path: root/eshtrans/backend/codegen.esh

# Shell command language code generator
#
# Copyright (C) 2016  Patrick "P. J." McDermott
#
# This file is part of the Eggshell Compiler.
#
# The Eggshell Compiler is free software: you can redistribute it
# and/or modify it under the terms of the GNU General Public License
# as published by the Free Software Foundation, either version 3 of
# the License, or (at your option) any later version.
#
# The Eggshell Compiler is distributed in the hope that it will be
# useful, but WITHOUT ANY WARRANTY; without even the implied warranty
# of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with the Eggshell Compiler.  If not, see
# <http://www.gnu.org/licenses/>.

sbufi=
sbufc=
sc=
tu_id=
static_fn_n=
pragma_nostack=false

sgetc()
{
	if [ ${sbufi} -ge ${sbufc} ]; then
		sc=''
	else
		eval "sc=\${sbufv_${sbufi}}"
		sbufi=$((${sbufi} + 1))
	fi
}

add_static_fn()
{
	local fn="${1}"

	printf '__%s_static_fns="${__%s_static_fns} %s:__%s_fn%d"\n' \
		"${tu_id}" "${tu_id}" "${fn}" "${tu_id}" ${static_fn_n}
	printf '__%s_fn%d' "${tu_id}" ${static_fn_n}
	static_fn_n=$((${static_fn_n} + 1))
}

codegen_sub()
{
	local array="${1}"
	shift 1
	local print_spc=
	local ionum=
	local pragma=
	local fname=
	local static=
	local static_fname=
	local type=
	local params=
	local types=
	local t=

	print_spc=false
	ionum=false
	pragma=false
	fname=''
	static=false
	static_fname=false
	type=false
	params=0
	types=''

	IFS="${RS}"
	for t in ${array}; do
		if ${print_spc}; then
			case "${t%${US}*}" in
				T_NEWLINE)
					;;
				*)
				printf ' '
					;;
			esac
			print_spc=false
		fi

		# Handle I/O number false positives.
		if ${ionum}; then
			case "${t%${US}*}" in
				T_WORD|T_CMDNAME|T_FNAME)
					printf ' '
				;;
			esac
			ionum=false
		fi

		case "${t%${US}*}" in
			T_USE)
				pragma=true
				continue
				;;
			T_FNAME)
				fname="${t}"
				params=0
				types=''
				;;
			T_STATIC)
				static=true
				continue
				;;
			T_TYPE)
				type=true
				params=$((${params} + 1))
				types="${types}${t#*${US}}:"
				continue
				;;
			T_COMMA)
				continue
				;;
		esac
		if ${pragma}; then
			pragma=false
			eval "pragma_${t#*${US}}=true"
			continue
		fi
		# State machine for static variables and functions
		if ${static}; then
			static=false
			case "${t%${US}*}" in
				T_FNAME)
					static_fname=true
					continue
					;;
				*)
					printf '__static '
					;;
			esac
		elif ${static_fname}; then
			static_fname=false
			case "${t%${US}*}" in
				T_LPAREN)
					add_static_fn "${fname#*${US}}"
					;;
				*)
					printf '__static '
					toktext "${fname}"
					;;
			esac
		elif ${type}; then
			type=false
			types="${types}${t#*${US}} "
			continue
		fi

		# Function start/end tokens
		case "${t%${US}*}" in
			T_FN_START)
				if ! ${pragma_nostack}; then
					printf '__fn_start %s %s;' \
						"${tu_id}" "${fname#*${US}}"
				else
					printf ':;'
				fi
				if [ ${params} -gt 0 ]; then
					printf ' __check_args %s %d %s %s;' \
						"${fname#*${US}}" ${params} \
						"'${types% }'" '"${@}"'
				fi
				;;
			T_FN_END)
				if ! ${pragma_nostack}; then
					printf '__fn_end;'
				else
					printf ':;'
				fi
				;;
			T_RETURN)
				if ! ${pragma_nostack}; then
					printf '__fn_end; '
				fi
				;;
			T_STATIC)
				printf '__static '
				continue
				;;
			T_LOCAL)
				printf '__local '
				continue
				;;
		esac

		toktext "${t}"

		case "${t%${US}*}" in
			T_NEWLINE|T_LPAREN)
				# Indenting lines can mess up here-documents.
				# Adding a space after "(" can mess up function
				# definitions (on zsh at least, while other
				# shells accept the space).
				;;
			T_IO_NUMBER)
				ionum=true
				;;
			*)
				print_spc=true
				;;
		esac
	done
	unset IFS
}

# The token stack is encoded in a string in the following grammar:
#     Terminal symbols:
#         TOKEN
#     Production rules:
#         stack  = tokens [ '<SOH>' type '<STX>' stack '<ETX>' [ tokens ] ] ;
#         tokens = TOKEN { '<RS>' TOKEN } ;
#         type   = 'C' ;
# We need to recurse through this stack to get to all the tokens.
# Each element in the stack (an array of tokens) gets run through the codegen to
# become text that is inserted into the array below.
parse_stack()
{
	local array=
	local res=

	array=''
	while :; do
		sgetc
		case "${sc}" in
			'')
				# EOF
				break
				;;
			"${SOH}")
				# New stack element
				sgetc
				case "${sc}" in
					'C')
						# Command substitution
						sgetc  # STX
						res="$(parse_stack)"
						sbufi="${res%% *}"
						array="${array}${res#* }"
						;;
				esac
				;;
			"${ETX}")
				# End of stack element
				break
				;;
			*)
				# Token character
				array="${array}${sc}"
				;;
		esac
	done
	printf '%d ' ${sbufi}
	codegen_sub "${array}"
}

sh_parse_stack()
{
	local buf="${1}"
	shift 1
	local res=

	eval "$(printf '%s' "${buf}" | awk -v FS='' -v j=0 \
		-v squote="'" -v esc_squote="'\\\\''" '
		{
			for (i = 1; i <= NF; ++i) {
				sub(squote, esc_squote, $i);
				printf("sbufv_%d='\''%s'\''\n", j++, $i);
			};
			printf("sbufv_%d='\''\n'\''\n", j++);
		}
		')"
	sbufi=0
	sbufc=${#buf}
	res="$(parse_stack; printf '.')"
	res="${res%.}"
	printf '%s' "${res#* }"
}

sh_set_tu_id()
{
	local toks="${1}"

	tu_id="$(printf '%s' "${toks}" | sha256sum -)"
	tu_id="${tu_id%  -}"

	return 0
}

sh_start_tu()
{
	printf '__tu=%s\n' "${tu_id}"
	printf '__%s_static_vars=\n' "${tu_id}"
	printf '__%s_static_var_vals=\n' "${tu_id}"
	printf '__%s_static_fns=\n' "${tu_id}"
	static_fn_n=0
}

sh_end_tu()
{
	printf '__tu_end\n'
}