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         git -C pc1 rev-list HEAD --quiet --objects --missing=print \
  38                 | awk -f print_1.awk \
  39                 | sed "s/?//" \
  40                 | sort >observed.oids &&
  41         test_cmp expect_1.oids observed.oids &&
  42         test "$(git -C pc1 config --local core.repositoryformatversion)" = "1" &&
  43         test "$(git -C pc1 config --local extensions.partialclone)" = "origin" &&
  44         test "$(git -C pc1 config --local core.partialclonefilter)" = "blob:none"
  45 '
  46
  47 # checkout master to force dynamic object fetch of blobs at HEAD.
  48 test_expect_success 'verify checkout with dynamic object fetch' '
  49         git -C pc1 rev-list HEAD --quiet --objects --missing=print >observed &&
  50         test_line_count = 4 observed &&
  51         git -C pc1 checkout master &&
  52         git -C pc1 rev-list HEAD --quiet --objects --missing=print >observed &&
  53         test_line_count = 0 observed
  54 '
  55
  56 # create new commits in "src" repo to establish a blame history on file.1.txt
  57 # and push to "srv.bare".
  58 test_expect_success 'push new commits to server' '
  59         git -C src remote add srv "file://$(pwd)/srv.bare" &&
  60         for x in a b c d e
  61         do
  62                 echo "Mod file.1.txt $x" >>src/file.1.txt
  63                 git -C src add file.1.txt
  64                 git -C src commit -m "mod $x"
  65         done &&
  66         git -C src blame master -- file.1.txt >expect.blame &&
  67         git -C src push -u srv master
  68 '
  69
  70 # (partial) fetch in the partial clone repo from the promisor remote.
  71 # verify that fetch inherited the filter-spec from the config and DOES NOT
  72 # have the new blobs.
  73 test_expect_success 'partial fetch inherits filter settings' '
  74         git -C pc1 fetch origin &&
  75         git -C pc1 rev-list master..origin/master --quiet --objects --missing=print >observed &&
  76         test_line_count = 5 observed
  77 '
  78
  79 # force dynamic object fetch using diff.
  80 # we should only get 1 new blob (for the file in origin/master).
  81 test_expect_success 'verify diff causes dynamic object fetch' '
  82         git -C pc1 diff master..origin/master -- file.1.txt &&
  83         git -C pc1 rev-list master..origin/master --quiet --objects --missing=print >observed &&
  84         test_line_count = 4 observed
  85 '
  86
  87 # force full dynamic object fetch of the file's history using blame.
  88 # we should get the intermediate blobs for the file.
  89 test_expect_success 'verify blame causes dynamic object fetch' '
  90         git -C pc1 blame origin/master -- file.1.txt >observed.blame &&
  91         test_cmp expect.blame observed.blame &&
  92         git -C pc1 rev-list master..origin/master --quiet --objects --missing=print >observed &&
  93         test_line_count = 0 observed
  94 '
  95
  96 # create new commits in "src" repo to establish a history on file.2.txt
  97 # and push to "srv.bare".
  98 test_expect_success 'push new commits to server for file.2.txt' '
  99         for x in a b c d e f
 100         do
 101                 echo "Mod file.2.txt $x" >>src/file.2.txt
 102                 git -C src add file.2.txt
 103                 git -C src commit -m "mod $x"
 104         done &&
 105         git -C src push -u srv master
 106 '
 107
 108 # Do FULL fetch by disabling inherited filter-spec using --no-filter.
 109 # Verify we have all the new blobs.
 110 test_expect_success 'override inherited filter-spec using --no-filter' '
 111         git -C pc1 fetch --no-filter origin &&
 112         git -C pc1 rev-list master..origin/master --quiet --objects --missing=print >observed &&
 113         test_line_count = 0 observed
 114 '
 115
 116 # create new commits in "src" repo to establish a history on file.3.txt
 117 # and push to "srv.bare".
 118 test_expect_success 'push new commits to server for file.3.txt' '
 119         for x in a b c d e f
 120         do
 121                 echo "Mod file.3.txt $x" >>src/file.3.txt
 122                 git -C src add file.3.txt
 123                 git -C src commit -m "mod $x"
 124         done &&
 125         git -C src push -u srv master
 126 '
 127
 128 # Do a partial fetch and then try to manually fetch the missing objects.
 129 # This can be used as the basis of a pre-command hook to bulk fetch objects
 130 # perhaps combined with a command in dry-run mode.
 131 test_expect_success 'manual prefetch of missing objects' '
 132         git -C pc1 fetch --filter=blob:none origin &&
 133         git -C pc1 rev-list master..origin/master --quiet --objects --missing=print \
 134                 | awk -f print_1.awk \
 135                 | sed "s/?//" \
 136                 | sort >observed.oids &&
 137         test_line_count = 6 observed.oids &&
 138         git -C pc1 fetch-pack --stdin "file://$(pwd)/srv.bare" <observed.oids &&
 139         git -C pc1 rev-list master..origin/master --quiet --objects --missing=print \
 140                 | awk -f print_1.awk \
 141                 | sed "s/?//" \
 142                 | sort >observed.oids &&
 143         test_line_count = 0 observed.oids
 144 '
 145
 146 test_expect_success 'partial clone with transfer.fsckobjects=1 uses index-pack --fsck-objects' '
 147         git init src &&
 148         test_commit -C src x &&
 149         test_config -C src uploadpack.allowfilter 1 &&
 150         test_config -C src uploadpack.allowanysha1inwant 1 &&
 151
 152         GIT_TRACE="$(pwd)/trace" git -c transfer.fsckobjects=1 \
 153                 clone --filter="blob:none" "file://$(pwd)/src" dst &&
 154         grep "git index-pack.*--fsck-objects" trace
 155 '
 156
 157 test_expect_success 'use fsck before and after manually fetching a missing subtree' '
 158         # push new commit so server has a subtree
 159         mkdir src/dir &&
 160         echo "in dir" >src/dir/file.txt &&
 161         git -C src add dir/file.txt &&
 162         git -C src commit -m "file in dir" &&
 163         git -C src push -u srv master &&
 164         SUBTREE=$(git -C src rev-parse HEAD:dir) &&
 165
 166         rm -rf dst &&
 167         git clone --no-checkout --filter=tree:0 "file://$(pwd)/srv.bare" dst &&
 168         git -C dst fsck &&
 169
 170         # Make sure we only have commits, and all trees and blobs are missing.
 171         git -C dst rev-list --missing=allow-any --objects master \
 172                 >fetched_objects &&
 173         awk -f print_1.awk fetched_objects |
 174         xargs -n1 git -C dst cat-file -t >fetched_types &&
 175
 176         sort -u fetched_types >unique_types.observed &&
 177         echo commit >unique_types.expected &&
 178         test_cmp unique_types.expected unique_types.observed &&
 179
 180         # Auto-fetch a tree with cat-file.
 181         git -C dst cat-file -p $SUBTREE >tree_contents &&
 182         grep file.txt tree_contents &&
 183
 184         # fsck still works after an auto-fetch of a tree.
 185         git -C dst fsck &&
 186
 187         # Auto-fetch all remaining trees and blobs with --missing=error
 188         git -C dst rev-list --missing=error --objects master >fetched_objects &&
 189         test_line_count = 70 fetched_objects &&
 190
 191         awk -f print_1.awk fetched_objects |
 192         xargs -n1 git -C dst cat-file -t >fetched_types &&
 193
 194         sort -u fetched_types >unique_types.observed &&
 195         printf "blob\ncommit\ntree\n" >unique_types.expected &&
 196         test_cmp unique_types.expected unique_types.observed
 197 '
 198
 199 test_expect_success 'partial clone fetches blobs pointed to by refs even if normally filtered out' '
 200         rm -rf src dst &&
 201         git init src &&
 202         test_commit -C src x &&
 203         test_config -C src uploadpack.allowfilter 1 &&
 204         test_config -C src uploadpack.allowanysha1inwant 1 &&
 205
 206         # Create a tag pointing to a blob.
 207         BLOB=$(echo blob-contents | git -C src hash-object --stdin -w) &&
 208         git -C src tag myblob "$BLOB" &&
 209
 210         git clone --filter="blob:none" "file://$(pwd)/src" dst 2>err &&
 211         ! grep "does not point to a valid object" err &&
 212         git -C dst fsck
 213 '
 214
 215 . "$TEST_DIRECTORY"/lib-httpd.sh
 216 start_httpd
 217
 218 # Converts bytes into a form suitable for inclusion in a sed command. For
 219 # example, "printf 'ab\r\n' | hex_unpack" results in '\x61\x62\x0d\x0a'.
 220 sed_escape () {
 221         perl -e '$/ = undef; $input = <>; print unpack("H2" x length($input), $input)' |
 222                 sed 's/\(..\)/\\x\1/g'
 223 }
 224
 225 test_expect_success 'upon cloning, check that all refs point to objects' '
 226         SERVER="$HTTPD_DOCUMENT_ROOT_PATH/server" &&
 227         rm -rf "$SERVER" repo &&
 228         test_create_repo "$SERVER" &&
 229         test_commit -C "$SERVER" foo &&
 230         test_config -C "$SERVER" uploadpack.allowfilter 1 &&
 231         test_config -C "$SERVER" uploadpack.allowanysha1inwant 1 &&
 232
 233         # Create a tag pointing to a blob.
 234         BLOB=$(echo blob-contents | git -C "$SERVER" hash-object --stdin -w) &&
 235         git -C "$SERVER" tag myblob "$BLOB" &&
 236
 237         # Craft a packfile not including that blob.
 238         git -C "$SERVER" rev-parse HEAD |
 239                 git -C "$SERVER" pack-objects --stdout >incomplete.pack &&
 240
 241         # Replace the existing packfile with the crafted one. The protocol
 242         # requires that the packfile be sent in sideband 1, hence the extra
 243         # \x01 byte at the beginning.
 244         printf "1,/packfile/!c %04x\\\\x01%s0000" \
 245                 "$(($(wc -c <incomplete.pack) + 5))" \
 246                 "$(sed_escape <incomplete.pack)" \
 247                 >"$HTTPD_ROOT_PATH/one-time-sed" &&
 248
 249         # Use protocol v2 because the sed command looks for the "packfile"
 250         # section header.
 251         test_config -C "$SERVER" protocol.version 2 &&
 252         test_must_fail git -c protocol.version=2 clone \
 253                 --filter=blob:none $HTTPD_URL/one_time_sed/server repo 2>err &&
 254
 255         grep "did not send all necessary objects" err &&
 256
 257         # Ensure that the one-time-sed script was used.
 258         ! test -e "$HTTPD_ROOT_PATH/one-time-sed"
 259 '
 260
 261 test_expect_success 'when partial cloning, tolerate server not sending target of tag' '
 262         SERVER="$HTTPD_DOCUMENT_ROOT_PATH/server" &&
 263         rm -rf "$SERVER" repo &&
 264         test_create_repo "$SERVER" &&
 265         test_commit -C "$SERVER" foo &&
 266         test_config -C "$SERVER" uploadpack.allowfilter 1 &&
 267         test_config -C "$SERVER" uploadpack.allowanysha1inwant 1 &&
 268
 269         # Create an annotated tag pointing to a blob.
 270         BLOB=$(echo blob-contents | git -C "$SERVER" hash-object --stdin -w) &&
 271         git -C "$SERVER" tag -m message -a myblob "$BLOB" &&
 272
 273         # Craft a packfile including the tag, but not the blob it points to.
 274         # Also, omit objects referenced from HEAD in order to force a second
 275         # fetch (to fetch missing objects) upon the automatic checkout that
 276         # happens after a clone.
 277         printf "%s\n%s\n--not\n%s\n%s\n" \
 278                 $(git -C "$SERVER" rev-parse HEAD) \
 279                 $(git -C "$SERVER" rev-parse myblob) \
 280                 $(git -C "$SERVER" rev-parse HEAD^{tree}) \
 281                 $(git -C "$SERVER" rev-parse myblob^{blob}) |
 282                 git -C "$SERVER" pack-objects --thin --stdout >incomplete.pack &&
 283
 284         # Replace the existing packfile with the crafted one. The protocol
 285         # requires that the packfile be sent in sideband 1, hence the extra
 286         # \x01 byte at the beginning.
 287         printf "1,/packfile/!c %04x\\\\x01%s0000" \
 288                 "$(($(wc -c <incomplete.pack) + 5))" \
 289                 "$(sed_escape <incomplete.pack)" \
 290                 >"$HTTPD_ROOT_PATH/one-time-sed" &&
 291
 292         # Use protocol v2 because the sed command looks for the "packfile"
 293         # section header.
 294         test_config -C "$SERVER" protocol.version 2 &&
 295
 296         # Exercise to make sure it works.
 297         git -c protocol.version=2 clone \
 298                 --filter=blob:none $HTTPD_URL/one_time_sed/server repo 2> err &&
 299         ! grep "missing object referenced by" err &&
 300
 301         # Ensure that the one-time-sed script was used.
 302         ! test -e "$HTTPD_ROOT_PATH/one-time-sed"
 303 '
 304
 305 stop_httpd
 306
 307 test_done