From 4faa985e6953b2c6d1cb94148daf12780b84a811 Mon Sep 17 00:00:00 2001
From: =?utf8?q?Paul=20H=C3=A4nsch?= <paul@plutz.net>
Date: Wed, 21 Jun 2023 19:50:26 +0200
Subject: [PATCH] Squashed 'cgilite/' changes from 8ce6dce8..d45722a7

d45722a7 Omit link IDs in nested headings, to prevent ambiguous IDs
e6442fb6 EXPERIMENTAL: <section> wrapping for Headlines
d0b1c70c bugfix: HEX_DECODE for prefixes looking like shell patterns

git-subtree-dir: cgilite
git-subtree-split: d45722a72e37c6b214877a81d53b22bfb422a900
---
 cgilite.sh   |  2 +-
 markdown.awk | 87 ++++++++++++++++++++++++++++++++++++----------------
 2 files changed, 61 insertions(+), 28 deletions(-)

diff --git a/cgilite.sh b/cgilite.sh
index 76c91b2..23f5a1c 100755
--- a/cgilite.sh
+++ b/cgilite.sh
@@ -92,7 +92,7 @@ HEX_DECODE(){
     esac
 
     case $in in
-      "$pfx"[0-9a-fA-F][0-9a-fA-F]*) in="${in#${pfx}}";;
+      "$pfx"[0-9a-fA-F][0-9a-fA-F]*) in="${in#"${pfx}"}";;
       \\*) in="${in#?}"; out="${out}\\\\"; continue;;
        %*) in="${in#?}"; out="${out}%%";  continue;;
         *) att="${in%%"${pfx}"*}"; att="${att%%%*}"; att="${att%%\\*}"
diff --git a/markdown.awk b/markdown.awk
index 3fa248d..29dc64d 100755
--- a/markdown.awk
+++ b/markdown.awk
@@ -38,6 +38,7 @@
 #
 # Extensions - Block elements:
 # ----------------------------
+# - [x] Automatic <section>-wrapping (custom)
 # -  ?  Heading identifiers (php md, pandoc)
 # - [x] Automatic heading identifiers (custom)
 # - [x] Fenced code blocks (php md, pandoc)
@@ -306,7 +307,23 @@ function inline( line, LOCAL, len, code, href, guard ) {
   }
 }
 
-function _block( block, LOCAL, st, len, hlvl, htxt, guard, code, indent, attrib ) {
+# Nested Block, resets heading counters
+function _nblock( block, LOCAL, hlsav, sec ) {
+  # Keeping arrays in a local scope is awkward, so we serialize the HL array
+  # into the scalar hlsav
+  hlsav = HL[1] " " HL[2] " " HL[3] " " HL[4] " " HL[5] " " HL[6];
+  HL[1] = 0; HL[2] = 0; HL[3] = 0; HL[4] = 0; HL[5] = 0; HL[6] = 0;
+
+  # Block Level
+  blvl++;
+
+  block = _block( block );
+  sec = ""; for ( n = 1; n <= 6; n++ ) { sec = sec (HL[n]?"</section>":""); }
+  split(hlsav, HL); blvl--;
+  return block sec;
+}
+
+function _block( block, LOCAL, st, len, hlvl, htxt, sec, guard, code, indent, attrib ) {
   gsub( /^\n+|\n+$/, "", block );
 
   if ( block == "" ) {
@@ -344,7 +361,7 @@ function _block( block, LOCAL, st, len, hlvl, htxt, guard, code, indent, attrib
   } else if ( match( block, /^> /) ) {
     match( block, /(^|\n)[[:space:]]*(\n|$)/ ) || match(block, /$/);
     len = RLENGTH; st = RSTART;
-    return "<blockquote>\n" _block( gensub( /(^|\n)> /, "\n", "g", substr(block, 1, st - 1) ) ) "</blockquote>\n\n" \
+    return "<blockquote>\n" _nblock( gensub( /(^|\n)> /, "\n", "g", substr(block, 1, st - 1) ) ) "</blockquote>\n\n" \
            _block( substr(block, st + len) );
 
   # Pipe Tables (pandoc / php md / gfm )
@@ -447,7 +464,7 @@ function _block( block, LOCAL, st, len, hlvl, htxt, guard, code, indent, attrib
 
       ttext = "<thead>\n<tr>"
       for (cnt = 1; cnt <= cols; cnt++)
-        ttext = ttext "<th align=\"" talign[cnt] "\">" _block(tarray[cnt]) "</th>"
+        ttext = ttext "<th align=\"" talign[cnt] "\">" _nblock(tarray[cnt]) "</th>"
       ttext = ttext "</tr>\n</thead>"
     }
 
@@ -470,10 +487,10 @@ function _block( block, LOCAL, st, len, hlvl, htxt, guard, code, indent, attrib
 
       ttext = ttext "<tr>"
       for (cnt = 1; cnt <= cols; cnt++)
-        ttext = ttext "<td align=\"" talign[cnt] "\">" _block(tarray[cnt]) "</td>"
+        ttext = ttext "<td align=\"" talign[cnt] "\">" _nblock(tarray[cnt]) "</td>"
       ttext = ttext "</tr>\n"
     }
-    return "<table>" ttext "</tbody></table>\n" _block(block);
+    return "<table>" ttext "</tbody></table>\n" _nblock(block);
 
   # Line Blocks (pandoc)
   } else if ( match(block, /^\| [^\n]*(\n|$)(\| [^\n]*(\n|$)|[ \t]+[^\n[:space:]][^\n]*(\n|$))*/) ) {
@@ -503,7 +520,7 @@ function _block( block, LOCAL, st, len, hlvl, htxt, guard, code, indent, attrib
     gsub(/(^ | $)/, "", attrib);
     if ( match(code, "(^|\n)" guard "+(\n|$)" ) ) {
       len = RLENGTH; st = RSTART;
-      return "<div class=\"" attrib "\">" _block( substr(code, 1, st - 1) ) "</div>\n" \
+      return "<div class=\"" attrib "\">" _nblock( substr(code, 1, st - 1) ) "</div>\n" \
              _block( substr( code, st + len ) );
     } else {
       match( block, /(^|\n)[[:space:]]*(\n|$)/ ) || match( block, /$/ );
@@ -554,35 +571,50 @@ function _block( block, LOCAL, st, len, hlvl, htxt, guard, code, indent, attrib
   gsub("(^|\n) {0," indent "}", "\n", list);
   return "\n<ol>\n" _list( substr(list, 2) ) "</ol>\n" _block( block );
 
-  # First Order Heading
+  # First Order Heading H1
   } else if ( match( block, /^[^\n]+\n===+(\n|$)/ ) ) {
-    len = RLENGTH;
+    len = RLENGTH; sec = "";
+
+    for ( n = 1; n <= 6; n++ ) { sec = sec (HL[n]?"</section>":""); }
     HL[1]++; HL[2] = 0; HL[3] = 0; HL[4] = 0; HL[5] = 0; HL[6] = 0;
-    return "<h1 id=\"" HL[1] ":" URL(gensub( /\n.*$/, "", "g", block )) "\">" \
-           inline( gensub( /\n.*$/, "", "g", block ) ) \
-           "<a class=\"anchor\" href=\"#" HL[1] ":" \
-           URL(gensub( /\n.*$/, "", "g", block )) "\"></a></h1>\n\n" \
+    hid = HL[1] ":" URL(gensub( /\n.*$/, "", "g", block ));
+
+    return sec "<section class=\"h1\"" ((blvl > 1)?"":" id=\"" hid "\"") ">" \
+           "<h1>" inline( gensub( /\n.*$/, "", "g", block ) ) \
+           ((blvl > 1)?"":"<a class=\"anchor\" href=\"#" hid "\"></a>") \
+           "</h1>\n\n" \
            _block( substr( block, len + 1 ) );
 
-  # Second Order Heading
+  # Second Order Heading H2
   } else if ( match( block, /^[^\n]+\n---+(\n|$)/ ) ) {
-    len = RLENGTH;
+    len = RLENGTH; sec = "";
+
+    for ( n = 2; n <= 6; n++ ) { sec = sec (HL[n]?"</section>":""); }
     HL[2]++; HL[3] = 0; HL[4] = 0; HL[5] = 0; HL[6] = 0;
-    return "<h2 id=\"" HL[1] "." HL[2] ":" URL(gensub( /\n.*$/, "", "g", block )) "\">" \
-           inline( gensub( /\n.*$/, "", "g", block ) ) \
-           "<a class=\"anchor\" href=\"#" HL[1] "." HL[2] ":" \
-           URL(gensub( /\n.*$/, "", "g", block )) "\"></a></h2>\n\n" \
+    hid= HL[1] "." HL[2] ":" URL(gensub( /\n.*$/, "", "g", block ));
+
+    return sec "<section class=\"h2\"" ((blvl > 1)?"":" id=\"" hid "\"") ">" \
+           "<h2>" inline( gensub( /\n.*$/, "", "g", block ) ) \
+           ((blvl > 1)?"":"<a class=\"anchor\" href=\"#" hid "\"></a>") \
+           "</h2>\n\n" \
            _block( substr( block, len + 1) );
 
-  # Nth Order Heading
+  # Nth Order Heading H1 H2 H3 H4 H5 H6
   } else if ( match( block, /^#{1,6}[ \t]*[^\n]+([ \t]*#*)(\n|$)/ ) ) {
-    len = RLENGTH;
+    len = RLENGTH; sec = "";
+
     hlvl = length( gensub( /^(#{1,6}).*$/, "\\1", "g", block ) );
     htxt = gensub(/^#{1,6}[ \t]*(([^ \t\n]+|[ \t]+[^ \t\n#]|[ \t]+#+[^\n#])+)([ \t]*#*)(\n.*)?$/, "\\1", 1, block);
-    HL[hlvl]++; for ( n = hlvl + 1; n < 7; n++) { HL[n] = 0;}
+
+    for ( n = hlvl; n <= 6; n++ ) { sec = sec (HL[n]?"</section>":""); }
+    HL[hlvl]++; for ( n = hlvl + 1; n <= 6; n++) { HL[n] = 0;}
     hid = HL[1]; for ( n = 2; n <= hlvl; n++) { hid = hid "." HL[n] ; }
-    return "<h" hlvl " id=\"" hid ":" URL(htxt) "\">" inline( htxt ) \
-           "<a class=\"anchor\" href=\"#" hid "\"></a></h" hlvl ">\n\n" \
+    hid = hid ":" URL(htxt);
+
+    return sec "<section class=\"h" hlvl "\"" ((blvl > 1)?"":" id=\"" hid "\"") ">" \
+           "<h" hlvl ">" inline( htxt ) \
+           ((blvl > 1)?"":"<a class=\"anchor\" href=\"#" hid "\"></a>") \
+           "</h" hlvl ">\n\n" \
            _block( substr( block, len + 1) );
 
   # block images (wrapped in <figure>)
@@ -641,7 +673,7 @@ function _block( block, LOCAL, st, len, hlvl, htxt, guard, code, indent, attrib
   } else if ( AllowMacros && match( block, /^<<(([^>]|>[^>])+)>>(\n|$)/) ) {
     len = RLENGTH;
     text = gensub(/^<<(([^>]|>[^>])+)>>(\n.*)?$/, "\\1", 1, block);
-    return macro(text) _block(substr(block, len + 1));
+    return macro(text) _block(substr(block, len + 1) );
 
   # Split paragraphs
   } else if ( match( block, /(^|\n)[[:space:]]*(\n|$)/) ) {
@@ -676,9 +708,9 @@ function _list( block, last, LOCAL, p) {
   # if this should be a paragraph item
   # either previous item (last) or current item (p) contains blank lines
   if (match(last, /\n[[:space:]]*\n/) || match(p, /\n[[:space:]]*\n/) ) {
-    last = p; p = _block(p);
+    last = p; p = _nblock(p);
   } else {
-    last = p; p = _block(p);
+    last = p; p = _nblock(p);
     sub( /^<p>/, "", p );
     sub( /<\/p>\n/, "", p );
   }
@@ -713,6 +745,7 @@ BEGIN {
   file = ""; rl_href[""] = ""; rl_title[""] = "";
   if (ENVIRON["MD_HTML"] == "true") { AllowHTML = "true"; }
   HL[1] = 0; HL[2] = 0; HL[3] = 0; HL[4] = 0; HL[5] = 0; HL[6] = 0;
+  # hls = "0 0 0 0 0 0";
 
   # Buffering of full file ist necessary, e.g. to find reference links
   while (getline) { file = file $0 "\n"; }
@@ -736,5 +769,5 @@ BEGIN {
   # for (n in rl_href) { debug(n " | " rl_href[n] " | " rl_title[n] ); }
 
   # Run Block Processing -> The Actual Markdown!
-  printf "%s", _block( file );
+  printf "%s", _nblock( file );
 }
-- 
2.39.2