From 8bdf66a1f5f21c05414ab48c294ca13d2f91a1b3 Mon Sep 17 00:00:00 2001
From: leitner <leitner>
Date: Wed, 17 Sep 2014 13:58:59 +0000
Subject: [PATCH] add scan_html_tagarg

---
 CHANGES              |  2 +-
 textcode.h           | 31 ++++++++++++++++++-------------
 textcode/fmt_html.c  |  6 +++++-
 textcode/scan_html.c | 12 ++++++++++--
 4 files changed, 34 insertions(+), 17 deletions(-)

diff --git a/CHANGES b/CHANGES
index 30fb65f..4c3d9a3 100644
--- a/CHANGES
+++ b/CHANGES
@@ -16,7 +16,7 @@
   switch epoll from level triggering to edge triggering
   introduce io_eagain_read and io_eagain_write (discontinue using io_eagain plz)
   fix buffer_get
-  add fmt_html_tagarg, fmt_xml
+  add fmt_html_tagarg, fmt_xml, scan_html_tagarg
 
 0.29:
   save 8 bytes in taia.h for 64-bit systems
diff --git a/textcode.h b/textcode.h
index 336668a..2bb1132 100644
--- a/textcode.h
+++ b/textcode.h
@@ -57,19 +57,24 @@ size_t fmt_base85(char* dest,const char* src,size_t len);
 /* These read one line from src, decode it, and write the result to
  * dest.  The number of decoded bytes is written to destlen.  dest
  * should be able to hold strlen(src) bytes as a rule of thumb. */
-size_t scan_uuencoded(const char *src,char *dest,size_t *destlen);
-size_t scan_base64(const char *src,char *dest,size_t *destlen);
-size_t scan_quotedprintable(const char *src,char *dest,size_t *destlen);
-size_t scan_urlencoded(const char *src,char *dest,size_t *destlen);
-size_t scan_urlencoded2(const char *src,char *dest,size_t *destlen);
-size_t scan_yenc(const char *src,char *dest,size_t *destlen);
-size_t scan_hexdump(const char *src,char *dest,size_t *destlen);
-size_t scan_html(const char *src,char *dest,size_t *destlen);
-size_t scan_cescape(const char *src,char *dest,size_t *destlen);
-size_t scan_ldapescape(const char* src,char* dest,size_t *destlen);
-size_t scan_jsonescape(const char* src,char* dest,size_t *destlen);
-
-size_t scan_base85(const char* src,char* dest,size_t *destlen);
+size_t scan_uuencoded(const char* src,char* dest,size_t* destlen);
+size_t scan_base64(const char* src,char* dest,size_t* destlen);
+size_t scan_quotedprintable(const char* src,char* dest,size_t* destlen);
+size_t scan_urlencoded(const char* src,char* dest,size_t* destlen);
+size_t scan_urlencoded2(const char* src,char* dest,size_t* destlen);
+size_t scan_yenc(const char* src,char* dest,size_t* destlen);
+size_t scan_hexdump(const char* src,char* dest,size_t* destlen);
+/* decodes all html5-standardized &foo; escapes, and also
+ * "<br>" to "\n" and "<p>" to "\n\n", leaves the rest of the tags alone */
+size_t scan_html(const char* src,char* dest,size_t* destlen);
+/* decodes all html5-standardized &foo; escapes, but leaves all tags
+ * alone */
+size_t scan_html_tagarg(const char* src,char* dest,size_t* destlen);
+size_t scan_cescape(const char* src,char* dest,size_t* destlen);
+size_t scan_ldapescape(const char* src,char* dest,size_t* destlen);
+size_t scan_jsonescape(const char* src,char* dest,size_t* destlen);
+
+size_t scan_base85(const char* src,char* dest,size_t* destlen);
 
 #ifdef STRALLOC_H
 /* WARNING: these functions _append_ to the stralloc, not overwrite! */
diff --git a/textcode/fmt_html.c b/textcode/fmt_html.c
index bd3e238..4847d90 100644
--- a/textcode/fmt_html.c
+++ b/textcode/fmt_html.c
@@ -13,7 +13,11 @@ size_t fmt_html(char* dest,const char* src,size_t len) {
     case '<': seq="&lt;"; goto doit;
     case '>': seq="&gt;"; goto doit;
     case '\n':
-	seq="<br>";
+	if (i<len && s[i+1]=='\n') {
+	  seq="<p>";
+	  ++i;
+	} else
+	  seq="<br>";
       doit:
 	written+=fmt_str(dest?dest+written:0,seq);
 	break;
diff --git a/textcode/scan_html.c b/textcode/scan_html.c
index 89fee30..97e9122 100644
--- a/textcode/scan_html.c
+++ b/textcode/scan_html.c
@@ -25,7 +25,7 @@ static const char* lookup(size_t ofs,const char* t) {
   return NULL;
 }
 
-size_t scan_html(const char *src,char *dest,size_t *destlen) {
+static size_t scan_html_inner(const char *src,char *dest,size_t *destlen,int flag) {
   register const unsigned char* s=(const unsigned char*) src;
   size_t written=0,i;
   for (i=0; s[i]; ++i) {
@@ -58,7 +58,7 @@ size_t scan_html(const char *src,char *dest,size_t *destlen) {
 	continue;
       } else
 	dest[written]='&';
-    } else if (s[i]=='<') {
+    } else if (flag && s[i]=='<') {
       if (case_starts((const char*)s+i+1,"br>")) {
 	dest[written]='\n';
 	i+=3;
@@ -75,3 +75,11 @@ size_t scan_html(const char *src,char *dest,size_t *destlen) {
   *destlen=written;
   return i;
 }
+
+size_t scan_html_tagarg(const char *src,char *dest,size_t *destlen) {
+  return scan_html_inner(src,dest,destlen,1);
+}
+
+size_t scan_html(const char *src,char *dest,size_t *destlen) {
+  return scan_html_inner(src,dest,destlen,0);
+}