rev |
line source |
bos@559
|
1 <!-- vim: set filetype=docbkxml shiftwidth=2 autoindent expandtab tw=77 : -->
|
bos@559
|
2
|
bos@559
|
3 <chapter id="chap:concepts">
|
bos@572
|
4 <?dbhtml filename="behind-the-scenes.html"?>
|
bos@559
|
5 <title>Behind the scenes</title>
|
bos@559
|
6
|
bos@620
|
7 <para id="x_2e8">Unlike many revision control systems, the concepts
|
bos@620
|
8 upon which Mercurial is built are simple enough that it's easy to
|
bos@620
|
9 understand how the software really works. Knowing these details
|
bos@620
|
10 certainly isn't necessary, so it is certainly safe to skip this
|
bos@620
|
11 chapter. However, I think you will get more out of the software
|
bos@620
|
12 with a <quote>mental model</quote> of what's going on.</para>
|
bos@620
|
13
|
bos@620
|
14 <para id="x_2e9">Being able to understand what's going on behind the
|
bos@620
|
15 scenes gives me confidence that Mercurial has been carefully
|
bos@620
|
16 designed to be both <emphasis>safe</emphasis> and
|
bos@559
|
17 <emphasis>efficient</emphasis>. And just as importantly, if it's
|
bos@559
|
18 easy for me to retain a good idea of what the software is doing
|
bos@559
|
19 when I perform a revision control task, I'm less likely to be
|
bos@559
|
20 surprised by its behaviour.</para>
|
bos@559
|
21
|
bos@584
|
22 <para id="x_2ea">In this chapter, we'll initially cover the core concepts
|
bos@559
|
23 behind Mercurial's design, then continue to discuss some of the
|
bos@559
|
24 interesting details of its implementation.</para>
|
bos@559
|
25
|
bos@559
|
26 <sect1>
|
bos@559
|
27 <title>Mercurial's historical record</title>
|
bos@559
|
28
|
bos@559
|
29 <sect2>
|
bos@559
|
30 <title>Tracking the history of a single file</title>
|
bos@559
|
31
|
bos@584
|
32 <para id="x_2eb">When Mercurial tracks modifications to a file, it stores
|
bos@559
|
33 the history of that file in a metadata object called a
|
bos@559
|
34 <emphasis>filelog</emphasis>. Each entry in the filelog
|
bos@559
|
35 contains enough information to reconstruct one revision of the
|
bos@559
|
36 file that is being tracked. Filelogs are stored as files in
|
bos@559
|
37 the <filename role="special"
|
bos@559
|
38 class="directory">.hg/store/data</filename> directory. A
|
bos@559
|
39 filelog contains two kinds of information: revision data, and
|
bos@559
|
40 an index to help Mercurial to find a revision
|
bos@559
|
41 efficiently.</para>
|
bos@559
|
42
|
bos@584
|
43 <para id="x_2ec">A file that is large, or has a lot of history, has its
|
bos@559
|
44 filelog stored in separate data
|
bos@559
|
45 (<quote><literal>.d</literal></quote> suffix) and index
|
bos@559
|
46 (<quote><literal>.i</literal></quote> suffix) files. For
|
bos@559
|
47 small files without much history, the revision data and index
|
bos@559
|
48 are combined in a single <quote><literal>.i</literal></quote>
|
bos@559
|
49 file. The correspondence between a file in the working
|
bos@559
|
50 directory and the filelog that tracks its history in the
|
bos@592
|
51 repository is illustrated in <xref
|
bos@559
|
52 linkend="fig:concepts:filelog"/>.</para>
|
bos@559
|
53
|
bos@591
|
54 <figure id="fig:concepts:filelog">
|
bos@591
|
55 <title>Relationships between files in working directory and
|
bos@591
|
56 filelogs in repository</title>
|
bos@591
|
57 <mediaobject>
|
bos@594
|
58 <imageobject><imagedata fileref="figs/filelog.png"/></imageobject>
|
bos@591
|
59 <textobject><phrase>XXX add text</phrase></textobject>
|
bos@591
|
60 </mediaobject>
|
bos@591
|
61 </figure>
|
bos@559
|
62
|
bos@559
|
63 </sect2>
|
bos@559
|
64 <sect2>
|
bos@559
|
65 <title>Managing tracked files</title>
|
bos@559
|
66
|
bos@584
|
67 <para id="x_2ee">Mercurial uses a structure called a
|
bos@559
|
68 <emphasis>manifest</emphasis> to collect together information
|
bos@559
|
69 about the files that it tracks. Each entry in the manifest
|
bos@559
|
70 contains information about the files present in a single
|
bos@559
|
71 changeset. An entry records which files are present in the
|
bos@559
|
72 changeset, the revision of each file, and a few other pieces
|
bos@559
|
73 of file metadata.</para>
|
bos@559
|
74
|
bos@559
|
75 </sect2>
|
bos@559
|
76 <sect2>
|
bos@559
|
77 <title>Recording changeset information</title>
|
bos@559
|
78
|
bos@584
|
79 <para id="x_2ef">The <emphasis>changelog</emphasis> contains information
|
bos@559
|
80 about each changeset. Each revision records who committed a
|
bos@559
|
81 change, the changeset comment, other pieces of
|
bos@559
|
82 changeset-related information, and the revision of the
|
bos@559
|
83 manifest to use.</para>
|
bos@559
|
84
|
bos@559
|
85 </sect2>
|
bos@559
|
86 <sect2>
|
bos@559
|
87 <title>Relationships between revisions</title>
|
bos@559
|
88
|
bos@584
|
89 <para id="x_2f0">Within a changelog, a manifest, or a filelog, each
|
bos@559
|
90 revision stores a pointer to its immediate parent (or to its
|
bos@559
|
91 two parents, if it's a merge revision). As I mentioned above,
|
bos@559
|
92 there are also relationships between revisions
|
bos@559
|
93 <emphasis>across</emphasis> these structures, and they are
|
bos@559
|
94 hierarchical in nature.</para>
|
bos@559
|
95
|
bos@584
|
96 <para id="x_2f1">For every changeset in a repository, there is exactly one
|
bos@559
|
97 revision stored in the changelog. Each revision of the
|
bos@559
|
98 changelog contains a pointer to a single revision of the
|
bos@559
|
99 manifest. A revision of the manifest stores a pointer to a
|
bos@559
|
100 single revision of each filelog tracked when that changeset
|
bos@592
|
101 was created. These relationships are illustrated in
|
bos@559
|
102 <xref linkend="fig:concepts:metadata"/>.</para>
|
bos@559
|
103
|
bos@591
|
104 <figure id="fig:concepts:metadata">
|
bos@591
|
105 <title>Metadata relationships</title>
|
bos@591
|
106 <mediaobject>
|
bos@594
|
107 <imageobject><imagedata fileref="figs/metadata.png"/></imageobject>
|
bos@591
|
108 <textobject><phrase>XXX add text</phrase></textobject>
|
bos@559
|
109 </mediaobject>
|
bos@591
|
110 </figure>
|
bos@559
|
111
|
bos@584
|
112 <para id="x_2f3">As the illustration shows, there is
|
bos@559
|
113 <emphasis>not</emphasis> a <quote>one to one</quote>
|
bos@559
|
114 relationship between revisions in the changelog, manifest, or
|
bos@559
|
115 filelog. If the manifest hasn't changed between two
|
bos@559
|
116 changesets, the changelog entries for those changesets will
|
bos@559
|
117 point to the same revision of the manifest. If a file that
|
bos@559
|
118 Mercurial tracks hasn't changed between two changesets, the
|
bos@559
|
119 entry for that file in the two revisions of the manifest will
|
bos@559
|
120 point to the same revision of its filelog.</para>
|
bos@559
|
121
|
bos@559
|
122 </sect2>
|
bos@559
|
123 </sect1>
|
bos@559
|
124 <sect1>
|
bos@559
|
125 <title>Safe, efficient storage</title>
|
bos@559
|
126
|
bos@584
|
127 <para id="x_2f4">The underpinnings of changelogs, manifests, and filelogs are
|
bos@559
|
128 provided by a single structure called the
|
bos@559
|
129 <emphasis>revlog</emphasis>.</para>
|
bos@559
|
130
|
bos@559
|
131 <sect2>
|
bos@559
|
132 <title>Efficient storage</title>
|
bos@559
|
133
|
bos@584
|
134 <para id="x_2f5">The revlog provides efficient storage of revisions using a
|
bos@559
|
135 <emphasis>delta</emphasis> mechanism. Instead of storing a
|
bos@559
|
136 complete copy of a file for each revision, it stores the
|
bos@559
|
137 changes needed to transform an older revision into the new
|
bos@559
|
138 revision. For many kinds of file data, these deltas are
|
bos@559
|
139 typically a fraction of a percent of the size of a full copy
|
bos@559
|
140 of a file.</para>
|
bos@559
|
141
|
bos@584
|
142 <para id="x_2f6">Some obsolete revision control systems can only work with
|
bos@559
|
143 deltas of text files. They must either store binary files as
|
bos@559
|
144 complete snapshots or encoded into a text representation, both
|
bos@559
|
145 of which are wasteful approaches. Mercurial can efficiently
|
bos@559
|
146 handle deltas of files with arbitrary binary contents; it
|
bos@559
|
147 doesn't need to treat text as special.</para>
|
bos@559
|
148
|
bos@559
|
149 </sect2>
|
bos@559
|
150 <sect2 id="sec:concepts:txn">
|
bos@559
|
151 <title>Safe operation</title>
|
bos@559
|
152
|
bos@584
|
153 <para id="x_2f7">Mercurial only ever <emphasis>appends</emphasis> data to
|
bos@559
|
154 the end of a revlog file. It never modifies a section of a
|
bos@559
|
155 file after it has written it. This is both more robust and
|
bos@559
|
156 efficient than schemes that need to modify or rewrite
|
bos@559
|
157 data.</para>
|
bos@559
|
158
|
bos@584
|
159 <para id="x_2f8">In addition, Mercurial treats every write as part of a
|
bos@559
|
160 <emphasis>transaction</emphasis> that can span a number of
|
bos@559
|
161 files. A transaction is <emphasis>atomic</emphasis>: either
|
bos@559
|
162 the entire transaction succeeds and its effects are all
|
bos@559
|
163 visible to readers in one go, or the whole thing is undone.
|
bos@559
|
164 This guarantee of atomicity means that if you're running two
|
bos@559
|
165 copies of Mercurial, where one is reading data and one is
|
bos@559
|
166 writing it, the reader will never see a partially written
|
bos@559
|
167 result that might confuse it.</para>
|
bos@559
|
168
|
bos@584
|
169 <para id="x_2f9">The fact that Mercurial only appends to files makes it
|
bos@559
|
170 easier to provide this transactional guarantee. The easier it
|
bos@559
|
171 is to do stuff like this, the more confident you should be
|
bos@559
|
172 that it's done correctly.</para>
|
bos@559
|
173
|
bos@559
|
174 </sect2>
|
bos@559
|
175 <sect2>
|
bos@559
|
176 <title>Fast retrieval</title>
|
bos@559
|
177
|
bos@584
|
178 <para id="x_2fa">Mercurial cleverly avoids a pitfall common to all earlier
|
bos@559
|
179 revision control systems: the problem of <emphasis>inefficient
|
bos@559
|
180 retrieval</emphasis>. Most revision control systems store
|
bos@559
|
181 the contents of a revision as an incremental series of
|
bos@559
|
182 modifications against a <quote>snapshot</quote>. To
|
bos@559
|
183 reconstruct a specific revision, you must first read the
|
bos@559
|
184 snapshot, and then every one of the revisions between the
|
bos@559
|
185 snapshot and your target revision. The more history that a
|
bos@559
|
186 file accumulates, the more revisions you must read, hence the
|
bos@559
|
187 longer it takes to reconstruct a particular revision.</para>
|
bos@559
|
188
|
bos@591
|
189 <figure id="fig:concepts:snapshot">
|
bos@591
|
190 <title>Snapshot of a revlog, with incremental deltas</title>
|
bos@591
|
191 <mediaobject>
|
bos@594
|
192 <imageobject><imagedata fileref="figs/snapshot.png"/></imageobject>
|
bos@591
|
193 <textobject><phrase>XXX add text</phrase></textobject>
|
bos@591
|
194 </mediaobject>
|
bos@591
|
195 </figure>
|
bos@559
|
196
|
bos@584
|
197 <para id="x_2fc">The innovation that Mercurial applies to this problem is
|
bos@559
|
198 simple but effective. Once the cumulative amount of delta
|
bos@559
|
199 information stored since the last snapshot exceeds a fixed
|
bos@559
|
200 threshold, it stores a new snapshot (compressed, of course),
|
bos@559
|
201 instead of another delta. This makes it possible to
|
bos@559
|
202 reconstruct <emphasis>any</emphasis> revision of a file
|
bos@559
|
203 quickly. This approach works so well that it has since been
|
bos@559
|
204 copied by several other revision control systems.</para>
|
bos@559
|
205
|
bos@592
|
206 <para id="x_2fd"><xref linkend="fig:concepts:snapshot"/> illustrates
|
bos@559
|
207 the idea. In an entry in a revlog's index file, Mercurial
|
bos@559
|
208 stores the range of entries from the data file that it must
|
bos@559
|
209 read to reconstruct a particular revision.</para>
|
bos@559
|
210
|
bos@559
|
211 <sect3>
|
bos@559
|
212 <title>Aside: the influence of video compression</title>
|
bos@559
|
213
|
bos@584
|
214 <para id="x_2fe">If you're familiar with video compression or have ever
|
bos@559
|
215 watched a TV feed through a digital cable or satellite
|
bos@559
|
216 service, you may know that most video compression schemes
|
bos@559
|
217 store each frame of video as a delta against its predecessor
|
bos@559
|
218 frame. In addition, these schemes use <quote>lossy</quote>
|
bos@559
|
219 compression techniques to increase the compression ratio, so
|
bos@559
|
220 visual errors accumulate over the course of a number of
|
bos@559
|
221 inter-frame deltas.</para>
|
bos@559
|
222
|
bos@584
|
223 <para id="x_2ff">Because it's possible for a video stream to <quote>drop
|
bos@559
|
224 out</quote> occasionally due to signal glitches, and to
|
bos@559
|
225 limit the accumulation of artefacts introduced by the lossy
|
bos@559
|
226 compression process, video encoders periodically insert a
|
bos@559
|
227 complete frame (called a <quote>key frame</quote>) into the
|
bos@559
|
228 video stream; the next delta is generated against that
|
bos@559
|
229 frame. This means that if the video signal gets
|
bos@559
|
230 interrupted, it will resume once the next key frame is
|
bos@559
|
231 received. Also, the accumulation of encoding errors
|
bos@559
|
232 restarts anew with each key frame.</para>
|
bos@559
|
233
|
bos@559
|
234 </sect3>
|
bos@559
|
235 </sect2>
|
bos@559
|
236 <sect2>
|
bos@559
|
237 <title>Identification and strong integrity</title>
|
bos@559
|
238
|
bos@584
|
239 <para id="x_300">Along with delta or snapshot information, a revlog entry
|
bos@559
|
240 contains a cryptographic hash of the data that it represents.
|
bos@559
|
241 This makes it difficult to forge the contents of a revision,
|
bos@559
|
242 and easy to detect accidental corruption.</para>
|
bos@559
|
243
|
bos@584
|
244 <para id="x_301">Hashes provide more than a mere check against corruption;
|
bos@559
|
245 they are used as the identifiers for revisions. The changeset
|
bos@559
|
246 identification hashes that you see as an end user are from
|
bos@559
|
247 revisions of the changelog. Although filelogs and the
|
bos@559
|
248 manifest also use hashes, Mercurial only uses these behind the
|
bos@559
|
249 scenes.</para>
|
bos@559
|
250
|
bos@584
|
251 <para id="x_302">Mercurial verifies that hashes are correct when it
|
bos@559
|
252 retrieves file revisions and when it pulls changes from
|
bos@559
|
253 another repository. If it encounters an integrity problem, it
|
bos@559
|
254 will complain and stop whatever it's doing.</para>
|
bos@559
|
255
|
bos@584
|
256 <para id="x_303">In addition to the effect it has on retrieval efficiency,
|
bos@559
|
257 Mercurial's use of periodic snapshots makes it more robust
|
bos@559
|
258 against partial data corruption. If a revlog becomes partly
|
bos@559
|
259 corrupted due to a hardware error or system bug, it's often
|
bos@559
|
260 possible to reconstruct some or most revisions from the
|
bos@559
|
261 uncorrupted sections of the revlog, both before and after the
|
bos@559
|
262 corrupted section. This would not be possible with a
|
bos@559
|
263 delta-only storage model.</para>
|
bos@559
|
264
|
bos@559
|
265 </sect2>
|
bos@559
|
266 </sect1>
|
bos@559
|
267 <sect1>
|
bos@559
|
268 <title>Revision history, branching, and merging</title>
|
bos@559
|
269
|
bos@584
|
270 <para id="x_304">Every entry in a Mercurial revlog knows the identity of its
|
bos@559
|
271 immediate ancestor revision, usually referred to as its
|
bos@559
|
272 <emphasis>parent</emphasis>. In fact, a revision contains room
|
bos@559
|
273 for not one parent, but two. Mercurial uses a special hash,
|
bos@559
|
274 called the <quote>null ID</quote>, to represent the idea
|
bos@559
|
275 <quote>there is no parent here</quote>. This hash is simply a
|
bos@559
|
276 string of zeroes.</para>
|
bos@559
|
277
|
bos@592
|
278 <para id="x_305">In <xref linkend="fig:concepts:revlog"/>, you can see
|
bos@559
|
279 an example of the conceptual structure of a revlog. Filelogs,
|
bos@559
|
280 manifests, and changelogs all have this same structure; they
|
bos@559
|
281 differ only in the kind of data stored in each delta or
|
bos@559
|
282 snapshot.</para>
|
bos@559
|
283
|
bos@584
|
284 <para id="x_306">The first revision in a revlog (at the bottom of the image)
|
bos@559
|
285 has the null ID in both of its parent slots. For a
|
bos@559
|
286 <quote>normal</quote> revision, its first parent slot contains
|
bos@559
|
287 the ID of its parent revision, and its second contains the null
|
bos@559
|
288 ID, indicating that the revision has only one real parent. Any
|
bos@559
|
289 two revisions that have the same parent ID are branches. A
|
bos@559
|
290 revision that represents a merge between branches has two normal
|
bos@559
|
291 revision IDs in its parent slots.</para>
|
bos@559
|
292
|
bos@591
|
293 <figure id="fig:concepts:revlog">
|
bos@591
|
294 <title>The conceptual structure of a revlog</title>
|
bos@591
|
295 <mediaobject>
|
bos@594
|
296 <imageobject><imagedata fileref="figs/revlog.png"/></imageobject>
|
bos@591
|
297 <textobject><phrase>XXX add text</phrase></textobject>
|
bos@591
|
298 </mediaobject>
|
bos@591
|
299 </figure>
|
bos@559
|
300
|
bos@559
|
301 </sect1>
|
bos@559
|
302 <sect1>
|
bos@559
|
303 <title>The working directory</title>
|
bos@559
|
304
|
bos@584
|
305 <para id="x_307">In the working directory, Mercurial stores a snapshot of the
|
bos@559
|
306 files from the repository as of a particular changeset.</para>
|
bos@559
|
307
|
bos@584
|
308 <para id="x_308">The working directory <quote>knows</quote> which changeset
|
bos@559
|
309 it contains. When you update the working directory to contain a
|
bos@559
|
310 particular changeset, Mercurial looks up the appropriate
|
bos@559
|
311 revision of the manifest to find out which files it was tracking
|
bos@559
|
312 at the time that changeset was committed, and which revision of
|
bos@559
|
313 each file was then current. It then recreates a copy of each of
|
bos@559
|
314 those files, with the same contents it had when the changeset
|
bos@559
|
315 was committed.</para>
|
bos@559
|
316
|
bos@584
|
317 <para id="x_309">The <emphasis>dirstate</emphasis> contains Mercurial's
|
bos@559
|
318 knowledge of the working directory. This details which
|
bos@559
|
319 changeset the working directory is updated to, and all of the
|
bos@559
|
320 files that Mercurial is tracking in the working
|
bos@559
|
321 directory.</para>
|
bos@559
|
322
|
bos@584
|
323 <para id="x_30a">Just as a revision of a revlog has room for two parents, so
|
bos@559
|
324 that it can represent either a normal revision (with one parent)
|
bos@559
|
325 or a merge of two earlier revisions, the dirstate has slots for
|
bos@559
|
326 two parents. When you use the <command role="hg-cmd">hg
|
bos@559
|
327 update</command> command, the changeset that you update to is
|
bos@559
|
328 stored in the <quote>first parent</quote> slot, and the null ID
|
bos@559
|
329 in the second. When you <command role="hg-cmd">hg
|
bos@559
|
330 merge</command> with another changeset, the first parent
|
bos@559
|
331 remains unchanged, and the second parent is filled in with the
|
bos@559
|
332 changeset you're merging with. The <command role="hg-cmd">hg
|
bos@559
|
333 parents</command> command tells you what the parents of the
|
bos@559
|
334 dirstate are.</para>
|
bos@559
|
335
|
bos@559
|
336 <sect2>
|
bos@559
|
337 <title>What happens when you commit</title>
|
bos@559
|
338
|
bos@584
|
339 <para id="x_30b">The dirstate stores parent information for more than just
|
bos@559
|
340 book-keeping purposes. Mercurial uses the parents of the
|
bos@559
|
341 dirstate as <emphasis>the parents of a new
|
bos@559
|
342 changeset</emphasis> when you perform a commit.</para>
|
bos@559
|
343
|
bos@591
|
344 <figure id="fig:concepts:wdir">
|
bos@591
|
345 <title>The working directory can have two parents</title>
|
bos@591
|
346 <mediaobject>
|
bos@594
|
347 <imageobject><imagedata fileref="figs/wdir.png"/></imageobject>
|
bos@591
|
348 <textobject><phrase>XXX add text</phrase></textobject>
|
bos@591
|
349 </mediaobject>
|
bos@591
|
350 </figure>
|
bos@559
|
351
|
bos@592
|
352 <para id="x_30d"><xref linkend="fig:concepts:wdir"/> shows the
|
bos@559
|
353 normal state of the working directory, where it has a single
|
bos@559
|
354 changeset as parent. That changeset is the
|
bos@559
|
355 <emphasis>tip</emphasis>, the newest changeset in the
|
bos@559
|
356 repository that has no children.</para>
|
bos@559
|
357
|
bos@591
|
358 <figure id="fig:concepts:wdir-after-commit">
|
bos@591
|
359 <title>The working directory gains new parents after a
|
bos@591
|
360 commit</title>
|
bos@591
|
361 <mediaobject>
|
bos@594
|
362 <imageobject><imagedata fileref="figs/wdir-after-commit.png"/></imageobject>
|
bos@591
|
363 <textobject><phrase>XXX add text</phrase></textobject>
|
bos@591
|
364 </mediaobject>
|
bos@591
|
365 </figure>
|
bos@559
|
366
|
bos@584
|
367 <para id="x_30f">It's useful to think of the working directory as
|
bos@559
|
368 <quote>the changeset I'm about to commit</quote>. Any files
|
bos@559
|
369 that you tell Mercurial that you've added, removed, renamed,
|
bos@559
|
370 or copied will be reflected in that changeset, as will
|
bos@559
|
371 modifications to any files that Mercurial is already tracking;
|
bos@559
|
372 the new changeset will have the parents of the working
|
bos@559
|
373 directory as its parents.</para>
|
bos@559
|
374
|
bos@592
|
375 <para id="x_310">After a commit, Mercurial will update the
|
bos@592
|
376 parents of the working directory, so that the first parent is
|
bos@592
|
377 the ID of the new changeset, and the second is the null ID.
|
bos@592
|
378 This is shown in <xref
|
bos@592
|
379 linkend="fig:concepts:wdir-after-commit"/>. Mercurial
|
bos@559
|
380 doesn't touch any of the files in the working directory when
|
bos@559
|
381 you commit; it just modifies the dirstate to note its new
|
bos@559
|
382 parents.</para>
|
bos@559
|
383
|
bos@559
|
384 </sect2>
|
bos@559
|
385 <sect2>
|
bos@559
|
386 <title>Creating a new head</title>
|
bos@559
|
387
|
bos@584
|
388 <para id="x_311">It's perfectly normal to update the working directory to a
|
bos@559
|
389 changeset other than the current tip. For example, you might
|
bos@559
|
390 want to know what your project looked like last Tuesday, or
|
bos@559
|
391 you could be looking through changesets to see which one
|
bos@559
|
392 introduced a bug. In cases like this, the natural thing to do
|
bos@559
|
393 is update the working directory to the changeset you're
|
bos@559
|
394 interested in, and then examine the files in the working
|
bos@559
|
395 directory directly to see their contents as they were when you
|
bos@559
|
396 committed that changeset. The effect of this is shown in
|
bos@592
|
397 <xref linkend="fig:concepts:wdir-pre-branch"/>.</para>
|
bos@559
|
398
|
bos@591
|
399 <figure id="fig:concepts:wdir-pre-branch">
|
bos@591
|
400 <title>The working directory, updated to an older
|
bos@591
|
401 changeset</title>
|
bos@591
|
402 <mediaobject>
|
bos@594
|
403 <imageobject><imagedata fileref="figs/wdir-pre-branch.png"/></imageobject>
|
bos@591
|
404 <textobject><phrase>XXX add text</phrase></textobject>
|
bos@591
|
405 </mediaobject>
|
bos@591
|
406 </figure>
|
bos@559
|
407
|
bos@592
|
408 <para id="x_313">Having updated the working directory to an
|
bos@592
|
409 older changeset, what happens if you make some changes, and
|
bos@592
|
410 then commit? Mercurial behaves in the same way as I outlined
|
bos@559
|
411 above. The parents of the working directory become the
|
bos@559
|
412 parents of the new changeset. This new changeset has no
|
bos@559
|
413 children, so it becomes the new tip. And the repository now
|
bos@559
|
414 contains two changesets that have no children; we call these
|
bos@559
|
415 <emphasis>heads</emphasis>. You can see the structure that
|
bos@592
|
416 this creates in <xref
|
bos@559
|
417 linkend="fig:concepts:wdir-branch"/>.</para>
|
bos@559
|
418
|
bos@591
|
419 <figure id="fig:concepts:wdir-branch">
|
bos@591
|
420 <title>After a commit made while synced to an older
|
bos@591
|
421 changeset</title>
|
bos@591
|
422 <mediaobject>
|
bos@594
|
423 <imageobject><imagedata fileref="figs/wdir-branch.png"/></imageobject>
|
bos@591
|
424 <textobject><phrase>XXX add text</phrase></textobject>
|
bos@591
|
425 </mediaobject>
|
bos@591
|
426 </figure>
|
bos@559
|
427
|
bos@559
|
428 <note>
|
bos@584
|
429 <para id="x_315"> If you're new to Mercurial, you should keep in mind a
|
bos@559
|
430 common <quote>error</quote>, which is to use the <command
|
bos@559
|
431 role="hg-cmd">hg pull</command> command without any
|
bos@559
|
432 options. By default, the <command role="hg-cmd">hg
|
bos@559
|
433 pull</command> command <emphasis>does not</emphasis>
|
bos@559
|
434 update the working directory, so you'll bring new changesets
|
bos@559
|
435 into your repository, but the working directory will stay
|
bos@559
|
436 synced at the same changeset as before the pull. If you
|
bos@559
|
437 make some changes and commit afterwards, you'll thus create
|
bos@559
|
438 a new head, because your working directory isn't synced to
|
bos@559
|
439 whatever the current tip is.</para>
|
bos@559
|
440
|
bos@620
|
441 <para id="x_316"> I put the word <quote>error</quote> in
|
bos@620
|
442 quotes because all that you need to do to rectify this
|
bos@620
|
443 situation is <command role="hg-cmd">hg merge</command>, then
|
bos@620
|
444 <command role="hg-cmd">hg commit</command>. In other words,
|
bos@620
|
445 this almost never has negative consequences; it's just
|
bos@620
|
446 something of a surprise for newcomers. I'll discuss other
|
bos@620
|
447 ways to avoid this behaviour, and why Mercurial behaves in
|
bos@620
|
448 this initially surprising way, later on.</para>
|
bos@559
|
449 </note>
|
bos@559
|
450
|
bos@559
|
451 </sect2>
|
bos@559
|
452 <sect2>
|
bos@620
|
453 <title>Merging changes</title>
|
bos@559
|
454
|
bos@592
|
455 <para id="x_317">When you run the <command role="hg-cmd">hg
|
bos@592
|
456 merge</command> command, Mercurial leaves the first parent
|
bos@592
|
457 of the working directory unchanged, and sets the second parent
|
bos@592
|
458 to the changeset you're merging with, as shown in <xref
|
bos@559
|
459 linkend="fig:concepts:wdir-merge"/>.</para>
|
bos@559
|
460
|
bos@591
|
461 <figure id="fig:concepts:wdir-merge">
|
bos@591
|
462 <title>Merging two heads</title>
|
bos@591
|
463 <mediaobject>
|
bos@591
|
464 <imageobject>
|
bos@594
|
465 <imagedata fileref="figs/wdir-merge.png"/>
|
bos@591
|
466 </imageobject>
|
bos@591
|
467 <textobject><phrase>XXX add text</phrase></textobject>
|
bos@591
|
468 </mediaobject>
|
bos@591
|
469 </figure>
|
bos@559
|
470
|
bos@584
|
471 <para id="x_319">Mercurial also has to modify the working directory, to
|
bos@559
|
472 merge the files managed in the two changesets. Simplified a
|
bos@559
|
473 little, the merging process goes like this, for every file in
|
bos@559
|
474 the manifests of both changesets.</para>
|
bos@559
|
475 <itemizedlist>
|
bos@584
|
476 <listitem><para id="x_31a">If neither changeset has modified a file, do
|
bos@559
|
477 nothing with that file.</para>
|
bos@559
|
478 </listitem>
|
bos@584
|
479 <listitem><para id="x_31b">If one changeset has modified a file, and the
|
bos@559
|
480 other hasn't, create the modified copy of the file in the
|
bos@559
|
481 working directory.</para>
|
bos@559
|
482 </listitem>
|
bos@584
|
483 <listitem><para id="x_31c">If one changeset has removed a file, and the
|
bos@559
|
484 other hasn't (or has also deleted it), delete the file
|
bos@559
|
485 from the working directory.</para>
|
bos@559
|
486 </listitem>
|
bos@584
|
487 <listitem><para id="x_31d">If one changeset has removed a file, but the
|
bos@559
|
488 other has modified the file, ask the user what to do: keep
|
bos@559
|
489 the modified file, or remove it?</para>
|
bos@559
|
490 </listitem>
|
bos@584
|
491 <listitem><para id="x_31e">If both changesets have modified a file,
|
bos@559
|
492 invoke an external merge program to choose the new
|
bos@559
|
493 contents for the merged file. This may require input from
|
bos@559
|
494 the user.</para>
|
bos@559
|
495 </listitem>
|
bos@584
|
496 <listitem><para id="x_31f">If one changeset has modified a file, and the
|
bos@559
|
497 other has renamed or copied the file, make sure that the
|
bos@559
|
498 changes follow the new name of the file.</para>
|
bos@559
|
499 </listitem></itemizedlist>
|
bos@584
|
500 <para id="x_320">There are more details&emdash;merging has plenty of corner
|
bos@559
|
501 cases&emdash;but these are the most common choices that are
|
bos@559
|
502 involved in a merge. As you can see, most cases are
|
bos@559
|
503 completely automatic, and indeed most merges finish
|
bos@559
|
504 automatically, without requiring your input to resolve any
|
bos@559
|
505 conflicts.</para>
|
bos@559
|
506
|
bos@584
|
507 <para id="x_321">When you're thinking about what happens when you commit
|
bos@559
|
508 after a merge, once again the working directory is <quote>the
|
bos@559
|
509 changeset I'm about to commit</quote>. After the <command
|
bos@559
|
510 role="hg-cmd">hg merge</command> command completes, the
|
bos@559
|
511 working directory has two parents; these will become the
|
bos@559
|
512 parents of the new changeset.</para>
|
bos@559
|
513
|
bos@584
|
514 <para id="x_322">Mercurial lets you perform multiple merges, but you must
|
bos@559
|
515 commit the results of each individual merge as you go. This
|
bos@559
|
516 is necessary because Mercurial only tracks two parents for
|
bos@559
|
517 both revisions and the working directory. While it would be
|
bos@559
|
518 technically possible to merge multiple changesets at once, the
|
bos@559
|
519 prospect of user confusion and making a terrible mess of a
|
bos@559
|
520 merge immediately becomes overwhelming.</para>
|
bos@559
|
521
|
bos@559
|
522 </sect2>
|
bos@620
|
523
|
bos@620
|
524 <sect2>
|
bos@620
|
525 <title>Merging and renames</title>
|
bos@620
|
526
|
bos@620
|
527 <para>A surprising number of revision control systems pay little
|
bos@620
|
528 or no attention to a file's <emphasis>name</emphasis> over
|
bos@620
|
529 time. For instance, it used to be common that if a file got
|
bos@620
|
530 renamed on one side of a merge, the changes from the other
|
bos@620
|
531 side would be silently dropped.</para>
|
bos@620
|
532
|
bos@620
|
533 <para>Mercurial records metadata when you tell it to perform a
|
bos@620
|
534 rename or copy. It uses this metadata during a merge to do the
|
bos@620
|
535 right thing in the case of a merge. For instance, if I rename
|
bos@620
|
536 a file, and you edit it without renaming it, when we merge our
|
bos@620
|
537 work the file will be renamed and have your edits
|
bos@620
|
538 applied.</para>
|
bos@620
|
539 </sect2>
|
bos@559
|
540 </sect1>
|
bos@620
|
541
|
bos@559
|
542 <sect1>
|
bos@559
|
543 <title>Other interesting design features</title>
|
bos@559
|
544
|
bos@584
|
545 <para id="x_323">In the sections above, I've tried to highlight some of the
|
bos@559
|
546 most important aspects of Mercurial's design, to illustrate that
|
bos@559
|
547 it pays careful attention to reliability and performance.
|
bos@559
|
548 However, the attention to detail doesn't stop there. There are
|
bos@559
|
549 a number of other aspects of Mercurial's construction that I
|
bos@559
|
550 personally find interesting. I'll detail a few of them here,
|
bos@559
|
551 separate from the <quote>big ticket</quote> items above, so that
|
bos@559
|
552 if you're interested, you can gain a better idea of the amount
|
bos@559
|
553 of thinking that goes into a well-designed system.</para>
|
bos@559
|
554
|
bos@559
|
555 <sect2>
|
bos@559
|
556 <title>Clever compression</title>
|
bos@559
|
557
|
bos@584
|
558 <para id="x_324">When appropriate, Mercurial will store both snapshots and
|
bos@559
|
559 deltas in compressed form. It does this by always
|
bos@559
|
560 <emphasis>trying to</emphasis> compress a snapshot or delta,
|
bos@559
|
561 but only storing the compressed version if it's smaller than
|
bos@559
|
562 the uncompressed version.</para>
|
bos@559
|
563
|
bos@584
|
564 <para id="x_325">This means that Mercurial does <quote>the right
|
bos@559
|
565 thing</quote> when storing a file whose native form is
|
bos@559
|
566 compressed, such as a <literal>zip</literal> archive or a JPEG
|
bos@559
|
567 image. When these types of files are compressed a second
|
bos@559
|
568 time, the resulting file is usually bigger than the
|
bos@559
|
569 once-compressed form, and so Mercurial will store the plain
|
bos@559
|
570 <literal>zip</literal> or JPEG.</para>
|
bos@559
|
571
|
bos@584
|
572 <para id="x_326">Deltas between revisions of a compressed file are usually
|
bos@559
|
573 larger than snapshots of the file, and Mercurial again does
|
bos@559
|
574 <quote>the right thing</quote> in these cases. It finds that
|
bos@559
|
575 such a delta exceeds the threshold at which it should store a
|
bos@559
|
576 complete snapshot of the file, so it stores the snapshot,
|
bos@559
|
577 again saving space compared to a naive delta-only
|
bos@559
|
578 approach.</para>
|
bos@559
|
579
|
bos@559
|
580 <sect3>
|
bos@559
|
581 <title>Network recompression</title>
|
bos@559
|
582
|
bos@584
|
583 <para id="x_327">When storing revisions on disk, Mercurial uses the
|
bos@559
|
584 <quote>deflate</quote> compression algorithm (the same one
|
bos@559
|
585 used by the popular <literal>zip</literal> archive format),
|
bos@559
|
586 which balances good speed with a respectable compression
|
bos@559
|
587 ratio. However, when transmitting revision data over a
|
bos@559
|
588 network connection, Mercurial uncompresses the compressed
|
bos@559
|
589 revision data.</para>
|
bos@559
|
590
|
bos@584
|
591 <para id="x_328">If the connection is over HTTP, Mercurial recompresses
|
bos@559
|
592 the entire stream of data using a compression algorithm that
|
bos@559
|
593 gives a better compression ratio (the Burrows-Wheeler
|
bos@559
|
594 algorithm from the widely used <literal>bzip2</literal>
|
bos@559
|
595 compression package). This combination of algorithm and
|
bos@559
|
596 compression of the entire stream (instead of a revision at a
|
bos@559
|
597 time) substantially reduces the number of bytes to be
|
bos@620
|
598 transferred, yielding better network performance over most
|
bos@620
|
599 kinds of network.</para>
|
bos@559
|
600
|
bos@584
|
601 <para id="x_329">(If the connection is over <command>ssh</command>,
|
bos@559
|
602 Mercurial <emphasis>doesn't</emphasis> recompress the
|
bos@559
|
603 stream, because <command>ssh</command> can already do this
|
bos@559
|
604 itself.)</para>
|
bos@559
|
605
|
bos@559
|
606 </sect3>
|
bos@559
|
607 </sect2>
|
bos@559
|
608 <sect2>
|
bos@559
|
609 <title>Read/write ordering and atomicity</title>
|
bos@559
|
610
|
bos@592
|
611 <para id="x_32a">Appending to files isn't the whole story when
|
bos@592
|
612 it comes to guaranteeing that a reader won't see a partial
|
bos@592
|
613 write. If you recall <xref linkend="fig:concepts:metadata"/>,
|
bos@592
|
614 revisions in
|
bos@592
|
615 the changelog point to revisions in the manifest, and
|
bos@592
|
616 revisions in the manifest point to revisions in filelogs.
|
bos@592
|
617 This hierarchy is deliberate.</para>
|
bos@559
|
618
|
bos@584
|
619 <para id="x_32b">A writer starts a transaction by writing filelog and
|
bos@559
|
620 manifest data, and doesn't write any changelog data until
|
bos@559
|
621 those are finished. A reader starts by reading changelog
|
bos@559
|
622 data, then manifest data, followed by filelog data.</para>
|
bos@559
|
623
|
bos@584
|
624 <para id="x_32c">Since the writer has always finished writing filelog and
|
bos@559
|
625 manifest data before it writes to the changelog, a reader will
|
bos@559
|
626 never read a pointer to a partially written manifest revision
|
bos@559
|
627 from the changelog, and it will never read a pointer to a
|
bos@559
|
628 partially written filelog revision from the manifest.</para>
|
bos@559
|
629
|
bos@559
|
630 </sect2>
|
bos@559
|
631 <sect2>
|
bos@559
|
632 <title>Concurrent access</title>
|
bos@559
|
633
|
bos@584
|
634 <para id="x_32d">The read/write ordering and atomicity guarantees mean that
|
bos@559
|
635 Mercurial never needs to <emphasis>lock</emphasis> a
|
bos@559
|
636 repository when it's reading data, even if the repository is
|
bos@559
|
637 being written to while the read is occurring. This has a big
|
bos@559
|
638 effect on scalability; you can have an arbitrary number of
|
bos@559
|
639 Mercurial processes safely reading data from a repository
|
bos@559
|
640 safely all at once, no matter whether it's being written to or
|
bos@559
|
641 not.</para>
|
bos@559
|
642
|
bos@584
|
643 <para id="x_32e">The lockless nature of reading means that if you're
|
bos@559
|
644 sharing a repository on a multi-user system, you don't need to
|
bos@559
|
645 grant other local users permission to
|
bos@559
|
646 <emphasis>write</emphasis> to your repository in order for
|
bos@559
|
647 them to be able to clone it or pull changes from it; they only
|
bos@559
|
648 need <emphasis>read</emphasis> permission. (This is
|
bos@559
|
649 <emphasis>not</emphasis> a common feature among revision
|
bos@559
|
650 control systems, so don't take it for granted! Most require
|
bos@559
|
651 readers to be able to lock a repository to access it safely,
|
bos@559
|
652 and this requires write permission on at least one directory,
|
bos@559
|
653 which of course makes for all kinds of nasty and annoying
|
bos@559
|
654 security and administrative problems.)</para>
|
bos@559
|
655
|
bos@584
|
656 <para id="x_32f">Mercurial uses locks to ensure that only one process can
|
bos@559
|
657 write to a repository at a time (the locking mechanism is safe
|
bos@559
|
658 even over filesystems that are notoriously hostile to locking,
|
bos@559
|
659 such as NFS). If a repository is locked, a writer will wait
|
bos@559
|
660 for a while to retry if the repository becomes unlocked, but
|
bos@559
|
661 if the repository remains locked for too long, the process
|
bos@559
|
662 attempting to write will time out after a while. This means
|
bos@559
|
663 that your daily automated scripts won't get stuck forever and
|
bos@559
|
664 pile up if a system crashes unnoticed, for example. (Yes, the
|
bos@559
|
665 timeout is configurable, from zero to infinity.)</para>
|
bos@559
|
666
|
bos@559
|
667 <sect3>
|
bos@559
|
668 <title>Safe dirstate access</title>
|
bos@559
|
669
|
bos@584
|
670 <para id="x_330">As with revision data, Mercurial doesn't take a lock to
|
bos@559
|
671 read the dirstate file; it does acquire a lock to write it.
|
bos@559
|
672 To avoid the possibility of reading a partially written copy
|
bos@559
|
673 of the dirstate file, Mercurial writes to a file with a
|
bos@559
|
674 unique name in the same directory as the dirstate file, then
|
bos@559
|
675 renames the temporary file atomically to
|
bos@559
|
676 <filename>dirstate</filename>. The file named
|
bos@559
|
677 <filename>dirstate</filename> is thus guaranteed to be
|
bos@559
|
678 complete, not partially written.</para>
|
bos@559
|
679
|
bos@559
|
680 </sect3>
|
bos@559
|
681 </sect2>
|
bos@559
|
682 <sect2>
|
bos@559
|
683 <title>Avoiding seeks</title>
|
bos@559
|
684
|
bos@584
|
685 <para id="x_331">Critical to Mercurial's performance is the avoidance of
|
bos@559
|
686 seeks of the disk head, since any seek is far more expensive
|
bos@559
|
687 than even a comparatively large read operation.</para>
|
bos@559
|
688
|
bos@584
|
689 <para id="x_332">This is why, for example, the dirstate is stored in a
|
bos@559
|
690 single file. If there were a dirstate file per directory that
|
bos@559
|
691 Mercurial tracked, the disk would seek once per directory.
|
bos@559
|
692 Instead, Mercurial reads the entire single dirstate file in
|
bos@559
|
693 one step.</para>
|
bos@559
|
694
|
bos@584
|
695 <para id="x_333">Mercurial also uses a <quote>copy on write</quote> scheme
|
bos@559
|
696 when cloning a repository on local storage. Instead of
|
bos@559
|
697 copying every revlog file from the old repository into the new
|
bos@559
|
698 repository, it makes a <quote>hard link</quote>, which is a
|
bos@559
|
699 shorthand way to say <quote>these two names point to the same
|
bos@559
|
700 file</quote>. When Mercurial is about to write to one of a
|
bos@559
|
701 revlog's files, it checks to see if the number of names
|
bos@559
|
702 pointing at the file is greater than one. If it is, more than
|
bos@559
|
703 one repository is using the file, so Mercurial makes a new
|
bos@559
|
704 copy of the file that is private to this repository.</para>
|
bos@559
|
705
|
bos@584
|
706 <para id="x_334">A few revision control developers have pointed out that
|
bos@559
|
707 this idea of making a complete private copy of a file is not
|
bos@559
|
708 very efficient in its use of storage. While this is true,
|
bos@559
|
709 storage is cheap, and this method gives the highest
|
bos@559
|
710 performance while deferring most book-keeping to the operating
|
bos@559
|
711 system. An alternative scheme would most likely reduce
|
bos@559
|
712 performance and increase the complexity of the software, each
|
bos@559
|
713 of which is much more important to the <quote>feel</quote> of
|
bos@559
|
714 day-to-day use.</para>
|
bos@559
|
715
|
bos@559
|
716 </sect2>
|
bos@559
|
717 <sect2>
|
bos@559
|
718 <title>Other contents of the dirstate</title>
|
bos@559
|
719
|
bos@584
|
720 <para id="x_335">Because Mercurial doesn't force you to tell it when you're
|
bos@559
|
721 modifying a file, it uses the dirstate to store some extra
|
bos@559
|
722 information so it can determine efficiently whether you have
|
bos@559
|
723 modified a file. For each file in the working directory, it
|
bos@559
|
724 stores the time that it last modified the file itself, and the
|
bos@559
|
725 size of the file at that time.</para>
|
bos@559
|
726
|
bos@584
|
727 <para id="x_336">When you explicitly <command role="hg-cmd">hg
|
bos@559
|
728 add</command>, <command role="hg-cmd">hg remove</command>,
|
bos@559
|
729 <command role="hg-cmd">hg rename</command> or <command
|
bos@559
|
730 role="hg-cmd">hg copy</command> files, Mercurial updates the
|
bos@559
|
731 dirstate so that it knows what to do with those files when you
|
bos@559
|
732 commit.</para>
|
bos@559
|
733
|
bos@584
|
734 <para id="x_337">When Mercurial is checking the states of files in the
|
bos@559
|
735 working directory, it first checks a file's modification time.
|
bos@559
|
736 If that has not changed, the file must not have been modified.
|
bos@559
|
737 If the file's size has changed, the file must have been
|
bos@559
|
738 modified. If the modification time has changed, but the size
|
bos@559
|
739 has not, only then does Mercurial need to read the actual
|
bos@559
|
740 contents of the file to see if they've changed. Storing these
|
bos@559
|
741 few extra pieces of information dramatically reduces the
|
bos@559
|
742 amount of data that Mercurial needs to read, which yields
|
bos@559
|
743 large performance improvements compared to other revision
|
bos@559
|
744 control systems.</para>
|
bos@559
|
745
|
bos@559
|
746 </sect2>
|
bos@559
|
747 </sect1>
|
bos@559
|
748 </chapter>
|
bos@559
|
749
|
bos@559
|
750 <!--
|
bos@559
|
751 local variables:
|
bos@559
|
752 sgml-parent-document: ("00book.xml" "book" "chapter")
|
bos@559
|
753 end:
|
bos@559
|
754 -->
|