hgbook

annotate es/tour-merge.tex @ 364:932c10e8c225

find changesets by author, revision, files, or words in the commit message
now, this has section 1.5 finished. Last commit had 1.4. Typo in commit message
author Javier Rojas <jerojasro@devnull.li>
date Sat Oct 25 15:41:52 2008 -0500 (2008-10-25)
parents 04c08ad7e92e
children 7ca1186c422f
rev   line source
jerojasro@336 1 \chapter{A tour of Mercurial: merging work}
jerojasro@336 2 \label{chap:tour-merge}
jerojasro@336 3
jerojasro@336 4 We've now covered cloning a repository, making changes in a
jerojasro@336 5 repository, and pulling or pushing changes from one repository into
jerojasro@336 6 another. Our next step is \emph{merging} changes from separate
jerojasro@336 7 repositories.
jerojasro@336 8
jerojasro@336 9 \section{Merging streams of work}
jerojasro@336 10
jerojasro@336 11 Merging is a fundamental part of working with a distributed revision
jerojasro@336 12 control tool.
jerojasro@336 13 \begin{itemize}
jerojasro@336 14 \item Alice and Bob each have a personal copy of a repository for a
jerojasro@336 15 project they're collaborating on. Alice fixes a bug in her
jerojasro@336 16 repository; Bob adds a new feature in his. They want the shared
jerojasro@336 17 repository to contain both the bug fix and the new feature.
jerojasro@336 18 \item I frequently work on several different tasks for a single
jerojasro@336 19 project at once, each safely isolated in its own repository.
jerojasro@336 20 Working this way means that I often need to merge one piece of my
jerojasro@336 21 own work with another.
jerojasro@336 22 \end{itemize}
jerojasro@336 23
jerojasro@336 24 Because merging is such a common thing to need to do, Mercurial makes
jerojasro@336 25 it easy. Let's walk through the process. We'll begin by cloning yet
jerojasro@336 26 another repository (see how often they spring up?) and making a change
jerojasro@336 27 in it.
jerojasro@336 28 \interaction{tour.merge.clone}
jerojasro@336 29 We should now have two copies of \filename{hello.c} with different
jerojasro@336 30 contents. The histories of the two repositories have also diverged,
jerojasro@336 31 as illustrated in figure~\ref{fig:tour-merge:sep-repos}.
jerojasro@336 32 \interaction{tour.merge.cat}
jerojasro@336 33
jerojasro@336 34 \begin{figure}[ht]
jerojasro@336 35 \centering
jerojasro@336 36 \grafix{tour-merge-sep-repos}
jerojasro@336 37 \caption{Divergent recent histories of the \dirname{my-hello} and
jerojasro@336 38 \dirname{my-new-hello} repositories}
jerojasro@336 39 \label{fig:tour-merge:sep-repos}
jerojasro@336 40 \end{figure}
jerojasro@336 41
jerojasro@336 42 We already know that pulling changes from our \dirname{my-hello}
jerojasro@336 43 repository will have no effect on the working directory.
jerojasro@336 44 \interaction{tour.merge.pull}
jerojasro@336 45 However, the \hgcmd{pull} command says something about ``heads''.
jerojasro@336 46
jerojasro@336 47 \subsection{Head changesets}
jerojasro@336 48
jerojasro@336 49 A head is a change that has no descendants, or children, as they're
jerojasro@336 50 also known. The tip revision is thus a head, because the newest
jerojasro@336 51 revision in a repository doesn't have any children, but a repository
jerojasro@336 52 can contain more than one head.
jerojasro@336 53
jerojasro@336 54 \begin{figure}[ht]
jerojasro@336 55 \centering
jerojasro@336 56 \grafix{tour-merge-pull}
jerojasro@336 57 \caption{Repository contents after pulling from \dirname{my-hello} into
jerojasro@336 58 \dirname{my-new-hello}}
jerojasro@336 59 \label{fig:tour-merge:pull}
jerojasro@336 60 \end{figure}
jerojasro@336 61
jerojasro@336 62 In figure~\ref{fig:tour-merge:pull}, you can see the effect of the
jerojasro@336 63 pull from \dirname{my-hello} into \dirname{my-new-hello}. The history
jerojasro@336 64 that was already present in \dirname{my-new-hello} is untouched, but a
jerojasro@336 65 new revision has been added. By referring to
jerojasro@336 66 figure~\ref{fig:tour-merge:sep-repos}, we can see that the
jerojasro@336 67 \emph{changeset ID} remains the same in the new repository, but the
jerojasro@336 68 \emph{revision number} has changed. (This, incidentally, is a fine
jerojasro@336 69 example of why it's not safe to use revision numbers when discussing
jerojasro@336 70 changesets.) We can view the heads in a repository using the
jerojasro@336 71 \hgcmd{heads} command.
jerojasro@336 72 \interaction{tour.merge.heads}
jerojasro@336 73
jerojasro@336 74 \subsection{Performing the merge}
jerojasro@336 75
jerojasro@336 76 What happens if we try to use the normal \hgcmd{update} command to
jerojasro@336 77 update to the new tip?
jerojasro@336 78 \interaction{tour.merge.update}
jerojasro@336 79 Mercurial is telling us that the \hgcmd{update} command won't do a
jerojasro@336 80 merge; it won't update the working directory when it thinks we might
jerojasro@336 81 be wanting to do a merge, unless we force it to do so. Instead, we
jerojasro@336 82 use the \hgcmd{merge} command to merge the two heads.
jerojasro@336 83 \interaction{tour.merge.merge}
jerojasro@336 84
jerojasro@336 85 \begin{figure}[ht]
jerojasro@336 86 \centering
jerojasro@336 87 \grafix{tour-merge-merge}
jerojasro@336 88 \caption{Working directory and repository during merge, and
jerojasro@336 89 following commit}
jerojasro@336 90 \label{fig:tour-merge:merge}
jerojasro@336 91 \end{figure}
jerojasro@336 92
jerojasro@336 93 This updates the working directory so that it contains changes from
jerojasro@336 94 \emph{both} heads, which is reflected in both the output of
jerojasro@336 95 \hgcmd{parents} and the contents of \filename{hello.c}.
jerojasro@336 96 \interaction{tour.merge.parents}
jerojasro@336 97
jerojasro@336 98 \subsection{Committing the results of the merge}
jerojasro@336 99
jerojasro@336 100 Whenever we've done a merge, \hgcmd{parents} will display two parents
jerojasro@336 101 until we \hgcmd{commit} the results of the merge.
jerojasro@336 102 \interaction{tour.merge.commit}
jerojasro@336 103 We now have a new tip revision; notice that it has \emph{both} of
jerojasro@336 104 our former heads as its parents. These are the same revisions that
jerojasro@336 105 were previously displayed by \hgcmd{parents}.
jerojasro@336 106 \interaction{tour.merge.tip}
jerojasro@336 107 In figure~\ref{fig:tour-merge:merge}, you can see a representation of
jerojasro@336 108 what happens to the working directory during the merge, and how this
jerojasro@336 109 affects the repository when the commit happens. During the merge, the
jerojasro@336 110 working directory has two parent changesets, and these become the
jerojasro@336 111 parents of the new changeset.
jerojasro@336 112
jerojasro@336 113 \section{Merging conflicting changes}
jerojasro@336 114
jerojasro@336 115 Most merges are simple affairs, but sometimes you'll find yourself
jerojasro@336 116 merging changes where each modifies the same portions of the same
jerojasro@336 117 files. Unless both modifications are identical, this results in a
jerojasro@336 118 \emph{conflict}, where you have to decide how to reconcile the
jerojasro@336 119 different changes into something coherent.
jerojasro@336 120
jerojasro@336 121 \begin{figure}[ht]
jerojasro@336 122 \centering
jerojasro@336 123 \grafix{tour-merge-conflict}
jerojasro@336 124 \caption{Conflicting changes to a document}
jerojasro@336 125 \label{fig:tour-merge:conflict}
jerojasro@336 126 \end{figure}
jerojasro@336 127
jerojasro@336 128 Figure~\ref{fig:tour-merge:conflict} illustrates an instance of two
jerojasro@336 129 conflicting changes to a document. We started with a single version
jerojasro@336 130 of the file; then we made some changes; while someone else made
jerojasro@336 131 different changes to the same text. Our task in resolving the
jerojasro@336 132 conflicting changes is to decide what the file should look like.
jerojasro@336 133
jerojasro@336 134 Mercurial doesn't have a built-in facility for handling conflicts.
jerojasro@336 135 Instead, it runs an external program called \command{hgmerge}. This
jerojasro@336 136 is a shell script that is bundled with Mercurial; you can change it to
jerojasro@336 137 behave however you please. What it does by default is try to find one
jerojasro@336 138 of several different merging tools that are likely to be installed on
jerojasro@336 139 your system. It first tries a few fully automatic merging tools; if
jerojasro@336 140 these don't succeed (because the resolution process requires human
jerojasro@336 141 guidance) or aren't present, the script tries a few different
jerojasro@336 142 graphical merging tools.
jerojasro@336 143
jerojasro@336 144 It's also possible to get Mercurial to run another program or script
jerojasro@336 145 instead of \command{hgmerge}, by setting the \envar{HGMERGE}
jerojasro@336 146 environment variable to the name of your preferred program.
jerojasro@336 147
jerojasro@336 148 \subsection{Using a graphical merge tool}
jerojasro@336 149
jerojasro@336 150 My preferred graphical merge tool is \command{kdiff3}, which I'll use
jerojasro@336 151 to describe the features that are common to graphical file merging
jerojasro@336 152 tools. You can see a screenshot of \command{kdiff3} in action in
jerojasro@336 153 figure~\ref{fig:tour-merge:kdiff3}. The kind of merge it is
jerojasro@336 154 performing is called a \emph{three-way merge}, because there are three
jerojasro@336 155 different versions of the file of interest to us. The tool thus
jerojasro@336 156 splits the upper portion of the window into three panes:
jerojasro@336 157 \begin{itemize}
jerojasro@336 158 \item At the left is the \emph{base} version of the file, i.e.~the
jerojasro@336 159 most recent version from which the two versions we're trying to
jerojasro@336 160 merge are descended.
jerojasro@336 161 \item In the middle is ``our'' version of the file, with the contents
jerojasro@336 162 that we modified.
jerojasro@336 163 \item On the right is ``their'' version of the file, the one that
jerojasro@336 164 from the changeset that we're trying to merge with.
jerojasro@336 165 \end{itemize}
jerojasro@336 166 In the pane below these is the current \emph{result} of the merge.
jerojasro@336 167 Our task is to replace all of the red text, which indicates unresolved
jerojasro@336 168 conflicts, with some sensible merger of the ``ours'' and ``theirs''
jerojasro@336 169 versions of the file.
jerojasro@336 170
jerojasro@336 171 All four of these panes are \emph{locked together}; if we scroll
jerojasro@336 172 vertically or horizontally in any of them, the others are updated to
jerojasro@336 173 display the corresponding sections of their respective files.
jerojasro@336 174
jerojasro@336 175 \begin{figure}[ht]
jerojasro@336 176 \centering
jerojasro@336 177 \grafix{kdiff3}
jerojasro@336 178 \caption{Using \command{kdiff3} to merge versions of a file}
jerojasro@336 179 \label{fig:tour-merge:kdiff3}
jerojasro@336 180 \end{figure}
jerojasro@336 181
jerojasro@336 182 For each conflicting portion of the file, we can choose to resolve
jerojasro@336 183 the conflict using some combination of text from the base version,
jerojasro@336 184 ours, or theirs. We can also manually edit the merged file at any
jerojasro@336 185 time, in case we need to make further modifications.
jerojasro@336 186
jerojasro@336 187 There are \emph{many} file merging tools available, too many to cover
jerojasro@336 188 here. They vary in which platforms they are available for, and in
jerojasro@336 189 their particular strengths and weaknesses. Most are tuned for merging
jerojasro@336 190 files containing plain text, while a few are aimed at specialised file
jerojasro@336 191 formats (generally XML).
jerojasro@336 192
jerojasro@336 193 \subsection{A worked example}
jerojasro@336 194
jerojasro@336 195 In this example, we will reproduce the file modification history of
jerojasro@336 196 figure~\ref{fig:tour-merge:conflict} above. Let's begin by creating a
jerojasro@336 197 repository with a base version of our document.
jerojasro@336 198 \interaction{tour-merge-conflict.wife}
jerojasro@336 199 We'll clone the repository and make a change to the file.
jerojasro@336 200 \interaction{tour-merge-conflict.cousin}
jerojasro@336 201 And another clone, to simulate someone else making a change to the
jerojasro@336 202 file. (This hints at the idea that it's not all that unusual to merge
jerojasro@336 203 with yourself when you isolate tasks in separate repositories, and
jerojasro@336 204 indeed to find and resolve conflicts while doing so.)
jerojasro@336 205 \interaction{tour-merge-conflict.son}
jerojasro@336 206 Having created two different versions of the file, we'll set up an
jerojasro@336 207 environment suitable for running our merge.
jerojasro@336 208 \interaction{tour-merge-conflict.pull}
jerojasro@336 209
jerojasro@336 210 In this example, I won't use Mercurial's normal \command{hgmerge}
jerojasro@336 211 program to do the merge, because it would drop my nice automated
jerojasro@336 212 example-running tool into a graphical user interface. Instead, I'll
jerojasro@336 213 set \envar{HGMERGE} to tell Mercurial to use the non-interactive
jerojasro@336 214 \command{merge} command. This is bundled with many Unix-like systems.
jerojasro@336 215 If you're following this example on your computer, don't bother
jerojasro@336 216 setting \envar{HGMERGE}.
jerojasro@336 217 \interaction{tour-merge-conflict.merge}
jerojasro@336 218 Because \command{merge} can't resolve the conflicting changes, it
jerojasro@336 219 leaves \emph{merge markers} inside the file that has conflicts,
jerojasro@336 220 indicating which lines have conflicts, and whether they came from our
jerojasro@336 221 version of the file or theirs.
jerojasro@336 222
jerojasro@336 223 Mercurial can tell from the way \command{merge} exits that it wasn't
jerojasro@336 224 able to merge successfully, so it tells us what commands we'll need to
jerojasro@336 225 run if we want to redo the merging operation. This could be useful
jerojasro@336 226 if, for example, we were running a graphical merge tool and quit
jerojasro@336 227 because we were confused or realised we had made a mistake.
jerojasro@336 228
jerojasro@336 229 If automatic or manual merges fail, there's nothing to prevent us from
jerojasro@336 230 ``fixing up'' the affected files ourselves, and committing the results
jerojasro@336 231 of our merge:
jerojasro@336 232 \interaction{tour-merge-conflict.commit}
jerojasro@336 233
jerojasro@336 234 \section{Simplifying the pull-merge-commit sequence}
jerojasro@336 235 \label{sec:tour-merge:fetch}
jerojasro@336 236
jerojasro@336 237 The process of merging changes as outlined above is straightforward,
jerojasro@336 238 but requires running three commands in sequence.
jerojasro@336 239 \begin{codesample2}
jerojasro@336 240 hg pull
jerojasro@336 241 hg merge
jerojasro@336 242 hg commit -m 'Merged remote changes'
jerojasro@336 243 \end{codesample2}
jerojasro@336 244 In the case of the final commit, you also need to enter a commit
jerojasro@336 245 message, which is almost always going to be a piece of uninteresting
jerojasro@336 246 ``boilerplate'' text.
jerojasro@336 247
jerojasro@336 248 It would be nice to reduce the number of steps needed, if this were
jerojasro@336 249 possible. Indeed, Mercurial is distributed with an extension called
jerojasro@336 250 \hgext{fetch} that does just this.
jerojasro@336 251
jerojasro@336 252 Mercurial provides a flexible extension mechanism that lets people
jerojasro@336 253 extend its functionality, while keeping the core of Mercurial small
jerojasro@336 254 and easy to deal with. Some extensions add new commands that you can
jerojasro@336 255 use from the command line, while others work ``behind the scenes,''
jerojasro@336 256 for example adding capabilities to the server.
jerojasro@336 257
jerojasro@336 258 The \hgext{fetch} extension adds a new command called, not
jerojasro@336 259 surprisingly, \hgcmd{fetch}. This extension acts as a combination of
jerojasro@336 260 \hgcmd{pull}, \hgcmd{update} and \hgcmd{merge}. It begins by pulling
jerojasro@336 261 changes from another repository into the current repository. If it
jerojasro@336 262 finds that the changes added a new head to the repository, it begins a
jerojasro@336 263 merge, then commits the result of the merge with an
jerojasro@336 264 automatically-generated commit message. If no new heads were added,
jerojasro@336 265 it updates the working directory to the new tip changeset.
jerojasro@336 266
jerojasro@336 267 Enabling the \hgext{fetch} extension is easy. Edit your
jerojasro@336 268 \sfilename{.hgrc}, and either go to the \rcsection{extensions} section
jerojasro@336 269 or create an \rcsection{extensions} section. Then add a line that
jerojasro@336 270 simply reads ``\Verb+fetch +''.
jerojasro@336 271 \begin{codesample2}
jerojasro@336 272 [extensions]
jerojasro@336 273 fetch =
jerojasro@336 274 \end{codesample2}
jerojasro@336 275 (Normally, on the right-hand side of the ``\texttt{=}'' would appear
jerojasro@336 276 the location of the extension, but since the \hgext{fetch} extension
jerojasro@336 277 is in the standard distribution, Mercurial knows where to search for
jerojasro@336 278 it.)
jerojasro@336 279
jerojasro@336 280 %%% Local Variables:
jerojasro@336 281 %%% mode: latex
jerojasro@336 282 %%% TeX-master: "00book"
jerojasro@336 283 %%% End: