tcsh syntax highlighting partially broken

[vvzen]

Summary

Hi all!
And thanks for this beatiful project. I absolutely love it!

While editing a tcsh file with hx, I noticed that the syntax highlighting was having troubles with an unescaped ' character.
This is what I see in hx:

This is what I see in neovim, instead:

I suppose that's because hx doesn't support tcsh natively?
If so, I'd like to contribute to it but I don't know where to start.
Thanks!

PS: don't ask why I still need to use tcsh in 2022, I'm forced to since it's what we have at work :|

Platform

CentOS Linux release 7.9.2009 (Core)

Terminal Emulator

iterm2

Helix Version

helix 22.03

[the-mikedavis]

I believe this was fixed in #1917

Can you try updating to 22.05?

[vvzen]

ah, great! will give it a go and let you know, thanks!

[vvzen]

tried in 22.05 - same behaviour. I think that it's just that tcsh isn't supported, right now?

$ hx --version
helix 22.05
$ hx --health tcsh
Language 'tcsh' not found
Did you mean one of these: toml, typescript, tsx, twig, tsq, tablegen, tfvars 

[the-mikedavis]

There isn't any support at the moment but it might be possible to use tree-sitter-bash for it like we do with zsh. Could you post the entire file from the screenshot or a minimal version? Is there a shebang at the top that's selecting the language?

[vvzen]

It does have a shebang, yes. This is another.tcsh file, but the gist is the same:

#!/bin/tcsh -e

set python_version_no_dots=`echo ${CM_PYTHON_VERSION} | tr -d "."`
set modulename="third-party-libs-devel-python-${python_version_no_dots}"
printf "Loading $modulename module.."
module load "$modulename"
printf "..done!\n"

# Load custom python
printf "Loading custom python/$CM_PYTHON_VERSION module.."
module load "/net/global/opt/pipeline/python-payloads/modules/python/$CM_PYTHON_VERSION"
printf "..done!\n"

module list
printf "\n"

echo "Current PATH:"
echo "$PATH" | tr ":" "\n"

echo "Current PYTHONPATH:"
echo "$PYTHONPATH" | tr ":" "\n"

echo "Current PL_CONFIG_PATH:"
echo $PL_CONFIG_PATH | tr ":" "\n"

set python="python${CM_PYTHON_VERSION}" 

printf "Running default Unit Test suite for Python $CM_PYTHON_VERSION with pytest..\n"
"$python" -m pytest "./$CM_UNIT_TESTS_DIR" -m "$CM_PYTEST_MARKS" \
    -vvv \
    --cov=$CM_PACKAGE_NAME \
    --junitxml=$CM_JUNIT_REPORT_NAME

# See https://docs.pytest.org/en/latest/reference/exit-codes.html
set code = $?
if $code != 0 then
    printf "Unit Tests didn't pass!\n"
    exit 1
endif
printf "..done!\n"

printf "Running additional custom tests..\n"

echo "--> Running tests using $python.."
"$python" ./ci-scripts/run_tests.py
set code = $?
if $code != 0 then
    echo "Additional unit tests didn't pass.\n"
    exit 1
endif

# This is used by https://docs.gitlab.com/13.12/ee/user/project/merge_requests/test_coverage_visualization.html
echo "--> Running coverage report.."
coverage report
coverage xml

echo "All done!"

[the-mikedavis]

It looks like the

if $code != 0 then

line is what puts the parser into error mode. I don't write that much shell but I think bash needs brackets around if conditions and then a newline or ; between the condition and the then, right? So I think we can't use tree-sitter-bash for tcsh

[lebakasable]

https://github.com/hyperupcall/tree-sitter-tcsh

[victororlyk]

I have the same problem with next script.sh file

#!/usr/bin/env bash
set -x
set -eo pipefail
if ! [ -x "$(command -v psql)" ]; then
  echo >&2 "Error: psql is not installed."
  exit 1
fi
if ! [ -x "$(command -v sqlx)" ]; then
  echo >&2 "Error: sqlx is not installed."
  echo >&2 "Use:"
  echo >&2 "    cargo install --version=0.5.7 sqlx-cli --no-default-features --features postgres"
  echo >&2 "to install it."
  exit 1
fi
DB_USER=${POSTGRES_USER:=postgres}
DB_PASSWORD="${POSTGRES_PASSWORD:=password}"
DB_NAME="${POSTGRES_DB:=newsletter}"
DB_PORT="${POSTGRES_PORT:=5432}"

# Allow to skip Docker if a dockerized Postgres database is already running
if [[ -z "${SKIP_DOCKER}" ]]
then
  docker run \
      -e POSTGRES_USER=${DB_USER} \
      -e POSTGRES_PASSWORD=${DB_PASSWORD} \
      -e POSTGRES_DB=${DB_NAME} \
      -p "${DB_PORT}":5432 \
      -d postgres \
      postgres -N 1000
fi
export PGPASSWORD="${DB_PASSWORD}"
until psql -h "localhost" -U "${DB_USER}" -p "${DB_PORT}" -d "postgres" -c '\q'; do
  >&2 echo "Postgres is still unavailable - sleeping"
sleep 1 done

>&2 echo "Postgres is up and running on port ${DB_PORT} - running migrations now!"

export DATABASE_URL=postgres://${DB_USER}:${DB_PASSWORD}@localhost:${DB_PORT}/${DB_NAME}
sqlx database create
sqlx migrate run

>&2 echo "Postgres has been migrated, ready to go!"