/* (C) Copyright 1993,1994 by Carnegie Mellon University * All Rights Reserved. * * Permission to use, copy, modify, distribute, and sell this software * and its documentation for any purpose is hereby granted without * fee, provided that the above copyright notice appear in all copies * and that both that copyright notice and this permission notice * appear in supporting documentation, and that the name of Carnegie * Mellon University not be used in advertising or publicity * pertaining to distribution of the software without specific, * written prior permission. Carnegie Mellon University makes no * representations about the suitability of this software for any * purpose. It is provided "as is" without express or implied * warranty. * * CARNEGIE MELLON UNIVERSITY DISCLAIMS ALL WARRANTIES WITH REGARD TO * THIS SOFTWARE, INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY * AND FITNESS, IN NO EVENT SHALL CARNEGIE MELLON UNIVERSITY BE LIABLE * FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN * AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING * OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS * SOFTWARE. */ #include #include #include #include "xmalloc.h" #include "common.h" #include "part.h" extern char *os_idtodir(); extern FILE *os_newtypedfile(); static FILE *startDescFile(); /* Length of a normal uuencoded line, including newline */ #define UULENGTH 62 /* * Table of valid boundary characters * * XXX: Old versions of Mark Crispin's c-client library * generate boundaries which contain the syntactically * illegal character '#'. It is marked in this table with * a 2 in case we want to use this table in the future to * complain about bad syntax. * */ static char bchar[256] = { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 2, 0, 0, 0, 1, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 1, 0, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, }; /* * Read an input file, looking for data in split-uuencode format */ handleUuencode(struct part *inpart, char *subject, int extractText) { char *fname = 0, *tmpfname; int part, nparts; int tmppart, tmpnparts; char buf[1024], buf2[1024]; char fnamebuf[80]; char *boundary_end, *p; int wantdescfile = 0; FILE *descfile = 0; /* Scan "Subject:" header for filename/part information */ if (parseSubject(subject, &fname, &part, &nparts) != 0) { part = -1; } if (part == 0) { return saveUuFile(inpart, fname, part, nparts, (char *)0); } if (part == 1) { wantdescfile = 1; } /* Scan body for interesting lines */ while (part_gets(buf, sizeof(buf), inpart)) { /* Uuencode "begin" line */ if (!strncmp(buf, "begin ", 6) && isdigit(buf[6]) && isdigit(buf[7]) && isdigit(buf[8]) && buf[9] == ' ') { if (part == -1) { /* * We have no part N of M information. Perhaps it is * a single-part uuencoded file. */ return saveUuFile(inpart, (char *)0, 1, 0, buf); } else { if (descfile) fclose(descfile); return saveUuFile(inpart, fname, part, nparts, buf); } } else if (!strncmp(buf, "section ", 8) && isdigit(buf[8])) { tmppart = 0; for (p = buf+8; isdigit(*p); p++) tmppart = tmppart*10 + *p - '0'; if (tmppart == 0) continue; if (strncmp(p, " of ", 4) == 0) { /* * "section N of ... of file F ..." */ for (p += 4; *p && strncmp(p, " of file ", 9) != 0; p++); if (!*p) continue; p += 9; tmpfname = p; p = strchr(p, ' '); if (!p) continue; *p = '\0'; if (descfile) fclose(descfile); return saveUuFile(inpart, tmpfname, tmppart, 0, (char *)0); } else if (*p == '/' && isdigit(p[1])) { /* * "section N/M file F ..." */ tmpnparts = 0; for (p++; isdigit(*p); p++) { tmpnparts = tmpnparts*10 + *p - '0'; } while (*p && isspace(*p)) p++; if (tmppart > tmpnparts || strncmp(p, "file ", 5) != 0) { continue; } tmpfname = p+5; p = strchr(tmpfname, ' '); if (!p) continue; *p = '\0'; if (descfile) fclose(descfile); return saveUuFile(inpart, tmpfname, tmppart, tmpnparts, (char *)0); } } else if (!strncmp(buf, "POST V", 6)) { /* * "POST Vd.d.d F (Part N/M)" */ p = strchr(buf+6, ' '); if (!p) continue; tmpfname = p+1; p = strchr(tmpfname, ' '); if (!p || strncmp(p, " (Part ", 7) != 0) continue; *p = '\0'; p += 7; tmppart = 0; while (isdigit(*p)) tmppart = tmppart*10 + *p++ - '0'; if (tmppart == 0 || *p++ != '/') continue; tmpnparts = 0; while (isdigit(*p)) tmpnparts = tmpnparts*10 + *p++ - '0'; if (tmppart > tmpnparts || *p != ')') continue; if (descfile) fclose(descfile); return saveUuFile(inpart, tmpfname, tmppart, tmpnparts, (char *)0); } else if (!strncmp(buf, "File: ", 6)) { /* * "File: F -- part N of M -- ... */ tmpfname = buf+6; p = strchr(tmpfname, ' '); if (!p || strncmp(p, " -- part ", 9) != 0) continue; *p = '\0'; p += 9; tmppart = 0; while (isdigit(*p)) tmppart = tmppart*10 + *p++ - '0'; if (tmppart == 0 || strncmp(p, " of ", 4) != 0) continue; p += 4; tmpnparts = 0; while (isdigit(*p)) tmpnparts = tmpnparts*10 + *p++ - '0'; if (tmppart > tmpnparts || strncmp(p, " -- ", 4) != 0) continue; if (descfile) fclose(descfile); return saveUuFile(inpart, tmpfname, tmppart, tmpnparts, (char *)0); } else if (!strncmp(buf, "[Section: ", 10)) { /* * "[Section: N/M File: F ..." */ tmppart = 0; for (p = buf+10; isdigit(*p); p++) tmppart = tmppart*10 + *p - '0'; if (tmppart == 0) continue; tmpnparts = 0; for (p++; isdigit(*p); p++) { tmpnparts = tmpnparts*10 + *p - '0'; } while (*p && isspace(*p)) p++; if (tmppart > tmpnparts || strncmp(p, "File: ", 6) != 0) { continue; } tmpfname = p+6; p = strchr(tmpfname, ' '); if (!p) continue; *p = '\0'; if (descfile) fclose(descfile); return saveUuFile(inpart, tmpfname, tmppart, tmpnparts, (char *)0); } else if (*buf == '[') { /* * "[F ... - part N of M]" * (usual BinHex practice) */ tmpfname = buf+1; p = strchr(tmpfname, ' '); if (!p) continue; *p++ = '\0'; while (p && strncmp(p, "- part ", 7) != 0) { p = strchr(p+1, '-'); } if (!p) continue; p += 7; tmppart = 0; while (isdigit(*p)) tmppart = tmppart*10 + *p++ - '0'; if (tmppart == 0 || strncmp(p, " of ", 4) != 0) continue; p += 4; tmpnparts = 0; while (isdigit(*p)) tmpnparts = tmpnparts*10 + *p++ - '0'; if (tmppart > tmpnparts || *p != ']') continue; if (descfile) fclose(descfile); return saveUuFile(inpart, tmpfname, tmppart, tmpnparts, (char *)0); } else if (fname && part > 0 && nparts > 0 && part <= nparts && (!strncmp(buf, "BEGIN", 5) || !strncmp(buf, "--- BEGIN ---", 12) || (buf[0] == 'M' && strlen(buf) == UULENGTH))) { /* * Found the start of a section of uuencoded data * and have the part N of M information. */ if (descfile) fclose(descfile); return saveUuFile(inpart, fname, part, nparts, buf); } else if (!cistrncmp(buf, "x-file-name: ", 13)) { for (p = buf + 13; *p && !isspace(*p); p++); *p = '\0'; strncpy(fnamebuf, buf+13, sizeof(fnamebuf)-1); fnamebuf[sizeof(fnamebuf)-1] = '\0'; fname = fnamebuf; continue; } else if (!cistrncmp(buf, "x-part: ", 8)) { tmppart = atoi(buf+8); if (tmppart > 0) part = tmppart; continue; } else if (!cistrncmp(buf, "x-part-total: ", 14)) { tmpnparts = atoi(buf+14); if (tmpnparts > 0) nparts = tmpnparts; continue; } else if (part == 1 && fname && !descfile && !cistrncmp(buf, "x-file-desc: ", 13)) { if (descfile = startDescFile(fname)) { fputs(buf+13, descfile); if (descfile) fclose(descfile); descfile = 0; } continue; } else if (!strcmp(buf, "(This file must be converted with BinHex 4.0)\n")) { if (descfile) fclose(descfile); return os_binhex(inpart, 1, 1); } else if (!cistrncmp(buf, "content-", 8)) { /* * HEURISTIC: If we see something that looks like a content-* * header, push it back and call the message parser. */ p = buf+8; /* Check to see if header's field-name is syntactically valid */ while (*p) { if (*p == ':' || *p <= ' ' || *p >= '\177') break; p++; } if (*p == ':') { part_ungets(buf, inpart); if (descfile) fclose(descfile); return handleMessage(inpart, "text/plain", 0, extractText); } } if (buf[0] == '-' && buf[1] == '-') { /* * Heuristic: If we see something that looks like a * multipart boundary, followed by something that looks * like a header, push them back and parse as a multipart. */ p = buf+2; while (*p) { if (!bchar[(unsigned char)*p]) break; p++; } if (*p != '\n') { /* * We found an invalid boundary character. * Move 'p' such that it will fail all subsequent checks. */ p = buf + 2; } /* Back up to ignore trailing whitespace */ while (p > buf+2 && p[-1] == ' ') p--; /* * Check that boundary is within legal size limits * If so, peek at next line */ if (p - buf > 2 && p - buf <= 72 && part_gets(buf2, sizeof(buf2), inpart)) { boundary_end = p; p = buf2; /* * Check to see if a syntactically valid header follows * what looks to be a boundary. * * XXX: Unfortunately, we can't check for "Content-"; * it is syntactically valid to have a body-part * header that doesn't start with that and ZMail * takes advantage of that. If this heuristic starts * causing problems, we could keep looking ahead until * we find a "Content-" header or find something that's * not a header. */ while (*p) { if (*p == ':' || *p <= ' ' || *p >= '\177') break; p++; } /* Push back the lookahead line */ part_ungets(buf2, inpart); if (p > buf2 && *p == ':') { /* Push back the boundary */ part_ungets(buf, inpart); /* * Generate and push back a header to get us into * the multipart parser. */ *boundary_end = '\0'; sprintf(buf2, "Content-type: multipart/mixed; boundary=\"%s\"\n\n", buf+2); part_ungets(buf2, inpart); if (descfile) fclose(descfile); return handleMessage(inpart, "text/plain", 0, extractText); } } } /* * Save useful-looking text that is before a "part 1 of N" * in a description file. */ if (wantdescfile && !descfile) { for (p = buf; *p && isspace(*p); p++); if (*p) { if (!cistrncmp(p, "x-", 2)) { /* * Check for "X-foobar:" * If so, there probably will be a "X-File-Desc:" line * later, so ignore this line. */ while (*p != ':' && *p > ' ' && *p < '\177') p++; if (*p == ':') continue; } if (!descEnd(buf) && (descfile = startDescFile(fname))) { fputs(buf, descfile); } wantdescfile = 0; } } else if (descfile) { if (descEnd(buf)) { if (descfile) fclose(descfile); descfile = 0; } else { fputs(buf, descfile); } } } if (descfile) fclose(descfile); return 0; } /* * Handle a split-uuencode part * If nparts is 0, then look for an "end" line to detect the last part. * If fname is null, then we are attempting to decode a single-part message. * If firstline is non-null, it is written as the first line of the saved part */ int saveUuFile(struct part *inpart, char *fname, int part, int nparts, char *firstline) { char buf[1024]; char *dir; FILE *partfile; if (fname) { sprintf(buf, "Saving part %d ", part); if (nparts) sprintf(buf+strlen(buf), "of %d ", nparts); strcat(buf, fname); chat(buf); } else fname = "unknown"; /* Create directory to store parts and copy this part there. */ dir = os_idtodir(fname); if (!dir) return 1; sprintf(buf, "%s%d", dir, part); partfile = fopen(buf, "w"); if (!partfile) { os_perror(buf); return 1; } if (firstline) fputs(firstline, partfile); while (part_gets(buf, sizeof(buf), inpart)) { fputs(buf, partfile); if (nparts == 0 && strcmp(buf, "end\n") == 0) { /* This is the last part. Remember the fact */ nparts = part; if (partfile) fclose(partfile); sprintf(buf, "%sCT", dir); partfile = fopen(buf, "w"); if (!partfile) { os_perror(buf); } else { fprintf(partfile, "%d\n", nparts); } break; } } if (partfile) fclose(partfile); /* Retrieve any previously saved number of the last part */ if (nparts == 0) { sprintf(buf, "%sCT", dir); if (partfile = fopen(buf, "r")) { if (fgets(buf, sizeof(buf), partfile)) { nparts = atoi(buf); if (nparts < 0) nparts = 0; } if (partfile) fclose(partfile); } } if (nparts == 0) return 0; /* Check to see if we have all parts. Start from the highest numbers * as we are more likely not to have them. */ for (part = nparts; part; part--) { sprintf(buf, "%s%d", dir, part); partfile = fopen(buf, "r"); if (partfile) { fclose(partfile); } else { return 0; } } return uudecodefiles(dir, nparts); } /* * Parse a Subject: header, looking for clues with which to decode * split-uuencoded data. */ int parseSubject(char *subject, char **fnamep, int *partp, int *npartsp) { char *scan, *bak, *start; int part = -1, nparts = 0, hasdot = 0; /* No subject header */ if (!subject) return 1; /* Skip leading whitespace and other garbage */ scan = subject; while (*scan == ' ' || *scan == '\t' || *scan == '-') scan++; if (!cistrncmp(scan, "repost", 6)) { for (scan += 6; *scan == ' ' || *scan == '\t' || *scan == ':' || *scan == '-'; scan++); } /* Replies aren't usually data */ if (!cistrncmp(scan, "re:", 3)) return 1; /* Get filename */ /* Grab the first filename-like string. Explicitly ignore strings with * prefix "v" ending in ":", since that is a popular volume/issue * representation syntax */ do { while (*scan != '\n' && isprint(*scan) && !isalnum(*scan) && *scan != '_') ++scan; *fnamep = start = scan; while (isalnum(*scan) || *scan == '-' || *scan == '+' || *scan == '&' || *scan == '_' || *scan == '.') { if (*scan++ == '.') hasdot = 1; } if (!*scan || *scan == '\n') return 1; } while (start == scan || (start[0] == 'v' && isdigit(start[1]) && *scan == ':')); *scan++ = '\0'; /* Try looking for a filename with a "." in it later in the subject line. * Exclude ., since that is usually a version number. */ if (!hasdot) { while (*(start = scan) != '\0' && *scan != '\n') { while (isspace(*start)) ++start; for (scan = start; isalnum(*scan) || *scan == '-' || *scan == '+' || *scan == '&' || *scan == '_' || *scan == '.'; ++scan) { if (*scan == '.' && (!isdigit(scan[-1]) || !isdigit(scan[1]))) { hasdot = 1; } } if (hasdot && scan > start) { *fnamep = start; *scan++ = '\0'; break; } while (*scan && *scan != '\n' && !isalnum(*scan)) ++scan; } scan = *fnamep + strlen(*fnamep) + 1; } /* Get part number */ while (*scan && *scan != '\n') { /* skip over versioning */ if (*scan == 'v' && isdigit(scan[1])) { ++scan; while (isdigit(*scan)) ++scan; } /* look for "1/6" or "1 / 6" or "1 of 6" or "1-of-6" or "1o6" */ if (isdigit(*scan) && (scan[1] == '/' || (scan[1] == ' ' && scan[2] == '/') || (scan[1] == ' ' && scan[2] == 'o' && scan[3] == 'f') || (scan[1] == '-' && scan[2] == 'o' && scan[3] == 'f') || (scan[1] == 'o' && isdigit(scan[2])))) { while (isdigit(scan[-1])) scan--; part = 0; while (isdigit(*scan)) { part = part * 10 + *scan++ - '0'; } while (*scan != '\0' && *scan != '\n' && !isdigit(*scan)) scan++; if (isdigit(*scan)) { nparts = 0; while (isdigit(*scan)) { nparts = nparts * 10 + *scan++ - '0'; } } break; } /* look for "6 parts" or "part 1" */ if (!cistrncmp("part", scan, 4)) { if (scan[4] == 's') { for (bak = scan; bak >= subject && !isdigit(*bak); bak--); if (bak > subject) { while (bak > subject && isdigit(bak[-1])) bak--; nparts = 0; while (isdigit(*bak)) { nparts = nparts * 10 + *bak++ - '0'; } } } else { while (*scan && *scan != '\n' && !isdigit(*scan)) scan++; bak = scan - 1; if (isdigit(*scan)) { part = 0; do { part = part * 10 + *scan++ - '0'; } while (isdigit(*scan)); } scan = bak; } } scan++; } if (nparts == 0 || part == -1 || part > nparts) return 1; *partp = part; *npartsp = nparts; return 0; } /* * Return nonzero if 'line' should mark the end of a part-1 description */ int descEnd(char *line) { return !strncmp(line, "---", 3) || !strncmp(line, "#!", 2) || !cistrncmp(line, "part=", 5) || !cistrncmp(line, "begin", 5); } /* * Open and return a file pointer for a description file for 'fname'. * If a description file for 'fname' already exists, or if there is an * error, return a null pointer. */ static FILE *startDescFile(char *fname) { char buf[1024]; char *dir; FILE *descfile; /* Create directory to store parts and copy this part there. */ dir = os_idtodir(fname); if (!dir) return 0; sprintf(buf, "%s0", dir); /* See if part 0 already exists, return failure if so */ descfile = fopen(buf, "r"); if (descfile) { fclose(descfile); return 0; } descfile = fopen(buf, "w"); if (!descfile) { os_perror(buf); return 0; } return descfile; } /* * Decode the uuencoded file that is in 'nparts' pieces in 'dir'. */ int uudecodefiles(char *dir, int nparts) { int part; enum {st_start, st_inactive, st_decode, st_nextlast, st_last, st_binhex} state; FILE *infile; FILE *outfile = NULL; struct part *inpart; char buf[1024]; char lastline[UULENGTH+1]; char *fname, *p; char *contentType = "application/octet-stream"; int line_length = 0; /* If a part 0, copy to description filename */ sprintf(buf, "%s0", dir); infile = fopen(buf, "r"); if (infile) { outfile = fopen(TEMPFILENAME, "w"); if (outfile) { while (fgets(buf, sizeof(buf), infile)) { fputs(buf, outfile); } if (outfile) fclose(outfile); outfile = NULL; } if (infile) fclose(infile); sprintf(buf, "%s0", dir); remove(buf); } state = st_start; /* Handle each part in order */ for (part = 1; part <= nparts; part++) { sprintf(buf, "%s%d", dir, part); infile = fopen(buf, "r"); if (!infile) { os_perror(buf); if (outfile) fclose(outfile); remove(TEMPFILENAME); return 1; } while (fgets(buf, sizeof(buf), infile)) { switch (state) { case st_start: /* Looking for start of uuencoded * or binhex'ed file */ if (!strcmp(buf, "(This file must be converted with BinHex 4.0)\n")) { state = st_binhex; inpart = part_init(infile); os_binhex(inpart, part, nparts); part_close(inpart); goto endbinhex; } if (strncmp(buf, "begin ", 6)) break; /* skip mode */ p = buf + 6; while (*p && !isspace(*p)) p++; while (*p && isspace(*p)) p++; fname = p; while (*p && !isspace(*p)) p++; *p = '\0'; if (!*fname) return 1; /* Guess the content-type of common filename extensions */ if (p = strrchr(fname, '.')) { if (!cistrcmp(p, ".gif")) contentType = "image/gif"; if (!cistrcmp(p, ".jpg")) contentType = "image/jpeg"; if (!cistrcmp(p, ".jpeg")) contentType = "image/jpeg"; if (!cistrcmp(p, ".mpg")) contentType = "video/mpeg"; if (!cistrcmp(p, ".mpeg")) contentType = "video/mpeg"; } /* Create output file and start decoding */ outfile = os_newtypedfile(fname, contentType, FILE_BINARY, (params) 0); if (!outfile) { if (infile) fclose(infile); return 1; } state = st_decode; break; case st_inactive: /* Looking for uuencoded data to resume */ if (*buf != 'M' || strlen(buf) != line_length) { if (*buf == 'B' && !strncmp(buf, "BEGIN", 5)) { state = st_decode; } break; } state = st_decode; /* FALL THROUGH */ case st_decode: /* Decoding data */ if (line_length == 0) line_length = strlen(buf); if (*buf == 'M' && strlen(buf) == line_length) { uudecodeline(buf, outfile); break; } if (strlen(buf) > line_length) { state = st_inactive; break; } /* * May be on nearing end of file. * Save this line in case we are. */ strcpy(lastline, buf); if (*buf == ' ' || *buf == '`') { state = st_last; } else { state = st_nextlast; } break; case st_nextlast: /* May be nearing end of file */ if (*buf == ' ' || *buf == '`') { state = st_last; } else { state = st_inactive; } break; case st_last: /* Should be at end of file */ if (!strncmp(buf, "end", 3) && isspace(buf[3])) { /* Handle that last line we saved */ uudecodeline(lastline, outfile); if (infile) fclose(infile); os_closetypedfile(outfile); for (;part <= nparts; part++) { sprintf(buf, "%s%d", dir, part); remove(buf); } sprintf(buf, "%sCT", dir); remove(buf); os_donewithdir(dir); return 0; } state = st_inactive; break; case st_binhex: if (strncmp(buf, "---", 3)) break; inpart = part_init(infile); os_binhex(inpart, part, nparts); part_close(inpart); goto endbinhex; } } if (state != st_binhex) state = st_inactive; if (infile) fclose(infile); endbinhex: sprintf(buf, "%s%d", dir, part); remove(buf); } if (outfile) os_closetypedfile(outfile); if (state == st_binhex) os_binhex(0, 0, 0); sprintf(buf, "%sCT", dir); remove(buf); os_donewithdir(dir); return 0; } #define DEC(c) (((c) - ' ') & 077) /* * Decode a uuencoded line to 'outfile' */ uudecodeline(char *line, FILE *outfile) { int c, len; len = DEC(*line++); while (len) { c = DEC(*line) << 2 | DEC(line[1]) >> 4; putc(c, outfile); if (--len) { c = DEC(line[1]) << 4 | DEC(line[2]) >> 2; putc(c, outfile); if (--len) { c = DEC(line[2]) << 6 | DEC(line[3]); putc(c, outfile); len--; } } line += 4; } return; }