

tests/test-import-eol

@@ -1,70 +1,85 b''
 #!/bin/sh
 
 cat > makepatch.py <<EOF
 f = file('eol.diff', 'wb')
 w = f.write
 w('test message\n')
 w('diff --git a/a b/a\n')
 w('--- a/a\n')
 w('+++ b/a\n')
 w('@@ -1,5 +1,5 @@\n')
 w(' a\n')
 w('-bbb\r\n')
 w('+yyyy\r\n')
 w(' cc\r\n')
 w(' \n')
 w(' d\n')
 w('-e\n')
 w('\ No newline at end of file\n')
 w('+z\r\n')
 w('\ No newline at end of file\r\n')
 EOF
 
 hg init repo
 cd repo
 echo '\.diff' > .hgignore
 
 # Test different --eol values
 python -c 'file("a", "wb").write("a\nbbb\ncc\n\nd\ne")'
 hg ci -Am adda
 python ../makepatch.py
 
 echo % invalid eol
 hg --config patch.eol='LFCR' import eol.diff
 hg revert -a
 
 echo % force LF
 hg --traceback --config patch.eol='LF' import eol.diff
 python -c 'print repr(file("a","rb").read())'
 hg st
 
 echo % force CRLF
 hg up -C 0
 hg --traceback --config patch.eol='CRLF' import eol.diff
 python -c 'print repr(file("a","rb").read())'
 hg st
 
 echo % auto EOL on LF file
 hg up -C 0
 hg --traceback --config patch.eol='auto' import eol.diff
 python -c 'print repr(file("a","rb").read())'
 hg st
 
 echo % auto EOL on CRLF file
 python -c 'file("a", "wb").write("a\r\nbbb\r\ncc\r\n\r\nd\r\ne")'
 hg commit -m 'switch EOLs in a'
 hg --traceback --config patch.eol='auto' import eol.diff
 python -c 'print repr(file("a","rb").read())'
 hg st
 
+echo % auto EOL on new file or source without any EOL
+python -c 'file("noeol", "wb").write("noeol")'
+hg add noeol
+hg commit -m 'add noeol'
+python -c 'file("noeol", "wb").write("noeol\r\nnoeol\n")'
+python -c 'file("neweol", "wb").write("neweol\nneweol\r\n")'
+hg add neweol
+hg diff --git > noeol.diff
+hg revert --no-backup noeol neweol
+rm neweol
+hg --traceback --config patch.eol='auto' import -m noeol noeol.diff
+python -c 'print repr(file("noeol","rb").read())'
+python -c 'print repr(file("neweol","rb").read())'
+hg st
+
 # Test --eol and binary patches
 python -c 'file("b", "wb").write("a\x00\nb")'
 hg ci -Am addb
 python -c 'file("b", "wb").write("a\x00\nc")'
 hg diff --git > bin.diff
 hg revert --no-backup b
 echo % binary patch with --eol
 hg import --config patch.eol='CRLF' -m changeb bin.diff
 python -c 'print repr(file("b","rb").read())'
 hg st
 cd ..

tests/test-import-eol.out

@@ -1,23 +1,27 b''
 adding .hgignore
 adding a
 % invalid eol
 applying eol.diff
 abort: Unsupported line endings type: LFCR
 % force LF
 applying eol.diff
 'a\nyyyy\ncc\n\nd\ne'
 % force CRLF
 1 files updated, 0 files merged, 0 files removed, 0 files unresolved
 applying eol.diff
 'a\r\nyyyy\r\ncc\r\n\r\nd\r\ne'
 % auto EOL on LF file
 1 files updated, 0 files merged, 0 files removed, 0 files unresolved
 applying eol.diff
 'a\nyyyy\ncc\n\nd\ne'
 % auto EOL on CRLF file
 applying eol.diff
 'a\r\nyyyy\r\ncc\r\n\r\nd\r\ne'
+% auto EOL on new file or source without any EOL
+applying noeol.diff
+'noeol\r\nnoeol\n'
+'neweol\nneweol\r\n'
 adding b
 % binary patch with --eol
 applying bin.diff
 'a\x00\nc'