Re: [PATCH] Use wc instead of awk to count subtrees in t0090-cache-tree

Junio C Hamano [off-list ref] writes:

From: Ben Walton <redacted>

The awk statements previously used in this test weren't compatible
with the native versions of awk on Solaris:

    echo "dir" | /bin/awk -v c=0 '$1 {++c} END {print c}'
    awk: syntax error near line 1
    awk: bailing out near line 1

    echo "dir" | /usr/xpg4/bin/awk -v c=0 '$1 {++c} END {print c}'
    0

And with GNU awk for comparison:

    echo "dir" | /opt/csw/gnu/awk -v c=0 '$1 {++c} END {print c}'
    1

Work it around by using $1 != "" to state more explicitly that we
are skipping empty lines.

By the way, I was hoping (eh, what kind of hope is that???) that $1
alone is not a kosher POSIX way but a GNUism, but that does not seem
to be the case.  POSIX has this [*1*]

    When an expression is used in a Boolean context, if it has a
    numeric value, a value of zero shall be treated as false and any
    other value shall be treated as true. Otherwise, a string value
    of the null string shall be treated as false and any other value
    shall be treated as true. A Boolean context shall be one of the
    following:

and among the "Boolean context" listed is:

    * An expression used as a pattern (as in Overall Program Structure)

So the example with /usr/xpg4/bin/awk does not seem to be a
behaviour from a conformant implementationd, and it seems to be
correct to label this as "work it around by ..." (not "avoid using
GNUism").

We learn new things every day (not that I really wanted to learn
glitches in various implementations of awk) ;-).

Thanks.


[Reference]

*1* http://pubs.opengroup.org/onlinepubs/9699919799/utilities/awk.html