Add an option in psql to avoid encoding issues on some platforms

[gfphoenix78]

When running ICW tests, the following command has a strange behavior.

psql ... <<EOF
$(cat prehook inputfile)
EOF

The command has an issue on some platforms that will input unexpected data stream to psql. For example, The following first line will become to the second line:
"колонка 6" int, "колонка 7" int, "колонка 8" int, "колонка 9" int, "колонка 10" int, "колонка 11" int,

"колонка 6" int, "колонка 7" int, "колонка 8" int, "колонка 9" int, "колонЀ�а 10" int, "колонка 11" int,

It's unknown why the input stream is changed. The other similar commands will have correct data stream to psql,
like

cat prehook inputfile | psql ...

or for a single input file

psql ... < inputfile

In this commit, we convert the original command to

psql ... -f prehook -f inputfile --ignore-log-file

The new option --ignore-log-file will ignore the logging prefix for psql and filename, so the program psql will log like no input filename is provided.

fix #482

Contributor's Checklist

Here are some reminders and checklists before/when submitting your pull request, please check them:

• Make sure your Pull Request has a clear title and commit message. You can take git-commit template as a reference.
• Sign the Contributor License Agreement as prompted for your first-time contribution(One-time setup).
• Learn the coding contribution guide, including our code conventions, workflow and more.
• List your communication in the GitHub Issues or Discussions (if has or needed).
• Document changes.
• Add tests for the change
• Pass make installcheck
• Pass make -C src/test installcheck-cbdb-parallel
• Feel free to request cloudberrydb/dev team for review and approval when your PR is ready🥳