Tests: Include bash_tap.sh for quick Bash tests

test/bash_tap.sh

@@ -0,0 +1,108 @@
+#!/usr/bin/env bash
+# For more information, see https://github.com/wbsch/bash_tap
+# Subject to the MIT License. See LICENSE file or http://opensource.org/licenses/MIT
+# Copyright (c) 2015 Wilhelm Schürmann
+
+function bashtap_on_error {
+    # A command in the parent script failed, interpret this as a test failure.
+    # $bashtap_line contains the last executed line, or an error.
+    echo -n "$bashtap_output"
+    echo "not ok 1 - ${bashtap_line}"
+    bashtap_clean_tmpdir
+}
+
+function bashtap_run_testcase {
+    # Run each line in the parent script up to the first "exit".
+    bashtap_output=""
+    while IFS= read -r bashtap_line && [ "${bashtap_line:0:4}" != "exit" ]; do
+        # Skip shebang.
+        if [ "${bashtap_line:0:2}" == "#!" ]; then
+            continue
+        fi
+
+        # Avoid recursively sourcing this script, and any helper scripts.
+        if [[ "$bashtap_line" =~ ^(\.|source).*(/|[[:blank:]])bash_tap[^/]*$ ]]; then
+            continue
+        fi
+
+        # Include comments as-is.
+        if [ "${bashtap_line:0:1}" == "#" ]; then
+            bashtap_output+="$bashtap_line"
+            bashtap_output+=$'\n'
+            continue
+        fi
+
+        # Run file line by line.
+        if [ ! -z "$bashtap_line" ] && [ "${bashtap_line:0:2}" != "#!" ]; then
+            bashtap_output+="# $ $bashtap_line"
+            bashtap_output+=$'\n'
+            local cmd_output
+            local cmd_ret
+            cmd_output=$(eval "$bashtap_line" 2>&1 | sed 's/^/# >>> /')
+            cmd_ret=$?
+            if [ ! -z "$cmd_output" ]; then
+                bashtap_output+="$cmd_output"
+                bashtap_output+=$'\n'
+            fi
+            if [ "$cmd_ret" -ne 0 ]; then
+                exit $cmd_ret
+            fi
+        fi
+    done <"$bashtap_org_script"
+}
+
+function bashtap_clean_tmpdir {
+    if [ ! -z "$bashtap_tmpdir" ] && [ -d "$bashtap_tmpdir" ]; then
+        cd "$bashtap_org_pwd"
+        rm -rf "$bashtap_tmpdir"
+    fi
+}
+
+function bashtap_get_absolute_path {
+    # NOTE: No actual thought put into this. Might break. Horribly.
+    # Using this instead of readlink/realpath for OSX compatibility.
+    echo $(cd "$(dirname "$1")" && pwd)/$(basename "$1")
+}
+
+
+bashtap_org_pwd=$(pwd)
+bashtap_org_script=$(bashtap_get_absolute_path "$0")
+
+if [ "${0:(-2)}" == ".t" ] || [ "$1" == "-t" ]; then
+    # Make sure any failing commands are caught.
+    set -e
+    set -o pipefail
+
+    # TAP header. Hardcoded number of tests, 1.
+    echo "1..1"
+
+    # Output TAP failure on early exit.
+    trap bashtap_on_error EXIT
+
+    # The different calls to mktemp are necessary for OSX compatibility.
+    bashtap_tmpdir=$(mktemp -d 2>/dev/null || mktemp -d -t 'bash_tap')
+    if [ ! -z "$bashtap_tmpdir" ]; then
+        cd "$bashtap_tmpdir"
+    else
+        bashtap_line="Unable to create temporary directory."
+        exit 1
+    fi
+
+    # Scripts sourced before bash_tap.sh may declare this function.
+    if declare -f bashtap_setup >/dev/null; then
+        bashtap_setup
+    fi
+
+    # Run test file interpreting failing commands as a test failure.
+    bashtap_run_testcase && echo "ok 1"
+
+    # Since we're in a sourced file and just ran the parent script,
+    # exit without running it a second time.
+    trap - EXIT
+    bashtap_clean_tmpdir
+    exit
+else
+    if declare -f bashtap_setup >/dev/null; then
+        bashtap_setup
+    fi
+fi

test/bash_tap_tw.sh

@@ -0,0 +1,63 @@
+#!/usr/bin/env bash
+# This file only contains helper functions for making Taskwarrior testing
+# easier. The magic happens in bash_tap.sh sourced at the end of this file.
+#
+# "task" is a bash function calling "/path/to/compiled/task rc:taskrc"
+# Only local paths are searched, see bash_tap_tw.sh:find_task_binary().
+#
+# "taskrc" is a file set up in bash_tap_tw.sh:setup_taskrc(), and can be
+# appended to or changed as needed.
+#
+# Subject to the MIT License. See LICENSE file or http://opensource.org/licenses/MIT
+# Copyright (c) 2015 Wilhelm Schürmann
+
+function setup_taskrc {
+    # Configuration
+    for i in pending.data completed.data undo.data backlog.data taskrc; do
+       if [ -f "$i" ]; then
+           rm "$i" 2>&1 >/dev/null
+       fi
+    done
+
+    echo 'data.location=.'                 > taskrc
+    echo 'confirmation=off'               >> taskrc
+    echo 'dateformat=m/d/Y'               >> taskrc
+    echo 'color.debug=rgb025'             >> taskrc
+    echo 'color.header=rgb025'            >> taskrc
+    echo 'color.footer=rgb025'            >> taskrc
+    echo 'color.error=bold white on red'  >> taskrc
+}
+
+function find_task_binary {
+    # $bashtap_org_pwd is set in bash_tap.sh. It is the directory the parent script is
+    # run from. Check for the task binary relative to that directory.
+    # Do not use the system "task" if no local one is found, error out instead.
+    for t in "${bashtap_org_pwd}/task" "${bashtap_org_pwd}/src/task" "${bashtap_org_pwd}/../task" "${bashtap_org_pwd}/../src/task" "${bashtap_org_pwd}/../build/src/task"; do
+        if [ -f "$t" ] && [ -x "$t" ]; then
+            t_abs=$(bashtap_get_absolute_path "$t")
+            eval "function task { ${t_abs} rc:taskrc \$@; }"
+            return 0
+        fi
+    done
+
+    echo "# ERROR: Could not find task binary!"
+
+    # Needed for normal, i.e. "non-test" mode.
+    eval "function task { exit; }"
+
+    # Set $line so we can have useful TAP output.
+    line="bash_tap.sh:find_task_binary()"
+
+    return 1
+}
+
+function bashtap_setup {
+    # This function is called by bash_tap.sh before running tests, or before
+    # running the parent script normally.
+    find_task_binary
+    setup_taskrc
+}
+
+
+# Include the base script that does the actual work.
+source bash_tap.sh