X-Git-Url: https://diplodocus.org/git/nmh/blobdiff_plain/d3ecbe4bae7a3f1972b553d4b1efab3079ef0bae..c36ae8a76cf2ab8065fa996bb182e286880d63c0:/uip/mhfixmsg.c diff --git a/uip/mhfixmsg.c b/uip/mhfixmsg.c index 2bc65a9d..6195b316 100644 --- a/uip/mhfixmsg.c +++ b/uip/mhfixmsg.c @@ -27,8 +27,8 @@ X("noreplacetextplain", 0, NREPLACETEXTPLAINSW) \ X("fixboundary", 0, FIXBOUNDARYSW) \ X("nofixboundary", 0, NFIXBOUNDARYSW) \ - X("fixcte", 0, FIXCTESW) \ - X("nofixcte", 0, NFIXCTESW) \ + X("fixcte", 0, FIXCOMPOSITECTESW) \ + X("nofixcte", 0, NFIXCOMPOSITECTESW) \ X("fixtype mimetype", 0, FIXTYPESW) \ X("file file", 0, FILESW) \ X("outfile file", 0, OUTFILESW) \ @@ -78,7 +78,7 @@ void freects_done (int) NORETURN; */ typedef struct fix_transformations { int fixboundary; - int fixcte; + int fixcompositecte; svector_t fixtypes; int reformat; int replacetextplain; @@ -97,7 +97,7 @@ static int replace_boundary (CT, char *, char *); static int fix_types (CT, svector_t, int *); static char *replace_substring (char **, const char *, const char *); static char *remove_parameter (char *, const char *); -static int fix_multipart_cte (CT, int *); +static int fix_composite_cte (CT, int *); static int set_ce (CT, int); static int ensure_text_plain (CT *, CT, int *, int); static int find_textplain_sibling (CT, int, int *); @@ -108,7 +108,6 @@ static CT divide_part (CT); static void copy_ctinfo (CI, CI); static int decode_part (CT); static int reformat_part (CT, char *, char *, char *, int); -static int charset_encoding (CT); static CT build_multipart_alt (CT, CT, int, int); static int boundary_in_content (FILE **, char *, const char *); static void transfer_noncontent_headers (CT, CT); @@ -117,6 +116,9 @@ static int decode_text_parts (CT, int, const char *, int *); static int should_decode(const char *, const char *, const char *); static int content_encoding (CT, const char **); static int strip_crs (CT, int *); +static void update_cte (CT); +static int least_restrictive_encoding (CT); +static int less_restrictive (int, int); static int convert_charsets (CT, char *, int *); static int fix_always (CT, int *); static int fix_filename_param (char *, char *, PM *, PM *); @@ -142,7 +144,7 @@ main (int argc, char **argv) { int chgflag = 1; int status = OK; fix_transformations fx; - fx.reformat = fx.fixcte = fx.fixboundary = 1; + fx.reformat = fx.fixcompositecte = fx.fixboundary = 1; fx.fixtypes = NULL; fx.replacetextplain = 0; fx.decodetext = CE_8BIT; @@ -150,7 +152,7 @@ main (int argc, char **argv) { fx.lf_line_endings = 0; fx.textcharset = NULL; - if (nmh_init(argv[0], 1)) { return 1; } + if (nmh_init(argv[0], 2)) { return 1; } done = freects_done; @@ -222,11 +224,11 @@ main (int argc, char **argv) { case NFIXBOUNDARYSW: fx.fixboundary = 0; continue; - case FIXCTESW: - fx.fixcte = 1; + case FIXCOMPOSITECTESW: + fx.fixcompositecte = 1; continue; - case NFIXCTESW: - fx.fixcte = 0; + case NFIXCOMPOSITECTESW: + fx.fixcompositecte = 0; continue; case FIXTYPESW: if (! (cp = *argp++) || (*cp == '-' && cp[1])) { @@ -518,15 +520,17 @@ mhfixmsgsbr (CT *ctp, const fix_transformations *fx, char *outfile) { if (status == OK && fx->fixtypes != NULL) { status = fix_types (*ctp, fx->fixtypes, &message_mods); } - if (status == OK && fx->fixcte) { - status = fix_multipart_cte (*ctp, &message_mods); + if (status == OK && fx->fixcompositecte) { + status = fix_composite_cte (*ctp, &message_mods); } if (status == OK && fx->reformat) { status = ensure_text_plain (ctp, NULL, &message_mods, fx->replacetextplain); } if (status == OK && fx->decodetext) { - status = decode_text_parts (*ctp, fx->decodetext, fx->decodetypes, &message_mods); + status = decode_text_parts (*ctp, fx->decodetext, fx->decodetypes, + &message_mods); + update_cte (*ctp); } if (status == OK && fx->textcharset != NULL) { status = convert_charsets (*ctp, fx->textcharset, &message_mods); @@ -1057,16 +1061,15 @@ remove_parameter (char *str, const char *name) { /* - * Fix Content-Transfer-Encoding. + * Fix Content-Transfer-Encoding of composite,, e.g., message or multipart, part. + * According to RFC 2045 Sec. 6.4, it must be 7bit, 8bit, or binary. Set it to + * 8 bit. */ static int -fix_multipart_cte (CT ct, int *message_mods) { +fix_composite_cte (CT ct, int *message_mods) { int status = OK; - if (ct->c_type == CT_MULTIPART) { - struct multipart *m; - struct part *part; - + if (ct->c_type == CT_MESSAGE || ct->c_type == CT_MULTIPART) { if (ct->c_encoding != CE_7BIT && ct->c_encoding != CE_8BIT && ct->c_encoding != CE_BINARY) { HF hf; @@ -1112,11 +1115,16 @@ fix_multipart_cte (CT ct, int *message_mods) { set_ce (ct, CE_8BIT); } - m = (struct multipart *) ct->c_ctparams; - for (part = m->mp_parts; part; part = part->mp_next) { - if (fix_multipart_cte (part->mp_part, message_mods) != OK) { - status = NOTOK; - break; + if (ct->c_type == CT_MULTIPART) { + struct multipart *m; + struct part *part; + + m = (struct multipart *) ct->c_ctparams; + for (part = m->mp_parts; part; part = part->mp_next) { + if (fix_composite_cte (part->mp_part, message_mods) != OK) { + status = NOTOK; + break; + } } } } @@ -1152,6 +1160,10 @@ set_ce (CT ct, int encoding) { ct->c_cefile.ce_file to the name of the file containing the contents. */ + if (ct->c_ceclosefnx) { + (*ct->c_ceclosefnx) (ct); + } + /* Restore the cefile. */ ct->c_cefile = decoded_content_info; @@ -1577,6 +1589,7 @@ decode_part (CT ct) { static int reformat_part (CT ct, char *file, char *type, char *subtype, int c_type) { int output_subtype, output_encoding; + const char *reason = NULL; char *cp, *cf; int status; @@ -1622,8 +1635,8 @@ reformat_part (CT ct, char *file, char *type, char *subtype, int c_type) { /* Set subtype to 0, which is always an UNKNOWN subtype. */ output_subtype = 0; } - output_encoding = charset_encoding (ct); + output_encoding = content_encoding (ct, &reason); if (set_ct_type (ct, c_type, output_subtype, output_encoding) == OK) { ct->c_cefile.ce_file = file; ct->c_cefile.ce_unlink = 1; @@ -1636,20 +1649,6 @@ reformat_part (CT ct, char *file, char *type, char *subtype, int c_type) { } -/* - * Identifies 7bit or 8bit content based on charset. - */ -static int -charset_encoding (CT ct) { - char *ct_charset = content_charset (ct); - int encoding = strcasecmp (ct_charset, "US-ASCII") ? CE_8BIT : CE_7BIT; - - free (ct_charset); - - return encoding; -} - - /* * Fill in a multipart/alternative part. */ @@ -1921,7 +1920,8 @@ set_ct_type (CT ct, int type, int subtype, int encoding) { * that character set again after decoding." */ static int -decode_text_parts (CT ct, int encoding, const char *decodetypes, int *message_mods) { +decode_text_parts (CT ct, int encoding, const char *decodetypes, + int *message_mods) { int status = OK; int lf_line_endings = 0; @@ -1933,7 +1933,8 @@ decode_text_parts (CT ct, int encoding, const char *decodetypes, int *message_mo /* Should check to see if the body for this part is encoded? For now, it gets passed along as-is by InitMultiPart(). */ for (part = m->mp_parts; status == OK && part; part = part->mp_next) { - status = decode_text_parts (part->mp_part, encoding, decodetypes, message_mods); + status = decode_text_parts (part->mp_part, encoding, decodetypes, + message_mods); } break; } @@ -1942,7 +1943,8 @@ decode_text_parts (CT ct, int encoding, const char *decodetypes, int *message_mo if (ct->c_subtype == MESSAGE_EXTERNAL) { struct exbody *e = (struct exbody *) ct->c_ctparams; - status = decode_text_parts (e->eb_content, encoding, decodetypes, message_mods); + status = decode_text_parts (e->eb_content, encoding, decodetypes, + message_mods); } break; @@ -1993,12 +1995,13 @@ decode_text_parts (CT ct, int encoding, const char *decodetypes, int *message_mo ct->c_cefile.ce_file = NULL; } else { int enc; + if (ct_encoding == CE_BINARY) { enc = CE_BINARY; } else if (ct_encoding == CE_8BIT && encoding == CE_7BIT) { enc = CE_QUOTED; } else { - enc = charset_encoding (ct); + enc = ct_encoding; } if (set_ce (ct, enc) == OK) { ++*message_mods; @@ -2283,6 +2286,103 @@ strip_crs (CT ct, int *message_mods) { } +/* + * Add/update, if necessary, the message C-T-E, based on the least restrictive + * of the part C-T-E's. + */ +static void +update_cte (CT ct) { + const int least_restrictive_enc = least_restrictive_encoding (ct); + + if (least_restrictive_enc != CE_UNKNOWN && + least_restrictive_enc != CE_7BIT) { + char *cte = concat (" ", ce_str (least_restrictive_enc), "\n", NULL); + HF hf; + int found_cte = 0; + + /* Update/add Content-Transfer-Encoding header field. */ + for (hf = ct->c_first_hf; hf; hf = hf->next) { + if (! strcasecmp (ENCODING_FIELD, hf->name)) { + found_cte = 1; + free (hf->value); + hf->value = cte; + } + } + if (! found_cte) { + add_header (ct, add (ENCODING_FIELD, NULL), cte); + } + } +} + + +/* + * Find the least restrictive encoding (7bit, 8bit, binary) of the parts + * within a message. + */ +static int +least_restrictive_encoding (CT ct) { + int encoding = CE_UNKNOWN; + + switch (ct->c_type) { + case CT_MULTIPART: { + struct multipart *m = (struct multipart *) ct->c_ctparams; + struct part *part; + + for (part = m->mp_parts; part; part = part->mp_next) { + const int part_encoding = + least_restrictive_encoding (part->mp_part); + + if (less_restrictive (encoding, part_encoding)) { + encoding = part_encoding; + } + } + break; + } + + case CT_MESSAGE: + if (ct->c_subtype == MESSAGE_EXTERNAL) { + struct exbody *e = (struct exbody *) ct->c_ctparams; + const int part_encoding = + least_restrictive_encoding (e->eb_content); + + if (less_restrictive (encoding, part_encoding)) { + encoding = part_encoding; + } + } + break; + + default: { + if (less_restrictive (encoding, ct->c_encoding)) { + encoding = ct->c_encoding; + } + }} + + return encoding; +} + + +/* + * Return whether the second encoding is less restrictive than the first, where + * "less restrictive" is in the sense used by RFC 2045 Secs. 6.1 and 6.4. So, + * CE_BINARY is less restrictive than CE_8BIT and + * CE_8BIT is less restrictive than CE_7BIT. + */ +static int +less_restrictive (int encoding, int second_encoding) { + switch (second_encoding) { + case CE_BINARY: + return encoding != CE_BINARY; + case CE_8BIT: + return encoding != CE_BINARY && encoding != CE_8BIT; + case CE_7BIT: + return encoding != CE_BINARY && encoding != CE_8BIT && + encoding != CE_7BIT; + default : + return 0; + } +} + + /* * Convert character set of each part. */