(csg) Roughly reworked pipeline with multiple types of scripts HEAD main

Now, rather than having all scripts at once, there are 6 types of script, running at 6 different "stages"

1 files changed, 117 insertions, 106 deletions

diff --git a/csg b/csg
index eb2a7c8..de8b6bd 100755
--- a/csg
+++ b/csg
@@ -65,8 +65,28 @@ URL="$4"
 [ -z "$URL" ] && echo "[Error] <base_url> cannot be an empty string!" \
 	&& csg_help 3
 
-AWKSCRIPTS="$5"
-[ -n "$AWKSCRIPTS" -a ! -d "$AWKSCRIPTS" ] && echo "[Error] $AWKSCRIPTS (<scripts_dir>) isn't a directory or doesn't exist!" \
+LIB_SCRIPT="$5"
+[ -n "$LIB_SCRIPT" -a ! -d "$LIB_SCRIPT" ] && echo "[Error] $LIB_SCRIPT (<scripts_dir>) isn't a directory or doesn't exist!" \
+	&& csg_help 1
+
+PREV_ALL="$6"
+[ -n "$PREV_ALL" -a ! -d "$PREV_ALL" ] && echo "[Error] $PREV_ALL (<scripts_dir>) isn't a directory or doesn't exist!" \
+	&& csg_help 1
+
+PREV_MD="$7"
+[ -n "$PREV_MD" -a ! -d "$PREV_MD" ] && echo "[Error] $PREV_MD (<scripts_dir>) isn't a directory or doesn't exist!" \
+	&& csg_help 1
+
+ON_MD="$8"
+[ -n "$ON_MD" -a ! -d "$ON_MD" ] && echo "[Error] $ON_MD (<scripts_dir>) isn't a directory or doesn't exist!" \
+	&& csg_help 1
+
+ON_HTML="$9"
+[ -n "$ON_HTML" -a ! -d "$ON_HTML" ] && echo "[Error] $ON_HTML (<scripts_dir>) isn't a directory or doesn't exist!" \
+	&& csg_help 1
+
+ON_OTHER="${10}"
+[ -n "$ON_OTHER" -a ! -d "$ON_OTHER" ] && echo "[Error] $ON_OTHER (<scripts_dir>) isn't a directory or doesn't exist!" \
 	&& csg_help 1
 
 ### Preprocessing 3
@@ -89,7 +109,8 @@ if [ -n "$CSG_AWK" ]
 then :
 elif isCommand 'gawk'
 then
-	CSG_AWK='gawk --posix'
+	CSG_AWK='gawk
+		--posix'
 else
 	echo '[Error] Couldn'\''t find an awk implementation!'
 	csg_help 4
@@ -99,10 +120,10 @@ if [ -n "$CSG_MD_TO_HTML" ]
 then :
 elif isCommand 'lowdown'
 then
-	CSG_MD_TO_HTML="lowdown \
-		--html-no-escapehtml \
-		--html-no-skiphtml \
-		--parse-no-metadata \
+	CSG_MD_TO_HTML="lowdown
+		--html-no-escapehtml
+		--html-no-skiphtml
+		--parse-no-metadata
 		--parse-no-autolink"
 else
 	echo '[Error] Couldn'\''t find a Markdown to HTML convertion program!'
@@ -168,89 +189,59 @@ done << EOF
 $(find "$DESTINATION" -type f -name "*index.md" -links +1)
 EOF
 
-if [ -n "$AWKSCRIPTS" ] && [ -n "$(find $AWKSCRIPTS -type f -name "[A-Z]*.awk")" ]
-then
-	overwriteLine "[Stage 1] Combining awk scripts"
+overwriteLine "[Stage 1] Combining awk scripts"
 
-	awk_scr="$($CSG_AWK -v TMP_DIR="$CSG_TMP_DIR" '
-	{
-		# Remove first field
-		$1=""; sub(FS, "")
-		# Associate current file with regex
-		if (awks[$0] == "") {
-			awks[$0] = FILENAME
-			regexes[regexcount++] = $0
-		}
-		else {
-			awks[$0] = awks[$0] " " FILENAME
-		}
+combine_scripts() {
+	rm "$CSG_TMP_DIR/$2" >/dev/null 2>&1
 
-		nextfile
-	}
-	END {
-		# Regexes are outputed in file alphabetic order
-		# Since certain scripts may require specific ordering
-		# between their rules, this is a crude way to allow that
-		for (regi = 0; regi < regexcount; regi++) {
-			print regexes[regi]
-			outf = TMP_DIR "/csg-" (awkcount++) ".awk"
-			print outf
-			system("cat " awks[regexes[regi]] " > " outf)
-		}
-	}
-	' $(find $AWKSCRIPTS -type f -name "[A-Z]*.awk" | sort))"
+	[ -z "$1" ] && echo "" && return
 
-	overwriteLine "[Stage 1] Finding awk libraries"
+	scripts=$(find "$1" -type f -name "*.awk" | sort)
+	[ -z "$scripts" ] && echo "" && return
 
-	scr_libs=""
-	while read -r file
-	do
-		[ -n "$file" ] && scr_libs="$scr_libs -f $file "
-	done << EOF
-$(find $AWKSCRIPTS -type f -name "#*.awk")
-EOF
-fi
+	cat $scripts > "$CSG_TMP_DIR/$2"
+	echo "-f	$CSG_TMP_DIR/$2"
+}
 
+lib_script="$(combine_scripts "$LIB_SCRIPT" "csg-lib_script.awk")"
+on_md="$(combine_scripts "$ON_MD" "csg-on_md.awk")"
+on_html="$(combine_scripts "$ON_HTML" "csg-on_html.awk")"
+on_other="$(combine_scripts "$ON_OTHER" "csg-on_other.awk")"
+prev_all="$(combine_scripts "$PREV_ALL" "csg-prev_all.awk")"
+prev_md="$(combine_scripts "$PREV_MD" "csg-prev_md.awk")"
+
+echo
 endTimer "[Stage 1] Done"
 
 ### Stage 2
 ### Transforming Markdown
 startTimer "[Stage 2] Transforming Markdown to HTML"
 
-transform_file() {
-	local file="$1"
-	local outf="${file%md}html"
-
-	# Gets special information from the file
-	local title descr words
-	{ read -r title; read -r descr; read -r words; } << EOF
-$($CSG_AWK -v titlef="$TITLEF" '
-	title == "" && /^# / {
-		# Remove first field
-		$1 = ""; sub(FS, "")
-		# Output title
-		sub(/.*/, titlef)
-		title = $0
-		next
+# prev_all
+cat << EOF > "$CSG_TMP_DIR/csg-export.awk"
+function export(name, value) {
+	if (name == "" || value == "") {
+		return
 	}
-	stopdesc == 0 && /^[[:alpha:]]/ {
-		startedGathering = 1
-		# Take only anchor text
-		gsub(/\]\([^)]*\)/, "")
-		gsub(/\[/, "")
-		# Append line
-		if (desc == "") desc = $0
-		else desc = desc " " $0
-		# Trim line to 200 characters
-		if (length(desc) > 200) {
-			desc = substr(desc, 1, 197) "..."
-			stopdesc = 1
-		}
+	# if is string
+	if (value + 0 != value) {
+		gsub(/"/, "\\\\\\"", value)
 	}
-	stopdesc == 0 && startedGathering == 1 && /^$/ { stopdesc = 1 }
-	{ words += NF }
-	END { print title ORS desc ORS words }' "$file")
+
+	printf "	-v	" name "=" ((value + 0 == value) ? value : ("\\"" value "\\"")) "	"
+}
 EOF
+exports=$($CSG_AWK \
+          	-f "$CSG_TMP_DIR/csg-export.awk" \
+          	$lib_script $prev_all \
+          	$(find "$DESTINATION" -type f -name "*.md"))
+
+echo '$0 != 0 {print}' > "$CSG_TMP_DIR/csg-print.awk"
+echo 'BEGIN { FILENAME = truefilename } NR == 1 { FILENAME = truefilename }' > "$CSG_TMP_DIR/csg-filename.awk"
+
+transform_file() {
+	local file="$1"
+	local outf="${file%md}html"
 
 	local created edited
 	if [ -f "$CSG_TIMESTAMP_FILE" ]
@@ -265,50 +256,70 @@ $($CSG_AWK -v file="$outf" '
 EOF
 	fi
 
-	# Generate
-	$CSG_MD_TO_HTML "$file" > "$outf"
-
-	# Run custom awk scripts
-	local scripts="$(printf "$awk_scr" | $CSG_AWK -v file="$outf" '
-		capture == 1 { capture = 0; printf " -f " $0; next; }
-		capture == 2 { capture = 0; next; }
-		{ capture = 1 + !match(file, $0) }
-	')"
+	# prev_md
+	local_exports=$($CSG_AWK\
+					-f	"$CSG_TMP_DIR/csg-export.awk"\
+					$lib_script	$prev_md\
+					"$file")
 
+	# on_md and on_html
 	$CSG_AWK \
-		-v title="$title" \
-		-v description="$descr" \
-		-v words="$words" \
 		-v created="$created" \
 		-v edited="$edited" \
-		$scr_libs $scripts -f "$CSG_TMP_DIR/csg-print.awk" \
-		"$outf" > "$outf.tmp"
-	mv "$outf.tmp" "$outf"
+		$exports $local_exports \
+		$lib_script $on_md -f "$CSG_TMP_DIR/csg-print.awk" \
+		"$file" \
+	| $CSG_MD_TO_HTML \
+	| cat "$SOURCE/_header.html" - "$SOURCE/_footer.html" \
+	| $CSG_AWK \
+		-v created="$created" \
+		-v edited="$edited" \
+		$exports $local_exports \
+		-v truefilename="$outf" -f "$CSG_TMP_DIR/csg-filename.awk" \
+		$lib_script $on_html -f "$CSG_TMP_DIR/csg-print.awk" \
+	> "$outf"
 
-	# Create final file
-	$CSG_AWK \
-		-v title="$title" \
-		-v description="$descr" '
-	BEGIN { gsub(/&/, "\\\\&", title) }
-	/\[DESCRIPTION\]/ {
-		gsub(/\[DESCRIPTION\]/, description)
-	}
-	/\[TITLE\]/ {
-		gsub(/\[TITLE\]/, title)
-	}
-	{print}' "$SOURCE/_header.html" > "$file"
-	cat "$outf" "$SOURCE/_footer.html" >> "$file"
+	# For hardlinks
+	cat "$outf" > "$file"
 	mv "$file" "$outf"
 }
 
-echo '$0 != 0 {print}' > "$CSG_TMP_DIR/csg-print.awk"
+tempifs=$IFS
+IFS='
+	'
 
 while read -r file
 do
-	transform_file "$file" "$scripts" &
+	transform_file "$file" &
 done << EOF
 $(find "$DESTINATION" -type f -name "*.md")
 EOF
+
+IFS=$tempifs
+
+if [ -n "$on_other" ]
+then
+	transform_other() {
+		local file="$1"
+
+		$CSG_AWK \
+			$exports $local_exports \
+			$lib_script $on_other -f "$CSG_TMP_DIR/csg-print.awk" \
+			"$file" > temp
+		mv temp "$file"
+	}
+
+	while read -r file
+	do
+		if [ -f "$file" ]
+		then
+			transform_other "$file" &
+		fi
+	done << EOF
+$(find "$DESTINATION" -type f -not -name "*.md")
+EOF
+fi
+
 wait
 
 overwriteLine "[Stage 2] Generating sitemap"