From ac186529b6b5f80906c3215a67c98505db7bb156 Mon Sep 17 00:00:00 2001
From: Kaz Kylheku <kaz@kylheku.com>
Date: Tue, 13 Apr 2021 19:13:58 -0700
Subject: Implement .M2HT request for HTML passthrough.

---
 man2html/man2html.1 |  3 +++
 man2html/man2html.c | 42 ++++++++++++++++++++++++++++++++++++++++--
 2 files changed, 43 insertions(+), 2 deletions(-)

diff --git a/man2html/man2html.1 b/man2html/man2html.1
index 089edf4..b374a96 100644
--- a/man2html/man2html.1
+++ b/man2html/man2html.1
@@ -149,6 +149,9 @@ The datum is the heading to be indexed and emitted into the HTML output.
 The standard .SH and .SS requests use levels 0 and 1, respectively,
 and HTML tags H2 and H3, respectively.
 
+The request ".M2HT argument" can be used to emit raw HTML. The
+argument is passed through as HTML without any escaping.
+
 .SH BUGS
 There are many heuristics.  The output will not always be perfect.
 The lynxcgi method will not work if lynx was compiled without
diff --git a/man2html/man2html.c b/man2html/man2html.c
index 8f9b0e5..3a06408 100644
--- a/man2html/man2html.c
+++ b/man2html/man2html.c
@@ -50,6 +50,7 @@ int maxtstop=12;
 int curpos=0;
 
 static char *scan_troff(char *c, int san, char **result);
+static char *scan_troff_raw(char *c, int san, char **result);
 static char *scan_troff_mandoc(char *c, int san, char **result);
 
 static char **argument=NULL;
@@ -2286,6 +2287,18 @@ scan_request(char *c) {
 		free(wordlist[2]);
 	    }
 	    break;
+	case V4('M','2','H','T'):
+	    {
+		char *h = NULL;
+		c=c+j;
+		c = fill_words(c, wordlist, SIZE(wordlist), &words, 0);
+		if (words!=1)
+		    abort();
+		scan_troff_raw(wordlist[0], 1, &h);
+		out_html(h);
+		free(h);
+	    }
+	    break;
 	case V('T','S'):
 	    c=scan_table(c);
 	    nofillout=1;
@@ -3002,7 +3015,7 @@ static int contained_tab=0;
 static int mandoc_line=0;/* Signals whether to look for embedded mandoc cmds */
 
 static char *
-scan_troff(char *c, int san, char **result) {   /* san : stop at newline */
+do_scan_troff(char *c, int san, int htmlesc, char **result) { /* san : stop at newline */
     char *h;
     char intbuff[500];
     int ibp=0;
@@ -3070,7 +3083,7 @@ scan_troff(char *c, int san, char **result) {   /* san : stop at newline */
 		curpos=0;
 		still_dd=0;
 	    }
-	    switch (*h) {
+	    if (htmlesc) switch (*h) {
 	    case '&':
 		intbuff[ibp++]='&';
 		intbuff[ibp++]='a';
@@ -3177,6 +3190,21 @@ scan_troff(char *c, int san, char **result) {   /* san : stop at newline */
 		}
 		curpos++;
 		break;
+	    } else {
+		if (*h > 31 && *h < 127) {
+		    intbuff[ibp++]=*h;
+		} else if (((unsigned char)(*h)) > 127) {
+#ifdef NO_8BIT
+		    intbuff[ibp++]='&';
+		    intbuff[ibp++]='#';
+		    intbuff[ibp++]='0'+((unsigned char)(*h))/100;
+		    intbuff[ibp++]='0'+(((unsigned char)(*h))%100)/10;
+		    intbuff[ibp++]='0'+((unsigned char)(*h))%10;
+		    intbuff[ibp++]=';';
+#else
+		    intbuff[ibp++]=*h;
+#endif
+		}
 	    }
 	    if (ibp>480) FLUSHIBP;
 	    h++;
@@ -3196,6 +3224,16 @@ scan_troff(char *c, int san, char **result) {   /* san : stop at newline */
     return h;
 }
 
+static char *
+scan_troff(char *c, int san, char **result) {
+    return do_scan_troff(c, san, 1, result);
+}
+
+static char *
+scan_troff_raw(char *c, int san, char **result) {
+    return do_scan_troff(c, san, 0, result);
+}
+
 static char *scan_troff_mandoc(char *c, int san, char **result) {
      char *ret, *end = c;
      int oldval = mandoc_line;
-- 
cgit v1.2.3