X-Git-Url: https://diplodocus.org/git/nmh/blobdiff_plain/337b4e616e8f53ba06285b1645e1df9918ed5c16..6e50e1df759610bcdd9ffbb26ae405f2289288b9:/uip/mhparse.c

diff --git a/uip/mhparse.c b/uip/mhparse.c
index e7782f14..1193d564 100644
--- a/uip/mhparse.c
+++ b/uip/mhparse.c
@@ -9,7 +9,6 @@
 
 #include <h/mh.h>
 #include <fcntl.h>
-#include <h/signals.h>
 #include <h/md5.h>
 #include <h/mts.h>
 #include <h/tws.h>
@@ -34,10 +33,21 @@ int checksw = 0;	/* check Content-MD5 field */
  * 1) Instruct parser not to detect invalid Content-Transfer-Encoding
  *    in a multipart.
  * 2) Suppress the warning about bogus multipart content, and report it.
+ * 3) Suppress the warning about extraneous trailing ';' in header parameter
+ *    lists, and report it.
  */
 int skip_mp_cte_check;
 int suppress_bogus_mp_content_warning;
 int bogus_mp_content;
+int suppress_extraneous_trailing_semicolon_warning;
+int extraneous_trailing_semicolon;
+int suppress_multiple_mime_version_warning = 1;
+
+/* list of preferred type/subtype pairs, for -prefer */
+char *preferred_types[NPREFS],
+     *preferred_subtypes[NPREFS];
+int npreferred;
+
 
 /*
  * Structures for TEXT messages
@@ -59,6 +69,7 @@ struct k2v SubMultiPart[] = {
     { "alternative", MULTI_ALTERNATE },
     { "digest",      MULTI_DIGEST },
     { "parallel",    MULTI_PARALLEL },
+    { "related",     MULTI_RELATED },
     { NULL,          MULTI_UNKNOWN }    /* this one must be last! */
 };
 
@@ -99,7 +110,7 @@ static struct k2v EncodingType[] = {
 int find_cache (CT, int, int *, char *, char *, int);
 
 /* mhmisc.c */
-int part_ok (CT, int);
+int part_ok (CT);
 int type_ok (CT, int);
 void content_error (char *, CT, char *, ...);
 
@@ -116,6 +127,7 @@ static int InitGeneric (CT);
 static int InitText (CT);
 static int InitMultiPart (CT);
 static void reverse_parts (CT);
+static void prefer_parts(CT ct);
 static int InitMessage (CT);
 static int InitApplication (CT);
 static int init_encoding (CT, OpenCEFunc);
@@ -283,7 +295,7 @@ get_content (FILE *in, char *file, int toplevel)
     m_getfld_state_t gstate = 0;
 
     /* allocate the content structure */
-    if (!(ct = (CT) calloc (1, sizeof(*ct))))
+    if (!(ct = (CT) mh_xcalloc (1, sizeof(*ct))))
 	adios (NULL, "out of memory");
 
     ct->c_fp = in;
@@ -354,16 +366,12 @@ get_content (FILE *in, char *file, int toplevel)
 	if (!strcasecmp (hp->name, VRSN_FIELD)) {
 	    int ucmp;
 	    char c, *cp, *dp;
+	    char *vrsn;
 
-	    if (ct->c_vrsn) {
-		advise (NULL, "message %s has multiple %s: fields",
-			ct->c_file, VRSN_FIELD);
-		goto next_header;
-	    }
-	    ct->c_vrsn = add (hp->value, NULL);
+	    vrsn = add (hp->value, NULL);
 
 	    /* Now, cleanup this field */
-	    cp = ct->c_vrsn;
+	    cp = vrsn;
 
 	    while (isspace ((unsigned char) *cp))
 		cp++;
@@ -390,6 +398,14 @@ get_content (FILE *in, char *file, int toplevel)
 		admonish (NULL, "message %s has unknown value for %s: field (%s)",
 		ct->c_file, VRSN_FIELD, cp);
 	    }
+	    if (!ct->c_vrsn) {
+		ct->c_vrsn = vrsn;
+	    } else {
+		if (! suppress_multiple_mime_version_warning)
+		    advise (NULL, "message %s has multiple %s: fields",
+			    ct->c_file, VRSN_FIELD);
+		free(vrsn);
+	    }
 	}
 	else if (!strcasecmp (hp->name, TYPE_FIELD)) {
 	/* Get Content-Type field */
@@ -1003,7 +1019,6 @@ InitText (CT ct)
     char *chset = NULL;
     char *cp;
     PM pm;
-    struct k2v *kv;
     struct text *t;
     CI ci = &ct->c_ctinfo;
 
@@ -1012,13 +1027,10 @@ InitText (CT ct)
 	ci->ci_subtype = add ("plain", ci->ci_subtype);
 
     /* match subtype */
-    for (kv = SubText; kv->kv_key; kv++)
-	if (!strcasecmp (ci->ci_subtype, kv->kv_key))
-	    break;
-    ct->c_subtype = kv->kv_value;
+    ct->c_subtype = ct_str_subtype (CT_TEXT, ci->ci_subtype);
 
     /* allocate text character set structure */
-    if ((t = (struct text *) calloc (1, sizeof(*t))) == NULL)
+    if ((t = (struct text *) mh_xcalloc (1, sizeof(*t))) == NULL)
 	adios (NULL, "out of memory");
     ct->c_ctparams = (void *) t;
 
@@ -1068,7 +1080,6 @@ InitMultiPart (CT ct)
     size_t buflen;
     ssize_t gotlen;
     struct multipart *m;
-    struct k2v *kv;
     struct part *part, **next;
     CI ci = &ct->c_ctinfo;
     CT p;
@@ -1090,8 +1101,9 @@ InitMultiPart (CT ct)
 
 	admonish (NULL,
 		  "\"%s/%s\" type in message %s must be encoded in\n"
-		  "7bit, 8bit, or binary, per RFC 2045 (6.4).  One workaround "
-		  "is to\nmanually edit the file and change the \"%s\"\n"
+		  "7bit, 8bit, or binary, per RFC 2045 (6.4).  "
+                  "mhfixmsg -fixcte can fix it, or\n"
+                  "manually edit the file and change the \"%s\"\n"
 		  "Content-Transfer-Encoding to one of those.  For now",
 		  ci->ci_type, ci->ci_subtype, ct->c_file, bp);
 	free (cte);
@@ -1100,10 +1112,7 @@ InitMultiPart (CT ct)
     }
 
     /* match subtype */
-    for (kv = SubMultiPart; kv->kv_key; kv++)
-	if (!strcasecmp (ci->ci_subtype, kv->kv_key))
-	    break;
-    ct->c_subtype = kv->kv_value;
+    ct->c_subtype = ct_str_subtype (CT_MULTIPART, ci->ci_subtype);
 
     /*
      * Check for "boundary" parameter, which is
@@ -1126,7 +1135,7 @@ InitMultiPart (CT ct)
     }
 
     /* allocate primary structure for multipart info */
-    if ((m = (struct multipart *) calloc (1, sizeof(*m))) == NULL)
+    if ((m = (struct multipart *) mh_xcalloc (1, sizeof(*m))) == NULL)
 	adios (NULL, "out of memory");
     ct->c_ctparams = (void *) m;
 
@@ -1171,7 +1180,7 @@ InitMultiPart (CT ct)
 	    if (strcmp (bufp + 2, m->mp_start))
 		continue;
 next_part:
-	    if ((part = (struct part *) calloc (1, sizeof(*part))) == NULL)
+	    if ((part = (struct part *) mh_xcalloc (1, sizeof(*part))) == NULL)
 		adios (NULL, "out of memory");
 	    *next = part;
 	    next = &part->mp_next;
@@ -1179,7 +1188,6 @@ next_part:
 	    if (!(p = get_content (fp, ct->c_file,
 			ct->c_subtype == MULTI_DIGEST ? -1 : 0))) {
 		free(bufp);
-		fclose (ct->c_fp);
 		ct->c_fp = NULL;
 		return NOTOK;
 	    }
@@ -1227,8 +1235,10 @@ end_part:
 
 last_part:
     /* reverse the order of the parts for multipart/alternative */
-    if (ct->c_subtype == MULTI_ALTERNATE)
+    if (ct->c_subtype == MULTI_ALTERNATE) {
 	reverse_parts (ct);
+	prefer_parts (ct);
+    }
 
     /*
      * label all subparts with part number, and
@@ -1274,9 +1284,15 @@ last_part:
 
 
 /*
- * reverse the order of the parts of a multipart/alternative
+ * reverse the order of the parts of a multipart/alternative,
+ * presumably to put the "most favored" alternative first, for
+ * ease of choosing/displaying it later on.  from a mail message on
+ * nmh-workers, from kenh:
+ *  "Stock" MH 6.8.5 did not have a reverse_parts() function, but I
+ *  see code in mhn that did the same thing...  Acccording to the RCS
+ *  logs, that code was around from the initial checkin of mhn.c by
+ *  John Romine in 1992, which is as far back as we have."
  */
-
 static void
 reverse_parts (CT ct)
 {
@@ -1293,6 +1309,60 @@ reverse_parts (CT ct)
     }
 }
 
+static void
+move_preferred_part (CT ct, char *type, char *subtype)
+{
+    struct multipart *m = (struct multipart *) ct->c_ctparams;
+    struct part *part, *prev, *head, *nhead, *ntail;
+    struct part h, n;
+    CI ci;
+
+    /* move the matching part(s) to the head of the list:  walk the
+     * list of parts, move matching parts to a new list (maintaining
+     * their order), and finally, concatenate the old list onto the
+     * new.
+     */
+
+    head = &h;
+    nhead = &n;
+
+    head->mp_next = m->mp_parts;
+    nhead->mp_next = NULL;
+    ntail = nhead;
+
+    prev = head;
+    part = head->mp_next;
+    while (part != NULL) {
+	ci = &part->mp_part->c_ctinfo;
+	if (!strcasecmp(ci->ci_type, type) &&
+		(!subtype || !strcasecmp(ci->ci_subtype, subtype))) {
+	    prev->mp_next = part->mp_next;
+	    part->mp_next = NULL;
+	    ntail->mp_next = part;
+	    ntail = part;
+	    part = prev->mp_next;
+	} else {
+	    prev = part;
+	    part = prev->mp_next;
+	}
+    }
+    ntail->mp_next = head->mp_next;
+    m->mp_parts = nhead->mp_next;
+
+}
+
+/*
+ * move parts that match the user's preferences (-prefer) to the head
+ * of the line.  process preferences in reverse so first one given
+ * ends up first in line
+ */
+static void
+prefer_parts(CT ct)
+{
+    int i;
+    for (i = npreferred-1; i >= 0; i--)
+	move_preferred_part(ct, preferred_types[i], preferred_subtypes[i]);
+}
 
 
 
@@ -1325,7 +1395,6 @@ reverse_alternative_parts (CT ct) {
 static int
 InitMessage (CT ct)
 {
-    struct k2v *kv;
     CI ci = &ct->c_ctinfo;
 
     if ((ct->c_encoding != CE_7BIT) && (ct->c_encoding != CE_8BIT)) {
@@ -1340,10 +1409,7 @@ InitMessage (CT ct)
 	ci->ci_subtype = add ("rfc822", ci->ci_subtype);
 
     /* match subtype */
-    for (kv = SubMessage; kv->kv_key; kv++)
-	if (!strcasecmp (ci->ci_subtype, kv->kv_key))
-	    break;
-    ct->c_subtype = kv->kv_value;
+    ct->c_subtype = ct_str_subtype (CT_MESSAGE, ci->ci_subtype);
 
     switch (ct->c_subtype) {
 	case MESSAGE_RFC822:
@@ -1354,7 +1420,7 @@ InitMessage (CT ct)
 		PM pm;
 		struct partial *p;
 
-		if ((p = (struct partial *) calloc (1, sizeof(*p))) == NULL)
+		if ((p = (struct partial *) mh_xcalloc (1, sizeof(*p))) == NULL)
 		    adios (NULL, "out of memory");
 		ct->c_ctparams = (void *) p;
 
@@ -1403,7 +1469,7 @@ invalid_param:
 		CT p;
 		FILE *fp;
 
-		if ((e = (struct exbody *) calloc (1, sizeof(*e))) == NULL)
+		if ((e = (struct exbody *) mh_xcalloc (1, sizeof(*e))) == NULL)
 		    adios (NULL, "out of memory");
 		ct->c_ctparams = (void *) e;
 
@@ -1593,14 +1659,10 @@ params_external (CT ct, int composing)
 static int
 InitApplication (CT ct)
 {
-    struct k2v *kv;
     CI ci = &ct->c_ctinfo;
 
     /* match subtype */
-    for (kv = SubApplication; kv->kv_key; kv++)
-	if (!strcasecmp (ci->ci_subtype, kv->kv_key))
-	    break;
-    ct->c_subtype = kv->kv_value;
+    ct->c_subtype = ct_str_subtype (CT_APPLICATION, ci->ci_subtype);
 
     return OK;
 }
@@ -2127,6 +2189,7 @@ ready_to_go:
       fclose (ct->c_fp);
       ct->c_fp = NULL;
     }
+    free (bufp);
     return fileno (ce->ce_fp);
 
 clean_up:
@@ -2135,6 +2198,7 @@ clean_up:
       fclose (ct->c_fp);
       ct->c_fp = NULL;
     }
+    free (bufp);
     return NOTOK;
 }
 
@@ -2274,7 +2338,9 @@ open7Bit (CT ct, char **file)
 		cc = len;
 	    len -= cc;
 
-	    fwrite (buffer, sizeof(*buffer), cc, ce->ce_fp);
+	    if ((int) fwrite (buffer, sizeof(*buffer), cc, ce->ce_fp) < cc) {
+		advise ("open7Bit", "fwrite");
+	    }
 	    if (ferror (ce->ce_fp)) {
 		content_error (ce->ce_file, ct, "error writing to");
 		goto clean_up;
@@ -2409,7 +2475,9 @@ openFile (CT ct, char **file)
 
 	    while ((cc = fread (buffer, sizeof(*buffer), sizeof(buffer), gp))
 		       > 0)
-		fwrite (buffer, sizeof(*buffer), cc, fp);
+		if ((int) fwrite (buffer, sizeof(*buffer), cc, fp) < cc) {
+		    advise ("openFile", "fwrite");
+		}
 	    fflush (fp);
 
 	    if (ferror (gp)) {
@@ -2618,7 +2686,9 @@ openFTP (CT ct, char **file)
 
 		while ((cc= fread (buffer, sizeof(*buffer), sizeof(buffer), gp))
 		           > 0)
-		    fwrite (buffer, sizeof(*buffer), cc, fp);
+		    if ((int) fwrite (buffer, sizeof(*buffer), cc, fp) < cc) {
+			advise ("openFTP", "fwrite");
+		    }
 		fflush (fp);
 
 		if (ferror (gp)) {
@@ -2878,7 +2948,9 @@ openURL (CT ct, char **file)
 
 		while ((cc = fread(buffer, sizeof(*buffer),
 				   sizeof(buffer), gp)) > 0)
-		    fwrite(buffer, sizeof(*buffer), cc, fp);
+		    if ((int) fwrite(buffer, sizeof(*buffer), cc, fp) < cc) {
+			advise ("openURL", "fwrite");
+		    }
 
 		fflush(fp);
 
@@ -3140,6 +3212,8 @@ ct_subtype_str (int type, int subtype) {
             return "digest";
         case MULTI_PARALLEL:
             return "parallel";
+        case MULTI_RELATED:
+            return "related";
         default:
             return "unknown_multipart_subtype";
         }
@@ -3160,6 +3234,62 @@ ct_subtype_str (int type, int subtype) {
 }
 
 
+int
+ct_str_type (const char *type) {
+    struct str2init *s2i;
+
+    for (s2i = str2cts; s2i->si_key; ++s2i) {
+        if (! strcasecmp (type, s2i->si_key)) {
+            break;
+        }
+    }
+    if (! s2i->si_key  &&  ! uprf (type, "X-")) {
+        ++s2i;
+    }
+
+    return s2i->si_val;
+}
+
+
+int
+ct_str_subtype (int type, const char *subtype) {
+    struct k2v *kv;
+
+    switch (type) {
+    case CT_APPLICATION:
+        for (kv = SubApplication; kv->kv_key; ++kv) {
+            if (! strcasecmp (subtype, kv->kv_key)) {
+                break;
+            }
+        }
+        return kv->kv_value;
+    case CT_MESSAGE:
+        for (kv = SubMessage; kv->kv_key; ++kv) {
+            if (! strcasecmp (subtype, kv->kv_key)) {
+                break;
+            }
+        }
+        return kv->kv_value;
+    case CT_MULTIPART:
+        for (kv = SubMultiPart; kv->kv_key; ++kv) {
+            if (! strcasecmp (subtype, kv->kv_key)) {
+                break;
+            }
+        }
+        return kv->kv_value;
+    case CT_TEXT:
+        for (kv = SubText; kv->kv_key; ++kv) {
+            if (! strcasecmp (subtype, kv->kv_key)) {
+                break;
+            }
+        }
+        return kv->kv_value;
+    default:
+        return 0;
+    }
+}
+
+
 /* Find the content type and InitFunc for the CT. */
 const struct str2init *
 get_ct_init (int type) {
@@ -3264,10 +3394,13 @@ parse_header_attrs (const char *filename, const char *fieldname,
         }
 
 	if (*cp == 0) {
-	    advise (NULL,
-		    "extraneous trailing ';' in message %s's %s: "
-                    "parameter list",
-		    filename, fieldname);
+	    if (! suppress_extraneous_trailing_semicolon_warning) {
+		advise (NULL,
+			"extraneous trailing ';' in message %s's %s: "
+			"parameter list",
+			filename, fieldname);
+	    }
+	    extraneous_trailing_semicolon = 1;
 	    return DONE;
 	}
 
@@ -3542,6 +3675,7 @@ bad_quote:
 				"%s's %s: field\n%*s(parameter %s)", sp->index,
 				filename, fieldname, strlen(invo_name) + 2, "",
 				nameptr);
+			free (nameptr);
 			return NOTOK;
 		    }
 		    if (sp2->index < sp->index &&
@@ -3557,6 +3691,7 @@ bad_quote:
 		    	   "param in message %s's %s: field\n%*s(parameter %s)",
 			   filename, fieldname, strlen(invo_name) + 2, "",
 			   nameptr);
+		    free (nameptr);
 		    return NOTOK;
 		}
 	    }
@@ -3632,21 +3767,16 @@ bad_quote:
 }
 
 /*
- * Return the charset for a particular content type.  Return pointer is
- * only valid until the next call to content_charset().
+ * Return the charset for a particular content type.
  */
 
 char *
 content_charset (CT ct) {
-    static char *ret_charset = NULL;
-
-    if (ret_charset != NULL) {
-	free(ret_charset);
-    }
+    char *ret_charset = NULL;
 
     ret_charset = get_param(ct->c_ctinfo.ci_first_pm, "charset", '?', 0);
 
-    return ret_charset ? ret_charset : "US-ASCII";
+    return ret_charset ? ret_charset : getcpy ("US-ASCII");
 }
 
 
@@ -4204,7 +4334,7 @@ char *get_param_value(PM pm, char replace)
 	    }
 	    if (utf8) {
 		for (++p, --inbytes;
-		     inbytes > 0 && (((unsigned char) *q) & 0xc0) == 0x80;
+		     inbytes > 0 && (((unsigned char) *p) & 0xc0) == 0x80;
 		     ++p, --inbytes)
 		    continue;
 	    } else {