feat(parser): Rework completion

1 files changed, 160 insertions, 164 deletions

diff --git a/md-parser.awk b/md-parser.awk
index 7f3519b..2161482 100644
--- a/md-parser.awk
+++ b/md-parser.awk
@@ -61,16 +61,51 @@ BEGIN {
 
 #== Block markup ==#
 
-# Inline HTML
-/^[ \t]?[ \t]?[ \t]?<\/?[0-9A-Za-z]+/ {
-	# if it isn't inside a list
-	if (indentation() - 2 * list_level < 4) {
-		STATE_set(STATE_HTML)
-		soft_next()
+#-- Blockquotes --#
+blockquote_level > 0 && empty0 {
+	STATE_unset(STATE_PARAGRAPH_P1)
+	if (STATE_thenunset(STATE_CODEBLOCK)) {
+		PNappnd("</code></pre>")
+	}
+	for (P1 = P1 ORS; blockquote_level > 0; --blockquote_level) {
+		P1 = P1 "</blockquote>"
+	}
+}
+/^>[ \t]/ {
+	STATE_set(STATE_BLOCKQUOTE)
+
+	match($0, /^[> \t]*> /)
+	blevel = substr($0, RSTART, RLENGTH)
+
+	$0 = substr($0, RSTART + RLENGTH)
+	gsub(/[ \t]/, "", blevel)
+	blevel = length(blevel)
+
+	for (;blockquote_level < blevel; ++blockquote_level) {
+		P1 = P1 "<blockquote>"
+	}
+	for (;blockquote_level > blevel; --blockquote_level) {
+		P2 = P2 "</blockquote>"
 	}
 }
+STATE_isset(STATE_BLOCKQUOTE) && $0 == ">" {
+	$0 = ""
+	empty0 = TRUE
+}
+
+#-- Remove list indentation --#
+list_level > 0 && !empty0 && !is_blank() {
+	level = int((indentation() + 2) / 2)
+	sub("^[ \t]{1," lists_shift[list_level] "}", "")
+}
+
+#-- Inline HTML --#
+!STATE_isset(STATE_CODEBLOCK) && /^[ \t]{0,3}<\/?[0-9A-Za-z]+/ {
+	STATE_set(STATE_HTML)
+	soft_next()
+}
 STATE_isset(STATE_HTML) {
-	if (/^[ \t]*$/) {
+	if (is_blank()) {
 		STATE_unset(STATE_HTML)
 	}
 	else {
@@ -78,101 +113,88 @@ STATE_isset(STATE_HTML) {
 	}
 }
 
-# Fenced code blocks
+#-- Fenced code blocks --#
 STATE_isset(STATE_FENCED_CODEBLOCK) {
-	if (fenced_class) {
-		$0 = fenced_class $0
-		fenced_class = ""
+	if (fenced_start) {
+		$0 = fenced_start $0
+		fenced_start = ""
 		soft_next()
 	}
-	$0 = "    " $0
+	if (/^```/) {
+		PNappnd("</code></pre>")
+		STATE_unset(STATE_CODEBLOCK)
+		STATE_unset(STATE_FENCED_CODEBLOCK)
+		next
+	}
 }
 
-# Code blocks (ending condition)
+#-- Code blocks --#
 STATE_isset(STATE_CODEBLOCK) {
-	# If ended codeblock
-	if ((STATE_isset(STATE_FENCED_CODEBLOCK) && /    ```$/) ||
-	    ! /^(    |\t|[ \t]*$)/ ||
-	    (blockquote_level > 0 && empty0))
-	{
+	if (!empty0 && !STATE_isset(STATE_FENCED_CODEBLOCK) && ! /^(    |\t)/) {
 		PNappnd("</code></pre>")
 		STATE_unset(STATE_CODEBLOCK)
-		emptyP1 = 1
-		if (STATE_thenunset(STATE_FENCED_CODEBLOCK)) {
-			next
-		}
 	}
-	else if (!STATE_isset(STATE_BLOCKQUOTE)) {
-		handle_codeblock()
+	else {
+		escape_in_codeblock()
 		soft_next()
 	}
 }
 
-# Blockquotes
-blockquote_level > 0 && empty0 {
-	STATE_unset(STATE_PARAGRAPH_P1)
-	for (P1 = P1 ORS; blockquote_level > 0; --blockquote_level) {
-		P1 = P1 "</blockquote>"
-	}
-}
-/^>[ \t]/ {
-	match($0, /^[> \t]*> /)
-	level = substr($0, RSTART, RLENGTH)
-
-	$0 = substr($0, RSTART + RLENGTH)
-	gsub(/[ \t]/, "", level)
-	level = length(level)
-
-	STATE_set(STATE_BLOCKQUOTE)
-	for (;blockquote_level < level; ++blockquote_level) {
-		P1 = P1 "<blockquote>"
-	}
-	for (;blockquote_level > level; --blockquote_level) {
-		P2 = P2 "</blockquote>"
-	}
-}
-STATE_isset(STATE_BLOCKQUOTE) && $0 == ">" { $0 = "" }
-
-# Code block (end)
-STATE_isset(STATE_CODEBLOCK) { ## when in blockquote
-	handle_codeblock()
-	soft_next()
-}
-
-# Link definition
-/^\[[^\]]+\]:[ \t]+.*/ { # line starts link definition
+#-- Link definition (ignore) --#
+/^\[[^\]]+\]:[ \t]+.*/ {
 	next
 }
 
-# Escapes
+#-- Escapes --#
 /\\[-\\`*_{}\[\]()#+.!^~]/ { # has escaped char
 	while (match($0, /\\[-\\`*_{}\[\]()#+.!^~]/)) {
 		sub_match(escapes[char($0, RSTART + 1)])
 	}
 }
 
-# Horizontal rule
-/^(---+|- - -( -)*|\*\*\*+|\* \* \*( \*)*|___+|_ _ _( _)*)$/ {
+#-- Horizontal rule --#
+/^(---+|- -( -)+|\*\*\*+|\* \*( \*)+|___+|_ _( _)+)$/ {
 	STATE_unset(STATE_PARAGRAPH)
 	$0 = "<hr/>"
 }
 
-# Lists
-function end_list() {
-	STATE_unset(STATE_LIST)
+#-- Lists --#
+/^[ \t]*[*+-] / { ## Unordered list
+	level = level + int(indentation() / 2)
 
-	if (list_level == 0) return
-	PNappnd("</li></" (lists[list_level] == "." ? "o" : "u") "l>")
-	--list_level
+	match($0, /^[ \t]*[*+-] /)
+	type  = char($0, RSTART + RLENGTH - 2)
+	shift = RLENGTH
+	sub_match("")
+
+	start_list(level, type, shift)
+	level = 0
 }
+/^[ \t]*[0-9]+\. / { ## Ordered list
+	level = level + int(indentation() / 2)
+
+	type  = "."
+	match($0, /^[^.]+\. /)
+	shift = RLENGTH
+	sub_match("")
+
+	start_list(level, type, shift)
+	level = 0
+}
+level > 0 && list_level > 0 && !is_blank() && emptyP1 { ## Non-list line
+	while (list_level >= level) {
+		end_list()
+	}
+}
+list_level > 0 {
+	STATE_unset(STATE_PARAGRAPH)
+}
+
 function start_list(level, type, shift) {
 	STATE_set(STATE_LIST)
-	brand_new = TRUE
 
 	## Every two columns we define a new level
 	level = int((level + 2) / 2)
-	## If list definition is the first line inside a code block
-	if (level - list_level >= 3) return
 
 	## Make list_level = level - 1
 	while (list_level < level - 1) {
@@ -197,85 +219,55 @@ function start_list(level, type, shift) {
 		P1 = P1 ORS "<li>"
 	}
 }
-/^[ \t]*[*+-] / { ## Unordered list
-	level = indentation()
-
-	match($0, /^[^*+-]*[*+-] /)
-	type  = char($0, RSTART + RLENGTH - 2)
-	shift = RLENGTH
-	sub_match("")
-
-	start_list(level, type, shift)
-}
-/^[ \t]*[0-9]+\. / { ## Ordered list
-	level = indentation()
-
-	type  = "."
-	match($0, /^[^.]+\. /)
-	shift = RLENGTH
-	sub_match("")
-
-	start_list(level, type, shift)
-}
-!brand_new && list_level > 0 && /[^ \t]/ { ## Non-list line
-	level = int((indentation() + 2) / 2)
-	sub("^[ \t]{1," lists_shift[list_level] "}", "")
+function end_list() {
+	STATE_unset(STATE_LIST)
 
-	while (list_level >= level && emptyP1) {
-		end_list()
-	}
-}
-list_level > 0 {
-	STATE_unset(STATE_PARAGRAPH)
-	brand_new = FALSE
+	if (list_level == 0) return
+	PNappnd("</li></" (lists[list_level] == "." ? "o" : "u") "l>")
+	--list_level
 }
 
-# Checklists
-STATE_isset(STATE_LIST) && /^\[[x ]\][ \t]/ {
+#-- Checklists --#
+STATE_isset(STATE_LIST) && level == 0 && /^\[[x ]\][ \t]/ {
 	checked = (char($0, 2) == "x") ? " checked=\"checked\"" : "";
 	$0 = "<input type=\"checkbox\" disabled=\"disabled\"" checked \
 	     "><label>" substr($0, 5) "</label>"
 }
 
-# Fenced code blocks
+#-- Fenced code blocks (new) --#
 /^[ \t]*```/ {
 	STATE_set(STATE_FENCED_CODEBLOCK)
 	STATE_set(STATE_CODEBLOCK)
 
 	match(P1, /```[^ \t]*/)
 	if (RLENGTH > 3) {
-		fenced_class = " class=\"language-" tolower(substr(P1, RSTART + 3, RLENGTH - 3)) "\""
+		fenced_start = " class=\"language-" tolower(substr(P1, RSTART + 3, RLENGTH - 3)) "\""
 	}
 	else {
-		fenced_class = ""
+		fenced_start = ""
 	}
 
-	fenced_class = "<pre><code" fenced_class ">"
+	fenced_start = "<pre><code" fenced_start ">"
 	next
 }
 
-# Code blocks
-function handle_codeblock() {
-	## Autoescape & and <
+#-- Code blocks (new) --#
+/^(    |\t)/ && /[^ \t]/ {
+	STATE_set(STATE_CODEBLOCK)
+	escape_in_codeblock()
+	$0 = "<pre><code>" $0
+	soft_next()
+}
+function escape_in_codeblock() {
 	for (i = 1; match(substr($0, i), /&/); i += RSTART) {
 		RSTART += i - 1
 		sub_match("&amp;")
 	}
 	gsub(/</, "\\&lt;")
 	gsub(/>/, "\\&gt;")
-
-	if (STATE_isset(STATE_FENCED_CODEBLOCK)) {
-		$0 = substr($0, 5)
-	}
-}
-/^(    |\t)/ && /[^ \t]/ { ## New codeblock
-	STATE_set(STATE_CODEBLOCK)
-	handle_codeblock()
-	$0 = "<pre><code>" $0
-	soft_next()
 }
 
-# Definition lists
+#-- Definition lists --#
 STATE_isset(STATE_DEFINITION_LIST) {
 	if (!empty0) {
 		STATE_unset(STATE_PARAGRAPH)
@@ -319,7 +311,20 @@ STATE_isset(STATE_DEFINITION_LIST) {
 
 #== Inline markup ==#
 
-# Automatic links
+#-- Autoescape & --#
+/&/ {
+	for (i = 1; match(substr($0, i), /&/); i += RSTART) {
+		RSTART += i - 1
+		if (substr($0, RSTART) ~ /&[A-Za-z0-9#]+;/) { continue }
+		sub_match("&amp;")
+	}
+}
+#-- Autoescape < --#
+/ < / {
+	gsub(/ < /, "\\&lt;")
+}
+
+#-- Automatic links --#
 /<[^ \t<>]+>/ {
 	## URL: <uri_scheme:...>
 	while (match($0, /<[A-Za-z][-A-Za-z0-9+.]*:[^ \t<>]+>/)) {
@@ -343,25 +348,12 @@ STATE_isset(STATE_DEFINITION_LIST) {
 	}
 }
 
-# Autoescape &
-/&/ {
-	for (i = 1; match(substr($0, i), /&/); i += RSTART) {
-		RSTART += i - 1
-		if (substr($0, RSTART) ~ /&[A-Za-z0-9#]+;/) { continue }
-		sub_match("&amp;")
-	}
-}
-# Autoescape <
-/ < / {
-	gsub(/ < /, "\\&lt;")
-}
-
-# Line break
+#-- Line break --#
 /  $/ {
 	sub(/[ \t]*  $/, "</br>")
 }
 
-# atx-style headers
+#-- atx-style headers --#
 /^[ \t]*#+/ {
 	match($0, /#+/)
 	if (RLENGTH <= 6) {
@@ -374,7 +366,7 @@ STATE_isset(STATE_DEFINITION_LIST) {
 	}
 }
 
-# setex-style headers
+#-- setex-style headers --#
 !emptyP1 && /^[ \t]*(=|-)+[ \t]*$/ {
 	hlevel = 1 + ($0 ~ /-/)
 	P1 = "<h" hlevel ">" substr(P1, 4) "</h" hlevel ">"
@@ -382,7 +374,25 @@ STATE_isset(STATE_DEFINITION_LIST) {
 	STATE_unset(STATE_PARAGRAPH)
 }
 
-# Emphasis
+#-- Emphasis --#
+{
+	## Make sure <em> is on the inside for long chains
+	## E.g. *****a***** will be the same as ****_a_****
+	surround("___"      , "<strong><em>", "</em></strong>")
+	surround("\\*\\*\\*", "<strong><em>", "</em></strong>", 3)
+
+	gsurround("__"    , "<strong>", "</strong>")
+	gsurround("_"     , "<em>"    , "</em>")
+	gsurround("\\*\\*", "<strong>", "</strong>", 2)
+	gsurround("\\*"   , "<em>"    , "</em>", 1)
+	gsurround("``"    , "<code>"  , "</code>")
+	gsurround("`"     , "<code>"  , "</code>")
+
+	gsurround("~~" , "<s>"     , "</s>")
+	gsurround("~"  , "<sub>"   , "</sub>")
+	gsurround("\\^", "<sup>"   , "</sup>", 1)
+}
+
 function surround(brace, left, right, blen) {
 	if (blen == "") { blen = length(brace) }
 	if (match($0, brace (brace == "``" ? "." : "[^" brace "]") "+" brace)) {
@@ -405,31 +415,13 @@ function gsurround(brace, left, right, blen) {
 	}
 }
 
-{
-	## Make sure <em> is on the inside for long chains
-	## E.g. *****a***** will be the same as ****_a_****
-	surround("___"      , "<strong><em>", "</em></strong>")
-	surround("\\*\\*\\*", "<strong><em>", "</em></strong>", 3)
-
-	gsurround("__"    , "<strong>", "</strong>")
-	gsurround("_"     , "<em>"    , "</em>")
-	gsurround("\\*\\*", "<strong>", "</strong>", 2)
-	gsurround("\\*"   , "<em>"    , "</em>", 1)
-	gsurround("``"    , "<code>"  , "</code>")
-	gsurround("`"     , "<code>"  , "</code>")
-
-	gsurround("~~" , "<s>"     , "</s>")
-	gsurround("~"  , "<sub>"   , "</sub>")
-	gsurround("\\^", "<sup>"   , "</sup>", 1)
-}
-
-# Footnote references
+#-- Footnote references --#
 /\[\^[^\]]+\]/ {
 	while (match($0, /\[\^[\]]+\]/)) {
 	}
 }
 
-# Links
+#-- Links --#
 /!?\[[^\]]*\] ?[(\[][^)\]]*[\])]/ {
 	while (match($0, /!?\[[^\]]*\] ?[(\[][^)\]]*[\])]/)) {
 		link = substr($0, RSTART, RLENGTH)
@@ -468,11 +460,11 @@ function gsurround(brace, left, right, blen) {
 	}
 }
 
-# Paragraph
+#-- Paragraph --#
 emptyP1 && !empty0 && STATE_isset(STATE_PARAGRAPH) { $0 = "<p>" $0 }
 !emptyP1 && empty0 && STATE_isset(STATE_PARAGRAPH_P1) { P1 = P1 "</p>" }
 
-# After all $0 parsing
+#== Buffered printing ==#
 {
 	__buffered_print()
 }
@@ -538,6 +530,7 @@ END {
 }
 
 #== Boolean "constants" ==#
+
 BEGIN {
 	FALSE = 0
 	TRUE = 1
@@ -599,17 +592,25 @@ function indentation(    len) {
 	return len
 }
 
+function is_blank() {
+	return ! /[^ \t]/
+}
+
 #== Match handling ==#
 
 function save_match() {
 	backup_rstart = RSTART
 	backup_rlength = RLENGTH
 }
+
 function restore_match() {
 	RSTART = backup_rstart
 	RLENGTH = backup_rlength
 }
 
+function sub_match(replacement) {
+	$0 = substr($0, 1, RSTART - 1) replacement substr($0, RSTART + RLENGTH)
+}
 
 #== Other ==#
 
@@ -629,14 +630,9 @@ function soft_next() {
 	next
 }
 
-function sub_match(replacement) {
-	$0 = substr($0, 1, RSTART - 1) replacement substr($0, RSTART + RLENGTH)
-}
-
 function char(string, ind) {
 	if (1 > ind || ind > length(string)) {
 		return "";
 	}
 	return substr(string, ind, 1)
 }
-