X-Git-Url: https://diplodocus.org/git/nmh/blobdiff_plain/7cb4e52b1051d8f5eab4ebc157831bbfb084d22e..d2520ac7054ad75d60342606bf13c821305d958c:/uip/mhfixmsg.c diff --git a/uip/mhfixmsg.c b/uip/mhfixmsg.c index 07433377..f00fc97d 100644 --- a/uip/mhfixmsg.c +++ b/uip/mhfixmsg.c @@ -12,15 +12,12 @@ #include #include #include -#ifdef HAVE_ICONV -# include -#endif #define MHFIXMSG_SWITCHES \ X("decodetext 8bit|7bit", 0, DECODETEXTSW) \ X("nodecodetext", 0, NDECODETEXTSW) \ - X("textcodeset", 0, TEXTCODESETSW) \ - X("notextcodeset", 0, NTEXTCODESETSW) \ + X("textcharset", 0, TEXTCHARSETSW) \ + X("notextcharset", 0, NTEXTCHARSETSW) \ X("reformat", 0, REFORMATSW) \ X("noreformat", 0, NREFORMATSW) \ X("replacetextplain", 0, REPLACETEXTPLAINSW) \ @@ -81,15 +78,14 @@ typedef struct fix_transformations { int reformat; int replacetextplain; int decodetext; - char *textcodeset; + char *textcharset; } fix_transformations; int mhfixmsgsbr (CT *, const fix_transformations *, char *); static void reverse_alternative_parts (CT); static int fix_boundary (CT *, int *); static int get_multipart_boundary (CT, char **); -static int replace_boundary (CT, char *, const char *); -static char *update_attr (char *, const char *, const char *e); +static int replace_boundary (CT, char *, char *); static int fix_multipart_cte (CT, int *); static int set_ce (CT, int); static int ensure_text_plain (CT *, CT, int *, int); @@ -104,15 +100,12 @@ static int boundary_in_content (FILE **, char *, const char *); static void transfer_noncontent_headers (CT, CT); static int set_ct_type (CT, int type, int subtype, int encoding); static int decode_text_parts (CT, int, int *); -static int content_encoding (CT); +static int content_encoding (CT, const char **); static int strip_crs (CT, int *); -static int convert_codesets (CT, char *, int *); -static int convert_codeset (CT, char *, int *); -static char *content_codeset (CT); +static int convert_charsets (CT, char *, int *); static int write_content (CT, char *, char *, int, int); static int remove_file (char *); static void report (char *, char *, char *, ...); -static char *upcase (char *); static void pipeser (int); @@ -132,7 +125,7 @@ main (int argc, char **argv) { fx.reformat = fx.fixcte = fx.fixboundary = 1; fx.replacetextplain = 0; fx.decodetext = CE_8BIT; - fx.textcodeset = NULL; + fx.textcharset = NULL; if (nmh_init(argv[0], 1)) { return 1; } @@ -176,13 +169,13 @@ main (int argc, char **argv) { case NDECODETEXTSW: fx.decodetext = 0; continue; - case TEXTCODESETSW: + case TEXTCHARSETSW: if (! (cp = *argp++) || (*cp == '-' && cp[1])) adios (NULL, "missing argument to %s", argp[-2]); - fx.textcodeset = cp; + fx.textcharset = cp; continue; - case NTEXTCODESETSW: - fx.textcodeset = 0; + case NTEXTCHARSETSW: + fx.textcharset = 0; continue; case FIXBOUNDARYSW: fx.fixboundary = 1; @@ -415,8 +408,8 @@ mhfixmsgsbr (CT *ctp, const fix_transformations *fx, char *outfile) { if (status == OK && fx->decodetext) { status = decode_text_parts (*ctp, fx->decodetext, &message_mods); } - if (status == OK && fx->textcodeset != NULL) { - status = convert_codesets (*ctp, fx->textcodeset, &message_mods); + if (status == OK && fx->textcharset != NULL) { + status = convert_charsets (*ctp, fx->textcharset, &message_mods); } if (! (*ctp)->c_umask) { @@ -629,7 +622,7 @@ get_multipart_boundary (CT ct, char **part_boundary) { /* Open and copy ct->c_file to file, replacing the multipart boundary. */ static int -replace_boundary (CT ct, char *file, const char *boundary) { +replace_boundary (CT ct, char *file, char *boundary) { FILE *fpin, *fpout; int compnum, state; char buf[BUFSIZ], name[NAMESZ]; @@ -675,10 +668,22 @@ replace_boundary (CT ct, char *file, const char *boundary) { if (strcasecmp (TYPE_FIELD, np)) { fprintf (fpout, "%s:%s", np, vp); } else { - char *new_boundary = update_attr (vp, "boundary=", boundary); - - fprintf (fpout, "%s:%s\n", np, new_boundary); - free (new_boundary); + char *new_ctline, *new_params; + + replace_param(&ct->c_ctinfo.ci_first_pm, + &ct->c_ctinfo.ci_last_pm, "boundary", + boundary, 0); + + new_ctline = concat(" ", ct->c_ctinfo.ci_type, "/", + ct->c_ctinfo.ci_subtype, NULL); + new_params = output_params(strlen(TYPE_FIELD) + + strlen(new_ctline) + 1, + ct->c_ctinfo.ci_first_pm, NULL, 0); + fprintf (fpout, "%s:%s%s\n", np, new_ctline, + new_params ? new_params : ""); + free(new_ctline); + if (new_params) + free(new_params); } free (vp); @@ -718,44 +723,6 @@ replace_boundary (CT ct, char *file, const char *boundary) { } -/* Change the value of a name=value pair in a header field body. - If the name isn't there, append them. In any case, a new - string will be allocated and must be free'd by the caller. - Trims any trailing newlines. */ -static char * -update_attr (char *body, const char *name, const char *value) { - char *bp = nmh_strcasestr (body, name); - char *new_body; - - if (bp) { - char *other_attrs = strchr (bp, ';'); - - *(bp + strlen (name)) = '\0'; - new_body = concat (body, "\"", value, "\"", NULL); - - if (other_attrs) { - char *cp; - - /* Trim any trailing newlines. */ - for (cp = &other_attrs[strlen (other_attrs) - 1]; - cp > other_attrs && *cp == '\n'; - *cp-- = '\0') continue; - new_body = add (other_attrs, new_body); - } - } else { - char *cp; - - /* Append name/value pair, after first removing a final newline - and (extraneous) semicolon. */ - if (*(cp = &body[strlen (body) - 1]) == '\n') *cp = '\0'; - if (*(cp = &body[strlen (body) - 1]) == ';') *cp = '\0'; - new_body = concat (body, "; ", name, "\"", value, "\"", NULL); - } - - return new_body; -} - - static int fix_multipart_cte (CT ct, int *message_mods) { int status = OK; @@ -1196,11 +1163,11 @@ reformat_part (CT ct, char *file, char *type, char *subtype, int c_type) { static int charset_encoding (CT ct) { /* norm_charmap() is case sensitive. */ - char *codeset = upcase (content_codeset (ct)); + char *charset = upcase (content_charset (ct)); int encoding = - strcmp (norm_charmap (codeset), "US-ASCII") ? CE_8BIT : CE_7BIT; + strcmp (norm_charmap (charset), "US-ASCII") ? CE_8BIT : CE_7BIT; - free (codeset); + free (charset); return encoding; } @@ -1468,16 +1435,19 @@ decode_text_parts (CT ct, int encoding, int *message_mods) { int ct_encoding; if (decode_part (ct) == OK && ct->c_cefile.ce_file) { - if ((ct_encoding = content_encoding (ct)) == CE_BINARY && - encoding != CE_BINARY) { + const char *reason = NULL; + + if ((ct_encoding = content_encoding (ct, &reason)) == CE_BINARY + && encoding != CE_BINARY) { /* The decoding isn't acceptable so discard it. Leave status as OK to allow other transformations. */ if (verbosw) { report (ct->c_partno, ct->c_file, - "will not decode%s because it is binary", + "will not decode%s because it is binary (%s)", ct->c_partno ? "" : ct->c_ctline ? ct->c_ctline - : ""); + : "", + reason); } (void) m_unlink (ct->c_cefile.ce_file); free (ct->c_cefile.ce_file); @@ -1561,9 +1531,10 @@ decode_text_parts (CT ct, int encoding, int *message_mods) { /* See if the decoded content is 7bit, 8bit, or binary. It's binary if it has any NUL characters, a CR not followed by a LF, or lines - greater than 998 characters in length. */ + greater than 998 characters in length. If binary, reason is set + to a string explaining why. */ static int -content_encoding (CT ct) { +content_encoding (CT ct, const char **reason) { CE ce = &ct->c_cefile; int encoding = CE_7BIT; @@ -1588,6 +1559,16 @@ content_encoding (CT ct) { if (*cp == '\0' || ++line_len > 998 || (*cp != '\n' && last_char_was_cr)) { encoding = CE_BINARY; + if (*cp == '\0') { + *reason = "null character"; + } else if (line_len > 998) { + *reason = "line length > 998"; + } else if (*cp != '\n' && last_char_was_cr) { + *reason = "CR not followed by LF"; + } else { + /* Should not reach this. */ + *reason = ""; + } break; } else if (*cp == '\n') { line_len = 0; @@ -1610,18 +1591,18 @@ content_encoding (CT ct) { static int strip_crs (CT ct, int *message_mods) { /* norm_charmap() is case sensitive. */ - char *codeset = upcase (content_codeset (ct)); + char *charset = upcase (content_charset (ct)); int status = OK; /* Only strip carriage returns if content is ASCII or another - codeset that has the same readily recognizable CR followed by a + charset that has the same readily recognizable CR followed by a LF. We can include UTF-8 here because if the high-order bit of a UTF-8 byte is 0, then it must be a single-byte ASCII character. */ - if (! strcmp (norm_charmap (codeset), "US-ASCII") || - ! strncmp (norm_charmap (codeset), "ISO-8859-", 9) || - ! strncmp (norm_charmap (codeset), "UTF-8", 5) || - ! strncmp (norm_charmap (codeset), "WINDOWS-12", 10)) { + if (! strcmp (norm_charmap (charset), "US-ASCII") || + ! strncmp (norm_charmap (charset), "ISO-8859-", 9) || + ! strncmp (norm_charmap (charset), "UTF-8", 5) || + ! strncmp (norm_charmap (charset), "WINDOWS-12", 10)) { char **file = NULL; FILE **fp = NULL; size_t begin; @@ -1683,7 +1664,7 @@ strip_crs (CT ct, int *message_mods) { if (has_crs) { int fd; char *stripped_content_file; - char *tempfile = m_mktemp2 (NULL, invo_name, &fd, NULL); + char *tempfile = m_mktemp2 (NULL, invo_name, &fd, NULL); if (tempfile == NULL) { adios (NULL, "unable to create temporary file in %s", @@ -1745,45 +1726,23 @@ strip_crs (CT ct, int *message_mods) { } } - free (codeset); + free (charset); return status; } -char * -content_codeset (CT ct) { - const char *const charset = "charset"; - char *default_codeset = NULL; - CI ctinfo = &ct->c_ctinfo; - char **src_codeset = NULL; - PM pm; - - for (pm = ctinfo->ci_first_pm; pm; pm = pm->pm_next) { - if (! strcasecmp (pm->pm_name, charset)) { - src_codeset = &pm->pm_value; - break; - } - } - - /* RFC 2045, Sec. 5.2: default to us-ascii. */ - if (src_codeset == NULL) { - src_codeset = &default_codeset; - if (*src_codeset == NULL) - *src_codeset = "US-ASCII"; - } - - return *src_codeset; -} - - static int -convert_codesets (CT ct, char *dest_codeset, int *message_mods) { +convert_charsets (CT ct, char *dest_charset, int *message_mods) { int status = OK; switch (ct->c_type) { case CT_TEXT: if (ct->c_subtype == TEXT_PLAIN) { - status = convert_codeset (ct, dest_codeset, message_mods); + status = convert_charset (ct, dest_charset, message_mods); + if (verbosw && status == OK) { + report (ct->c_partno, ct->c_file, "convert %s to %s", + content_charset(ct), dest_charset); + } } break; @@ -1795,7 +1754,7 @@ convert_codesets (CT ct, char *dest_codeset, int *message_mods) { For now, it gets passed along as-is by InitMultiPart(). */ for (part = m->mp_parts; status == OK && part; part = part->mp_next) { status = - convert_codesets (part->mp_part, dest_codeset, message_mods); + convert_charsets (part->mp_part, dest_charset, message_mods); } break; } @@ -1806,7 +1765,7 @@ convert_codesets (CT ct, char *dest_codeset, int *message_mods) { e = (struct exbody *) ct->c_ctparams; status = - convert_codesets (e->eb_content, dest_codeset, message_mods); + convert_charsets (e->eb_content, dest_charset, message_mods); } break; @@ -1818,161 +1777,6 @@ convert_codesets (CT ct, char *dest_codeset, int *message_mods) { } -static int -convert_codeset (CT ct, char *dest_codeset, int *message_mods) { - char *src_codeset = content_codeset (ct); - int status = OK; - - /* norm_charmap() is case sensitive. */ - char *src_codeset_u = upcase (src_codeset); - char *dest_codeset_u = upcase (dest_codeset); - int different_codesets = - strcmp (norm_charmap (src_codeset), norm_charmap (dest_codeset)); - - free (dest_codeset_u); - free (src_codeset_u); - - if (different_codesets) { -#ifdef HAVE_ICONV - iconv_t conv_desc = NULL; - char *dest; - int fd = -1; - char **file = NULL; - FILE **fp = NULL; - size_t begin; - size_t end; - int opened_input_file = 0; - char src_buffer[BUFSIZ]; - HF hf; - char *tempfile; - - if ((conv_desc = iconv_open (dest_codeset, src_codeset)) == - (iconv_t) -1) { - advise (NULL, "Can't convert %s to %s", src_codeset, dest_codeset); - return -1; - } - - if ((tempfile = m_mktemp2 (NULL, invo_name, &fd, NULL)) == NULL) { - adios (NULL, "unable to create temporary file in %s", - get_temp_dir()); - } - dest = add (tempfile, NULL); - - if (ct->c_cefile.ce_file) { - file = &ct->c_cefile.ce_file; - fp = &ct->c_cefile.ce_fp; - begin = end = 0; - } else if (ct->c_file) { - file = &ct->c_file; - fp = &ct->c_fp; - begin = (size_t) ct->c_begin; - end = (size_t) ct->c_end; - } /* else no input file: shouldn't happen */ - - if (file && *file && fp) { - if (! *fp) { - if ((*fp = fopen (*file, "r")) == NULL) { - advise (*file, "unable to open for reading"); - status = NOTOK; - } else { - opened_input_file = 1; - } - } - } - - if (fp && *fp) { - size_t inbytes; - size_t bytes_to_read = - end > 0 && end > begin ? end - begin : sizeof src_buffer; - - fseeko (*fp, begin, SEEK_SET); - while ((inbytes = fread (src_buffer, 1, - min (bytes_to_read, sizeof src_buffer), - *fp)) > 0) { - char dest_buffer[BUFSIZ]; - ICONV_CONST char *ib = src_buffer; - char *ob = dest_buffer; - size_t outbytes = sizeof dest_buffer; - size_t outbytes_before = outbytes; - - if (end > 0) bytes_to_read -= inbytes; - - if (iconv (conv_desc, &ib, &inbytes, &ob, &outbytes) == - (size_t) -1) { - status = NOTOK; - break; - } else { - write (fd, dest_buffer, outbytes_before - outbytes); - } - } - - if (opened_input_file) { - fclose (*fp); - *fp = NULL; - } - } - - iconv_close (conv_desc); - close (fd); - - if (status == OK) { - /* Replace the decoded file with the converted one. */ - if (ct->c_cefile.ce_file) { - if (ct->c_cefile.ce_unlink) { - (void) m_unlink (ct->c_cefile.ce_file); - } - free (ct->c_cefile.ce_file); - } - ct->c_cefile.ce_file = dest; - ct->c_cefile.ce_unlink = 1; - - ++*message_mods; - if (verbosw) { - report (ct->c_partno, ct->c_file, "convert %s to %s", - src_codeset, dest_codeset); - } - - /* Update ci_attrs. */ - src_codeset = dest_codeset; - - /* Update ct->c_ctline. */ - if (ct->c_ctline) { - char *ctline = - update_attr (ct->c_ctline, "charset=", dest_codeset); - - free (ct->c_ctline); - ct->c_ctline = ctline; - } /* else no CT line, which is odd */ - - /* Update Content-Type header field. */ - for (hf = ct->c_first_hf; hf; hf = hf->next) { - if (! strcasecmp (TYPE_FIELD, hf->name)) { - char *ctline_less_newline = - update_attr (hf->value, "charset=", dest_codeset); - char *ctline = concat (ctline_less_newline, "\n", NULL); - free (ctline_less_newline); - - free (hf->value); - hf->value = ctline; - break; - } - } - } else { - (void) m_unlink (dest); - } -#else /* ! HAVE_ICONV */ - NMH_UNUSED (message_mods); - - advise (NULL, "Can't convert %s to %s without iconv", src_codeset, - dest_codeset); - status = NOTOK; -#endif /* ! HAVE_ICONV */ - } - - return status; -} - - static int write_content (CT ct, char *input_filename, char *outfile, int modify_inplace, int message_mods) { @@ -2086,17 +1890,6 @@ report (char *partno, char *filename, char *message, ...) { } -static char * -upcase (char *str) { - char *up = cpytrim (str); - char *cp; - - for (cp = up; *cp; ++cp) *cp = toupper ((unsigned char) *cp); - - return up; -} - - static void pipeser (int i) {