reduce use of non-posix gensub

diff --git a/markdown.awk b/markdown.awk
index 86657adf28c4c347b49557bc77e8c2a9105efb42..eb40210c9bc78c8cacadda40389c73df2e8d6d6b 100755 (executable)
--- a/markdown.awk
+++ b/markdown.awk
@@ -107,7 +107,7 @@ function URL ( text, sharp ) {
   return text;
 }
 
-function inline( line, LOCAL, len, code, href, guard ) {
+function inline( line, LOCAL, len, text, code, href, guard ) {
   nu = "(\\\\\\\\|\\\\[^\\\\]|[^\\\\_]|_[[:alnum:]])*"    # not underline (except when escaped)
   na = "(\\\\\\\\|\\\\[^\\\\]|[^\\\\\\*])*"  # not asterisk (except when escaped)
   ieu =  "_([^_[:space:]]|[^_[:space:]]" nu "[^_[:space:]])_"                 # inner <em> (underline)
@@ -134,7 +134,7 @@ function inline( line, LOCAL, len, code, href, guard ) {
       code = substr( line, len + 1, match( substr(line, len + 1), guard ) - 1)
       len = 2 * length(guard) + length(code)
       #  strip single surrounding white spaces
-      code = gensub( /^ | $/, "", "g" , code)
+      gsub( /^ | $/, "", code)
       #  escape HTML within code span
       gsub( /&/, "\\&amp;", code ); gsub( /</, "\\&lt;", code ); gsub( />/, "\\&gt;", code );
       return "<code>" code "</code>" inline( substr( line, len + 1 ) )
@@ -310,7 +310,7 @@ function inline( line, LOCAL, len, code, href, guard ) {
 }
 
 function headline( hlvl, htxt, attrib, LOCAL, sec, n, HL) {
-  split( gensub( /^(.* )?([0-9]+( [0-9]+){5})$/, "\\2" ,"1", hstack),  HL);
+  match(hstack, /([0-9]+( [0-9]+){5})$/); split( substr(hstack, RSTART),  HL);
 
   for ( n = hlvl; n <= 6; n++ ) { sec = sec (HL[n]?"</section>":""); }
   HL[hlvl]++; for ( n = hlvl + 1; n <= 6; n++) { HL[n] = 0;}
@@ -319,8 +319,8 @@ function headline( hlvl, htxt, attrib, LOCAL, sec, n, HL) {
   hid = hid HL[1]; for ( n = 2; n <= hlvl; n++) { hid = hid "." HL[n] ; }
   hid = hid ":" URL(htxt, 1);
 
-  hstack = gensub( /^(.* )?([0-9]+( [0-9]+){5})$/, "\\1" ,"1", hstack) \
-           HL[1] " " HL[2] " " HL[3] " " HL[4] " " HL[5] " " HL[6];
+  sub(/([0-9]+( [0-9]+){5})$/, "", hstack);
+  hstack = hstack HL[1] " " HL[2] " " HL[3] " " HL[4] " " HL[5] " " HL[6];
 
   return sec "<section class=\"" (attrib ? "h" hlvl " " attrib : "h" hlvl)  "\" id=\"" hid "\">" \
          "<h" hlvl (attrib ? " class=\"" attrib "\"" : "") ">" inline( htxt ) \
@@ -337,7 +337,7 @@ function _nblock( block, LOCAL, sec, n ) {
   for ( n = blvl + 1; n in BL; n++) { delete BL[n]; }
 
   block = _block( block );
-  split( gensub( /^(.* )?([0-9]+( [0-9]+){5})$/, "\\2" ,"1", hstack),  HL);
+  match(hstack, /([0-9]+( [0-9]+){5})$/); split( substr(hstack, RSTART),  HL);
   sec = ""; for ( n = 1; n <= 6; n++ ) { sec = sec (HL[n]?"</section>":""); }
 
   sub("( +[0-9]+){6} *$", "", hstack); blvl--;
@@ -382,8 +382,9 @@ function _block( block, LOCAL, st, len, text, title, attrib, href, guard, code,
   } else if ( match( block, /^> /) ) {
     match( block, /(^|\n)[[:space:]]*(\n|$)/ ) || match(block, /$/);
     len = RLENGTH; st = RSTART;
-    return "<blockquote>" gensub( /^\n|\n$/, "", "g",  _nblock( gensub( /(^|\n)> /, "\n", "g", substr(block, 1, st - 1) ) ) ) "</blockquote>\n\n" \
-           _block( substr(block, st + len) );
+    text = substr(block, 1, st - 1); gsub( /(^|\n)> /, "\n", text );
+    text = _nblock( text ); gsub( /^\n|\n$/, "", text )
+    return "<blockquote>" text "</blockquote>\n\n" _block( substr(block, st + len) );
 
   # Pipe Tables (pandoc / php md / gfm )
   } else if ( match(block, "^((\\|)?([^\n]+\\|)+[^\n]+(\\|)?)\n" \
@@ -516,12 +517,12 @@ function _block( block, LOCAL, st, len, text, title, attrib, href, guard, code,
   # Line Blocks (pandoc)
   } else if ( match(block, /^\| [^\n]*(\n|$)(\| [^\n]*(\n|$)|[ \t]+[^\n[:space:]][^\n]*(\n|$))*/) ) {
     len = RLENGTH; st = RSTART;
-    text = substr(block, 1, len);
-    gsub(/\n[[:space:]]+/, " ", text);
-    gsub(/\n\| /, "\n", text);
-    gsub(/^\| |\n$/, "", text);
-    return "<div class=\"line-block\">" gensub(/\n/, "<br>\n", "g", inline( text )) "</div>\n" \
-           _block( substr( block, len + 1) );
+
+    text = substr(block, 1, len); gsub(/\n[[:space:]]+/, " ", text);
+    gsub(/\n\| /, "\n", text); gsub(/^\| |\n$/, "", text);
+    text = inline(text); gsub(/\n/, "<br>\n", text);
+
+    return "<div class=\"line-block\">" text "</div>\n" _block( substr( block, len + 1) );
 
   # Indented Code Block
   } else if ( match(block, /^(    |\t)( *\t*[^ \t\n]+ *\t*)+(\n|$)((    |\t)[^\n]+(\n|$)|[ \t]*(\n|$))*/) ) {
@@ -570,53 +571,55 @@ function _block( block, LOCAL, st, len, text, title, attrib, href, guard, code,
 
   # First Order Heading H1 + Attrib
   } else if ( match( block, /^([^\n]+)([ \t]*\{([^\}\n]+)\})\n===+(\n|$)/ ) ) {
-    len = RLENGTH;
-    text   = gensub(/^([^\n]+)([ \t]*\{([^\}\n]+)\})\n===+(\n.*)?$/, "\\1", 1, block)
-    attrib = gensub(/^([^\n]+)([ \t]*\{([^\}\n]+)\})\n===+(\n.*)?$/, "\\3", 1, block)
+    len = RLENGTH; text = attrib = block;
+    sub(/([ \t]*\{([^\}\n]+)\})\n===+(\n.*)?$/, "", text);
+    sub(/\}\n===+(\n.*)?$/, "", attrib); sub(/^([^\n]+)[ \t]*\{/, "", attrib);
     gsub(/[^a-zA-Z0-9_-]+/, " ", attrib); gsub(/(^ | $)/, "", attrib);
 
     return headline(1, text, attrib) _block( substr( block, len + 1 ) );
 
   # First Order Heading H1
   } else if ( match( block, /^([^\n]+)\n===+(\n|$)/ ) ) {
-    len = RLENGTH;
-    text   = gensub(/^([^\n]+)\n===+(\n.*)?$/, "\\1", 1, block)
+    len = RLENGTH; text = substr(block, 1, len);
+    sub(/\n===+(\n.*)?$/, "", text);
 
     return headline(1, text, 0) _block( substr( block, len + 1 ) );
 
   # Second Order Heading H2 + Attrib
   } else if ( match( block, /^([^\n]+)([ \t]*\{([^\}\n]+)\})\n---+(\n|$)/ ) ) {
-    len = RLENGTH;
-    text   = gensub(/^([^\n]+)([ \t]*\{([^\}\n]+)\})\n---+(\n.*)?$/, "\\1", 1, block)
-    attrib = gensub(/^([^\n]+)([ \t]*\{([^\}\n]+)\})\n---+(\n.*)?$/, "\\3", 1, block)
+    len = RLENGTH; text = attrib = block;
+    sub(/([ \t]*\{([^\}\n]+)\})\n---+(\n.*)?$/, "", text);
+    sub(/\}\n---+(\n.*)?$/, "", attrib); sub(/^([^\n]+)[ \t]*\{/, "", attrib);
     gsub(/[^a-zA-Z0-9_-]+/, " ", attrib); gsub(/(^ | $)/, "", attrib);
 
     return headline(2, text, attrib) _block( substr( block, len + 1) );
 
   # Second Order Heading H2
   } else if ( match( block, /^([^\n]+)\n---+(\n|$)/ ) ) {
-    len = RLENGTH;
-    text   = gensub(/^([^\n]+)\n---+(\n.*)?$/, "\\1", 1, block)
+    len = RLENGTH; text = substr(block, 1, len);
+    sub(/\n---+(\n.*)?$/, "", text);
 
     return headline(2, text, 0) _block( substr( block, len + 1) );
 
   # Nth Order Heading H1 H2 H3 H4 H5 H6 + Attrib
   } else if ( match( block, /^(#{1,6})[ \t]*(([^ \t\n]+|[ \t]+[^ \t\n#]|[ \t]+#+[ \t]*[^ \t\n#])+)[ \t]*#*([ \t]*\{([a-zA-Z \t-]*)\})(\n|$)/ ) ) {
-    len = RLENGTH;
-    n      = gensub(/^(#{1,6})[ \t]*(([^ \t\n]+|[ \t]+[^ \t\n#]|[ \t]+#+[ \t]*[^ \t\n#])+)[ \t]*#*([ \t]*\{([a-zA-Z \t-]*)\})(\n.*)?$/, "\\1", 1, block);
-    text   = gensub(/^(#{1,6})[ \t]*(([^ \t\n]+|[ \t]+[^ \t\n#]|[ \t]+#+[ \t]*[^ \t\n#])+)[ \t]*#*([ \t]*\{([a-zA-Z \t-]*)\})(\n.*)?$/, "\\2", 1, block);
-    attrib = gensub(/^(#{1,6})[ \t]*(([^ \t\n]+|[ \t]+[^ \t\n#]|[ \t]+#+[ \t]*[^ \t\n#])+)[ \t]*#*([ \t]*\{([a-zA-Z \t-]*)\})(\n.*)?$/, "\\5", 1, block);
+    len = RLENGTH; text = attrib = substr(block, 1, len);
+    match(block, /^#{1,6}/); n = RLENGTH;
+
+    sub(/^(#{1,6})[ \t]*/, "", text);   sub(/[ \t]*#*([ \t]*\{([a-zA-Z \t-]*)\})(\n.*)?$/, "", text);
+    sub(/^(#{1,6})[ \t]*(([^ \t\n]+|[ \t]+[^ \t\n#]|[ \t]+#+[ \t]*[^ \t\n#])+)[ \t]*#*[ \t]*\{/, "", attrib);
+    sub(/\})(\n.*)?$/, "", attrib);
     gsub(/[^a-zA-Z0-9_-]+/, " ", attrib); gsub(/(^ | $)/, "", attrib);
 
-    return headline( length(n), text, attrib ) _block( substr( block, len + 1) );
+    return headline( n, text, attrib ) _block( substr( block, len + 1) );
 
   # Nth Order Heading H1 H2 H3 H4 H5 H6
   } else if ( match( block, /^(#{1,6})[ \t]*(([^ \t\n]+|[ \t]+[^ \t\n#]|[ \t]+#+[ \t]*[^ \t\n#])+)[ \t]*#*(\n|$)/ ) ) {
-    len = RLENGTH;
-    n      = gensub(/^(#{1,6})[ \t]*(([^ \t\n]+|[ \t]+[^ \t\n#]|[ \t]+#+[ \t]*[^ \t\n#])+)[ \t]*#*(\n.*)?$/, "\\1", 1, block);
-    text   = gensub(/^(#{1,6})[ \t]*(([^ \t\n]+|[ \t]+[^ \t\n#]|[ \t]+#+[ \t]*[^ \t\n#])+)[ \t]*#*(\n.*)?$/, "\\2", 1, block);
+    len = RLENGTH; text = substr(block, 1, len);
+    match(block, /^#{1,6}/); n = RLENGTH;
+    sub(/^(#{1,6})[ \t]*/, "", text); sub(/[ \t]*#*(\n.*)?$/, "", text);
 
-    return headline( length(n), text, 0 ) _block( substr( block, len + 1) );
+    return headline( n, text, 0 ) _block( substr( block, len + 1) );
 
   # block images (wrapped in <figure>)
   } else if ( match(block, /^!\[([^]]*)\]\(([^"\)]+)([ \t]+"([^"]+)")?\)(\{([a-zA-Z \t-]*)\})?(\n|$)/) ) {