Andrew's git - gitweb.git/blob - t/t5616-partial-clone.sh

   1#!/bin/sh
   2
   3test_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.
   9test_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.
  26test_expect_success 'setup bare clone for server' '
  27        git clone --bare "file://$(pwd)/src" srv.bare &&
  28        git -C srv.bare config --local uploadpack.allowfilter 1 &&
  29        git -C srv.bare config --local uploadpack.allowanysha1inwant 1
  30'
  31
  32# do basic partial clone from "srv.bare"
  33# confirm we are missing all of the known blobs.
  34# confirm partial clone was registered in the local config.
  35test_expect_success 'do partial clone 1' '
  36        git clone --no-checkout --filter=blob:none "file://$(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 "$(git -C pc1 config --local core.partialclonefilter)" = "blob:none"
  45'
  46
  47# checkout master to force dynamic object fetch of blobs at HEAD.
  48test_expect_success 'verify checkout with dynamic object fetch' '
  49        git -C pc1 rev-list HEAD --quiet --objects --missing=print >observed &&
  50        test_line_count = 4 observed &&
  51        git -C pc1 checkout master &&
  52        git -C pc1 rev-list HEAD --quiet --objects --missing=print >observed &&
  53        test_line_count = 0 observed
  54'
  55
  56# create new commits in "src" repo to establish a blame history on file.1.txt
  57# and push to "srv.bare".
  58test_expect_success 'push new commits to server' '
  59        git -C src remote add srv "file://$(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.
  73test_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).
  81test_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.
  89test_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".
  98test_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.
 110test_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".
 118test_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 $x"
 124        done &&
 125        git -C src push -u srv master
 126'
 127
 128# Do a partial fetch and then try to manually fetch the missing objects.
 129# This can be used as the basis of a pre-command hook to bulk fetch objects
 130# perhaps combined with a command in dry-run mode.
 131test_expect_success 'manual prefetch of missing objects' '
 132        git -C pc1 fetch --filter=blob:none origin &&
 133        git -C pc1 rev-list master..origin/master --quiet --objects --missing=print \
 134                | awk -f print_1.awk \
 135                | sed "s/?//" \
 136                | sort >observed.oids &&
 137        test_line_count = 6 observed.oids &&
 138        git -C pc1 fetch-pack --stdin "file://$(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
 146test_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
 157test_expect_success 'partial clone fetches blobs pointed to by refs even if normally filtered out' '
 158        rm -rf src dst &&
 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        # Create a tag pointing to a blob.
 165        BLOB=$(echo blob-contents | git -C src hash-object --stdin -w) &&
 166        git -C src tag myblob "$BLOB" &&
 167
 168        git clone --filter="blob:none" "file://$(pwd)/src" dst 2>err &&
 169        ! grep "does not point to a valid object" err &&
 170        git -C dst fsck
 171'
 172
 173. "$TEST_DIRECTORY"/lib-httpd.sh
 174start_httpd
 175
 176# Converts bytes into a form suitable for inclusion in a sed command. For
 177# example, "printf 'ab\r\n' | hex_unpack" results in '\x61\x62\x0d\x0a'.
 178sed_escape () {
 179        perl -e '$/ = undef; $input = <>; print unpack("H2" x length($input), $input)' |
 180                sed 's/\(..\)/\\x\1/g'
 181}
 182
 183test_expect_success 'upon cloning, check that all refs point to objects' '
 184        SERVER="$HTTPD_DOCUMENT_ROOT_PATH/server" &&
 185        rm -rf "$SERVER" repo &&
 186        test_create_repo "$SERVER" &&
 187        test_commit -C "$SERVER" foo &&
 188        test_config -C "$SERVER" uploadpack.allowfilter 1 &&
 189        test_config -C "$SERVER" uploadpack.allowanysha1inwant 1 &&
 190
 191        # Create a tag pointing to a blob.
 192        BLOB=$(echo blob-contents | git -C "$SERVER" hash-object --stdin -w) &&
 193        git -C "$SERVER" tag myblob "$BLOB" &&
 194
 195        # Craft a packfile not including that blob.
 196        git -C "$SERVER" rev-parse HEAD |
 197                git -C "$SERVER" pack-objects --stdout >incomplete.pack &&
 198
 199        # Replace the existing packfile with the crafted one. The protocol
 200        # requires that the packfile be sent in sideband 1, hence the extra
 201        # \x01 byte at the beginning.
 202        printf "1,/packfile/!c %04x\\\\x01%s0000" \
 203                "$(($(wc -c <incomplete.pack) + 5))" \
 204                "$(sed_escape <incomplete.pack)" \
 205                >"$HTTPD_ROOT_PATH/one-time-sed" &&
 206
 207        # Use protocol v2 because the sed command looks for the "packfile"
 208        # section header.
 209        test_config -C "$SERVER" protocol.version 2 &&
 210        test_must_fail git -c protocol.version=2 clone \
 211                --filter=blob:none $HTTPD_URL/one_time_sed/server repo 2>err &&
 212
 213        grep "did not send all necessary objects" err &&
 214
 215        # Ensure that the one-time-sed script was used.
 216        ! test -e "$HTTPD_ROOT_PATH/one-time-sed"
 217'
 218
 219test_expect_success 'when partial cloning, tolerate server not sending target of tag' '
 220        SERVER="$HTTPD_DOCUMENT_ROOT_PATH/server" &&
 221        rm -rf "$SERVER" repo &&
 222        test_create_repo "$SERVER" &&
 223        test_commit -C "$SERVER" foo &&
 224        test_config -C "$SERVER" uploadpack.allowfilter 1 &&
 225        test_config -C "$SERVER" uploadpack.allowanysha1inwant 1 &&
 226
 227        # Create an annotated tag pointing to a blob.
 228        BLOB=$(echo blob-contents | git -C "$SERVER" hash-object --stdin -w) &&
 229        git -C "$SERVER" tag -m message -a myblob "$BLOB" &&
 230
 231        # Craft a packfile including the tag, but not the blob it points to.
 232        # Also, omit objects referenced from HEAD in order to force a second
 233        # fetch (to fetch missing objects) upon the automatic checkout that
 234        # happens after a clone.
 235        printf "%s\n%s\n--not\n%s\n%s\n" \
 236                $(git -C "$SERVER" rev-parse HEAD) \
 237                $(git -C "$SERVER" rev-parse myblob) \
 238                $(git -C "$SERVER" rev-parse HEAD^{tree}) \
 239                $(git -C "$SERVER" rev-parse myblob^{blob}) |
 240                git -C "$SERVER" pack-objects --thin --stdout >incomplete.pack &&
 241
 242        # Replace the existing packfile with the crafted one. The protocol
 243        # requires that the packfile be sent in sideband 1, hence the extra
 244        # \x01 byte at the beginning.
 245        printf "1,/packfile/!c %04x\\\\x01%s0000" \
 246                "$(($(wc -c <incomplete.pack) + 5))" \
 247                "$(sed_escape <incomplete.pack)" \
 248                >"$HTTPD_ROOT_PATH/one-time-sed" &&
 249
 250        # Use protocol v2 because the sed command looks for the "packfile"
 251        # section header.
 252        test_config -C "$SERVER" protocol.version 2 &&
 253
 254        # Exercise to make sure it works.
 255        git -c protocol.version=2 clone \
 256                --filter=blob:none $HTTPD_URL/one_time_sed/server repo 2> err &&
 257        ! grep "missing object referenced by" err &&
 258
 259        # Ensure that the one-time-sed script was used.
 260        ! test -e "$HTTPD_ROOT_PATH/one-time-sed"
 261'
 262
 263stop_httpd
 264
 265test_done