tests/test-extdata.t

worker: avoid potential partial write of pickled data

Previously, the code wrote the pickled data using os.write(). However,
os.write() can write less bytes than passed to it. To trigger the problem, the
pickled data had to be larger than 2147479552 bytes on my system.

Instead, open a file object and pass it to pickle.dump(). This also has the
advantage that it doesn’t buffer the whole pickled data in memory.

Note that the opened file must be buffered because pickle doesn’t support
unbuffered streams because unbuffered streams’ write() method might write less
bytes than passed to it (like os.write()) but pickle.dump() relies on that all
bytes are written (see https://github.com/python/cpython/issues/93050).

The side effect of using a file object and a with statement is that wfd is
explicitly closed now while it seems like before it was implicitly closed by
process exit.

  $ hg init repo
  $ cd repo
  $ for n in 0 1 2 3 4 5 6 7 8 9 10 11; do
  >   echo $n > $n
  >   hg ci -qAm $n
  > done

test revset support

  $ cat <<'EOF' >> .hg/hgrc
  > [extdata]
  > filedata = file:extdata.txt
  > notes = notes.txt
  > shelldata = shell:cat extdata.txt | grep 2
  > emptygrep = shell:cat extdata.txt | grep empty
  > badparse = shell:cat badparse.txt
  > EOF
  $ cat <<'EOF' > extdata.txt
  > 2 another comment on 2
  > 3
  > EOF
  $ cat <<'EOF' > notes.txt
  > f6ed this change is great!
  > e834 this is buggy :(
  > 0625 first post
  > bogusnode gives no error
  > a ambiguous node gives no error
  > EOF

  $ hg log -qr "extdata(filedata)"
  2:f6ed99a58333
  3:9de260b1e88e
  $ hg log -qr "extdata(shelldata)"
  2:f6ed99a58333

test weight of extdata() revset

  $ hg debugrevspec -p optimized "extdata(filedata) & 3"
  * optimized:
  (andsmally
    (func
      (symbol 'extdata')
      (symbol 'filedata'))
    (symbol '3'))
  3

test non-zero exit of shell command

  $ hg log -qr "extdata(emptygrep)"
  abort: extdata command 'cat extdata.txt | grep empty' failed: exited with status 1
  [255]

test bad extdata() revset source

  $ hg log -qr "extdata()"
  hg: parse error: extdata takes at least 1 string argument
  [10]
  $ hg log -qr "extdata(unknown)"
  abort: unknown extdata source 'unknown'
  [255]

test a zero-exiting source that emits garbage to confuse the revset parser

  $ cat > badparse.txt <<'EOF'
  > +---------------------------------------+
  > 9de260b1e88e
  > EOF

It might be nice if this error message mentioned where the bad string
came from (eg line X of extdata source S), but the important thing is
that we don't crash before we can print the parse error.
  $ hg log -qr "extdata(badparse)"
  hg: parse error at 0: not a prefix: +
  (+---------------------------------------+
   ^ here)
  [10]

test template support:

  $ hg log -r:3 -T "{node|short}{if(extdata('notes'), ' # {extdata('notes')}')}\n"
  06254b906311 # first post
  e8342c9a2ed1 # this is buggy :(
  f6ed99a58333 # this change is great!
  9de260b1e88e

test template cache:

  $ hg log -r:3 -T '{rev} "{extdata("notes")}" "{extdata("shelldata")}"\n'
  0 "first post" ""
  1 "this is buggy :(" ""
  2 "this change is great!" "another comment on 2"
  3 "" ""

test bad extdata() template source

  $ hg log -T "{extdata()}\n"
  hg: parse error: extdata expects one argument
  [10]
  $ hg log -T "{extdata('unknown')}\n"
  abort: unknown extdata source 'unknown'
  [255]
  $ hg log -T "{extdata(unknown)}\n"
  hg: parse error: empty data source specified
  (did you mean extdata('unknown')?)
  [10]
  $ hg log -T "{extdata('{unknown}')}\n"
  hg: parse error: empty data source specified
  [10]

we don't fix up relative file URLs, but we do run shell commands in repo root

  $ mkdir sub
  $ cd sub
  $ hg log -qr "extdata(filedata)"
  abort: error: $ENOENT$
  [100]
  $ hg log -qr "extdata(shelldata)"
  2:f6ed99a58333

  $ cd ..