Main Page | Namespace List | Alphabetical List | Data Structures | Directories | File List | Data Fields | Globals

readpst.c

Go to the documentation of this file.
00001 /***
00002  * readpst.c
00003  * Part of the LibPST project
00004  * Written by David Smith
00005  *            dave.s@earthcorp.com
00006  */
00007 
00008 #include "define.h"
00009 #include "lzfu.h"
00010 
00011 #define OUTPUT_TEMPLATE "%s"
00012 #define OUTPUT_KMAIL_DIR_TEMPLATE ".%s.directory"
00013 #define KMAIL_INDEX ".%s.index"
00014 #define SEP_MAIL_FILE_TEMPLATE "%i%s"
00015 
00016 // max size of the c_time char*. It will store the date of the email
00017 #define C_TIME_SIZE 500
00018 
00019 struct file_ll {
00020     char *name;
00021     char *dname;
00022     FILE * output;
00023     int32_t stored_count;
00024     int32_t item_count;
00025     int32_t skip_count;
00026     int32_t type;
00027 };
00028 
00029 int       grim_reaper();
00030 pid_t     try_fork(char* folder);
00031 void      process(pst_item *outeritem, pst_desc_tree *d_ptr);
00032 void      write_email_body(FILE *f, char *body);
00033 void      removeCR(char *c);
00034 void      usage();
00035 void      version();
00036 char*     mk_kmail_dir(char* fname);
00037 int       close_kmail_dir();
00038 char*     mk_recurse_dir(char* dir, int32_t folder_type);
00039 int       close_recurse_dir();
00040 char*     mk_separate_dir(char *dir);
00041 int       close_separate_dir();
00042 int       mk_separate_file(struct file_ll *f, char *extension);
00043 char*     my_stristr(char *haystack, char *needle);
00044 void      check_filename(char *fname);
00045 void      write_separate_attachment(char f_name[], pst_item_attach* attach, int attach_num, pst_file* pst);
00046 void      write_embedded_message(FILE* f_output, pst_item_attach* attach, char *boundary, pst_file* pf, char** extra_mime_headers);
00047 void      write_inline_attachment(FILE* f_output, pst_item_attach* attach, char *boundary, pst_file* pst);
00048 void      header_has_field(char *header, char *field, int *flag);
00049 void      header_get_subfield(char *field, const char *subfield, char *body_subfield, size_t size_subfield);
00050 char*     header_get_field(char *header, char *field);
00051 char*     header_end_field(char *field);
00052 void      header_strip_field(char *header, char *field);
00053 int       test_base64(char *body);
00054 void      find_html_charset(char *html, char *charset, size_t charsetlen);
00055 void      find_rfc822_headers(char** extra_mime_headers);
00056 void      write_body_part(FILE* f_output, pst_string *body, char *mime, char *charset, char *boundary, pst_file* pst);
00057 void      write_schedule_part_data(FILE* f_output, pst_item* item, const char* sender, const char* method);
00058 void      write_schedule_part(FILE* f_output, pst_item* item, const char* sender, const char* boundary);
00059 void      write_normal_email(FILE* f_output, char f_name[], pst_item* item, int mode, int mode_MH, pst_file* pst, int save_rtf, char** extra_mime_headers);
00060 void      write_vcard(FILE* f_output, pst_item *item, pst_item_contact* contact, char comment[]);
00061 int       write_extra_categories(FILE* f_output, pst_item* item);
00062 void      write_journal(FILE* f_output, pst_item* item);
00063 void      write_appointment(FILE* f_output, pst_item *item);
00064 void      create_enter_dir(struct file_ll* f, pst_item *item);
00065 void      close_enter_dir(struct file_ll *f);
00066 
00067 const char*  prog_name;
00068 char*  output_dir = ".";
00069 char*  kmail_chdir = NULL;
00070 
00071 // Normal mode just creates mbox format files in the current directory. Each file is named
00072 // the same as the folder's name that it represents
00073 #define MODE_NORMAL 0
00074 
00075 // KMail mode creates a directory structure suitable for being used directly
00076 // by the KMail application
00077 #define MODE_KMAIL 1
00078 
00079 // recurse mode creates a directory structure like the PST file. Each directory
00080 // contains only one file which stores the emails in mboxrd format.
00081 #define MODE_RECURSE 2
00082 
00083 // separate mode creates the same directory structure as recurse. The emails are stored in
00084 // separate files, numbering from 1 upward. Attachments belonging to the emails are
00085 // saved as email_no-filename (e.g. 1-samplefile.doc or 1-Attachment2.zip)
00086 #define MODE_SEPARATE 3
00087 
00088 
00089 // Output Normal just prints the standard information about what is going on
00090 #define OUTPUT_NORMAL 0
00091 
00092 // Output Quiet is provided so that only errors are printed
00093 #define OUTPUT_QUIET 1
00094 
00095 // default mime-type for attachments that have a null mime-type
00096 #define MIME_TYPE_DEFAULT "application/octet-stream"
00097 #define RFC822            "message/rfc822"
00098 
00099 // output mode for contacts
00100 #define CMODE_VCARD 0
00101 #define CMODE_LIST  1
00102 
00103 // output mode for deleted items
00104 #define DMODE_EXCLUDE 0
00105 #define DMODE_INCLUDE 1
00106 
00107 // Output type mode flags
00108 #define OTMODE_EMAIL        1
00109 #define OTMODE_APPOINTMENT  2
00110 #define OTMODE_JOURNAL      4
00111 #define OTMODE_CONTACT      8
00112 
00113 // output settings for RTF bodies
00114 // filename for the attachment
00115 #define RTF_ATTACH_NAME "rtf-body.rtf"
00116 // mime type for the attachment
00117 #define RTF_ATTACH_TYPE "application/rtf"
00118 
00119 // global settings
00120 int         mode         = MODE_NORMAL;
00121 int         mode_MH      = 0;   // a submode of MODE_SEPARATE
00122 int         mode_EX      = 0;   // a submode of MODE_SEPARATE
00123 int         mode_thunder = 0;   // a submode of MODE_RECURSE
00124 int         output_mode  = OUTPUT_NORMAL;
00125 int         contact_mode = CMODE_VCARD;
00126 int         deleted_mode = DMODE_EXCLUDE;
00127 int         output_type_mode = 0xff;    // Default to all.
00128 int         contact_mode_specified = 0;
00129 int         overwrite = 0;
00130 int         save_rtf_body = 1;
00131 int         file_name_len = 10;     // enough room for MODE_SPEARATE file name
00132 pst_file    pstfile;
00133 regex_t     meta_charset_pattern;
00134 
00135 int         number_processors = 1;  // number of cpus we have
00136 int         max_children  = 0;      // based on number of cpus and command line args
00137 int         max_child_specified = 0;// have command line arg -j
00138 int         active_children;        // number of children of this process, cannot be larger than max_children
00139 pid_t*      child_processes;        // setup by main(), and at the start of new child process
00140 
00141 #ifdef HAVE_SEMAPHORE_H
00142 int         shared_memory_id;
00143 sem_t*      global_children = NULL;
00144 sem_t*      output_mutex    = NULL;
00145 #endif
00146 
00147 
00148 int grim_reaper(int waitall)
00149 {
00150     int available = 0;
00151 #ifdef HAVE_FORK
00152 #ifdef HAVE_SEMAPHORE_H
00153     if (global_children) {
00154         sem_getvalue(global_children, &available);
00155         //printf("grim reaper %s for pid %d (parent %d) with %d children, %d available\n", (waitall) ? "all" : "", getpid(), getppid(), active_children, available);
00156         //fflush(stdout);
00157         int i,j;
00158         for (i=0; i<active_children; i++) {
00159             pid_t child = child_processes[i];
00160             pid_t ch = waitpid(child, NULL, ((waitall) ? 0 : WNOHANG));
00161             if (ch == child) {
00162                 // this has terminated, remove it from the list
00163                 for (j=i; j<active_children-1; j++) {
00164                     child_processes[j] = child_processes[j+1];
00165                 }
00166                 active_children--;
00167                 i--;
00168             }
00169         }
00170         sem_getvalue(global_children, &available);
00171         //printf("grim reaper %s for pid %d with %d children, %d available\n", (waitall) ? "all" : "", getpid(), active_children, available);
00172         //fflush(stdout);
00173     }
00174 #endif
00175 #endif
00176     return available;
00177 }
00178 
00179 
00180 pid_t try_fork(char *folder)
00181 {
00182 #ifdef HAVE_FORK
00183 #ifdef HAVE_SEMAPHORE_H
00184     int available = grim_reaper(0);
00185     if (available) {
00186         sem_wait(global_children);
00187         pid_t child = fork();
00188         if (child < 0) {
00189             // fork failed, pretend it worked and we are the child
00190             return 0;
00191         }
00192         else if (child == 0) {
00193             // fork worked, and we are the child, reinitialize *our* list of children
00194             active_children = 0;
00195             memset(child_processes, 0, sizeof(pid_t) * max_children);
00196             pst_reopen(&pstfile);   // close and reopen the pst file to get an independent file position pointer
00197         }
00198         else {
00199             // fork worked, and we are the parent, record this child that we need to wait for
00200             //pid_t me = getpid();
00201             //printf("parent %d forked child pid %d to process folder %s\n", me, child, folder);
00202             //fflush(stdout);
00203             child_processes[active_children++] = child;
00204         }
00205         return child;
00206     }
00207     else {
00208         return 0;   // pretend to have forked and we are the child
00209     }
00210 #endif
00211 #endif
00212     return 0;
00213 }
00214 
00215 
00216 void process(pst_item *outeritem, pst_desc_tree *d_ptr)
00217 {
00218     struct file_ll ff;
00219     pst_item *item = NULL;
00220 
00221     DEBUG_ENT("process");
00222     memset(&ff, 0, sizeof(ff));
00223     create_enter_dir(&ff, outeritem);
00224 
00225     for (; d_ptr; d_ptr = d_ptr->next) {
00226         DEBUG_INFO(("New item record\n"));
00227         if (!d_ptr->desc) {
00228             ff.skip_count++;
00229             DEBUG_WARN(("ERROR item's desc record is NULL\n"));
00230             continue;
00231         }
00232         DEBUG_INFO(("Desc Email ID %#"PRIx64" [d_ptr->d_id = %#"PRIx64"]\n", d_ptr->desc->i_id, d_ptr->d_id));
00233 
00234         item = pst_parse_item(&pstfile, d_ptr, NULL);
00235         DEBUG_INFO(("About to process item\n"));
00236 
00237         if (!item) {
00238             ff.skip_count++;
00239             DEBUG_INFO(("A NULL item was seen\n"));
00240             continue;
00241         }
00242 
00243         if (item->subject.str) {
00244             DEBUG_INFO(("item->subject = %s\n", item->subject.str));
00245         }
00246 
00247         if (item->folder && item->file_as.str) {
00248             DEBUG_INFO(("Processing Folder \"%s\"\n", item->file_as.str));
00249             if (output_mode != OUTPUT_QUIET) {
00250                 pst_debug_lock();
00251                     printf("Processing Folder \"%s\"\n", item->file_as.str);
00252                     fflush(stdout);
00253                 pst_debug_unlock();
00254             }
00255             ff.item_count++;
00256             if (d_ptr->child && (deleted_mode == DMODE_INCLUDE || strcasecmp(item->file_as.str, "Deleted Items"))) {
00257                 //if this is a non-empty folder other than deleted items, we want to recurse into it
00258                 pid_t parent = getpid();
00259                 pid_t child = try_fork(item->file_as.str);
00260                 if (child == 0) {
00261                     // we are the child process, or the original parent if no children were available
00262                     pid_t me = getpid();
00263                     process(item, d_ptr->child);
00264 #ifdef HAVE_FORK
00265 #ifdef HAVE_SEMAPHORE_H
00266                     if (me != parent) {
00267                         // we really were a child, forked for the sole purpose of processing this folder
00268                         // free my child count slot before really exiting, since
00269                         // all I am doing here is waiting for my children to exit
00270                         sem_post(global_children);
00271                         grim_reaper(1); // wait for all my child processes to exit
00272                         exit(0);        // really exit
00273                     }
00274 #endif
00275 #endif
00276                 }
00277             }
00278 
00279         } else if (item->contact && (item->type == PST_TYPE_CONTACT)) {
00280             DEBUG_INFO(("Processing Contact\n"));
00281             if (!(output_type_mode & OTMODE_CONTACT)) {
00282                 ff.skip_count++;
00283                 DEBUG_INFO(("skipping contact: not in output type list\n"));
00284             }
00285             else {
00286                 if (!ff.type) ff.type = item->type;
00287                 if ((ff.type != PST_TYPE_CONTACT) && (mode != MODE_SEPARATE)) {
00288                     ff.skip_count++;
00289                     DEBUG_INFO(("I have a contact, but the folder type %"PRIi32" isn't a contacts folder. Skipping it\n", ff.type));
00290                 }
00291                 else {
00292                     ff.item_count++;
00293                     if (mode == MODE_SEPARATE) mk_separate_file(&ff, (mode_EX) ? ".vcf" : "");
00294                     if (contact_mode == CMODE_VCARD) {
00295                         pst_convert_utf8_null(item, &item->comment);
00296                         write_vcard(ff.output, item, item->contact, item->comment.str);
00297                     }
00298                     else {
00299                         pst_convert_utf8(item, &item->contact->fullname);
00300                         pst_convert_utf8(item, &item->contact->address1);
00301                         fprintf(ff.output, "%s <%s>\n", item->contact->fullname.str, item->contact->address1.str);
00302                     }
00303                 }
00304             }
00305 
00306         } else if (item->email && ((item->type == PST_TYPE_NOTE) || (item->type == PST_TYPE_SCHEDULE) || (item->type == PST_TYPE_REPORT))) {
00307             DEBUG_INFO(("Processing Email\n"));
00308             if (!(output_type_mode & OTMODE_EMAIL)) {
00309                 ff.skip_count++;
00310                 DEBUG_INFO(("skipping email: not in output type list\n"));
00311             }
00312             else {
00313                 if (!ff.type) ff.type = item->type;
00314                 if ((ff.type != PST_TYPE_NOTE) && (ff.type != PST_TYPE_SCHEDULE) && (ff.type != PST_TYPE_REPORT) && (mode != MODE_SEPARATE)) {
00315                     ff.skip_count++;
00316                     DEBUG_INFO(("I have an email type %"PRIi32", but the folder type %"PRIi32" isn't an email folder. Skipping it\n", item->type, ff.type));
00317                 }
00318                 else {
00319                     char *extra_mime_headers = NULL;
00320                     ff.item_count++;
00321                     if (mode == MODE_SEPARATE) mk_separate_file(&ff, (mode_EX) ? ".eml" : "");
00322                     write_normal_email(ff.output, ff.name, item, mode, mode_MH, &pstfile, save_rtf_body, &extra_mime_headers);
00323                 }
00324             }
00325 
00326         } else if (item->journal && (item->type == PST_TYPE_JOURNAL)) {
00327             DEBUG_INFO(("Processing Journal Entry\n"));
00328             if (!(output_type_mode & OTMODE_JOURNAL)) {
00329                 ff.skip_count++;
00330                 DEBUG_INFO(("skipping journal entry: not in output type list\n"));
00331             }
00332             else {
00333                 if (!ff.type) ff.type = item->type;
00334                 if ((ff.type != PST_TYPE_JOURNAL) && (mode != MODE_SEPARATE)) {
00335                     ff.skip_count++;
00336                     DEBUG_INFO(("I have a journal entry, but the folder type %"PRIi32" isn't a journal folder. Skipping it\n", ff.type));
00337                 }
00338                 else {
00339                     ff.item_count++;
00340                     if (mode == MODE_SEPARATE) mk_separate_file(&ff, (mode_EX) ? ".ics" : "");
00341                     write_journal(ff.output, item);
00342                     fprintf(ff.output, "\n");
00343                 }
00344             }
00345 
00346         } else if (item->appointment && (item->type == PST_TYPE_APPOINTMENT)) {
00347             DEBUG_INFO(("Processing Appointment Entry\n"));
00348             if (!(output_type_mode & OTMODE_APPOINTMENT)) {
00349                 ff.skip_count++;
00350                 DEBUG_INFO(("skipping appointment: not in output type list\n"));
00351             }
00352             else {
00353                 if (!ff.type) ff.type = item->type;
00354                 if ((ff.type != PST_TYPE_APPOINTMENT) && (mode != MODE_SEPARATE)) {
00355                     ff.skip_count++;
00356                     DEBUG_INFO(("I have an appointment, but the folder type %"PRIi32" isn't an appointment folder. Skipping it\n", ff.type));
00357                 }
00358                 else {
00359                     ff.item_count++;
00360                     if (mode == MODE_SEPARATE) mk_separate_file(&ff, (mode_EX) ? ".ics" : "");
00361                     write_schedule_part_data(ff.output, item, NULL, NULL);
00362                     fprintf(ff.output, "\n");
00363                 }
00364             }
00365 
00366         } else if (item->message_store) {
00367             // there should only be one message_store, and we have already done it
00368             ff.skip_count++;
00369             DEBUG_INFO(("item with message store content, type %i %s folder type %i, skipping it\n", item->type, item->ascii_type, ff.type));
00370 
00371         } else {
00372             ff.skip_count++;
00373             DEBUG_INFO(("Unknown item type %i (%s) name (%s)\n",
00374                         item->type, item->ascii_type, item->file_as.str));
00375         }
00376         pst_freeItem(item);
00377     }
00378     close_enter_dir(&ff);
00379     DEBUG_RET();
00380 }
00381 
00382 
00383 
00384 int main(int argc, char* const* argv) {
00385     pst_item *item = NULL;
00386     pst_desc_tree *d_ptr;
00387     char * fname = NULL;
00388     char *d_log  = NULL;
00389     int c,x;
00390     char *temp = NULL;               //temporary char pointer
00391     prog_name = argv[0];
00392 
00393     time_t now = time(NULL);
00394     srand((unsigned)now);
00395 
00396     if (regcomp(&meta_charset_pattern, "<meta[^>]*content=\"[^>]*charset=([^>\";]*)[\";]", REG_ICASE | REG_EXTENDED)) {
00397         printf("cannot compile regex pattern to find content charset in html bodies\n");
00398         exit(3);
00399     }
00400 
00401     // command-line option handling
00402     while ((c = getopt(argc, argv, "bc:Dd:ehj:kMo:qrSt:uVw"))!= -1) {
00403         switch (c) {
00404         case 'b':
00405             save_rtf_body = 0;
00406             break;
00407         case 'c':
00408             if (optarg && optarg[0]=='v') {
00409                 contact_mode=CMODE_VCARD;
00410                 contact_mode_specified = 1;
00411             }
00412             else if (optarg && optarg[0]=='l') {
00413                 contact_mode=CMODE_LIST;
00414                 contact_mode_specified = 1;
00415             }
00416             else {
00417                 usage();
00418                 exit(0);
00419             }
00420             break;
00421         case 'D':
00422             deleted_mode = DMODE_INCLUDE;
00423             break;
00424         case 'd':
00425             d_log = optarg;
00426             break;
00427         case 'h':
00428             usage();
00429             exit(0);
00430             break;
00431         case 'j':
00432             max_children = atoi(optarg);
00433             max_child_specified = 1;
00434             break;
00435         case 'k':
00436             mode = MODE_KMAIL;
00437             break;
00438         case 'M':
00439             mode = MODE_SEPARATE;
00440             mode_MH = 1;
00441             mode_EX = 0;
00442             break;
00443         case 'e':
00444             mode = MODE_SEPARATE;
00445             mode_MH = 1;
00446             mode_EX = 1;
00447             file_name_len = 14;
00448             break;
00449         case 'o':
00450             output_dir = optarg;
00451             break;
00452         case 'q':
00453             output_mode = OUTPUT_QUIET;
00454             break;
00455         case 'r':
00456             mode = MODE_RECURSE;
00457             mode_thunder = 0;
00458             break;
00459         case 'S':
00460             mode = MODE_SEPARATE;
00461             mode_MH = 0;
00462             mode_EX = 0;
00463             break;
00464         case 't':
00465             // email, appointment, contact, other
00466             if (!optarg) {
00467                 usage();
00468                 exit(0);
00469             }
00470             temp = optarg;
00471             output_type_mode = 0;
00472             while (*temp > 0) {
00473               switch (temp[0]) {
00474                 case 'e':
00475                     output_type_mode |= OTMODE_EMAIL;
00476                     break;
00477                 case 'a':
00478                     output_type_mode |= OTMODE_APPOINTMENT;
00479                     break;
00480                 case 'j':
00481                     output_type_mode |= OTMODE_JOURNAL;
00482                     break;
00483                 case 'c':
00484                     output_type_mode |= OTMODE_CONTACT;
00485                     break;
00486                 default:
00487                     usage();
00488                     exit(0);
00489                     break;
00490               }
00491               temp++;
00492             }
00493             break;
00494         case 'u':
00495             mode = MODE_RECURSE;
00496             mode_thunder = 1;
00497             break;
00498         case 'V':
00499             version();
00500             exit(0);
00501             break;
00502         case 'w':
00503             overwrite = 1;
00504             break;
00505         default:
00506             usage();
00507             exit(1);
00508             break;
00509         }
00510     }
00511 
00512     if (argc > optind) {
00513         fname = argv[optind];
00514     } else {
00515         usage();
00516         exit(2);
00517     }
00518 
00519 #ifdef _SC_NPROCESSORS_ONLN
00520     number_processors =  sysconf(_SC_NPROCESSORS_ONLN);
00521 #endif
00522     max_children    = (max_child_specified) ? max_children : number_processors * 4;
00523     active_children = 0;
00524     child_processes = (pid_t *)pst_malloc(sizeof(pid_t) * max_children);
00525     memset(child_processes, 0, sizeof(pid_t) * max_children);
00526 
00527 #ifdef HAVE_SEMAPHORE_H
00528     if (max_children) {
00529         shared_memory_id = shmget(IPC_PRIVATE, sizeof(sem_t)*2, 0777);
00530         if (shared_memory_id >= 0) {
00531             global_children = (sem_t *)shmat(shared_memory_id, NULL, 0);
00532             if (global_children == (sem_t *)-1) global_children = NULL;
00533             if (global_children) {
00534                 output_mutex = &(global_children[1]);
00535                 sem_init(global_children, 1, max_children);
00536                 sem_init(output_mutex, 1, 1);
00537             }
00538             shmctl(shared_memory_id, IPC_RMID, NULL);
00539         }
00540     }
00541 #endif
00542 
00543     #ifdef DEBUG_ALL
00544         // force a log file
00545         if (!d_log) d_log = "readpst.log";
00546     #endif // defined DEBUG_ALL
00547     #ifdef HAVE_SEMAPHORE_H
00548         DEBUG_INIT(d_log, output_mutex);
00549     #else
00550         DEBUG_INIT(d_log, NULL);
00551     #endif
00552     DEBUG_ENT("main");
00553 
00554     if (output_mode != OUTPUT_QUIET) printf("Opening PST file and indexes...\n");
00555 
00556     RET_DERROR(pst_open(&pstfile, fname), 1, ("Error opening File\n"));
00557     RET_DERROR(pst_load_index(&pstfile), 2, ("Index Error\n"));
00558 
00559     pst_load_extended_attributes(&pstfile);
00560 
00561     if (chdir(output_dir)) {
00562         x = errno;
00563         pst_close(&pstfile);
00564         DEBUG_RET();
00565         DIE(("Cannot change to output dir %s: %s\n", output_dir, strerror(x)));
00566     }
00567 
00568     d_ptr = pstfile.d_head; // first record is main record
00569     item  = pst_parse_item(&pstfile, d_ptr, NULL);
00570     if (!item || !item->message_store) {
00571         DEBUG_RET();
00572         DIE(("Could not get root record\n"));
00573     }
00574 
00575     // default the file_as to the same as the main filename if it doesn't exist
00576     if (!item->file_as.str) {
00577         if (!(temp = strrchr(fname, '/')))
00578             if (!(temp = strrchr(fname, '\\')))
00579                 temp = fname;
00580             else
00581                 temp++; // get past the "\\"
00582         else
00583             temp++; // get past the "/"
00584         item->file_as.str = (char*)pst_malloc(strlen(temp)+1);
00585         strcpy(item->file_as.str, temp);
00586         item->file_as.is_utf8 = 1;
00587         DEBUG_INFO(("file_as was blank, so am using %s\n", item->file_as.str));
00588     }
00589     DEBUG_INFO(("Root Folder Name: %s\n", item->file_as.str));
00590 
00591     d_ptr = pst_getTopOfFolders(&pstfile, item);
00592     if (!d_ptr) {
00593         DEBUG_RET();
00594         DIE(("Top of folders record not found. Cannot continue\n"));
00595     }
00596 
00597     process(item, d_ptr->child);    // do the children of TOPF
00598     grim_reaper(1); // wait for all child processes
00599 
00600     pst_freeItem(item);
00601     pst_close(&pstfile);
00602     DEBUG_RET();
00603 
00604 #ifdef HAVE_SEMAPHORE_H
00605     if (global_children) {
00606         sem_destroy(global_children);
00607         sem_destroy(output_mutex);
00608         shmdt(global_children);
00609     }
00610 #endif
00611 
00612     regfree(&meta_charset_pattern);
00613     return 0;
00614 }
00615 
00616 
00617 void write_email_body(FILE *f, char *body) {
00618     char *n = body;
00619     DEBUG_ENT("write_email_body");
00620     if (mode != MODE_SEPARATE) {
00621         while (n) {
00622             char *p = body;
00623             while (*p == '>') p++;
00624             if (strncmp(p, "From ", 5) == 0) fprintf(f, ">");
00625             if ((n = strchr(body, '\n'))) {
00626                 n++;
00627                 pst_fwrite(body, n-body, 1, f); //write just a line
00628                 body = n;
00629             }
00630         }
00631     }
00632     pst_fwrite(body, strlen(body), 1, f);
00633     DEBUG_RET();
00634 }
00635 
00636 
00637 void removeCR (char *c) {
00638     // converts \r\n to \n
00639     char *a, *b;
00640     DEBUG_ENT("removeCR");
00641     a = b = c;
00642     while (*a != '\0') {
00643         *b = *a;
00644         if (*a != '\r') b++;
00645         a++;
00646     }
00647     *b = '\0';
00648     DEBUG_RET();
00649 }
00650 
00651 
00652 void usage() {
00653     DEBUG_ENT("usage");
00654     version();
00655     printf("Usage: %s [OPTIONS] {PST FILENAME}\n", prog_name);
00656     printf("OPTIONS:\n");
00657     printf("\t-V\t- Version. Display program version\n");
00658     printf("\t-D\t- Include deleted items in output\n");
00659     printf("\t-M\t- Write emails in the MH (rfc822) format\n");
00660     printf("\t-S\t- Separate. Write emails in the separate format\n");
00661     printf("\t-b\t- Don't save RTF-Body attachments\n");
00662     printf("\t-c[v|l]\t- Set the Contact output mode. -cv = VCard, -cl = EMail list\n");
00663     printf("\t-d <filename> \t- Debug to file.\n");
00664     printf("\t-e\t- As with -M, but include extensions on output files\n");
00665     printf("\t-h\t- Help. This screen\n");
00666     printf("\t-j <integer>\t- Number of parallel jobs to run\n");
00667     printf("\t-k\t- KMail. Output in kmail format\n");
00668     printf("\t-o <dirname>\t- Output directory to write files to. CWD is changed *after* opening pst file\n");
00669     printf("\t-q\t- Quiet. Only print error messages\n");
00670     printf("\t-r\t- Recursive. Output in a recursive format\n");
00671     printf("\t-t[eajc]\t- Set the output type list. e = email, a = attachment, j = journal, c = contact\n");
00672     printf("\t-u\t- Thunderbird mode. Write two extra .size and .type files\n");
00673     printf("\t-w\t- Overwrite any output mbox files\n");
00674     printf("\n");
00675     printf("Only one of -k -M -r -S should be specified\n");
00676     DEBUG_RET();
00677 }
00678 
00679 
00680 void version() {
00681     DEBUG_ENT("version");
00682     printf("ReadPST / LibPST v%s\n", VERSION);
00683 #if BYTE_ORDER == BIG_ENDIAN
00684     printf("Big Endian implementation being used.\n");
00685 #elif BYTE_ORDER == LITTLE_ENDIAN
00686     printf("Little Endian implementation being used.\n");
00687 #else
00688 #  error "Byte order not supported by this library"
00689 #endif
00690 #ifdef __GNUC__
00691     printf("GCC %d.%d : %s %s\n", __GNUC__, __GNUC_MINOR__, __DATE__, __TIME__);
00692 #endif
00693     DEBUG_RET();
00694 }
00695 
00696 
00697 char *mk_kmail_dir(char *fname) {
00698     //change to that directory
00699     //make a directory based on OUTPUT_KMAIL_DIR_TEMPLATE
00700     //allocate space for OUTPUT_TEMPLATE and form a char* with fname
00701     //return that value
00702     char *dir, *out_name, *index;
00703     int x;
00704     DEBUG_ENT("mk_kmail_dir");
00705     if (kmail_chdir && chdir(kmail_chdir)) {
00706         x = errno;
00707         DIE(("mk_kmail_dir: Cannot change to directory %s: %s\n", kmail_chdir, strerror(x)));
00708     }
00709     dir = malloc(strlen(fname)+strlen(OUTPUT_KMAIL_DIR_TEMPLATE)+1);
00710     sprintf(dir, OUTPUT_KMAIL_DIR_TEMPLATE, fname);
00711     check_filename(dir);
00712     if (D_MKDIR(dir)) {
00713         if (errno != EEXIST) {  // not an error because it exists
00714             x = errno;
00715             DIE(("mk_kmail_dir: Cannot create directory %s: %s\n", dir, strerror(x)));
00716         }
00717     }
00718     kmail_chdir = realloc(kmail_chdir, strlen(dir)+1);
00719     strcpy(kmail_chdir, dir);
00720     free (dir);
00721 
00722     //we should remove any existing indexes created by KMail, cause they might be different now
00723     index = malloc(strlen(fname)+strlen(KMAIL_INDEX)+1);
00724     sprintf(index, KMAIL_INDEX, fname);
00725     unlink(index);
00726     free(index);
00727 
00728     out_name = malloc(strlen(fname)+strlen(OUTPUT_TEMPLATE)+1);
00729     sprintf(out_name, OUTPUT_TEMPLATE, fname);
00730     DEBUG_RET();
00731     return out_name;
00732 }
00733 
00734 
00735 int close_kmail_dir() {
00736     // change ..
00737     int x;
00738     DEBUG_ENT("close_kmail_dir");
00739     if (kmail_chdir) { //only free kmail_chdir if not NULL. do not change directory
00740         free(kmail_chdir);
00741         kmail_chdir = NULL;
00742     } else {
00743         if (chdir("..")) {
00744             x = errno;
00745             DIE(("close_kmail_dir: Cannot move up dir (..): %s\n", strerror(x)));
00746         }
00747     }
00748     DEBUG_RET();
00749     return 0;
00750 }
00751 
00752 
00753 // this will create a directory by that name,
00754 // then make an mbox file inside that directory.
00755 char *mk_recurse_dir(char *dir, int32_t folder_type) {
00756     int x;
00757     char *out_name;
00758     DEBUG_ENT("mk_recurse_dir");
00759     check_filename(dir);
00760     if (D_MKDIR (dir)) {
00761         if (errno != EEXIST) {  // not an error because it exists
00762             x = errno;
00763             DIE(("mk_recurse_dir: Cannot create directory %s: %s\n", dir, strerror(x)));
00764         }
00765     }
00766     if (chdir (dir)) {
00767         x = errno;
00768         DIE(("mk_recurse_dir: Cannot change to directory %s: %s\n", dir, strerror(x)));
00769     }
00770     switch (folder_type) {
00771         case PST_TYPE_APPOINTMENT:
00772             out_name = strdup("calendar");
00773             break;
00774         case PST_TYPE_CONTACT:
00775             out_name = strdup("contacts");
00776             break;
00777         case PST_TYPE_JOURNAL:
00778             out_name = strdup("journal");
00779             break;
00780         case PST_TYPE_STICKYNOTE:
00781         case PST_TYPE_TASK:
00782         case PST_TYPE_NOTE:
00783         case PST_TYPE_OTHER:
00784         case PST_TYPE_REPORT:
00785         default:
00786             out_name = strdup("mbox");
00787             break;
00788     }
00789     DEBUG_RET();
00790     return out_name;
00791 }
00792 
00793 
00794 int close_recurse_dir() {
00795     int x;
00796     DEBUG_ENT("close_recurse_dir");
00797     if (chdir("..")) {
00798         x = errno;
00799         DIE(("close_recurse_dir: Cannot go up dir (..): %s\n", strerror(x)));
00800     }
00801     DEBUG_RET();
00802     return 0;
00803 }
00804 
00805 
00806 char *mk_separate_dir(char *dir) {
00807     size_t dirsize = strlen(dir) + 10;
00808     char dir_name[dirsize];
00809     int x = 0, y = 0;
00810 
00811     DEBUG_ENT("mk_separate_dir");
00812     do {
00813         if (y == 0)
00814             snprintf(dir_name, dirsize, "%s", dir);
00815         else
00816       snprintf(dir_name, dirsize, "%s" SEP_MAIL_FILE_TEMPLATE, dir, y, ""); // enough for 9 digits allocated above
00817 
00818         check_filename(dir_name);
00819         DEBUG_INFO(("about to try creating %s\n", dir_name));
00820         if (D_MKDIR(dir_name)) {
00821             if (errno != EEXIST) { // if there is an error, and it doesn't already exist
00822                 x = errno;
00823                 DIE(("mk_separate_dir: Cannot create directory %s: %s\n", dir, strerror(x)));
00824             }
00825         } else {
00826             break;
00827         }
00828         y++;
00829     } while (overwrite == 0);
00830 
00831     if (chdir(dir_name)) {
00832         x = errno;
00833         DIE(("mk_separate_dir: Cannot change to directory %s: %s\n", dir, strerror(x)));
00834     }
00835 
00836     if (overwrite) {
00837         // we should probably delete all files from this directory
00838 #if !defined(WIN32) && !defined(__CYGWIN__)
00839         DIR * sdir = NULL;
00840         struct dirent *dirent = NULL;
00841         struct stat filestat;
00842         if (!(sdir = opendir("./"))) {
00843             DEBUG_WARN(("mk_separate_dir: Cannot open dir \"%s\" for deletion of old contents\n", "./"));
00844         } else {
00845             while ((dirent = readdir(sdir))) {
00846                 if (lstat(dirent->d_name, &filestat) != -1)
00847                     if (S_ISREG(filestat.st_mode)) {
00848                         if (unlink(dirent->d_name)) {
00849                             y = errno;
00850                             DIE(("mk_separate_dir: unlink returned error on file %s: %s\n", dirent->d_name, strerror(y)));
00851                         }
00852                     }
00853             }
00854         }
00855 #endif
00856     }
00857 
00858     // we don't return a filename here cause it isn't necessary.
00859     DEBUG_RET();
00860     return NULL;
00861 }
00862 
00863 
00864 int close_separate_dir() {
00865     int x;
00866     DEBUG_ENT("close_separate_dir");
00867     if (chdir("..")) {
00868         x = errno;
00869         DIE(("close_separate_dir: Cannot go up dir (..): %s\n", strerror(x)));
00870     }
00871     DEBUG_RET();
00872     return 0;
00873 }
00874 
00875 
00876 int mk_separate_file(struct file_ll *f, char *extension) {
00877     DEBUG_ENT("mk_separate_file");
00878     DEBUG_INFO(("opening next file to save email\n"));
00879     if (f->item_count > 999999999) { // bigger than nine 9's
00880         DIE(("mk_separate_file: The number of emails in this folder has become too high to handle\n"));
00881     }
00882     sprintf(f->name, SEP_MAIL_FILE_TEMPLATE, f->item_count, extension);
00883     if (f->output) fclose(f->output);
00884     f->output = NULL;
00885     check_filename(f->name);
00886     if (!(f->output = fopen(f->name, "w"))) {
00887         DIE(("mk_separate_file: Cannot open file to save email \"%s\"\n", f->name));
00888     }
00889     DEBUG_RET();
00890     return 0;
00891 }
00892 
00893 
00894 char *my_stristr(char *haystack, char *needle) {
00895     // my_stristr varies from strstr in that its searches are case-insensitive
00896     char *x=haystack, *y=needle, *z = NULL;
00897     if (!haystack || !needle) {
00898         return NULL;
00899     }
00900     while (*y != '\0' && *x != '\0') {
00901         if (tolower(*y) == tolower(*x)) {
00902             // move y on one
00903             y++;
00904             if (!z) {
00905                 z = x; // store first position in haystack where a match is made
00906             }
00907         } else {
00908             y = needle; // reset y to the beginning of the needle
00909             z = NULL; // reset the haystack storage point
00910         }
00911         x++; // advance the search in the haystack
00912     }
00913     // If the haystack ended before our search finished, it's not a match.
00914     if (*y != '\0') return NULL;
00915     return z;
00916 }
00917 
00918 
00919 void check_filename(char *fname) {
00920     char *t = fname;
00921     DEBUG_ENT("check_filename");
00922     if (!t) {
00923         DEBUG_RET();
00924         return;
00925     }
00926     while ((t = strpbrk(t, "/\\:"))) {
00927         // while there are characters in the second string that we don't want
00928         *t = '_'; //replace them with an underscore
00929     }
00930     DEBUG_RET();
00931 }
00932 
00933 
00934 void write_separate_attachment(char f_name[], pst_item_attach* attach, int attach_num, pst_file* pst)
00935 {
00936     FILE *fp = NULL;
00937     int x = 0;
00938     char *temp = NULL;
00939 
00940     // If there is a long filename (filename2) use that, otherwise
00941     // use the 8.3 filename (filename1)
00942     char *attach_filename = (attach->filename2.str) ? attach->filename2.str
00943                                                     : attach->filename1.str;
00944     DEBUG_ENT("write_separate_attachment");
00945 
00946     if (!attach->data.data) {
00947         // make sure we can fetch data from the id
00948         pst_index_ll *ptr = pst_getID(pst, attach->i_id);
00949         if (!ptr) {
00950             DEBUG_WARN(("Couldn't find i_id %#"PRIx64". Cannot save attachment to file\n", attach->i_id));
00951             DEBUG_RET();
00952             return;
00953         }
00954     }
00955 
00956     check_filename(f_name);
00957     if (!attach_filename) {
00958         // generate our own (dummy) filename for the attachement
00959         temp = pst_malloc(strlen(f_name)+15);
00960         sprintf(temp, "%s-attach%i", f_name, attach_num);
00961     } else {
00962         // have an attachment name, make sure it's unique
00963         temp = pst_malloc(strlen(f_name)+strlen(attach_filename)+15);
00964         do {
00965             if (fp) fclose(fp);
00966             if (x == 0)
00967                 sprintf(temp, "%s-%s", f_name, attach_filename);
00968             else
00969                 sprintf(temp, "%s-%s-%i", f_name, attach_filename, x);
00970         } while ((fp = fopen(temp, "r")) && ++x < 99999999);
00971         if (x > 99999999) {
00972             DIE(("error finding attachment name. exhausted possibilities to %s\n", temp));
00973         }
00974     }
00975     DEBUG_INFO(("Saving attachment to %s\n", temp));
00976     if (!(fp = fopen(temp, "w"))) {
00977         DEBUG_WARN(("write_separate_attachment: Cannot open attachment save file \"%s\"\n", temp));
00978     } else {
00979         (void)pst_attach_to_file(pst, attach, fp);
00980         fclose(fp);
00981     }
00982     if (temp) free(temp);
00983     DEBUG_RET();
00984 }
00985 
00986 
00987 void write_embedded_message(FILE* f_output, pst_item_attach* attach, char *boundary, pst_file* pf, char** extra_mime_headers)
00988 {
00989     pst_index_ll *ptr;
00990     DEBUG_ENT("write_embedded_message");
00991     ptr = pst_getID(pf, attach->i_id);
00992 
00993     pst_desc_tree d_ptr;
00994     d_ptr.d_id        = 0;
00995     d_ptr.parent_d_id = 0;
00996     d_ptr.assoc_tree  = NULL;
00997     d_ptr.desc        = ptr;
00998     d_ptr.no_child    = 0;
00999     d_ptr.prev        = NULL;
01000     d_ptr.next        = NULL;
01001     d_ptr.parent      = NULL;
01002     d_ptr.child       = NULL;
01003     d_ptr.child_tail  = NULL;
01004 
01005     pst_item *item = pst_parse_item(pf, &d_ptr, attach->id2_head);
01006     // It appears that if the embedded message contains an appointment/
01007     // calendar item, pst_parse_item returns NULL due to the presence of
01008     // an unexpected reference type of 0x1048, which seems to represent
01009     // an array of GUIDs representing a CLSID. It's likely that this is
01010     // a reference to an internal Outlook COM class.
01011     //      Log the skipped item and continue on.
01012     if (!item) {
01013         DEBUG_WARN(("write_embedded_message: pst_parse_item was unable to parse the embedded message in attachment ID %llu", attach->i_id));
01014     } else {
01015         if (!item->email) {
01016             DEBUG_WARN(("write_embedded_message: pst_parse_item returned type %d, not an email message", item->type));
01017         } else {
01018             fprintf(f_output, "\n--%s\n", boundary);
01019             fprintf(f_output, "Content-Type: %s\n\n", attach->mimetype.str);
01020             write_normal_email(f_output, "", item, MODE_NORMAL, 0, pf, 0, extra_mime_headers);
01021         }
01022         pst_freeItem(item);
01023     }
01024 
01025     DEBUG_RET();
01026 }
01027 
01028 
01029 void write_inline_attachment(FILE* f_output, pst_item_attach* attach, char *boundary, pst_file* pst)
01030 {
01031     char *attach_filename;
01032     DEBUG_ENT("write_inline_attachment");
01033     DEBUG_INFO(("Attachment Size is %"PRIu64", id %#"PRIx64"\n", (uint64_t)attach->data.size, attach->i_id));
01034 
01035     if (!attach->data.data) {
01036         // make sure we can fetch data from the id
01037         pst_index_ll *ptr = pst_getID(pst, attach->i_id);
01038         if (!ptr) {
01039             DEBUG_WARN(("Couldn't find ID pointer. Cannot save attachment to file\n"));
01040             DEBUG_RET();
01041             return;
01042         }
01043     }
01044 
01045     fprintf(f_output, "\n--%s\n", boundary);
01046     if (!attach->mimetype.str) {
01047         fprintf(f_output, "Content-Type: %s\n", MIME_TYPE_DEFAULT);
01048     } else {
01049         fprintf(f_output, "Content-Type: %s\n", attach->mimetype.str);
01050     }
01051     fprintf(f_output, "Content-Transfer-Encoding: base64\n");
01052 
01053     // If there is a long filename (filename2) use that, otherwise
01054     // use the 8.3 filename (filename1)
01055     attach_filename = (attach->filename2.str) ? attach->filename2.str : attach->filename1.str;
01056     if (!attach_filename) {
01057         fprintf(f_output, "Content-Disposition: inline\n\n");
01058     } else {
01059         fprintf(f_output, "Content-Disposition: attachment; filename=\"%s\"\n\n", attach_filename);
01060     }
01061 
01062     (void)pst_attach_to_file_base64(pst, attach, f_output);
01063     fprintf(f_output, "\n\n");
01064     DEBUG_RET();
01065 }
01066 
01067 
01068 void header_has_field(char *header, char *field, int *flag)
01069 {
01070     DEBUG_ENT("header_has_field");
01071     if (my_stristr(header, field) || (strncasecmp(header, field+1, strlen(field)-1) == 0)) {
01072         DEBUG_INFO(("header block has %s header\n", field+1));
01073         *flag = 1;
01074     }
01075     DEBUG_RET();
01076 }
01077 
01078 
01079 void header_get_subfield(char *field, const char *subfield, char *body_subfield, size_t size_subfield)
01080 {
01081     if (!field) return;
01082     DEBUG_ENT("header_get_subfield");
01083     char search[60];
01084     snprintf(search, sizeof(search), " %s=", subfield);
01085     field++;
01086     char *n = header_end_field(field);
01087     char *s = my_stristr(field, search);
01088     if (n && s && (s < n)) {
01089         char *e, *f, save;
01090         s += strlen(search);    // skip over subfield=
01091         if (*s == '"') {
01092             s++;
01093             e = strchr(s, '"');
01094         }
01095         else {
01096             e = strchr(s, ';');
01097             f = strchr(s, '\n');
01098             if (e && f && (f < e)) e = f;
01099         }
01100         if (!e || (e > n)) e = n;   // use the trailing lf as terminator if nothing better
01101         save = *e;
01102         *e = '\0';
01103             snprintf(body_subfield, size_subfield, "%s", s);  // copy the subfield to our buffer
01104         *e = save;
01105         DEBUG_INFO(("body %s %s from headers\n", subfield, body_subfield));
01106     }
01107     DEBUG_RET();
01108 }
01109 
01110 char* header_get_field(char *header, char *field)
01111 {
01112     char *t = my_stristr(header, field);
01113     if (!t && (strncasecmp(header, field+1, strlen(field)-1) == 0)) t = header;
01114     return t;
01115 }
01116 
01117 
01118 // return pointer to \n at the end of this header field,
01119 // or NULL if this field goes to the end of the string.
01120 char *header_end_field(char *field)
01121 {
01122     char *e = strchr(field+1, '\n');
01123     while (e && ((e[1] == ' ') || (e[1] == '\t'))) {
01124         e = strchr(e+1, '\n');
01125     }
01126     return e;
01127 }
01128 
01129 
01130 void header_strip_field(char *header, char *field)
01131 {
01132     char *t = header_get_field(header, field);
01133     if (t) {
01134         char *e = header_end_field(t);
01135         if (e) {
01136             if (t == header) e++;   // if *t is not \n, we don't want to keep the \n at *e either.
01137             while (*e != '\0') {
01138                 *t = *e;
01139                 t++;
01140                 e++;
01141             }
01142             *t = '\0';
01143         }
01144         else {
01145             // this was the last header field, truncate the headers
01146             *t = '\0';
01147         }
01148     }
01149 }
01150 
01151 
01152 int  test_base64(char *body)
01153 {
01154     int b64 = 0;
01155     uint8_t *b = (uint8_t *)body;
01156     DEBUG_ENT("test_base64");
01157     while (*b != 0) {
01158         if ((*b < 32) && (*b != 9) && (*b != 10)) {
01159             DEBUG_INFO(("found base64 byte %d\n", (int)*b));
01160             DEBUG_HEXDUMPC(body, strlen(body), 0x10);
01161             b64 = 1;
01162             break;
01163         }
01164         b++;
01165     }
01166     DEBUG_RET();
01167     return b64;
01168 }
01169 
01170 
01171 void find_html_charset(char *html, char *charset, size_t charsetlen)
01172 {
01173     const int  index = 1;
01174     const int nmatch = index+1;
01175     regmatch_t match[nmatch];
01176     DEBUG_ENT("find_html_charset");
01177     int rc = regexec(&meta_charset_pattern, html, nmatch, match, 0);
01178     if (rc == 0) {
01179         int s = match[index].rm_so;
01180         int e = match[index].rm_eo;
01181         if (s != -1) {
01182             char save = html[e];
01183             html[e] = '\0';
01184                 snprintf(charset, charsetlen, "%s", html+s);    // copy the html charset
01185             html[e] = save;
01186             DEBUG_INFO(("charset %s from html text\n", charset));
01187         }
01188         else {
01189             DEBUG_INFO(("matching %d %d %d %d\n", match[0].rm_so, match[0].rm_eo, match[1].rm_so, match[1].rm_eo));
01190             DEBUG_HEXDUMPC(html, strlen(html), 0x10);
01191         }
01192     }
01193     else {
01194         DEBUG_INFO(("regexec returns %d\n", rc));
01195     }
01196     DEBUG_RET();
01197 }
01198 
01199 
01200 void find_rfc822_headers(char** extra_mime_headers)
01201 {
01202     DEBUG_ENT("find_rfc822_headers");
01203     char *headers = *extra_mime_headers;
01204     if (headers) {
01205         char *temp, *t;
01206         while ((temp = strstr(headers, "\n\n"))) {
01207             temp[1] = '\0';
01208             t = header_get_field(headers, "\nContent-Type: ");
01209             if (t) {
01210                 t++;
01211                 DEBUG_INFO(("found content type header\n"));
01212                 char *n = strchr(t, '\n');
01213                 char *s = strstr(t, ": ");
01214                 char *e = strchr(t, ';');
01215                 if (!e || (e > n)) e = n;
01216                 if (s && (s < e)) {
01217                     s += 2;
01218                     if (!strncasecmp(s, RFC822, e-s)) {
01219                         headers = temp+2;   // found rfc822 header
01220                         DEBUG_INFO(("found 822 headers\n%s\n", headers));
01221                         break;
01222                     }
01223                 }
01224             }
01225             //DEBUG_INFO(("skipping to next block after\n%s\n", headers));
01226             headers = temp+2;   // skip to next chunk of headers
01227         }
01228         *extra_mime_headers = headers;
01229     }
01230     DEBUG_RET();
01231 }
01232 
01233 
01234 void write_body_part(FILE* f_output, pst_string *body, char *mime, char *charset, char *boundary, pst_file* pst)
01235 {
01236     DEBUG_ENT("write_body_part");
01237     if (body->is_utf8 && (strcasecmp("utf-8", charset))) {
01238         // try to convert to the specified charset since the target
01239         // is not utf-8, and the data came from a unicode (utf16) field
01240         // and is now in utf-8.
01241         size_t rc;
01242         DEBUG_INFO(("Convert %s utf-8 to %s\n", mime, charset));
01243         pst_vbuf *newer = pst_vballoc(2);
01244         rc = pst_vb_utf8to8bit(newer, body->str, strlen(body->str), charset);
01245         if (rc == (size_t)-1) {
01246             // unable to convert, change the charset to utf8
01247             free(newer->b);
01248             DEBUG_INFO(("Failed to convert %s utf-8 to %s\n", mime, charset));
01249             charset = "utf-8";
01250         }
01251         else {
01252             // null terminate the output string
01253             pst_vbgrow(newer, 1);
01254             newer->b[newer->dlen] = '\0';
01255             free(body->str);
01256             body->str = newer->b;
01257         }
01258         free(newer);
01259     }
01260     removeCR(body->str);
01261     int base64 = test_base64(body->str);
01262     fprintf(f_output, "\n--%s\n", boundary);
01263     fprintf(f_output, "Content-Type: %s; charset=\"%s\"\n", mime, charset);
01264     if (base64) fprintf(f_output, "Content-Transfer-Encoding: base64\n");
01265     fprintf(f_output, "\n");
01266     if (base64) {
01267         char *enc = pst_base64_encode(body->str, strlen(body->str));
01268         if (enc) {
01269             write_email_body(f_output, enc);
01270             fprintf(f_output, "\n");
01271             free(enc);
01272         }
01273     }
01274     else {
01275         write_email_body(f_output, body->str);
01276     }
01277     DEBUG_RET();
01278 }
01279 
01280 
01281 void write_schedule_part_data(FILE* f_output, pst_item* item, const char* sender, const char* method)
01282 {
01283     fprintf(f_output, "BEGIN:VCALENDAR\n");
01284     fprintf(f_output, "VERSION:2.0\n");
01285     fprintf(f_output, "PRODID:LibPST v%s\n", VERSION);
01286     if (method) fprintf(f_output, "METHOD:%s\n", method);
01287     fprintf(f_output, "BEGIN:VEVENT\n");
01288     if (sender) {
01289         if (item->email->outlook_sender_name.str) {
01290         fprintf(f_output, "ORGANIZER;CN=\"%s\":MAILTO:%s\n", item->email->outlook_sender_name.str, sender);
01291     } else {
01292         fprintf(f_output, "ORGANIZER;CN=\"\":MAILTO:%s\n", sender);
01293     }
01294     }
01295     write_appointment(f_output, item);
01296     fprintf(f_output, "END:VCALENDAR\n");
01297 }
01298 
01299 
01300 void write_schedule_part(FILE* f_output, pst_item* item, const char* sender, const char* boundary)
01301 {
01302     const char* method  = "REQUEST";
01303     const char* charset = "utf-8";
01304     char fname[30];
01305     if (!item->appointment) return;
01306 
01307     // inline appointment request
01308     fprintf(f_output, "\n--%s\n", boundary);
01309     fprintf(f_output, "Content-Type: %s; method=\"%s\"; charset=\"%s\"\n\n", "text/calendar", method, charset);
01310     write_schedule_part_data(f_output, item, sender, method);
01311     fprintf(f_output, "\n");
01312 
01313     // attachment appointment request
01314     snprintf(fname, sizeof(fname), "i%i.ics", rand());
01315     fprintf(f_output, "\n--%s\n", boundary);
01316     fprintf(f_output, "Content-Type: %s; charset=\"%s\"; name=\"%s\"\n", "text/calendar", "utf-8", fname);
01317     fprintf(f_output, "Content-Disposition: attachment; filename=\"%s\"\n\n", fname);
01318     write_schedule_part_data(f_output, item, sender, method);
01319     fprintf(f_output, "\n");
01320 }
01321 
01322 
01323 void write_normal_email(FILE* f_output, char f_name[], pst_item* item, int mode, int mode_MH, pst_file* pst, int save_rtf, char** extra_mime_headers)
01324 {
01325     char boundary[60];
01326     char altboundary[66];
01327     char *altboundaryp = NULL;
01328     char body_charset[30];
01329     char buffer_charset[30];
01330     char body_report[60];
01331     char sender[60];
01332     int  sender_known = 0;
01333     char *temp = NULL;
01334     time_t em_time;
01335     char *c_time;
01336     char *headers = NULL;
01337     int has_from, has_subject, has_to, has_cc, has_date, has_msgid;
01338     has_from = has_subject = has_to = has_cc = has_date = has_msgid = 0;
01339     DEBUG_ENT("write_normal_email");
01340 
01341     pst_convert_utf8_null(item, &item->email->header);
01342     headers = (item->email->header.str) ? item->email->header.str : *extra_mime_headers;
01343 
01344     // setup default body character set and report type
01345     strncpy(body_charset, pst_default_charset(item, sizeof(buffer_charset), buffer_charset), sizeof(body_charset));
01346     body_charset[sizeof(body_charset)-1] = '\0';
01347     strncpy(body_report, "delivery-status", sizeof(body_report));
01348     body_report[sizeof(body_report)-1] = '\0';
01349 
01350     // setup default sender
01351     pst_convert_utf8(item, &item->email->sender_address);
01352     if (item->email->sender_address.str && strchr(item->email->sender_address.str, '@')) {
01353         temp = item->email->sender_address.str;
01354         sender_known = 1;
01355     }
01356     else {
01357         temp = "MAILER-DAEMON";
01358     }
01359     strncpy(sender, temp, sizeof(sender));
01360     sender[sizeof(sender)-1] = '\0';
01361 
01362     // convert the sent date if it exists, or set it to a fixed date
01363     if (item->email->sent_date) {
01364         em_time = pst_fileTimeToUnixTime(item->email->sent_date);
01365         c_time = ctime(&em_time);
01366         if (c_time)
01367             c_time[strlen(c_time)-1] = '\0'; //remove end \n
01368         else
01369             c_time = "Fri Dec 28 12:06:21 2001";
01370     } else
01371         c_time= "Fri Dec 28 12:06:21 2001";
01372 
01373     // create our MIME boundaries here.
01374     snprintf(boundary, sizeof(boundary), "--boundary-LibPST-iamunique-%i_-_-", rand());
01375     snprintf(altboundary, sizeof(altboundary), "alt-%s", boundary);
01376 
01377     // we will always look at the headers to discover some stuff
01378     if (headers ) {
01379         char *t;
01380         removeCR(headers);
01381 
01382         temp = strstr(headers, "\n\n");
01383         if (temp) {
01384             // cut off our real rfc822 headers here
01385             temp[1] = '\0';
01386             // pointer to all the embedded MIME headers.
01387             // we use these to find the actual rfc822 headers for embedded message/rfc822 mime parts
01388             *extra_mime_headers = temp+2;
01389             DEBUG_INFO(("Found extra mime headers\n%s\n", temp+2));
01390         }
01391 
01392         // Check if the headers have all the necessary fields
01393         header_has_field(headers, "\nFrom: ",        &has_from);
01394         header_has_field(headers, "\nTo: ",          &has_to);
01395         header_has_field(headers, "\nSubject: ",     &has_subject);
01396         header_has_field(headers, "\nDate: ",        &has_date);
01397         header_has_field(headers, "\nCC: ",          &has_cc);
01398         header_has_field(headers, "\nMessage-Id: ",  &has_msgid);
01399 
01400         // look for charset and report-type in Content-Type header
01401         t = header_get_field(headers, "\nContent-Type: ");
01402         header_get_subfield(t, "charset", body_charset, sizeof(body_charset));
01403         header_get_subfield(t, "report-type", body_report, sizeof(body_report));
01404 
01405         // derive a proper sender email address
01406         if (!sender_known) {
01407             t = header_get_field(headers, "\nFrom: ");
01408             if (t) {
01409                 // assume address is on the first line, rather than on a continuation line
01410                 t++;
01411                 char *n = strchr(t, '\n');
01412                 char *s = strchr(t, '<');
01413                 char *e = strchr(t, '>');
01414                 if (s && e && n && (s < e) && (e < n)) {
01415                 char save = *e;
01416                 *e = '\0';
01417                     snprintf(sender, sizeof(sender), "%s", s+1);
01418                 *e = save;
01419                 }
01420             }
01421         }
01422 
01423         // Strip out the mime headers and some others that we don't want to emit
01424         header_strip_field(headers, "\nMicrosoft Mail Internet Headers");
01425         header_strip_field(headers, "\nMIME-Version: ");
01426         header_strip_field(headers, "\nContent-Type: ");
01427         header_strip_field(headers, "\nContent-Transfer-Encoding: ");
01428         header_strip_field(headers, "\nContent-class: ");
01429         header_strip_field(headers, "\nX-MimeOLE: ");
01430         header_strip_field(headers, "\nBcc:");
01431         header_strip_field(headers, "\nX-From_: ");
01432     }
01433 
01434     DEBUG_INFO(("About to print Header\n"));
01435 
01436     if (item && item->subject.str) {
01437         pst_convert_utf8(item, &item->subject);
01438         DEBUG_INFO(("item->subject = %s\n", item->subject.str));
01439     }
01440 
01441     if (mode != MODE_SEPARATE) {
01442         // most modes need this separator line.
01443         // procmail produces this separator without the quotes around the
01444         // sender email address, but apparently some Mac email client needs
01445         // those quotes, and they don't seem to cause problems for anyone else.
01446         fprintf(f_output, "From \"%s\" %s\n", sender, c_time);
01447     }
01448 
01449     // print the supplied email headers
01450     if (headers) {
01451         int len = strlen(headers);
01452         if (len > 0) {
01453             fprintf(f_output, "%s", headers);
01454             // make sure the headers end with a \n
01455             if (headers[len-1] != '\n') fprintf(f_output, "\n");
01456         }
01457     }
01458 
01459     // create required header fields that are not already written
01460 
01461     if (!has_from) {
01462         if (item->email->outlook_sender_name.str){
01463             fprintf(f_output, "From: \"%s\" <%s>\n", item->email->outlook_sender_name.str, sender);
01464         } else {
01465             fprintf(f_output, "From: <%s>\n", sender);
01466         }
01467     }
01468 
01469     if (!has_subject) {
01470         if (item->subject.str) {
01471             fprintf(f_output, "Subject: %s\n", item->subject.str);
01472         } else {
01473             fprintf(f_output, "Subject: \n");
01474         }
01475     }
01476 
01477     if (!has_to && item->email->sentto_address.str) {
01478         pst_convert_utf8(item, &item->email->sentto_address);
01479         fprintf(f_output, "To: %s\n", item->email->sentto_address.str);
01480     }
01481 
01482     if (!has_cc && item->email->cc_address.str) {
01483         pst_convert_utf8(item, &item->email->cc_address);
01484         fprintf(f_output, "Cc: %s\n", item->email->cc_address.str);
01485     }
01486 
01487     if (!has_date && item->email->sent_date) {
01488         char c_time[C_TIME_SIZE];
01489         struct tm stm;
01490         gmtime_r(&em_time, &stm);
01491         strftime(c_time, C_TIME_SIZE, "%a, %d %b %Y %H:%M:%S %z", &stm);
01492         fprintf(f_output, "Date: %s\n", c_time);
01493     }
01494 
01495     if (!has_msgid && item->email->messageid.str) {
01496         pst_convert_utf8(item, &item->email->messageid);
01497         fprintf(f_output, "Message-Id: %s\n", item->email->messageid.str);
01498     }
01499 
01500     // add forensic headers to capture some .pst stuff that is not really
01501     // needed or used by mail clients
01502     pst_convert_utf8_null(item, &item->email->sender_address);
01503     if (item->email->sender_address.str && !strchr(item->email->sender_address.str, '@')
01504                                         && strcmp(item->email->sender_address.str, ".")
01505                                         && (strlen(item->email->sender_address.str) > 0)) {
01506         fprintf(f_output, "X-libpst-forensic-sender: %s\n", item->email->sender_address.str);
01507     }
01508 
01509     if (item->email->bcc_address.str) {
01510         pst_convert_utf8(item, &item->email->bcc_address);
01511         fprintf(f_output, "X-libpst-forensic-bcc: %s\n", item->email->bcc_address.str);
01512     }
01513 
01514     // add our own mime headers
01515     fprintf(f_output, "MIME-Version: 1.0\n");
01516     if (item->type == PST_TYPE_REPORT) {
01517         // multipart/report for DSN/MDN reports
01518         fprintf(f_output, "Content-Type: multipart/report; report-type=%s;\n\tboundary=\"%s\"\n", body_report, boundary);
01519     }
01520     else {
01521         fprintf(f_output, "Content-Type: multipart/mixed;\n\tboundary=\"%s\"\n", boundary);
01522     }
01523     fprintf(f_output, "\n");    // end of headers, start of body
01524 
01525     // now dump the body parts
01526     if ((item->type == PST_TYPE_REPORT) && (item->email->report_text.str)) {
01527         write_body_part(f_output, &item->email->report_text, "text/plain", body_charset, boundary, pst);
01528         fprintf(f_output, "\n");
01529     }
01530 
01531     if (item->body.str && item->email->htmlbody.str) {
01532         // start the nested alternative part
01533         fprintf(f_output, "\n--%s\n", boundary);
01534         fprintf(f_output, "Content-Type: multipart/alternative;\n\tboundary=\"%s\"\n", altboundary);
01535         altboundaryp = altboundary;
01536     }
01537     else {
01538         altboundaryp = boundary;
01539     }
01540 
01541     if (item->body.str) {
01542         write_body_part(f_output, &item->body, "text/plain", body_charset, altboundaryp, pst);
01543     }
01544 
01545     if (item->email->htmlbody.str) {
01546         find_html_charset(item->email->htmlbody.str, body_charset, sizeof(body_charset));
01547         write_body_part(f_output, &item->email->htmlbody, "text/html", body_charset, altboundaryp, pst);
01548     }
01549 
01550     if (item->body.str && item->email->htmlbody.str) {
01551         // end the nested alternative part
01552         fprintf(f_output, "\n--%s--\n", altboundary);
01553     }
01554 
01555     if (item->email->rtf_compressed.data && save_rtf) {
01556         pst_item_attach* attach = (pst_item_attach*)pst_malloc(sizeof(pst_item_attach));
01557         DEBUG_INFO(("Adding RTF body as attachment\n"));
01558         memset(attach, 0, sizeof(pst_item_attach));
01559         attach->next = item->attach;
01560         item->attach = attach;
01561         attach->data.data         = pst_lzfu_decompress(item->email->rtf_compressed.data, item->email->rtf_compressed.size, &attach->data.size);
01562         attach->filename2.str     = strdup(RTF_ATTACH_NAME);
01563         attach->filename2.is_utf8 = 1;
01564         attach->mimetype.str      = strdup(RTF_ATTACH_TYPE);
01565         attach->mimetype.is_utf8  = 1;
01566     }
01567 
01568     if (item->email->encrypted_body.data) {
01569         pst_item_attach* attach = (pst_item_attach*)pst_malloc(sizeof(pst_item_attach));
01570         DEBUG_INFO(("Adding encrypted text body as attachment\n"));
01571         attach = (pst_item_attach*) pst_malloc(sizeof(pst_item_attach));
01572         memset(attach, 0, sizeof(pst_item_attach));
01573         attach->next = item->attach;
01574         item->attach = attach;
01575         attach->data.data = item->email->encrypted_body.data;
01576         attach->data.size = item->email->encrypted_body.size;
01577         item->email->encrypted_body.data = NULL;
01578     }
01579 
01580     if (item->email->encrypted_htmlbody.data) {
01581         pst_item_attach* attach = (pst_item_attach*)pst_malloc(sizeof(pst_item_attach));
01582         DEBUG_INFO(("Adding encrypted HTML body as attachment\n"));
01583         attach = (pst_item_attach*) pst_malloc(sizeof(pst_item_attach));
01584         memset(attach, 0, sizeof(pst_item_attach));
01585         attach->next = item->attach;
01586         item->attach = attach;
01587         attach->data.data = item->email->encrypted_htmlbody.data;
01588         attach->data.size = item->email->encrypted_htmlbody.size;
01589         item->email->encrypted_htmlbody.data = NULL;
01590     }
01591 
01592     if (item->type == PST_TYPE_SCHEDULE) {
01593         write_schedule_part(f_output, item, sender, boundary);
01594     }
01595 
01596     // other attachments
01597     {
01598         pst_item_attach* attach;
01599         int attach_num = 0;
01600         for (attach = item->attach; attach; attach = attach->next) {
01601             pst_convert_utf8_null(item, &attach->filename1);
01602             pst_convert_utf8_null(item, &attach->filename2);
01603             pst_convert_utf8_null(item, &attach->mimetype);
01604             DEBUG_INFO(("Attempting Attachment encoding\n"));
01605             if (attach->method == PST_ATTACH_EMBEDDED) {
01606                 DEBUG_INFO(("have an embedded rfc822 message attachment\n"));
01607                 if (attach->mimetype.str) {
01608                     DEBUG_INFO(("which already has a mime-type of %s\n", attach->mimetype.str));
01609                     free(attach->mimetype.str);
01610                 }
01611                 attach->mimetype.str = strdup(RFC822);
01612                 attach->mimetype.is_utf8 = 1;
01613                 find_rfc822_headers(extra_mime_headers);
01614                 write_embedded_message(f_output, attach, boundary, pst, extra_mime_headers);
01615             }
01616             else if (attach->data.data || attach->i_id) {
01617                 if (mode == MODE_SEPARATE && !mode_MH)
01618                     write_separate_attachment(f_name, attach, ++attach_num, pst);
01619                 else
01620                     write_inline_attachment(f_output, attach, boundary, pst);
01621             }
01622         }
01623     }
01624 
01625     fprintf(f_output, "\n--%s--\n\n", boundary);
01626     DEBUG_RET();
01627 }
01628 
01629 
01630 void write_vcard(FILE* f_output, pst_item* item, pst_item_contact* contact, char comment[])
01631 {
01632     char*  result = NULL;
01633     size_t resultlen = 0;
01634     char   time_buffer[30];
01635     // We can only call rfc escape once per printf, since the second call
01636     // may free the buffer returned by the first call.
01637     // I had tried to place those into a single printf - Carl.
01638 
01639     DEBUG_ENT("write_vcard");
01640 
01641     // make everything utf8
01642     pst_convert_utf8_null(item, &contact->fullname);
01643     pst_convert_utf8_null(item, &contact->surname);
01644     pst_convert_utf8_null(item, &contact->first_name);
01645     pst_convert_utf8_null(item, &contact->middle_name);
01646     pst_convert_utf8_null(item, &contact->display_name_prefix);
01647     pst_convert_utf8_null(item, &contact->suffix);
01648     pst_convert_utf8_null(item, &contact->nickname);
01649     pst_convert_utf8_null(item, &contact->address1);
01650     pst_convert_utf8_null(item, &contact->address2);
01651     pst_convert_utf8_null(item, &contact->address3);
01652     pst_convert_utf8_null(item, &contact->home_po_box);
01653     pst_convert_utf8_null(item, &contact->home_street);
01654     pst_convert_utf8_null(item, &contact->home_city);
01655     pst_convert_utf8_null(item, &contact->home_state);
01656     pst_convert_utf8_null(item, &contact->home_postal_code);
01657     pst_convert_utf8_null(item, &contact->home_country);
01658     pst_convert_utf8_null(item, &contact->home_address);
01659     pst_convert_utf8_null(item, &contact->business_po_box);
01660     pst_convert_utf8_null(item, &contact->business_street);
01661     pst_convert_utf8_null(item, &contact->business_city);
01662     pst_convert_utf8_null(item, &contact->business_state);
01663     pst_convert_utf8_null(item, &contact->business_postal_code);
01664     pst_convert_utf8_null(item, &contact->business_country);
01665     pst_convert_utf8_null(item, &contact->business_address);
01666     pst_convert_utf8_null(item, &contact->other_po_box);
01667     pst_convert_utf8_null(item, &contact->other_street);
01668     pst_convert_utf8_null(item, &contact->other_city);
01669     pst_convert_utf8_null(item, &contact->other_state);
01670     pst_convert_utf8_null(item, &contact->other_postal_code);
01671     pst_convert_utf8_null(item, &contact->other_country);
01672     pst_convert_utf8_null(item, &contact->other_address);
01673     pst_convert_utf8_null(item, &contact->business_fax);
01674     pst_convert_utf8_null(item, &contact->business_phone);
01675     pst_convert_utf8_null(item, &contact->business_phone2);
01676     pst_convert_utf8_null(item, &contact->car_phone);
01677     pst_convert_utf8_null(item, &contact->home_fax);
01678     pst_convert_utf8_null(item, &contact->home_phone);
01679     pst_convert_utf8_null(item, &contact->home_phone2);
01680     pst_convert_utf8_null(item, &contact->isdn_phone);
01681     pst_convert_utf8_null(item, &contact->mobile_phone);
01682     pst_convert_utf8_null(item, &contact->other_phone);
01683     pst_convert_utf8_null(item, &contact->pager_phone);
01684     pst_convert_utf8_null(item, &contact->primary_fax);
01685     pst_convert_utf8_null(item, &contact->primary_phone);
01686     pst_convert_utf8_null(item, &contact->radio_phone);
01687     pst_convert_utf8_null(item, &contact->telex);
01688     pst_convert_utf8_null(item, &contact->job_title);
01689     pst_convert_utf8_null(item, &contact->profession);
01690     pst_convert_utf8_null(item, &contact->assistant_name);
01691     pst_convert_utf8_null(item, &contact->assistant_phone);
01692     pst_convert_utf8_null(item, &contact->company_name);
01693     pst_convert_utf8_null(item, &item->body);
01694 
01695     // the specification I am following is (hopefully) RFC2426 vCard Mime Directory Profile
01696     fprintf(f_output, "BEGIN:VCARD\n");
01697     fprintf(f_output, "FN:%s\n", pst_rfc2426_escape(contact->fullname.str, &result, &resultlen));
01698 
01699     //fprintf(f_output, "N:%s;%s;%s;%s;%s\n",
01700     fprintf(f_output, "N:%s;", (!contact->surname.str)             ? "" : pst_rfc2426_escape(contact->surname.str, &result, &resultlen));
01701     fprintf(f_output, "%s;",   (!contact->first_name.str)          ? "" : pst_rfc2426_escape(contact->first_name.str, &result, &resultlen));
01702     fprintf(f_output, "%s;",   (!contact->middle_name.str)         ? "" : pst_rfc2426_escape(contact->middle_name.str, &result, &resultlen));
01703     fprintf(f_output, "%s;",   (!contact->display_name_prefix.str) ? "" : pst_rfc2426_escape(contact->display_name_prefix.str, &result, &resultlen));
01704     fprintf(f_output, "%s\n",  (!contact->suffix.str)              ? "" : pst_rfc2426_escape(contact->suffix.str, &result, &resultlen));
01705 
01706     if (contact->nickname.str)
01707         fprintf(f_output, "NICKNAME:%s\n", pst_rfc2426_escape(contact->nickname.str, &result, &resultlen));
01708     if (contact->address1.str)
01709         fprintf(f_output, "EMAIL:%s\n", pst_rfc2426_escape(contact->address1.str, &result, &resultlen));
01710     if (contact->address2.str)
01711         fprintf(f_output, "EMAIL:%s\n", pst_rfc2426_escape(contact->address2.str, &result, &resultlen));
01712     if (contact->address3.str)
01713         fprintf(f_output, "EMAIL:%s\n", pst_rfc2426_escape(contact->address3.str, &result, &resultlen));
01714     if (contact->birthday)
01715         fprintf(f_output, "BDAY:%s\n", pst_rfc2425_datetime_format(contact->birthday, sizeof(time_buffer), time_buffer));
01716 
01717     if (contact->home_address.str) {
01718         //fprintf(f_output, "ADR;TYPE=home:%s;%s;%s;%s;%s;%s;%s\n",
01719         fprintf(f_output, "ADR;TYPE=home:%s;",  (!contact->home_po_box.str)      ? "" : pst_rfc2426_escape(contact->home_po_box.str, &result, &resultlen));
01720         fprintf(f_output, "%s;",                ""); // extended Address
01721         fprintf(f_output, "%s;",                (!contact->home_street.str)      ? "" : pst_rfc2426_escape(contact->home_street.str, &result, &resultlen));
01722         fprintf(f_output, "%s;",                (!contact->home_city.str)        ? "" : pst_rfc2426_escape(contact->home_city.str, &result, &resultlen));
01723         fprintf(f_output, "%s;",                (!contact->home_state.str)       ? "" : pst_rfc2426_escape(contact->home_state.str, &result, &resultlen));
01724         fprintf(f_output, "%s;",                (!contact->home_postal_code.str) ? "" : pst_rfc2426_escape(contact->home_postal_code.str, &result, &resultlen));
01725         fprintf(f_output, "%s\n",               (!contact->home_country.str)     ? "" : pst_rfc2426_escape(contact->home_country.str, &result, &resultlen));
01726         fprintf(f_output, "LABEL;TYPE=home:%s\n", pst_rfc2426_escape(contact->home_address.str, &result, &resultlen));
01727     }
01728 
01729     if (contact->business_address.str) {
01730         //fprintf(f_output, "ADR;TYPE=work:%s;%s;%s;%s;%s;%s;%s\n",
01731         fprintf(f_output, "ADR;TYPE=work:%s;",  (!contact->business_po_box.str)      ? "" : pst_rfc2426_escape(contact->business_po_box.str, &result, &resultlen));
01732         fprintf(f_output, "%s;",                ""); // extended Address
01733         fprintf(f_output, "%s;",                (!contact->business_street.str)      ? "" : pst_rfc2426_escape(contact->business_street.str, &result, &resultlen));
01734         fprintf(f_output, "%s;",                (!contact->business_city.str)        ? "" : pst_rfc2426_escape(contact->business_city.str, &result, &resultlen));
01735         fprintf(f_output, "%s;",                (!contact->business_state.str)       ? "" : pst_rfc2426_escape(contact->business_state.str, &result, &resultlen));
01736         fprintf(f_output, "%s;",                (!contact->business_postal_code.str) ? "" : pst_rfc2426_escape(contact->business_postal_code.str, &result, &resultlen));
01737         fprintf(f_output, "%s\n",               (!contact->business_country.str)     ? "" : pst_rfc2426_escape(contact->business_country.str, &result, &resultlen));
01738         fprintf(f_output, "LABEL;TYPE=work:%s\n", pst_rfc2426_escape(contact->business_address.str, &result, &resultlen));
01739     }
01740 
01741     if (contact->other_address.str) {
01742         //fprintf(f_output, "ADR;TYPE=postal:%s;%s;%s;%s;%s;%s;%s\n",
01743         fprintf(f_output, "ADR;TYPE=postal:%s;",(!contact->other_po_box.str)       ? "" : pst_rfc2426_escape(contact->other_po_box.str, &result, &resultlen));
01744         fprintf(f_output, "%s;",                ""); // extended Address
01745         fprintf(f_output, "%s;",                (!contact->other_street.str)       ? "" : pst_rfc2426_escape(contact->other_street.str, &result, &resultlen));
01746         fprintf(f_output, "%s;",                (!contact->other_city.str)         ? "" : pst_rfc2426_escape(contact->other_city.str, &result, &resultlen));
01747         fprintf(f_output, "%s;",                (!contact->other_state.str)        ? "" : pst_rfc2426_escape(contact->other_state.str, &result, &resultlen));
01748         fprintf(f_output, "%s;",                (!contact->other_postal_code.str)  ? "" : pst_rfc2426_escape(contact->other_postal_code.str, &result, &resultlen));
01749         fprintf(f_output, "%s\n",               (!contact->other_country.str)      ? "" : pst_rfc2426_escape(contact->other_country.str, &result, &resultlen));
01750         fprintf(f_output, "LABEL;TYPE=postal:%s\n", pst_rfc2426_escape(contact->other_address.str, &result, &resultlen));
01751     }
01752 
01753     if (contact->business_fax.str)      fprintf(f_output, "TEL;TYPE=work,fax:%s\n",         pst_rfc2426_escape(contact->business_fax.str, &result, &resultlen));
01754     if (contact->business_phone.str)    fprintf(f_output, "TEL;TYPE=work,voice:%s\n",       pst_rfc2426_escape(contact->business_phone.str, &result, &resultlen));
01755     if (contact->business_phone2.str)   fprintf(f_output, "TEL;TYPE=work,voice:%s\n",       pst_rfc2426_escape(contact->business_phone2.str, &result, &resultlen));
01756     if (contact->car_phone.str)         fprintf(f_output, "TEL;TYPE=car,voice:%s\n",        pst_rfc2426_escape(contact->car_phone.str, &result, &resultlen));
01757     if (contact->home_fax.str)          fprintf(f_output, "TEL;TYPE=home,fax:%s\n",         pst_rfc2426_escape(contact->home_fax.str, &result, &resultlen));
01758     if (contact->home_phone.str)        fprintf(f_output, "TEL;TYPE=home,voice:%s\n",       pst_rfc2426_escape(contact->home_phone.str, &result, &resultlen));
01759     if (contact->home_phone2.str)       fprintf(f_output, "TEL;TYPE=home,voice:%s\n",       pst_rfc2426_escape(contact->home_phone2.str, &result, &resultlen));
01760     if (contact->isdn_phone.str)        fprintf(f_output, "TEL;TYPE=isdn:%s\n",             pst_rfc2426_escape(contact->isdn_phone.str, &result, &resultlen));
01761     if (contact->mobile_phone.str)      fprintf(f_output, "TEL;TYPE=cell,voice:%s\n",       pst_rfc2426_escape(contact->mobile_phone.str, &result, &resultlen));
01762     if (contact->other_phone.str)       fprintf(f_output, "TEL;TYPE=msg:%s\n",              pst_rfc2426_escape(contact->other_phone.str, &result, &resultlen));
01763     if (contact->pager_phone.str)       fprintf(f_output, "TEL;TYPE=pager:%s\n",            pst_rfc2426_escape(contact->pager_phone.str, &result, &resultlen));
01764     if (contact->primary_fax.str)       fprintf(f_output, "TEL;TYPE=fax,pref:%s\n",         pst_rfc2426_escape(contact->primary_fax.str, &result, &resultlen));
01765     if (contact->primary_phone.str)     fprintf(f_output, "TEL;TYPE=phone,pref:%s\n",       pst_rfc2426_escape(contact->primary_phone.str, &result, &resultlen));
01766     if (contact->radio_phone.str)       fprintf(f_output, "TEL;TYPE=pcs:%s\n",              pst_rfc2426_escape(contact->radio_phone.str, &result, &resultlen));
01767     if (contact->telex.str)             fprintf(f_output, "TEL;TYPE=bbs:%s\n",              pst_rfc2426_escape(contact->telex.str, &result, &resultlen));
01768     if (contact->job_title.str)         fprintf(f_output, "TITLE:%s\n",                     pst_rfc2426_escape(contact->job_title.str, &result, &resultlen));
01769     if (contact->profession.str)        fprintf(f_output, "ROLE:%s\n",                      pst_rfc2426_escape(contact->profession.str, &result, &resultlen));
01770     if (contact->assistant_name.str || contact->assistant_phone.str) {
01771         fprintf(f_output, "AGENT:BEGIN:VCARD\n");
01772         if (contact->assistant_name.str)    fprintf(f_output, "FN:%s\n",                    pst_rfc2426_escape(contact->assistant_name.str, &result, &resultlen));
01773         if (contact->assistant_phone.str)   fprintf(f_output, "TEL:%s\n",                   pst_rfc2426_escape(contact->assistant_phone.str, &result, &resultlen));
01774     }
01775     if (contact->company_name.str)      fprintf(f_output, "ORG:%s\n",                       pst_rfc2426_escape(contact->company_name.str, &result, &resultlen));
01776     if (comment)                        fprintf(f_output, "NOTE:%s\n",                      pst_rfc2426_escape(comment, &result, &resultlen));
01777     if (item->body.str)                 fprintf(f_output, "NOTE:%s\n",                      pst_rfc2426_escape(item->body.str, &result, &resultlen));
01778 
01779     write_extra_categories(f_output, item);
01780 
01781     fprintf(f_output, "VERSION: 3.0\n");
01782     fprintf(f_output, "END:VCARD\n\n");
01783     if (result) free(result);
01784     DEBUG_RET();
01785 }
01786 
01787 
01795 int write_extra_categories(FILE* f_output, pst_item* item)
01796 {
01797     char*  result = NULL;
01798     size_t resultlen = 0;
01799     pst_item_extra_field *ef = item->extra_fields;
01800     const char *fmt = "CATEGORIES:%s";
01801     int category_started = 0;
01802     while (ef) {
01803         if (strcmp(ef->field_name, "Keywords") == 0) {
01804             fprintf(f_output, fmt, pst_rfc2426_escape(ef->value, &result, &resultlen));
01805             fmt = ", %s";
01806             category_started = 1;
01807         }
01808         ef = ef->next;
01809     }
01810     if (category_started) fprintf(f_output, "\n");
01811     if (result) free(result);
01812     return category_started;
01813 }
01814 
01815 
01816 void write_journal(FILE* f_output, pst_item* item)
01817 {
01818     char*  result = NULL;
01819     size_t resultlen = 0;
01820     char   time_buffer[30];
01821     pst_item_journal* journal = item->journal;
01822 
01823     // make everything utf8
01824     pst_convert_utf8_null(item, &item->subject);
01825     pst_convert_utf8_null(item, &item->body);
01826 
01827     fprintf(f_output, "BEGIN:VJOURNAL\n");
01828     fprintf(f_output, "DTSTAMP:%s\n",                     pst_rfc2445_datetime_format_now(sizeof(time_buffer), time_buffer));
01829     if (item->create_date)
01830         fprintf(f_output, "CREATED:%s\n",                 pst_rfc2445_datetime_format(item->create_date, sizeof(time_buffer), time_buffer));
01831     if (item->modify_date)
01832         fprintf(f_output, "LAST-MOD:%s\n",                pst_rfc2445_datetime_format(item->modify_date, sizeof(time_buffer), time_buffer));
01833     if (item->subject.str)
01834         fprintf(f_output, "SUMMARY:%s\n",                 pst_rfc2426_escape(item->subject.str, &result, &resultlen));
01835     if (item->body.str)
01836         fprintf(f_output, "DESCRIPTION:%s\n",             pst_rfc2426_escape(item->body.str, &result, &resultlen));
01837     if (journal && journal->start)
01838         fprintf(f_output, "DTSTART;VALUE=DATE-TIME:%s\n", pst_rfc2445_datetime_format(journal->start, sizeof(time_buffer), time_buffer));
01839     fprintf(f_output, "END:VJOURNAL\n");
01840     if (result) free(result);
01841 }
01842 
01843 
01844 void write_appointment(FILE* f_output, pst_item* item)
01845 {
01846     char*  result = NULL;
01847     size_t resultlen = 0;
01848     char   time_buffer[30];
01849     pst_item_appointment* appointment = item->appointment;
01850 
01851     // make everything utf8
01852     pst_convert_utf8_null(item, &item->subject);
01853     pst_convert_utf8_null(item, &item->body);
01854     pst_convert_utf8_null(item, &appointment->location);
01855 
01856     fprintf(f_output, "DTSTAMP:%s\n",                     pst_rfc2445_datetime_format_now(sizeof(time_buffer), time_buffer));
01857     if (item->create_date)
01858         fprintf(f_output, "CREATED:%s\n",                 pst_rfc2445_datetime_format(item->create_date, sizeof(time_buffer), time_buffer));
01859     if (item->modify_date)
01860         fprintf(f_output, "LAST-MOD:%s\n",                pst_rfc2445_datetime_format(item->modify_date, sizeof(time_buffer), time_buffer));
01861     if (item->subject.str)
01862         fprintf(f_output, "SUMMARY:%s\n",                 pst_rfc2426_escape(item->subject.str, &result, &resultlen));
01863     if (item->body.str)
01864         fprintf(f_output, "DESCRIPTION:%s\n",             pst_rfc2426_escape(item->body.str, &result, &resultlen));
01865     if (appointment && appointment->start)
01866         fprintf(f_output, "DTSTART;VALUE=DATE-TIME:%s\n", pst_rfc2445_datetime_format(appointment->start, sizeof(time_buffer), time_buffer));
01867     if (appointment && appointment->end)
01868         fprintf(f_output, "DTEND;VALUE=DATE-TIME:%s\n",   pst_rfc2445_datetime_format(appointment->end, sizeof(time_buffer), time_buffer));
01869     if (appointment && appointment->location.str)
01870         fprintf(f_output, "LOCATION:%s\n",                pst_rfc2426_escape(appointment->location.str, &result, &resultlen));
01871     if (appointment) {
01872         switch (appointment->showas) {
01873             case PST_FREEBUSY_TENTATIVE:
01874                 fprintf(f_output, "STATUS:TENTATIVE\n");
01875                 break;
01876             case PST_FREEBUSY_FREE:
01877                 // mark as transparent and as confirmed
01878                 fprintf(f_output, "TRANSP:TRANSPARENT\n");
01879             case PST_FREEBUSY_BUSY:
01880             case PST_FREEBUSY_OUT_OF_OFFICE:
01881                 fprintf(f_output, "STATUS:CONFIRMED\n");
01882                 break;
01883         }
01884         if (appointment->is_recurring) {
01885             const char* rules[] = {"DAILY", "WEEKLY", "MONTHLY", "YEARLY"};
01886             const char* days[]  = {"SU", "MO", "TU", "WE", "TH", "FR", "SA"};
01887             pst_recurrence *rdata = pst_convert_recurrence(appointment);
01888             fprintf(f_output, "RRULE:FREQ=%s", rules[rdata->type]);
01889             if (rdata->count)       fprintf(f_output, ";COUNT=%u",      rdata->count);
01890             if ((rdata->interval != 1) &&
01891                 (rdata->interval))  fprintf(f_output, ";INTERVAL=%u",   rdata->interval);
01892             if (rdata->dayofmonth)  fprintf(f_output, ";BYMONTHDAY=%d", rdata->dayofmonth);
01893             if (rdata->monthofyear) fprintf(f_output, ";BYMONTH=%d",    rdata->monthofyear);
01894             if (rdata->position)    fprintf(f_output, ";BYSETPOS=%d",   rdata->position);
01895             if (rdata->bydaymask) {
01896                 char byday[40];
01897                 int  empty = 1;
01898                 int i=0;
01899                 memset(byday, 0, sizeof(byday));
01900                 for (i=0; i<6; i++) {
01901                     int bit = 1 << i;
01902                     if (bit & rdata->bydaymask) {
01903                         char temp[40];
01904                         snprintf(temp, sizeof(temp), "%s%s%s", byday, (empty) ? ";BYDAY=" : ";", days[i]);
01905                         strcpy(byday, temp);
01906                         empty = 0;
01907                     }
01908                 }
01909                 fprintf(f_output, "%s", byday);
01910             }
01911             fprintf(f_output, "\n");
01912             pst_free_recurrence(rdata);
01913         }
01914         switch (appointment->label) {
01915             case PST_APP_LABEL_NONE:
01916                 if (!write_extra_categories(f_output, item)) fprintf(f_output, "CATEGORIES:NONE\n");
01917                 break;
01918             case PST_APP_LABEL_IMPORTANT:
01919                 fprintf(f_output, "CATEGORIES:IMPORTANT\n");
01920                 break;
01921             case PST_APP_LABEL_BUSINESS:
01922                 fprintf(f_output, "CATEGORIES:BUSINESS\n");
01923                 break;
01924             case PST_APP_LABEL_PERSONAL:
01925                 fprintf(f_output, "CATEGORIES:PERSONAL\n");
01926                 break;
01927             case PST_APP_LABEL_VACATION:
01928                 fprintf(f_output, "CATEGORIES:VACATION\n");
01929                 break;
01930             case PST_APP_LABEL_MUST_ATTEND:
01931                 fprintf(f_output, "CATEGORIES:MUST-ATTEND\n");
01932                 break;
01933             case PST_APP_LABEL_TRAVEL_REQ:
01934                 fprintf(f_output, "CATEGORIES:TRAVEL-REQUIRED\n");
01935                 break;
01936             case PST_APP_LABEL_NEEDS_PREP:
01937                 fprintf(f_output, "CATEGORIES:NEEDS-PREPARATION\n");
01938                 break;
01939             case PST_APP_LABEL_BIRTHDAY:
01940                 fprintf(f_output, "CATEGORIES:BIRTHDAY\n");
01941                 break;
01942             case PST_APP_LABEL_ANNIVERSARY:
01943                 fprintf(f_output, "CATEGORIES:ANNIVERSARY\n");
01944                 break;
01945             case PST_APP_LABEL_PHONE_CALL:
01946                 fprintf(f_output, "CATEGORIES:PHONE-CALL\n");
01947                 break;
01948         }
01949     }
01950     fprintf(f_output, "END:VEVENT\n");
01951     if (result) free(result);
01952 }
01953 
01954 
01955 void create_enter_dir(struct file_ll* f, pst_item *item)
01956 {
01957     pst_convert_utf8(item, &item->file_as);
01958     f->type         = item->type;
01959     f->stored_count = (item->folder) ? item->folder->item_count : 0;
01960 
01961     DEBUG_ENT("create_enter_dir");
01962     if (mode == MODE_KMAIL)
01963         f->name = mk_kmail_dir(item->file_as.str);
01964     else if (mode == MODE_RECURSE) {
01965         f->name = mk_recurse_dir(item->file_as.str, f->type);
01966         if (mode_thunder) {
01967             FILE *type_file = fopen(".type", "w");
01968             fprintf(type_file, "%d\n", item->type);
01969             fclose(type_file);
01970         }
01971     } else if (mode == MODE_SEPARATE) {
01972         // do similar stuff to recurse here.
01973         mk_separate_dir(item->file_as.str);
01974         f->name = (char*) pst_malloc(file_name_len);
01975         memset(f->name, 0, file_name_len);
01976     } else {
01977         f->name = (char*) pst_malloc(strlen(item->file_as.str)+strlen(OUTPUT_TEMPLATE)+1);
01978         sprintf(f->name, OUTPUT_TEMPLATE, item->file_as.str);
01979     }
01980 
01981     f->dname = (char*) pst_malloc(strlen(item->file_as.str)+1);
01982     strcpy(f->dname, item->file_as.str);
01983 
01984     if (overwrite != 1) {
01985         int x = 0;
01986         char *temp = (char*) pst_malloc (strlen(f->name)+10); //enough room for 10 digits
01987 
01988         sprintf(temp, "%s", f->name);
01989         check_filename(temp);
01990         while ((f->output = fopen(temp, "r"))) {
01991             DEBUG_INFO(("need to increase filename because one already exists with that name\n"));
01992             DEBUG_INFO(("- increasing it to %s%d\n", f->name, x));
01993             x++;
01994             sprintf(temp, "%s%08d", f->name, x);
01995             DEBUG_INFO(("- trying \"%s\"\n", f->name));
01996             if (x == 99999999) {
01997                 DIE(("create_enter_dir: Why can I not create a folder %s? I have tried %i extensions...\n", f->name, x));
01998             }
01999             fclose(f->output);
02000         }
02001         if (x > 0) { //then the f->name should change
02002             free (f->name);
02003             f->name = temp;
02004         } else {
02005             free(temp);
02006         }
02007     }
02008 
02009     DEBUG_INFO(("f->name = %s\nitem->folder_name = %s\n", f->name, item->file_as.str));
02010     if (mode != MODE_SEPARATE) {
02011         check_filename(f->name);
02012         if (!(f->output = fopen(f->name, "w"))) {
02013             DIE(("create_enter_dir: Could not open file \"%s\" for write\n", f->name));
02014         }
02015     }
02016     DEBUG_RET();
02017 }
02018 
02019 
02020 void close_enter_dir(struct file_ll *f)
02021 {
02022     DEBUG_INFO(("processed item count for folder %s is %i, skipped %i, total %i \n",
02023                 f->dname, f->item_count, f->skip_count, f->stored_count));
02024     if (output_mode != OUTPUT_QUIET) {
02025         pst_debug_lock();
02026             printf("\t\"%s\" - %i items done, %i items skipped.\n", f->dname, f->item_count, f->skip_count);
02027             fflush(stdout);
02028         pst_debug_unlock();
02029     }
02030     if (f->output) {
02031         struct stat st;
02032         fclose(f->output);
02033         stat(f->name, &st);
02034         if (!st.st_size) {
02035             DEBUG_WARN(("removing empty output file %s\n", f->name));
02036             remove(f->name);
02037         }
02038     }
02039     free(f->name);
02040     free(f->dname);
02041 
02042     if (mode == MODE_KMAIL)
02043         close_kmail_dir();
02044     else if (mode == MODE_RECURSE) {
02045         if (mode_thunder) {
02046             FILE *type_file = fopen(".size", "w");
02047             fprintf(type_file, "%i %i\n", f->item_count, f->stored_count);
02048             fclose(type_file);
02049         }
02050         close_recurse_dir();
02051     } else if (mode == MODE_SEPARATE)
02052         close_separate_dir();
02053 }
02054 

Generated on Mon Sep 13 14:51:49 2010 for 'LibPst' by  doxygen 1.3.9.1