From 0f0fc6dbb7e7354ac662951f124b0a64c54b741b Mon Sep 17 00:00:00 2001
From: John Denker <jsd@av8n.com>
Date: Mon, 23 Jul 2012 21:37:35 -0700
Subject: get line-counter to work

---
 tools/skrewt.c | 134 +++++++++++++++++++++++++++++++++------------------------
 1 file changed, 77 insertions(+), 57 deletions(-)

(limited to 'tools/skrewt.c')

diff --git a/tools/skrewt.c b/tools/skrewt.c
index ed0e627..d2e1bbc 100644
--- a/tools/skrewt.c
+++ b/tools/skrewt.c
@@ -145,56 +145,52 @@ string basename(const string path){
 string progname, progid;
 int mypid;
 
-
 /* Content-Type: text/plain; charset="us-ascii"                                         */
 /* Content-Type: multipart/mixed; boundary="1170861315-1262462055-1341954763=:92165"    */
+//
 void parse_content(const string type_spec_line, string &maintype, string &boundary) {
-      string mainline(type_spec_line);
-
-      string get_type(toLower(mainline));
-      size_t where = get_type.find_first_of(" \t;\n");
-      if (where == string::npos) {
-        // keep whole string
-      }
-      else {
-        get_type = get_type.substr(0,where);
-      }
-      where = get_type.find("/");
-      if (where == string::npos){
-        maintype = "";
-      } else {
-        maintype = get_type.substr(0, where);
-      }
-}
-
-#ifdef xxxxxxxxxxxxxxxxxxxx
-      if
-      if (0) cerr << "type (" << get_type << ")  "
-            << (text_type?"text":"nope") << endl;
-
-
-      string srch = "boundary=";
-      where = headrec.find(srch);
-      if (where != string::npos) {
-        where += srch.length();
-        boundary = headrec.substr(where);
-        if (boundary[0] == '"') {
-          boundary = boundary.substr(1);
-          where = boundary.find_first_of("\"");
-        } else {
-          where = boundary.find_first_of(" \t;\n");
-        }
-        if (where == string::npos) {
-          /* do nothing, boundary=boundary as a whole */
-        } else {
-          boundary = boundary.substr(0, where);
-        }
-      }
-    }
-#endif
+  cerr << "parser called with: " << type_spec_line << endl;
+  string get_type(type_spec_line);
 
+  size_t where = get_type.find_first_of(" \t;\n");
+  string rest;
+  if (where == string::npos) {
+    // keep whole string
+  }
+  else {
+    rest = get_type.substr(where+1);
+    get_type = get_type.substr(0,where);
+  }
+  where = get_type.find("/");
+  if (where == string::npos){
+    maintype = "";
+    cerr << "could not find / in " << get_type << endl;
+  } else {
+    maintype = get_type.substr(0, where);
+  }
 
+// now need to find boundary
 
+  string srch = "boundary=";
+  where = rest.find(srch);
+  if (where != string::npos) {
+    where += srch.length();
+    boundary = rest.substr(where);
+    if (boundary[0] == '"') {
+      boundary = boundary.substr(1);
+      where = boundary.find_first_of("\"");
+    } else {
+      where = boundary.find_first_of(" \t;\n");
+    }
+    if (where == string::npos) {
+      /* do nothing, boundary=boundary as a whole */
+    } else {
+      boundary = boundary.substr(0, where);
+    }
+  } else {
+    //xxxxxxx cerr << "boundary= not found in " << type_spec_line << endl;
+  }
+}
 
 
 ////////////////////////////////////////////////////////////
@@ -329,15 +325,17 @@ int main(int _argc, const char** _argv){
     exeunt(ex_spam);              // disallow mail with no date
   }
 
-  string contype;
-  int textlines(0);
-  parse_content(content_type, contype, boundary);
-  int is_text = contype == "text";
+  string main_contype;
+  parse_content(content_type, main_contype, boundary);
+  int currently_text = main_contype == "text";
+  int main_multipart = main_contype == "multipart";
 
 // early-stage thinking has been done.
 // Now spew the rest of the message
-  cerr << "body begins: " << contype << " " << is_text << endl;
-  int inheads(0);
+  cerr << "body begins: " << main_contype << " " << currently_text << " " << boundary << endl;
+  int in_subheads(0);
+  int textlines(0);
+
   for (;;){             // outer loop over all lines in the body
     if (cin.eof()) break;
     if (cin.bad()) return 1;
@@ -351,19 +349,41 @@ int main(int _argc, const char** _argv){
     }
     bigbuf.push_back(line);
     cout << line << endl;
-    if (line == "--" + boundary) {
-      inheads = 1;
+    if (in_subheads){
+      if (line == "" || line == "\r") in_subheads = 0;
+    }
+    if (in_subheads){
+        string sub_contype;
+        string junk;
+// in principle could worry about folded headers,
+// but in this application it doesn't actually matter
+        string headword;
+        string rest;
+        size_t where = line.find(":");
+        if (where != string::npos) {
+          headword = line.substr(0, where);
+          rest = ltrim(line.substr(1+where));
+        }
+        headword = toLower(headword);
+        if (headword == "content-type") {
+          parse_content(rest, sub_contype, junk);
+          currently_text = sub_contype == "text";
+          cerr << "setting contype '" << sub_contype << "' " << currently_text << " ... " << textlines << endl;
+        }
     } else {
-      if (is_text) {
-        if (ltrim(line).length()) textlines++;
+      if (main_multipart && line == "--" + boundary) {
+        //xxxx cerr << "found subhead boundary" << endl;
+        in_subheads = 1;
+        continue;
       }
+      if (currently_text) textlines++;
     }
   }
 
-  if (0) cerr << "textlines: " << textlines << endl;
-  if (000 && !textlines) {
+  if (1) cerr << "textlines: " << textlines << endl;
+  if (1 && !textlines) {
     cerr << progid << " rejection: no text" << endl;
-    exeunt(ex_spam);
+//    exeunt(ex_spam);
   }
   cerr << progid << " normal completion" << endl;
   exit(ex_good);
-- 
cgit v1.2.3