tests/pdiff
author Boris Feld <boris.feld@octobus.net>
Thu, 06 Jul 2017 14:50:17 +0200
changeset 34414 014d467f9d08
parent 33610 a2b55ee62803
permissions -rwxr-xr-x
effectflag: store an empty effect flag for the moment The idea behind effect flag is to store additional information in obs-markers about what changed between a changeset and its successor(s). It's a low-level information that comes without guarantees. This information can be computed a posteriori, but only if we have all changesets locally. This is not the case with distributed workflows where you work with several people or on several computers (eg: laptop + build server). Storing the effect-flag as a bitfield has several advantages: - It's compact, we are using one byte per obs-marker at most for the effect- flag. - It's compoundable, the obsfate log approach needs to display evolve history that could spans several obs-markers. Computing the effect-flag between a changeset and its grand-grand-grand-successor is simple thanks to the bitfield. The effect-flag design has also some limitations: - Evolving a changeset and reverting these changes just after would lead to two obs-markers with the same effect-flag without information that the first and third changesets are the same. The effect-flag current design is a trade-off between compactness and usefulness. Storing this information helps commands to display a more complete and understandable evolve history. For example, obslog (an Evolve command) use it to improve its output: x 62206adfd571 (34302) obscache: skip updating outdated obscache... | rewritten(parent) by Matthieu Laneuville <matthieu.laneuville@octobus... | rewritten(content) by Boris Feld <boris.feld@octobus.net> The effect flag is stored in obs-markers metadata while we iterate on the information we want to store. We plan to extend the existing obsmarkers bit-field when the effect flag design will be stabilized. It's different from the CommitCustody concept, effect-flag are not signed and can be forged. It's also different from the operation metadata as the command name (for example: amend) could alter a changeset in different ways (changing the content with hg amend, changing the description with hg amend -e, changing the user with hg amend -U). Also it's compatible with every custom command that writes obs-markers without needing to be updated. The effect-flag is placed behind an experimental flag set to off by default. Hook the saving of effect flag in create markers, but store only an empty one for the moment, I will refine the values in effect flag in following patches. For more information, see: https://www.mercurial-scm.org/wiki/ChangesetEvolutionDevel#Record_types_of_operation Differential Revision: https://phab.mercurial-scm.org/D533

#!/bin/sh

# Script to get stable diff output on any platform.
#
# Output of this script is almost equivalent to GNU diff with "-Nru".
#
# Use this script as "hg pdiff" via extdiff extension with preparation
# below in test scripts:
#
#   $ cat >> $HGRCPATH <<EOF
#   > [extdiff]
#   > pdiff = sh "$RUNTESTDIR/pdiff"
#   > EOF

filediff(){
    # USAGE: filediff file1 file2 [header]

    # compare with /dev/null if file doesn't exist (as "-N" option)
    file1="$1"
    if test ! -f "$file1"; then
        file1=/dev/null
    fi
    file2="$2"
    if test ! -f "$file2"; then
        file2=/dev/null
    fi

    if cmp -s "$file1" "$file2" 2> /dev/null; then
        # Return immediately, because comparison isn't needed. This
        # also avoids redundant message of diff like "No differences
        # encountered" (on Solaris)
        return
    fi

    if test -n "$3"; then
        # show header only in recursive case
        echo "$3"
    fi

    # replace "/dev/null" by corresponded filename (as "-N" option)
    diff -u "$file1" "$file2" |
    sed "s@^--- /dev/null\(.*\)\$@--- $1\1@" |
    sed "s@^\+\+\+ /dev/null\(.*\)\$@+++ $2\1@"

    # in this case, files differ from each other
    return 1
}

if test -d "$1" -o -d "$2"; then
    # ensure comparison in dictionary order
    (
    if test -d "$1"; then (cd "$1" && find . -type f); fi
    if test -d "$2"; then (cd "$2" && find . -type f); fi
    ) |
    sed 's@^\./@@g' | sort | uniq |
    while read file; do
        filediff "$1/$file" "$2/$file" "diff -Nru $1/$file $2/$file"
    done

    # TODO: there is no portable way for current while-read based
    # implementation to return 1 at detecting changes.
    #
    # On bash and dash, assignment to variable inside while-block
    # doesn't affect outside, because inside while-block is executed
    # in sub-shell. BTW, it affects outside while-block on ksh (as sh
    # on Solaris).
else
    filediff "$1" "$2"
fi