t/t5616-partial-clone.sh

   1 #!/bin/sh
   2
   3 test_description='git partial clone'
   4
   5 . ./test-lib.sh
   6
   7 # create a normal "src" repo where we can later create new commits.
   8 # expect_1.oids will contain a list of the OIDs of all blobs.
   9 test_expect_success 'setup normal src repo' '
  10         echo "{print \$1}" >print_1.awk &&
  11         echo "{print \$2}" >print_2.awk &&
  12
  13         git init src &&
  14         for n in 1 2 3 4
  15         do
  16                 echo "This is file: $n" > src/file.$n.txt
  17                 git -C src add file.$n.txt
  18                 git -C src commit -m "file $n"
  19                 git -C src ls-files -s file.$n.txt >>temp
  20         done &&
  21         awk -f print_2.awk <temp | sort >expect_1.oids &&
  22         test_line_count = 4 expect_1.oids
  23 '
  24
  25 # bare clone "src" giving "srv.bare" for use as our server.
  26 test_expect_success 'setup bare clone for server' '
  27         git clone --bare "file://$(pwd)/src" srv.bare &&
  28         git -C srv.bare config --local uploadpack.allowfilter 1 &&
  29         git -C srv.bare config --local uploadpack.allowanysha1inwant 1
  30 '
  31
  32 # do basic partial clone from "srv.bare"
  33 # confirm we are missing all of the known blobs.
  34 # confirm partial clone was registered in the local config.
  35 test_expect_success 'do partial clone 1' '
  36         git clone --no-checkout --filter=blob:none "file://$(pwd)/srv.bare" pc1 &&
  37
  38         git -C pc1 rev-list --quiet --objects --missing=print HEAD >revs &&
  39         awk -f print_1.awk revs |
  40         sed "s/?//" |
  41         sort >observed.oids &&
  42
  43         test_cmp expect_1.oids observed.oids &&
  44         test "$(git -C pc1 config --local core.repositoryformatversion)" = "1" &&
  45         test "$(git -C pc1 config --local extensions.partialclone)" = "origin" &&
  46         test "$(git -C pc1 config --local core.partialclonefilter)" = "blob:none"
  47 '
  48
  49 # checkout master to force dynamic object fetch of blobs at HEAD.
  50 test_expect_success 'verify checkout with dynamic object fetch' '
  51         git -C pc1 rev-list --quiet --objects --missing=print HEAD >observed &&
  52         test_line_count = 4 observed &&
  53         git -C pc1 checkout master &&
  54         git -C pc1 rev-list --quiet --objects --missing=print HEAD >observed &&
  55         test_line_count = 0 observed
  56 '
  57
  58 # create new commits in "src" repo to establish a blame history on file.1.txt
  59 # and push to "srv.bare".
  60 test_expect_success 'push new commits to server' '
  61         git -C src remote add srv "file://$(pwd)/srv.bare" &&
  62         for x in a b c d e
  63         do
  64                 echo "Mod file.1.txt $x" >>src/file.1.txt
  65                 git -C src add file.1.txt
  66                 git -C src commit -m "mod $x"
  67         done &&
  68         git -C src blame master -- file.1.txt >expect.blame &&
  69         git -C src push -u srv master
  70 '
  71
  72 # (partial) fetch in the partial clone repo from the promisor remote.
  73 # verify that fetch inherited the filter-spec from the config and DOES NOT
  74 # have the new blobs.
  75 test_expect_success 'partial fetch inherits filter settings' '
  76         git -C pc1 fetch origin &&
  77         git -C pc1 rev-list --quiet --objects --missing=print \
  78                 master..origin/master >observed &&
  79         test_line_count = 5 observed
  80 '
  81
  82 # force dynamic object fetch using diff.
  83 # we should only get 1 new blob (for the file in origin/master).
  84 test_expect_success 'verify diff causes dynamic object fetch' '
  85         git -C pc1 diff master..origin/master -- file.1.txt &&
  86         git -C pc1 rev-list --quiet --objects --missing=print \
  87                  master..origin/master >observed &&
  88         test_line_count = 4 observed
  89 '
  90
  91 # force full dynamic object fetch of the file's history using blame.
  92 # we should get the intermediate blobs for the file.
  93 test_expect_success 'verify blame causes dynamic object fetch' '
  94         git -C pc1 blame origin/master -- file.1.txt >observed.blame &&
  95         test_cmp expect.blame observed.blame &&
  96         git -C pc1 rev-list --quiet --objects --missing=print \
  97                 master..origin/master >observed &&
  98         test_line_count = 0 observed
  99 '
 100
 101 # create new commits in "src" repo to establish a history on file.2.txt
 102 # and push to "srv.bare".
 103 test_expect_success 'push new commits to server for file.2.txt' '
 104         for x in a b c d e f
 105         do
 106                 echo "Mod file.2.txt $x" >>src/file.2.txt
 107                 git -C src add file.2.txt
 108                 git -C src commit -m "mod $x"
 109         done &&
 110         git -C src push -u srv master
 111 '
 112
 113 # Do FULL fetch by disabling inherited filter-spec using --no-filter.
 114 # Verify we have all the new blobs.
 115 test_expect_success 'override inherited filter-spec using --no-filter' '
 116         git -C pc1 fetch --no-filter origin &&
 117         git -C pc1 rev-list --quiet --objects --missing=print \
 118                 master..origin/master >observed &&
 119         test_line_count = 0 observed
 120 '
 121
 122 # create new commits in "src" repo to establish a history on file.3.txt
 123 # and push to "srv.bare".
 124 test_expect_success 'push new commits to server for file.3.txt' '
 125         for x in a b c d e f
 126         do
 127                 echo "Mod file.3.txt $x" >>src/file.3.txt
 128                 git -C src add file.3.txt
 129                 git -C src commit -m "mod $x"
 130         done &&
 131         git -C src push -u srv master
 132 '
 133
 134 # Do a partial fetch and then try to manually fetch the missing objects.
 135 # This can be used as the basis of a pre-command hook to bulk fetch objects
 136 # perhaps combined with a command in dry-run mode.
 137 test_expect_success 'manual prefetch of missing objects' '
 138         git -C pc1 fetch --filter=blob:none origin &&
 139
 140         git -C pc1 rev-list --quiet --objects --missing=print \
 141                  master..origin/master >revs &&
 142         awk -f print_1.awk revs |
 143         sed "s/?//" |
 144         sort >observed.oids &&
 145
 146         test_line_count = 6 observed.oids &&
 147         git -C pc1 fetch-pack --stdin "file://$(pwd)/srv.bare" <observed.oids &&
 148
 149         git -C pc1 rev-list --quiet --objects --missing=print \
 150                 master..origin/master >revs &&
 151         awk -f print_1.awk revs |
 152         sed "s/?//" |
 153         sort >observed.oids &&
 154
 155         test_line_count = 0 observed.oids
 156 '
 157
 158 test_expect_success 'partial clone with transfer.fsckobjects=1 uses index-pack --fsck-objects' '
 159         git init src &&
 160         test_commit -C src x &&
 161         test_config -C src uploadpack.allowfilter 1 &&
 162         test_config -C src uploadpack.allowanysha1inwant 1 &&
 163
 164         GIT_TRACE="$(pwd)/trace" git -c transfer.fsckobjects=1 \
 165                 clone --filter="blob:none" "file://$(pwd)/src" dst &&
 166         grep "git index-pack.*--fsck-objects" trace
 167 '
 168
 169 test_expect_success 'use fsck before and after manually fetching a missing subtree' '
 170         # push new commit so server has a subtree
 171         mkdir src/dir &&
 172         echo "in dir" >src/dir/file.txt &&
 173         git -C src add dir/file.txt &&
 174         git -C src commit -m "file in dir" &&
 175         git -C src push -u srv master &&
 176         SUBTREE=$(git -C src rev-parse HEAD:dir) &&
 177
 178         rm -rf dst &&
 179         git clone --no-checkout --filter=tree:0 "file://$(pwd)/srv.bare" dst &&
 180         git -C dst fsck &&
 181
 182         # Make sure we only have commits, and all trees and blobs are missing.
 183         git -C dst rev-list --missing=allow-any --objects master \
 184                 >fetched_objects &&
 185         awk -f print_1.awk fetched_objects |
 186         xargs -n1 git -C dst cat-file -t >fetched_types &&
 187
 188         sort -u fetched_types >unique_types.observed &&
 189         echo commit >unique_types.expected &&
 190         test_cmp unique_types.expected unique_types.observed &&
 191
 192         # Auto-fetch a tree with cat-file.
 193         git -C dst cat-file -p $SUBTREE >tree_contents &&
 194         grep file.txt tree_contents &&
 195
 196         # fsck still works after an auto-fetch of a tree.
 197         git -C dst fsck &&
 198
 199         # Auto-fetch all remaining trees and blobs with --missing=error
 200         git -C dst rev-list --missing=error --objects master >fetched_objects &&
 201         test_line_count = 70 fetched_objects &&
 202
 203         awk -f print_1.awk fetched_objects |
 204         xargs -n1 git -C dst cat-file -t >fetched_types &&
 205
 206         sort -u fetched_types >unique_types.observed &&
 207         test_write_lines blob commit tree >unique_types.expected &&
 208         test_cmp unique_types.expected unique_types.observed
 209 '
 210
 211 test_expect_success 'partial clone fetches blobs pointed to by refs even if normally filtered out' '
 212         rm -rf src dst &&
 213         git init src &&
 214         test_commit -C src x &&
 215         test_config -C src uploadpack.allowfilter 1 &&
 216         test_config -C src uploadpack.allowanysha1inwant 1 &&
 217
 218         # Create a tag pointing to a blob.
 219         BLOB=$(echo blob-contents | git -C src hash-object --stdin -w) &&
 220         git -C src tag myblob "$BLOB" &&
 221
 222         git clone --filter="blob:none" "file://$(pwd)/src" dst 2>err &&
 223         ! grep "does not point to a valid object" err &&
 224         git -C dst fsck
 225 '
 226
 227 test_expect_success 'fetch what is specified on CLI even if already promised' '
 228         rm -rf src dst.git &&
 229         git init src &&
 230         test_commit -C src foo &&
 231         test_config -C src uploadpack.allowfilter 1 &&
 232         test_config -C src uploadpack.allowanysha1inwant 1 &&
 233
 234         git hash-object --stdin <src/foo.t >blob &&
 235
 236         git clone --bare --filter=blob:none "file://$(pwd)/src" dst.git &&
 237         git -C dst.git rev-list --objects --quiet --missing=print HEAD >missing_before &&
 238         grep "?$(cat blob)" missing_before &&
 239         git -C dst.git fetch origin $(cat blob) &&
 240         git -C dst.git rev-list --objects --quiet --missing=print HEAD >missing_after &&
 241         ! grep "?$(cat blob)" missing_after
 242 '
 243
 244 . "$TEST_DIRECTORY"/lib-httpd.sh
 245 start_httpd
 246
 247 # Converts bytes into their hexadecimal representation. For example,
 248 # "printf 'ab\r\n' | hex_unpack" results in '61620d0a'.
 249 hex_unpack () {
 250         perl -e '$/ = undef; $input = <>; print unpack("H2" x length($input), $input)'
 251 }
 252
 253 # Inserts $1 at the start of the string and every 2 characters thereafter.
 254 intersperse () {
 255         sed 's/\(..\)/'$1'\1/g'
 256 }
 257
 258 # Create a one-time-sed command to replace the existing packfile with $1.
 259 replace_packfile () {
 260         # The protocol requires that the packfile be sent in sideband 1, hence
 261         # the extra \x01 byte at the beginning.
 262         printf "1,/packfile/!c %04x\\\\x01%s0000" \
 263                 "$(($(wc -c <$1) + 5))" \
 264                 "$(hex_unpack <$1 | intersperse '\\x')" \
 265                 >"$HTTPD_ROOT_PATH/one-time-sed"
 266 }
 267
 268 test_expect_success 'upon cloning, check that all refs point to objects' '
 269         SERVER="$HTTPD_DOCUMENT_ROOT_PATH/server" &&
 270         rm -rf "$SERVER" repo &&
 271         test_create_repo "$SERVER" &&
 272         test_commit -C "$SERVER" foo &&
 273         test_config -C "$SERVER" uploadpack.allowfilter 1 &&
 274         test_config -C "$SERVER" uploadpack.allowanysha1inwant 1 &&
 275
 276         # Create a tag pointing to a blob.
 277         BLOB=$(echo blob-contents | git -C "$SERVER" hash-object --stdin -w) &&
 278         git -C "$SERVER" tag myblob "$BLOB" &&
 279
 280         # Craft a packfile not including that blob.
 281         git -C "$SERVER" rev-parse HEAD |
 282         git -C "$SERVER" pack-objects --stdout >incomplete.pack &&
 283
 284         # Replace the existing packfile with the crafted one. The protocol
 285         # requires that the packfile be sent in sideband 1, hence the extra
 286         # \x01 byte at the beginning.
 287         replace_packfile incomplete.pack &&
 288
 289         # Use protocol v2 because the sed command looks for the "packfile"
 290         # section header.
 291         test_config -C "$SERVER" protocol.version 2 &&
 292         test_must_fail git -c protocol.version=2 clone \
 293                 --filter=blob:none $HTTPD_URL/one_time_sed/server repo 2>err &&
 294
 295         test_i18ngrep "did not send all necessary objects" err &&
 296
 297         # Ensure that the one-time-sed script was used.
 298         ! test -e "$HTTPD_ROOT_PATH/one-time-sed"
 299 '
 300
 301 test_expect_success 'when partial cloning, tolerate server not sending target of tag' '
 302         SERVER="$HTTPD_DOCUMENT_ROOT_PATH/server" &&
 303         rm -rf "$SERVER" repo &&
 304         test_create_repo "$SERVER" &&
 305         test_commit -C "$SERVER" foo &&
 306         test_config -C "$SERVER" uploadpack.allowfilter 1 &&
 307         test_config -C "$SERVER" uploadpack.allowanysha1inwant 1 &&
 308
 309         # Create an annotated tag pointing to a blob.
 310         BLOB=$(echo blob-contents | git -C "$SERVER" hash-object --stdin -w) &&
 311         git -C "$SERVER" tag -m message -a myblob "$BLOB" &&
 312
 313         # Craft a packfile including the tag, but not the blob it points to.
 314         # Also, omit objects referenced from HEAD in order to force a second
 315         # fetch (to fetch missing objects) upon the automatic checkout that
 316         # happens after a clone.
 317         printf "%s\n%s\n--not\n%s\n%s\n" \
 318                 $(git -C "$SERVER" rev-parse HEAD) \
 319                 $(git -C "$SERVER" rev-parse myblob) \
 320                 $(git -C "$SERVER" rev-parse HEAD^{tree}) \
 321                 $(git -C "$SERVER" rev-parse myblob^{blob}) |
 322                 git -C "$SERVER" pack-objects --thin --stdout >incomplete.pack &&
 323
 324         # Replace the existing packfile with the crafted one. The protocol
 325         # requires that the packfile be sent in sideband 1, hence the extra
 326         # \x01 byte at the beginning.
 327         replace_packfile incomplete.pack &&
 328
 329         # Use protocol v2 because the sed command looks for the "packfile"
 330         # section header.
 331         test_config -C "$SERVER" protocol.version 2 &&
 332
 333         # Exercise to make sure it works.
 334         git -c protocol.version=2 clone \
 335                 --filter=blob:none $HTTPD_URL/one_time_sed/server repo 2> err &&
 336         ! grep "missing object referenced by" err &&
 337
 338         # Ensure that the one-time-sed script was used.
 339         ! test -e "$HTTPD_ROOT_PATH/one-time-sed"
 340 '
 341
 342 test_expect_success 'tolerate server sending REF_DELTA against missing promisor objects' '
 343         SERVER="$HTTPD_DOCUMENT_ROOT_PATH/server" &&
 344         rm -rf "$SERVER" repo &&
 345         test_create_repo "$SERVER" &&
 346         test_config -C "$SERVER" uploadpack.allowfilter 1 &&
 347         test_config -C "$SERVER" uploadpack.allowanysha1inwant 1 &&
 348
 349         # Create a commit with 2 blobs to be used as delta bases.
 350         for i in $(test_seq 10)
 351         do
 352                 echo "this is a line" >>"$SERVER/foo.txt" &&
 353                 echo "this is another line" >>"$SERVER/have.txt"
 354         done &&
 355         git -C "$SERVER" add foo.txt have.txt &&
 356         git -C "$SERVER" commit -m bar &&
 357         git -C "$SERVER" rev-parse HEAD:foo.txt >deltabase_missing &&
 358         git -C "$SERVER" rev-parse HEAD:have.txt >deltabase_have &&
 359
 360         # Clone. The client has deltabase_have but not deltabase_missing.
 361         git -c protocol.version=2 clone --no-checkout \
 362                 --filter=blob:none $HTTPD_URL/one_time_sed/server repo &&
 363         git -C repo hash-object -w -- "$SERVER/have.txt" &&
 364
 365         # Sanity check to ensure that the client does not have
 366         # deltabase_missing.
 367         git -C repo rev-list --objects --ignore-missing \
 368                 -- $(cat deltabase_missing) >objlist &&
 369         test_line_count = 0 objlist &&
 370
 371         # Another commit. This commit will be fetched by the client.
 372         echo "abcdefghijklmnopqrstuvwxyz" >>"$SERVER/foo.txt" &&
 373         echo "abcdefghijklmnopqrstuvwxyz" >>"$SERVER/have.txt" &&
 374         git -C "$SERVER" add foo.txt have.txt &&
 375         git -C "$SERVER" commit -m baz &&
 376
 377         # Pack a thin pack containing, among other things, HEAD:foo.txt
 378         # delta-ed against HEAD^:foo.txt and HEAD:have.txt delta-ed against
 379         # HEAD^:have.txt.
 380         printf "%s\n--not\n%s\n" \
 381                 $(git -C "$SERVER" rev-parse HEAD) \
 382                 $(git -C "$SERVER" rev-parse HEAD^) |
 383                 git -C "$SERVER" pack-objects --thin --stdout >thin.pack &&
 384
 385         # Ensure that the pack contains one delta against HEAD^:foo.txt. Since
 386         # the delta contains at least 26 novel characters, the size cannot be
 387         # contained in 4 bits, so the object header will take up 2 bytes. The
 388         # most significant nybble of the first byte is 0b1111 (0b1 to indicate
 389         # that the header continues, and 0b111 to indicate REF_DELTA), followed
 390         # by any 3 nybbles, then the OID of the delta base.
 391         printf "f.,..%s" $(intersperse "," <deltabase_missing) >want &&
 392         hex_unpack <thin.pack | intersperse "," >have &&
 393         grep $(cat want) have &&
 394
 395         # Ensure that the pack contains one delta against HEAD^:have.txt,
 396         # similar to the above.
 397         printf "f.,..%s" $(intersperse "," <deltabase_have) >want &&
 398         hex_unpack <thin.pack | intersperse "," >have &&
 399         grep $(cat want) have &&
 400
 401         replace_packfile thin.pack &&
 402
 403         # Use protocol v2 because the sed command looks for the "packfile"
 404         # section header.
 405         test_config -C "$SERVER" protocol.version 2 &&
 406
 407         # Fetch the thin pack and ensure that index-pack is able to handle the
 408         # REF_DELTA object with a missing promisor delta base.
 409         GIT_TRACE_PACKET="$(pwd)/trace" git -C repo -c protocol.version=2 fetch &&
 410
 411         # Ensure that the missing delta base was directly fetched, but not the
 412         # one that the client has.
 413         grep "want $(cat deltabase_missing)" trace &&
 414         ! grep "want $(cat deltabase_have)" trace &&
 415
 416         # Ensure that the one-time-sed script was used.
 417         ! test -e "$HTTPD_ROOT_PATH/one-time-sed"
 418 '
 419
 420 test_done