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 #include "msg.h"
00011 
00012 #define OUTPUT_TEMPLATE "%s"
00013 #define OUTPUT_KMAIL_DIR_TEMPLATE ".%s.directory"
00014 #define KMAIL_INDEX ".%s.index"
00015 #define SEP_MAIL_FILE_TEMPLATE "%i%s"
00016 
00017 // max size of the c_time char*. It will store the date of the email
00018 #define C_TIME_SIZE 500
00019 
00020 struct file_ll {
00021     char *name;
00022     char *dname;
00023     FILE * output;
00024     int32_t stored_count;
00025     int32_t item_count;
00026     int32_t skip_count;
00027     int32_t type;
00028 };
00029 
00030 int       grim_reaper();
00031 pid_t     try_fork(char* folder);
00032 void      process(pst_item *outeritem, pst_desc_tree *d_ptr);
00033 void      write_email_body(FILE *f, char *body);
00034 void      removeCR(char *c);
00035 void      usage();
00036 void      version();
00037 char*     mk_kmail_dir(char* fname);
00038 int       close_kmail_dir();
00039 char*     mk_recurse_dir(char* dir, int32_t folder_type);
00040 int       close_recurse_dir();
00041 char*     mk_separate_dir(char *dir);
00042 int       close_separate_dir();
00043 void      mk_separate_file(struct file_ll *f, char *extension, int openit);
00044 void      close_separate_file(struct file_ll *f);
00045 char*     my_stristr(char *haystack, char *needle);
00046 void      check_filename(char *fname);
00047 int       acceptable_ext(pst_item_attach* attach);
00048 void      write_separate_attachment(char f_name[], pst_item_attach* attach, int attach_num, pst_file* pst);
00049 void      write_embedded_message(FILE* f_output, pst_item_attach* attach, char *boundary, pst_file* pf, int save_rtf, char** extra_mime_headers);
00050 void      write_inline_attachment(FILE* f_output, pst_item_attach* attach, char *boundary, pst_file* pst);
00051 int       valid_headers(char *header);
00052 void      header_has_field(char *header, char *field, int *flag);
00053 void      header_get_subfield(char *field, const char *subfield, char *body_subfield, size_t size_subfield);
00054 char*     header_get_field(char *header, char *field);
00055 char*     header_end_field(char *field);
00056 void      header_strip_field(char *header, char *field);
00057 int       test_base64(char *body);
00058 void      find_html_charset(char *html, char *charset, size_t charsetlen);
00059 void      find_rfc822_headers(char** extra_mime_headers);
00060 void      write_body_part(FILE* f_output, pst_string *body, char *mime, char *charset, char *boundary, pst_file* pst);
00061 void      write_schedule_part_data(FILE* f_output, pst_item* item, const char* sender, const char* method);
00062 void      write_schedule_part(FILE* f_output, pst_item* item, const char* sender, const char* boundary);
00063 void      write_normal_email(FILE* f_output, char f_name[], pst_item* item, int mode, int mode_MH, pst_file* pst, int save_rtf, int embedding, char** extra_mime_headers);
00064 void      write_vcard(FILE* f_output, pst_item *item, pst_item_contact* contact, char comment[]);
00065 int       write_extra_categories(FILE* f_output, pst_item* item);
00066 void      write_journal(FILE* f_output, pst_item* item);
00067 void      write_appointment(FILE* f_output, pst_item *item);
00068 void      create_enter_dir(struct file_ll* f, pst_item *item);
00069 void      close_enter_dir(struct file_ll *f);
00070 
00071 const char*  prog_name;
00072 char*  output_dir = ".";
00073 char*  kmail_chdir = NULL;
00074 
00075 // Normal mode just creates mbox format files in the current directory. Each file is named
00076 // the same as the folder's name that it represents
00077 #define MODE_NORMAL 0
00078 
00079 // KMail mode creates a directory structure suitable for being used directly
00080 // by the KMail application
00081 #define MODE_KMAIL 1
00082 
00083 // recurse mode creates a directory structure like the PST file. Each directory
00084 // contains only one file which stores the emails in mboxrd format.
00085 #define MODE_RECURSE 2
00086 
00087 // separate mode creates the same directory structure as recurse. The emails are stored in
00088 // separate files, numbering from 1 upward. Attachments belonging to the emails are
00089 // saved as email_no-filename (e.g. 1-samplefile.doc or 1-Attachment2.zip)
00090 #define MODE_SEPARATE 3
00091 
00092 
00093 // Output Normal just prints the standard information about what is going on
00094 #define OUTPUT_NORMAL 0
00095 
00096 // Output Quiet is provided so that only errors are printed
00097 #define OUTPUT_QUIET 1
00098 
00099 // default mime-type for attachments that have a null mime-type
00100 #define MIME_TYPE_DEFAULT "application/octet-stream"
00101 #define RFC822            "message/rfc822"
00102 
00103 // output mode for contacts
00104 #define CMODE_VCARD 0
00105 #define CMODE_LIST  1
00106 
00107 // output mode for deleted items
00108 #define DMODE_EXCLUDE 0
00109 #define DMODE_INCLUDE 1
00110 
00111 // Output type mode flags
00112 #define OTMODE_EMAIL        1
00113 #define OTMODE_APPOINTMENT  2
00114 #define OTMODE_JOURNAL      4
00115 #define OTMODE_CONTACT      8
00116 
00117 // output settings for RTF bodies
00118 // filename for the attachment
00119 #define RTF_ATTACH_NAME "rtf-body.rtf"
00120 // mime type for the attachment
00121 #define RTF_ATTACH_TYPE "application/rtf"
00122 
00123 // global settings
00124 int         mode         = MODE_NORMAL;
00125 int         mode_MH      = 0;   // a submode of MODE_SEPARATE
00126 int         mode_EX      = 0;   // a submode of MODE_SEPARATE
00127 int         mode_MSG     = 0;   // a submode of MODE_SEPARATE
00128 int         mode_thunder = 0;   // a submode of MODE_RECURSE
00129 int         output_mode  = OUTPUT_NORMAL;
00130 int         contact_mode = CMODE_VCARD;
00131 int         deleted_mode = DMODE_EXCLUDE;
00132 int         output_type_mode = 0xff;    // Default to all.
00133 int         contact_mode_specified = 0;
00134 int         overwrite = 0;
00135 int         save_rtf_body = 1;
00136 int         file_name_len = 10;     // enough room for MODE_SPEARATE file name
00137 pst_file    pstfile;
00138 regex_t     meta_charset_pattern;
00139 char*       default_charset = NULL;
00140 char*       acceptable_extensions = NULL;
00141 
00142 int         number_processors = 1;  // number of cpus we have
00143 int         max_children  = 0;      // based on number of cpus and command line args
00144 int         max_child_specified = 0;// have command line arg -j
00145 int         active_children;        // number of children of this process, cannot be larger than max_children
00146 pid_t*      child_processes;        // setup by main(), and at the start of new child process
00147 
00148 #ifdef HAVE_SEMAPHORE_H
00149 int         shared_memory_id;
00150 sem_t*      global_children = NULL;
00151 sem_t*      output_mutex    = NULL;
00152 #endif
00153 
00154 
00155 int grim_reaper(int waitall)
00156 {
00157     int available = 0;
00158 #ifdef HAVE_FORK
00159 #ifdef HAVE_SEMAPHORE_H
00160     if (global_children) {
00161         //sem_getvalue(global_children, &available);
00162         //printf("grim reaper %s for pid %d (parent %d) with %d children, %d available\n", (waitall) ? "all" : "", getpid(), getppid(), active_children, available);
00163         //fflush(stdout);
00164         int i,j;
00165         for (i=0; i<active_children; i++) {
00166             int status;
00167             pid_t child = child_processes[i];
00168             pid_t ch = waitpid(child, &status, ((waitall) ? 0 : WNOHANG));
00169             if (ch == child) {
00170                 // check termination status
00171                 //if (WIFEXITED(status)) {
00172                 //    int ext = WEXITSTATUS(status);
00173                 //    printf("Process %d exited with status  %d\n", child, ext);
00174                 //    fflush(stdout);
00175                 //}
00176                 if (WIFSIGNALED(status)) {
00177                     int sig = WTERMSIG(status);
00178                     DEBUG_INFO(("Process %d terminated with signal %d\n", child, sig));
00179                     //printf("Process %d terminated with signal %d\n", child, sig);
00180                     //fflush(stdout);
00181                 }
00182                 // this has terminated, remove it from the list
00183                 for (j=i; j<active_children-1; j++) {
00184                     child_processes[j] = child_processes[j+1];
00185                 }
00186                 active_children--;
00187                 i--;
00188             }
00189         }
00190         sem_getvalue(global_children, &available);
00191         //printf("grim reaper %s for pid %d with %d children, %d available\n", (waitall) ? "all" : "", getpid(), active_children, available);
00192         //fflush(stdout);
00193     }
00194 #endif
00195 #endif
00196     return available;
00197 }
00198 
00199 
00200 pid_t try_fork(char *folder)
00201 {
00202 #ifdef HAVE_FORK
00203 #ifdef HAVE_SEMAPHORE_H
00204     int available = grim_reaper(0);
00205     if (available) {
00206         sem_wait(global_children);
00207         pid_t child = fork();
00208         if (child < 0) {
00209             // fork failed, pretend it worked and we are the child
00210             return 0;
00211         }
00212         else if (child == 0) {
00213             // fork worked, and we are the child, reinitialize *our* list of children
00214             active_children = 0;
00215             memset(child_processes, 0, sizeof(pid_t) * max_children);
00216             pst_reopen(&pstfile);   // close and reopen the pst file to get an independent file position pointer
00217         }
00218         else {
00219             // fork worked, and we are the parent, record this child that we need to wait for
00220             //pid_t me = getpid();
00221             //printf("parent %d forked child pid %d to process folder %s\n", me, child, folder);
00222             //fflush(stdout);
00223             child_processes[active_children++] = child;
00224         }
00225         return child;
00226     }
00227     else {
00228         return 0;   // pretend to have forked and we are the child
00229     }
00230 #endif
00231 #endif
00232     return 0;
00233 }
00234 
00235 
00236 void process(pst_item *outeritem, pst_desc_tree *d_ptr)
00237 {
00238     struct file_ll ff;
00239     pst_item *item = NULL;
00240 
00241     DEBUG_ENT("process");
00242     memset(&ff, 0, sizeof(ff));
00243     create_enter_dir(&ff, outeritem);
00244 
00245     for (; d_ptr; d_ptr = d_ptr->next) {
00246         DEBUG_INFO(("New item record\n"));
00247         if (!d_ptr->desc) {
00248             ff.skip_count++;
00249             DEBUG_WARN(("ERROR item's desc record is NULL\n"));
00250             continue;
00251         }
00252         DEBUG_INFO(("Desc Email ID %#"PRIx64" [d_ptr->d_id = %#"PRIx64"]\n", d_ptr->desc->i_id, d_ptr->d_id));
00253 
00254         item = pst_parse_item(&pstfile, d_ptr, NULL);
00255         DEBUG_INFO(("About to process item\n"));
00256 
00257         if (!item) {
00258             ff.skip_count++;
00259             DEBUG_INFO(("A NULL item was seen\n"));
00260             continue;
00261         }
00262 
00263         if (item->subject.str) {
00264             DEBUG_INFO(("item->subject = %s\n", item->subject.str));
00265         }
00266 
00267         if (item->folder && item->file_as.str) {
00268             DEBUG_INFO(("Processing Folder \"%s\"\n", item->file_as.str));
00269             if (output_mode != OUTPUT_QUIET) {
00270                 pst_debug_lock();
00271                     printf("Processing Folder \"%s\"\n", item->file_as.str);
00272                     fflush(stdout);
00273                 pst_debug_unlock();
00274             }
00275             ff.item_count++;
00276             if (d_ptr->child && (deleted_mode == DMODE_INCLUDE || strcasecmp(item->file_as.str, "Deleted Items"))) {
00277                 //if this is a non-empty folder other than deleted items, we want to recurse into it
00278                 pid_t parent = getpid();
00279                 pid_t child = try_fork(item->file_as.str);
00280                 if (child == 0) {
00281                     // we are the child process, or the original parent if no children were available
00282                     pid_t me = getpid();
00283                     process(item, d_ptr->child);
00284 #ifdef HAVE_FORK
00285 #ifdef HAVE_SEMAPHORE_H
00286                     if (me != parent) {
00287                         // we really were a child, forked for the sole purpose of processing this folder
00288                         // free my child count slot before really exiting, since
00289                         // all I am doing here is waiting for my children to exit
00290                         sem_post(global_children);
00291                         grim_reaper(1); // wait for all my child processes to exit
00292                         exit(0);        // really exit
00293                     }
00294 #endif
00295 #endif
00296                 }
00297             }
00298 
00299         } else if (item->contact && (item->type == PST_TYPE_CONTACT)) {
00300             DEBUG_INFO(("Processing Contact\n"));
00301             if (!(output_type_mode & OTMODE_CONTACT)) {
00302                 ff.skip_count++;
00303                 DEBUG_INFO(("skipping contact: not in output type list\n"));
00304             }
00305             else {
00306                 if (!ff.type) ff.type = item->type;
00307                 if ((ff.type != PST_TYPE_CONTACT) && (mode != MODE_SEPARATE)) {
00308                     ff.skip_count++;
00309                     DEBUG_INFO(("I have a contact, but the folder type %"PRIi32" isn't a contacts folder. Skipping it\n", ff.type));
00310                 }
00311                 else {
00312                     ff.item_count++;
00313                     if (mode == MODE_SEPARATE) mk_separate_file(&ff, (mode_EX) ? ".vcf" : "", 1);
00314                     if (contact_mode == CMODE_VCARD) {
00315                         pst_convert_utf8_null(item, &item->comment);
00316                         write_vcard(ff.output, item, item->contact, item->comment.str);
00317                     }
00318                     else {
00319                         pst_convert_utf8(item, &item->contact->fullname);
00320                         pst_convert_utf8(item, &item->contact->address1);
00321                         fprintf(ff.output, "%s <%s>\n", item->contact->fullname.str, item->contact->address1.str);
00322                     }
00323                     if (mode == MODE_SEPARATE) close_separate_file(&ff);
00324                 }
00325             }
00326 
00327         } else if (item->email && ((item->type == PST_TYPE_NOTE) || (item->type == PST_TYPE_SCHEDULE) || (item->type == PST_TYPE_REPORT))) {
00328             DEBUG_INFO(("Processing Email\n"));
00329             if (!(output_type_mode & OTMODE_EMAIL)) {
00330                 ff.skip_count++;
00331                 DEBUG_INFO(("skipping email: not in output type list\n"));
00332             }
00333             else {
00334                 if (!ff.type) ff.type = item->type;
00335                 if ((ff.type != PST_TYPE_NOTE) && (ff.type != PST_TYPE_SCHEDULE) && (ff.type != PST_TYPE_REPORT) && (mode != MODE_SEPARATE)) {
00336                     ff.skip_count++;
00337                     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));
00338                 }
00339                 else {
00340                     char *extra_mime_headers = NULL;
00341                     ff.item_count++;
00342                     if (mode == MODE_SEPARATE) {
00343                         // process this single email message, possibly forking
00344                         pid_t parent = getpid();
00345                         pid_t child = try_fork(item->file_as.str);
00346                         if (child == 0) {
00347                             // we are the child process, or the original parent if no children were available
00348                             pid_t me = getpid();
00349                             mk_separate_file(&ff, (mode_EX) ? ".eml" : "", 1);
00350                             write_normal_email(ff.output, ff.name, item, mode, mode_MH, &pstfile, save_rtf_body, 0, &extra_mime_headers);
00351                             close_separate_file(&ff);
00352                             if (mode_MSG) {
00353                                 mk_separate_file(&ff, ".msg", 0);
00354                                 write_msg_email(ff.name, item, &pstfile);
00355                             }
00356 #ifdef HAVE_FORK
00357 #ifdef HAVE_SEMAPHORE_H
00358                             if (me != parent) {
00359                                 // we really were a child, forked for the sole purpose of processing this message
00360                                 // free my child count slot before really exiting, since
00361                                 // all I am doing here is waiting for my children to exit
00362                                 sem_post(global_children);
00363                                 grim_reaper(1); // wait for all my child processes to exit - there should not be any
00364                                 exit(0);        // really exit
00365                             }
00366 #endif
00367 #endif
00368                         }
00369                     }
00370                     else {
00371                         // process this single email message, cannot fork since not separate mode
00372                         write_normal_email(ff.output, ff.name, item, mode, mode_MH, &pstfile, save_rtf_body, 0, &extra_mime_headers);
00373                     }
00374                 }
00375             }
00376 
00377         } else if (item->journal && (item->type == PST_TYPE_JOURNAL)) {
00378             DEBUG_INFO(("Processing Journal Entry\n"));
00379             if (!(output_type_mode & OTMODE_JOURNAL)) {
00380                 ff.skip_count++;
00381                 DEBUG_INFO(("skipping journal entry: not in output type list\n"));
00382             }
00383             else {
00384                 if (!ff.type) ff.type = item->type;
00385                 if ((ff.type != PST_TYPE_JOURNAL) && (mode != MODE_SEPARATE)) {
00386                     ff.skip_count++;
00387                     DEBUG_INFO(("I have a journal entry, but the folder type %"PRIi32" isn't a journal folder. Skipping it\n", ff.type));
00388                 }
00389                 else {
00390                     ff.item_count++;
00391                     if (mode == MODE_SEPARATE) mk_separate_file(&ff, (mode_EX) ? ".ics" : "", 1);
00392                     write_journal(ff.output, item);
00393                     fprintf(ff.output, "\n");
00394                     if (mode == MODE_SEPARATE) close_separate_file(&ff);
00395                 }
00396             }
00397 
00398         } else if (item->appointment && (item->type == PST_TYPE_APPOINTMENT)) {
00399             DEBUG_INFO(("Processing Appointment Entry\n"));
00400             if (!(output_type_mode & OTMODE_APPOINTMENT)) {
00401                 ff.skip_count++;
00402                 DEBUG_INFO(("skipping appointment: not in output type list\n"));
00403             }
00404             else {
00405                 if (!ff.type) ff.type = item->type;
00406                 if ((ff.type != PST_TYPE_APPOINTMENT) && (mode != MODE_SEPARATE)) {
00407                     ff.skip_count++;
00408                     DEBUG_INFO(("I have an appointment, but the folder type %"PRIi32" isn't an appointment folder. Skipping it\n", ff.type));
00409                 }
00410                 else {
00411                     ff.item_count++;
00412                     if (mode == MODE_SEPARATE) mk_separate_file(&ff, (mode_EX) ? ".ics" : "", 1);
00413                     write_schedule_part_data(ff.output, item, NULL, NULL);
00414                     fprintf(ff.output, "\n");
00415                     if (mode == MODE_SEPARATE) close_separate_file(&ff);
00416                 }
00417             }
00418 
00419         } else if (item->message_store) {
00420             // there should only be one message_store, and we have already done it
00421             ff.skip_count++;
00422             DEBUG_INFO(("item with message store content, type %i %s folder type %i, skipping it\n", item->type, item->ascii_type, ff.type));
00423 
00424         } else {
00425             ff.skip_count++;
00426             DEBUG_INFO(("Unknown item type %i (%s) name (%s)\n",
00427                         item->type, item->ascii_type, item->file_as.str));
00428         }
00429         pst_freeItem(item);
00430     }
00431     close_enter_dir(&ff);
00432     DEBUG_RET();
00433 }
00434 
00435 
00436 
00437 int main(int argc, char* const* argv) {
00438     pst_item *item = NULL;
00439     pst_desc_tree *d_ptr;
00440     char * fname = NULL;
00441     char *d_log  = NULL;
00442     int c,x;
00443     char *temp = NULL;               //temporary char pointer
00444     prog_name = argv[0];
00445 
00446     time_t now = time(NULL);
00447     srand((unsigned)now);
00448 
00449     if (regcomp(&meta_charset_pattern, "<meta[^>]*content=\"[^>]*charset=([^>\";]*)[\";]", REG_ICASE | REG_EXTENDED)) {
00450         printf("cannot compile regex pattern to find content charset in html bodies\n");
00451         exit(3);
00452     }
00453 
00454     // command-line option handling
00455     while ((c = getopt(argc, argv, "a:bC:c:Dd:emhj:kMo:qrSt:uVw"))!= -1) {
00456         switch (c) {
00457         case 'a':
00458             if (optarg) {
00459                 int n = strlen(optarg);
00460                 acceptable_extensions = (char*)pst_malloc(n+2);
00461                 strcpy(acceptable_extensions, optarg);
00462                 acceptable_extensions[n+1] = '\0';  // double null terminates array of non-empty null terminated strings.
00463                 char *p = acceptable_extensions;
00464                 while (*p) {
00465                     if (*p == ',') *p = '\0';
00466                     p++;
00467                 }
00468             }
00469             break;
00470         case 'b':
00471             save_rtf_body = 0;
00472             break;
00473         case 'C':
00474             if (optarg) {
00475                 default_charset = optarg;
00476             }
00477             else {
00478                 usage();
00479                 exit(0);
00480             }
00481             break;
00482         case 'c':
00483             if (optarg && optarg[0]=='v') {
00484                 contact_mode=CMODE_VCARD;
00485                 contact_mode_specified = 1;
00486             }
00487             else if (optarg && optarg[0]=='l') {
00488                 contact_mode=CMODE_LIST;
00489                 contact_mode_specified = 1;
00490             }
00491             else {
00492                 usage();
00493                 exit(0);
00494             }
00495             break;
00496         case 'D':
00497             deleted_mode = DMODE_INCLUDE;
00498             break;
00499         case 'd':
00500             d_log = optarg;
00501             break;
00502         case 'h':
00503             usage();
00504             exit(0);
00505             break;
00506         case 'j':
00507             max_children = atoi(optarg);
00508             max_child_specified = 1;
00509             break;
00510         case 'k':
00511             mode = MODE_KMAIL;
00512             break;
00513         case 'M':
00514             mode = MODE_SEPARATE;
00515             mode_MH  = 1;
00516             mode_EX  = 0;
00517             mode_MSG = 0;
00518             break;
00519         case 'e':
00520             mode = MODE_SEPARATE;
00521             mode_MH  = 1;
00522             mode_EX  = 1;
00523             mode_MSG = 0;
00524             file_name_len = 14;
00525             break;
00526         case 'm':
00527             mode = MODE_SEPARATE;
00528             mode_MH  = 1;
00529             mode_EX  = 1;
00530             mode_MSG = 1;
00531             file_name_len = 14;
00532             break;
00533         case 'o':
00534             output_dir = optarg;
00535             break;
00536         case 'q':
00537             output_mode = OUTPUT_QUIET;
00538             break;
00539         case 'r':
00540             mode = MODE_RECURSE;
00541             mode_thunder = 0;
00542             break;
00543         case 'S':
00544             mode = MODE_SEPARATE;
00545             mode_MH  = 0;
00546             mode_EX  = 0;
00547             mode_MSG = 0;
00548             break;
00549         case 't':
00550             // email, appointment, contact, other
00551             if (!optarg) {
00552                 usage();
00553                 exit(0);
00554             }
00555             temp = optarg;
00556             output_type_mode = 0;
00557             while (*temp > 0) {
00558               switch (temp[0]) {
00559                 case 'e':
00560                     output_type_mode |= OTMODE_EMAIL;
00561                     break;
00562                 case 'a':
00563                     output_type_mode |= OTMODE_APPOINTMENT;
00564                     break;
00565                 case 'j':
00566                     output_type_mode |= OTMODE_JOURNAL;
00567                     break;
00568                 case 'c':
00569                     output_type_mode |= OTMODE_CONTACT;
00570                     break;
00571                 default:
00572                     usage();
00573                     exit(0);
00574                     break;
00575               }
00576               temp++;
00577             }
00578             break;
00579         case 'u':
00580             mode = MODE_RECURSE;
00581             mode_thunder = 1;
00582             break;
00583         case 'V':
00584             version();
00585             exit(0);
00586             break;
00587         case 'w':
00588             overwrite = 1;
00589             break;
00590         default:
00591             usage();
00592             exit(1);
00593             break;
00594         }
00595     }
00596 
00597     if (argc > optind) {
00598         fname = argv[optind];
00599     } else {
00600         usage();
00601         exit(2);
00602     }
00603 
00604 #ifdef _SC_NPROCESSORS_ONLN
00605     number_processors =  sysconf(_SC_NPROCESSORS_ONLN);
00606 #endif
00607     max_children    = (max_child_specified) ? max_children : number_processors * 4;
00608     active_children = 0;
00609     child_processes = (pid_t *)pst_malloc(sizeof(pid_t) * max_children);
00610     memset(child_processes, 0, sizeof(pid_t) * max_children);
00611 
00612 #ifdef HAVE_SEMAPHORE_H
00613     if (max_children) {
00614         shared_memory_id = shmget(IPC_PRIVATE, sizeof(sem_t)*2, 0777);
00615         if (shared_memory_id >= 0) {
00616             global_children = (sem_t *)shmat(shared_memory_id, NULL, 0);
00617             if (global_children == (sem_t *)-1) global_children = NULL;
00618             if (global_children) {
00619                 output_mutex = &(global_children[1]);
00620                 sem_init(global_children, 1, max_children);
00621                 sem_init(output_mutex, 1, 1);
00622             }
00623             shmctl(shared_memory_id, IPC_RMID, NULL);
00624         }
00625     }
00626 #endif
00627 
00628     #ifdef DEBUG_ALL
00629         // force a log file
00630         if (!d_log) d_log = "readpst.log";
00631     #endif // defined DEBUG_ALL
00632     #ifdef HAVE_SEMAPHORE_H
00633         DEBUG_INIT(d_log, output_mutex);
00634     #else
00635         DEBUG_INIT(d_log, NULL);
00636     #endif
00637     DEBUG_ENT("main");
00638 
00639     if (output_mode != OUTPUT_QUIET) printf("Opening PST file and indexes...\n");
00640     RET_DERROR(pst_open(&pstfile, fname, default_charset), 1, ("Error opening File\n"));
00641     RET_DERROR(pst_load_index(&pstfile), 2, ("Index Error\n"));
00642 
00643     pst_load_extended_attributes(&pstfile);
00644 
00645     if (chdir(output_dir)) {
00646         x = errno;
00647         pst_close(&pstfile);
00648         DEBUG_RET();
00649         DIE(("Cannot change to output dir %s: %s\n", output_dir, strerror(x)));
00650     }
00651 
00652     d_ptr = pstfile.d_head; // first record is main record
00653     item  = pst_parse_item(&pstfile, d_ptr, NULL);
00654     if (!item || !item->message_store) {
00655         DEBUG_RET();
00656         DIE(("Could not get root record\n"));
00657     }
00658 
00659     // default the file_as to the same as the main filename if it doesn't exist
00660     if (!item->file_as.str) {
00661         if (!(temp = strrchr(fname, '/')))
00662             if (!(temp = strrchr(fname, '\\')))
00663                 temp = fname;
00664             else
00665                 temp++; // get past the "\\"
00666         else
00667             temp++; // get past the "/"
00668         item->file_as.str = (char*)pst_malloc(strlen(temp)+1);
00669         strcpy(item->file_as.str, temp);
00670         item->file_as.is_utf8 = 1;
00671         DEBUG_INFO(("file_as was blank, so am using %s\n", item->file_as.str));
00672     }
00673     DEBUG_INFO(("Root Folder Name: %s\n", item->file_as.str));
00674 
00675     d_ptr = pst_getTopOfFolders(&pstfile, item);
00676     if (!d_ptr) {
00677         DEBUG_RET();
00678         DIE(("Top of folders record not found. Cannot continue\n"));
00679     }
00680 
00681     process(item, d_ptr->child);    // do the children of TOPF
00682     grim_reaper(1); // wait for all child processes
00683 
00684     pst_freeItem(item);
00685     pst_close(&pstfile);
00686     DEBUG_RET();
00687 
00688 #ifdef HAVE_SEMAPHORE_H
00689     if (global_children) {
00690         sem_destroy(global_children);
00691         sem_destroy(output_mutex);
00692         shmdt(global_children);
00693     }
00694 #endif
00695 
00696     regfree(&meta_charset_pattern);
00697     return 0;
00698 }
00699 
00700 
00701 void write_email_body(FILE *f, char *body) {
00702     char *n = body;
00703     DEBUG_ENT("write_email_body");
00704     if (mode != MODE_SEPARATE) {
00705         while (n) {
00706             char *p = body;
00707             while (*p == '>') p++;
00708             if (strncmp(p, "From ", 5) == 0) fprintf(f, ">");
00709             if ((n = strchr(body, '\n'))) {
00710                 n++;
00711                 pst_fwrite(body, n-body, 1, f); //write just a line
00712                 body = n;
00713             }
00714         }
00715     }
00716     pst_fwrite(body, strlen(body), 1, f);
00717     DEBUG_RET();
00718 }
00719 
00720 
00721 void removeCR (char *c) {
00722     // converts \r\n to \n
00723     char *a, *b;
00724     DEBUG_ENT("removeCR");
00725     a = b = c;
00726     while (*a != '\0') {
00727         *b = *a;
00728         if (*a != '\r') b++;
00729         a++;
00730     }
00731     *b = '\0';
00732     DEBUG_RET();
00733 }
00734 
00735 
00736 void usage() {
00737     DEBUG_ENT("usage");
00738     version();
00739     printf("Usage: %s [OPTIONS] {PST FILENAME}\n", prog_name);
00740     printf("OPTIONS:\n");
00741     printf("\t-V\t- Version. Display program version\n");
00742     printf("\t-C charset\t- character set for items with an unspecified character set\n");
00743     printf("\t-D\t- Include deleted items in output\n");
00744     printf("\t-M\t- Write emails in the MH (rfc822) format\n");
00745     printf("\t-S\t- Separate. Write emails in the separate format\n");
00746     printf("\t-a <attachment-extension-list>\t- Discard any attachment without an extension on the list\n");
00747     printf("\t-b\t- Don't save RTF-Body attachments\n");
00748     printf("\t-c[v|l]\t- Set the Contact output mode. -cv = VCard, -cl = EMail list\n");
00749     printf("\t-d <filename> \t- Debug to file.\n");
00750     printf("\t-e\t- As with -M, but include extensions on output files\n");
00751     printf("\t-h\t- Help. This screen\n");
00752     printf("\t-j <integer>\t- Number of parallel jobs to run\n");
00753     printf("\t-k\t- KMail. Output in kmail format\n");
00754     printf("\t-m\t- As with -e, but write .msg files also\n");
00755     printf("\t-o <dirname>\t- Output directory to write files to. CWD is changed *after* opening pst file\n");
00756     printf("\t-q\t- Quiet. Only print error messages\n");
00757     printf("\t-r\t- Recursive. Output in a recursive format\n");
00758     printf("\t-t[eajc]\t- Set the output type list. e = email, a = attachment, j = journal, c = contact\n");
00759     printf("\t-u\t- Thunderbird mode. Write two extra .size and .type files\n");
00760     printf("\t-w\t- Overwrite any output mbox files\n");
00761     printf("\n");
00762     printf("Only one of -M -S -e -k -m -r should be specified\n");
00763     DEBUG_RET();
00764 }
00765 
00766 
00767 void version() {
00768     DEBUG_ENT("version");
00769     printf("ReadPST / LibPST v%s\n", VERSION);
00770 #if BYTE_ORDER == BIG_ENDIAN
00771     printf("Big Endian implementation being used.\n");
00772 #elif BYTE_ORDER == LITTLE_ENDIAN
00773     printf("Little Endian implementation being used.\n");
00774 #else
00775 #  error "Byte order not supported by this library"
00776 #endif
00777     DEBUG_RET();
00778 }
00779 
00780 
00781 char *mk_kmail_dir(char *fname) {
00782     //change to that directory
00783     //make a directory based on OUTPUT_KMAIL_DIR_TEMPLATE
00784     //allocate space for OUTPUT_TEMPLATE and form a char* with fname
00785     //return that value
00786     char *dir, *out_name, *index;
00787     int x;
00788     DEBUG_ENT("mk_kmail_dir");
00789     if (kmail_chdir && chdir(kmail_chdir)) {
00790         x = errno;
00791         DIE(("mk_kmail_dir: Cannot change to directory %s: %s\n", kmail_chdir, strerror(x)));
00792     }
00793     dir = pst_malloc(strlen(fname)+strlen(OUTPUT_KMAIL_DIR_TEMPLATE)+1);
00794     sprintf(dir, OUTPUT_KMAIL_DIR_TEMPLATE, fname);
00795     check_filename(dir);
00796     if (D_MKDIR(dir)) {
00797         if (errno != EEXIST) {  // not an error because it exists
00798             x = errno;
00799             DIE(("mk_kmail_dir: Cannot create directory %s: %s\n", dir, strerror(x)));
00800         }
00801     }
00802     kmail_chdir = pst_realloc(kmail_chdir, strlen(dir)+1);
00803     strcpy(kmail_chdir, dir);
00804     free (dir);
00805 
00806     //we should remove any existing indexes created by KMail, cause they might be different now
00807     index = pst_malloc(strlen(fname)+strlen(KMAIL_INDEX)+1);
00808     sprintf(index, KMAIL_INDEX, fname);
00809     unlink(index);
00810     free(index);
00811 
00812     out_name = pst_malloc(strlen(fname)+strlen(OUTPUT_TEMPLATE)+1);
00813     sprintf(out_name, OUTPUT_TEMPLATE, fname);
00814     DEBUG_RET();
00815     return out_name;
00816 }
00817 
00818 
00819 int close_kmail_dir() {
00820     // change ..
00821     int x;
00822     DEBUG_ENT("close_kmail_dir");
00823     if (kmail_chdir) { //only free kmail_chdir if not NULL. do not change directory
00824         free(kmail_chdir);
00825         kmail_chdir = NULL;
00826     } else {
00827         if (chdir("..")) {
00828             x = errno;
00829             DIE(("close_kmail_dir: Cannot move up dir (..): %s\n", strerror(x)));
00830         }
00831     }
00832     DEBUG_RET();
00833     return 0;
00834 }
00835 
00836 
00837 // this will create a directory by that name,
00838 // then make an mbox file inside that directory.
00839 char *mk_recurse_dir(char *dir, int32_t folder_type) {
00840     int x;
00841     char *out_name;
00842     DEBUG_ENT("mk_recurse_dir");
00843     check_filename(dir);
00844     if (D_MKDIR (dir)) {
00845         if (errno != EEXIST) {  // not an error because it exists
00846             x = errno;
00847             DIE(("mk_recurse_dir: Cannot create directory %s: %s\n", dir, strerror(x)));
00848         }
00849     }
00850     if (chdir (dir)) {
00851         x = errno;
00852         DIE(("mk_recurse_dir: Cannot change to directory %s: %s\n", dir, strerror(x)));
00853     }
00854     switch (folder_type) {
00855         case PST_TYPE_APPOINTMENT:
00856             out_name = strdup("calendar");
00857             break;
00858         case PST_TYPE_CONTACT:
00859             out_name = strdup("contacts");
00860             break;
00861         case PST_TYPE_JOURNAL:
00862             out_name = strdup("journal");
00863             break;
00864         case PST_TYPE_STICKYNOTE:
00865         case PST_TYPE_TASK:
00866         case PST_TYPE_NOTE:
00867         case PST_TYPE_OTHER:
00868         case PST_TYPE_REPORT:
00869         default:
00870             out_name = strdup("mbox");
00871             break;
00872     }
00873     DEBUG_RET();
00874     return out_name;
00875 }
00876 
00877 
00878 int close_recurse_dir() {
00879     int x;
00880     DEBUG_ENT("close_recurse_dir");
00881     if (chdir("..")) {
00882         x = errno;
00883         DIE(("close_recurse_dir: Cannot go up dir (..): %s\n", strerror(x)));
00884     }
00885     DEBUG_RET();
00886     return 0;
00887 }
00888 
00889 
00890 char *mk_separate_dir(char *dir) {
00891     size_t dirsize = strlen(dir) + 10;
00892     char dir_name[dirsize];
00893     int x = 0, y = 0;
00894 
00895     DEBUG_ENT("mk_separate_dir");
00896     do {
00897         if (y == 0)
00898             snprintf(dir_name, dirsize, "%s", dir);
00899         else
00900             snprintf(dir_name, dirsize, "%s" SEP_MAIL_FILE_TEMPLATE, dir, y, ""); // enough for 9 digits allocated above
00901 
00902         check_filename(dir_name);
00903         DEBUG_INFO(("about to try creating %s\n", dir_name));
00904         if (D_MKDIR(dir_name)) {
00905             if (errno != EEXIST) { // if there is an error, and it doesn't already exist
00906                 x = errno;
00907                 DIE(("mk_separate_dir: Cannot create directory %s: %s\n", dir, strerror(x)));
00908             }
00909         } else {
00910             break;
00911         }
00912         y++;
00913     } while (overwrite == 0);
00914 
00915     if (chdir(dir_name)) {
00916         x = errno;
00917         DIE(("mk_separate_dir: Cannot change to directory %s: %s\n", dir, strerror(x)));
00918     }
00919 
00920     if (overwrite) {
00921         // we should probably delete all files from this directory
00922 #if !defined(WIN32) && !defined(__CYGWIN__)
00923         DIR * sdir = NULL;
00924         struct dirent *dirent = NULL;
00925         struct stat filestat;
00926         if (!(sdir = opendir("./"))) {
00927             DEBUG_WARN(("mk_separate_dir: Cannot open dir \"%s\" for deletion of old contents\n", "./"));
00928         } else {
00929             while ((dirent = readdir(sdir))) {
00930                 if (lstat(dirent->d_name, &filestat) != -1)
00931                     if (S_ISREG(filestat.st_mode)) {
00932                         if (unlink(dirent->d_name)) {
00933                             y = errno;
00934                             DIE(("mk_separate_dir: unlink returned error on file %s: %s\n", dirent->d_name, strerror(y)));
00935                         }
00936                     }
00937             }
00938         }
00939 #endif
00940     }
00941 
00942     // we don't return a filename here cause it isn't necessary.
00943     DEBUG_RET();
00944     return NULL;
00945 }
00946 
00947 
00948 int close_separate_dir() {
00949     int x;
00950     DEBUG_ENT("close_separate_dir");
00951     if (chdir("..")) {
00952         x = errno;
00953         DIE(("close_separate_dir: Cannot go up dir (..): %s\n", strerror(x)));
00954     }
00955     DEBUG_RET();
00956     return 0;
00957 }
00958 
00959 
00960 void mk_separate_file(struct file_ll *f, char *extension, int openit) {
00961     DEBUG_ENT("mk_separate_file");
00962     DEBUG_INFO(("opening next file to save email\n"));
00963     if (f->item_count > 999999999) { // bigger than nine 9's
00964         DIE(("mk_separate_file: The number of emails in this folder has become too high to handle\n"));
00965     }
00966     sprintf(f->name, SEP_MAIL_FILE_TEMPLATE, f->item_count, extension);
00967     check_filename(f->name);
00968     if (openit) {
00969         if (!(f->output = fopen(f->name, "w"))) {
00970             DIE(("mk_separate_file: Cannot open file to save email \"%s\"\n", f->name));
00971         }
00972     }
00973     DEBUG_RET();
00974 }
00975 
00976 
00977 void close_separate_file(struct file_ll *f) {
00978     DEBUG_ENT("close_separate_file");
00979     if (f->output) {
00980         struct stat st;
00981         fclose(f->output);
00982         stat(f->name, &st);
00983         if (!st.st_size) {
00984             DEBUG_WARN(("removing empty output file %s\n", f->name));
00985             remove(f->name);
00986         }
00987         f->output = NULL;
00988     }
00989     DEBUG_RET();
00990 }
00991 
00992 
00993 char *my_stristr(char *haystack, char *needle) {
00994     // my_stristr varies from strstr in that its searches are case-insensitive
00995     char *x=haystack, *y=needle, *z = NULL;
00996     if (!haystack || !needle) {
00997         return NULL;
00998     }
00999     while (*y != '\0' && *x != '\0') {
01000         if (tolower(*y) == tolower(*x)) {
01001             // move y on one
01002             y++;
01003             if (!z) {
01004                 z = x; // store first position in haystack where a match is made
01005             }
01006         } else {
01007             y = needle; // reset y to the beginning of the needle
01008             z = NULL; // reset the haystack storage point
01009         }
01010         x++; // advance the search in the haystack
01011     }
01012     // If the haystack ended before our search finished, it's not a match.
01013     if (*y != '\0') return NULL;
01014     return z;
01015 }
01016 
01017 
01018 void check_filename(char *fname) {
01019     char *t = fname;
01020     DEBUG_ENT("check_filename");
01021     if (!t) {
01022         DEBUG_RET();
01023         return;
01024     }
01025     while ((t = strpbrk(t, "/\\:"))) {
01026         // while there are characters in the second string that we don't want
01027         *t = '_'; //replace them with an underscore
01028     }
01029     DEBUG_RET();
01030 }
01031 
01032 
01039 int  acceptable_ext(pst_item_attach* attach)
01040 {
01041     if (!acceptable_extensions || *acceptable_extensions == '\0') return 1;     // acceptable list missing or empty
01042     char *attach_filename = (attach->filename2.str) ? attach->filename2.str
01043                                                     : attach->filename1.str;
01044     if (!attach_filename) return 1; // attachment with no name is always acceptable
01045     char *e = strrchr(attach_filename, '.');
01046     if (!e) return 1;               // attachment with no extension is always acceptable.
01047     DEBUG_ENT("acceptable_ext");
01048     DEBUG_INFO(("attachment extension %s\n", e));
01049     int rc = 0;
01050     char *a = acceptable_extensions;
01051     while (*a) {
01052         if (pst_stricmp(a, e) == 0) {
01053             rc = 1;
01054             break;
01055         }
01056         a += strlen(a) + 1;
01057     }
01058     DEBUG_INFO(("attachment acceptable returns %d\n", rc));
01059     DEBUG_RET();
01060     return rc;
01061 }
01062 
01063 
01064 void write_separate_attachment(char f_name[], pst_item_attach* attach, int attach_num, pst_file* pst)
01065 {
01066     FILE *fp = NULL;
01067     int x = 0;
01068     char *temp = NULL;
01069 
01070     // If there is a long filename (filename2) use that, otherwise
01071     // use the 8.3 filename (filename1)
01072     char *attach_filename = (attach->filename2.str) ? attach->filename2.str
01073                                                     : attach->filename1.str;
01074     DEBUG_ENT("write_separate_attachment");
01075     DEBUG_INFO(("Attachment %s Size is %#"PRIx64", data = %#"PRIxPTR", id %#"PRIx64"\n", attach_filename, (uint64_t)attach->data.size, attach->data.data, attach->i_id));
01076 
01077     if (!attach->data.data) {
01078         // make sure we can fetch data from the id
01079         pst_index_ll *ptr = pst_getID(pst, attach->i_id);
01080         if (!ptr) {
01081             DEBUG_WARN(("Couldn't find i_id %#"PRIx64". Cannot save attachment to file\n", attach->i_id));
01082             DEBUG_RET();
01083             return;
01084         }
01085     }
01086 
01087     check_filename(f_name);
01088     if (!attach_filename) {
01089         // generate our own (dummy) filename for the attachement
01090         temp = pst_malloc(strlen(f_name)+15);
01091         sprintf(temp, "%s-attach%i", f_name, attach_num);
01092     } else {
01093         // have an attachment name, make sure it's unique
01094         temp = pst_malloc(strlen(f_name)+strlen(attach_filename)+15);
01095         do {
01096             if (fp) fclose(fp);
01097             if (x == 0)
01098                 sprintf(temp, "%s-%s", f_name, attach_filename);
01099             else
01100                 sprintf(temp, "%s-%s-%i", f_name, attach_filename, x);
01101         } while ((fp = fopen(temp, "r")) && ++x < 99999999);
01102         if (x > 99999999) {
01103             DIE(("error finding attachment name. exhausted possibilities to %s\n", temp));
01104         }
01105     }
01106     DEBUG_INFO(("Saving attachment to %s\n", temp));
01107     if (!(fp = fopen(temp, "w"))) {
01108         DEBUG_WARN(("write_separate_attachment: Cannot open attachment save file \"%s\"\n", temp));
01109     } else {
01110         (void)pst_attach_to_file(pst, attach, fp);
01111         fclose(fp);
01112     }
01113     if (temp) free(temp);
01114     DEBUG_RET();
01115 }
01116 
01117 
01118 void write_embedded_message(FILE* f_output, pst_item_attach* attach, char *boundary, pst_file* pf, int save_rtf, char** extra_mime_headers)
01119 {
01120     pst_index_ll *ptr;
01121     DEBUG_ENT("write_embedded_message");
01122     ptr = pst_getID(pf, attach->i_id);
01123 
01124     pst_desc_tree d_ptr;
01125     d_ptr.d_id        = 0;
01126     d_ptr.parent_d_id = 0;
01127     d_ptr.assoc_tree  = NULL;
01128     d_ptr.desc        = ptr;
01129     d_ptr.no_child    = 0;
01130     d_ptr.prev        = NULL;
01131     d_ptr.next        = NULL;
01132     d_ptr.parent      = NULL;
01133     d_ptr.child       = NULL;
01134     d_ptr.child_tail  = NULL;
01135 
01136     pst_item *item = pst_parse_item(pf, &d_ptr, attach->id2_head);
01137     // It appears that if the embedded message contains an appointment/
01138     // calendar item, pst_parse_item returns NULL due to the presence of
01139     // an unexpected reference type of 0x1048, which seems to represent
01140     // an array of GUIDs representing a CLSID. It's likely that this is
01141     // a reference to an internal Outlook COM class.
01142     //      Log the skipped item and continue on.
01143     if (!item) {
01144         DEBUG_WARN(("write_embedded_message: pst_parse_item was unable to parse the embedded message in attachment ID %llu", attach->i_id));
01145     } else {
01146         if (!item->email) {
01147             DEBUG_WARN(("write_embedded_message: pst_parse_item returned type %d, not an email message", item->type));
01148         } else {
01149             fprintf(f_output, "\n--%s\n", boundary);
01150             fprintf(f_output, "Content-Type: %s\n\n", attach->mimetype.str);
01151             write_normal_email(f_output, "", item, MODE_NORMAL, 0, pf, save_rtf, 1, extra_mime_headers);
01152         }
01153         pst_freeItem(item);
01154     }
01155 
01156     DEBUG_RET();
01157 }
01158 
01159 
01160 void write_inline_attachment(FILE* f_output, pst_item_attach* attach, char *boundary, pst_file* pst)
01161 {
01162     DEBUG_ENT("write_inline_attachment");
01163     DEBUG_INFO(("Attachment Size is %#"PRIx64", data = %#"PRIxPTR", id %#"PRIx64"\n", (uint64_t)attach->data.size, attach->data.data, attach->i_id));
01164 
01165     if (!attach->data.data) {
01166         // make sure we can fetch data from the id
01167         pst_index_ll *ptr = pst_getID(pst, attach->i_id);
01168         if (!ptr) {
01169             DEBUG_WARN(("Couldn't find ID pointer. Cannot save attachment to file\n"));
01170             DEBUG_RET();
01171             return;
01172         }
01173     }
01174 
01175     fprintf(f_output, "\n--%s\n", boundary);
01176     if (!attach->mimetype.str) {
01177         fprintf(f_output, "Content-Type: %s\n", MIME_TYPE_DEFAULT);
01178     } else {
01179         fprintf(f_output, "Content-Type: %s\n", attach->mimetype.str);
01180     }
01181     fprintf(f_output, "Content-Transfer-Encoding: base64\n");
01182 
01183     if (attach->filename2.str) {
01184         // use the long filename, converted to proper encoding if needed.
01185         // it is already utf8
01186         pst_rfc2231(&attach->filename2);
01187         fprintf(f_output, "Content-Disposition: attachment; \n        filename*=%s\n\n", attach->filename2.str);
01188     }
01189     else if (attach->filename1.str) {
01190         // short filename never needs encoding
01191         fprintf(f_output, "Content-Disposition: attachment; filename=\"%s\"\n\n", attach->filename1.str);
01192     }
01193     else {
01194         // no filename is inline
01195         fprintf(f_output, "Content-Disposition: inline\n\n");
01196     }
01197 
01198     (void)pst_attach_to_file_base64(pst, attach, f_output);
01199     fprintf(f_output, "\n\n");
01200     DEBUG_RET();
01201 }
01202 
01203 
01204 int  valid_headers(char *header)
01205 {
01206     // headers are sometimes really bogus - they seem to be fragments of the
01207     // message body, so we only use them if they seem to be real rfc822 headers.
01208     // this list is composed of ones that we have seen in real pst files.
01209     // there are surely others. the problem is - given an arbitrary character
01210     // string, is it a valid (or even reasonable) set of rfc822 headers?
01211     if (header) {
01212         if ((strncasecmp(header, "X-Barracuda-URL: ", 17) == 0) ||
01213             (strncasecmp(header, "X-ASG-Debug-ID: ",  16) == 0) ||
01214             (strncasecmp(header, "Return-Path: ",     13) == 0) ||
01215             (strncasecmp(header, "Received: ",        10) == 0) ||
01216             (strncasecmp(header, "Subject: ",          9) == 0) ||
01217             (strncasecmp(header, "Date: ",             6) == 0) ||
01218             (strncasecmp(header, "From: ",             6) == 0) ||
01219             (strncasecmp(header, "X-x: ",              5) == 0) ||
01220             (strncasecmp(header, "Microsoft Mail Internet Headers", 31) == 0)) {
01221             return 1;
01222         }
01223         else {
01224             if (strlen(header) > 2) {
01225                 DEBUG_INFO(("Ignore bogus headers = %s\n", header));
01226             }
01227             return 0;
01228         }
01229     }
01230     else return 0;
01231 }
01232 
01233 
01234 void header_has_field(char *header, char *field, int *flag)
01235 {
01236     DEBUG_ENT("header_has_field");
01237     if (my_stristr(header, field) || (strncasecmp(header, field+1, strlen(field)-1) == 0)) {
01238         DEBUG_INFO(("header block has %s header\n", field+1));
01239         *flag = 1;
01240     }
01241     DEBUG_RET();
01242 }
01243 
01244 
01245 void header_get_subfield(char *field, const char *subfield, char *body_subfield, size_t size_subfield)
01246 {
01247     if (!field) return;
01248     DEBUG_ENT("header_get_subfield");
01249     char search[60];
01250     snprintf(search, sizeof(search), " %s=", subfield);
01251     field++;
01252     char *n = header_end_field(field);
01253     char *s = my_stristr(field, search);
01254     if (n && s && (s < n)) {
01255         char *e, *f, save;
01256         s += strlen(search);    // skip over subfield=
01257         if (*s == '"') {
01258             s++;
01259             e = strchr(s, '"');
01260         }
01261         else {
01262             e = strchr(s, ';');
01263             f = strchr(s, '\n');
01264             if (e && f && (f < e)) e = f;
01265         }
01266         if (!e || (e > n)) e = n;   // use the trailing lf as terminator if nothing better
01267         save = *e;
01268         *e = '\0';
01269             snprintf(body_subfield, size_subfield, "%s", s);  // copy the subfield to our buffer
01270         *e = save;
01271         DEBUG_INFO(("body %s %s from headers\n", subfield, body_subfield));
01272     }
01273     DEBUG_RET();
01274 }
01275 
01276 char* header_get_field(char *header, char *field)
01277 {
01278     char *t = my_stristr(header, field);
01279     if (!t && (strncasecmp(header, field+1, strlen(field)-1) == 0)) t = header;
01280     return t;
01281 }
01282 
01283 
01284 // return pointer to \n at the end of this header field,
01285 // or NULL if this field goes to the end of the string.
01286 char *header_end_field(char *field)
01287 {
01288     char *e = strchr(field+1, '\n');
01289     while (e && ((e[1] == ' ') || (e[1] == '\t'))) {
01290         e = strchr(e+1, '\n');
01291     }
01292     return e;
01293 }
01294 
01295 
01296 void header_strip_field(char *header, char *field)
01297 {
01298     char *t = header_get_field(header, field);
01299     if (t) {
01300         char *e = header_end_field(t);
01301         if (e) {
01302             if (t == header) e++;   // if *t is not \n, we don't want to keep the \n at *e either.
01303             while (*e != '\0') {
01304                 *t = *e;
01305                 t++;
01306                 e++;
01307             }
01308             *t = '\0';
01309         }
01310         else {
01311             // this was the last header field, truncate the headers
01312             *t = '\0';
01313         }
01314     }
01315 }
01316 
01317 
01318 int  test_base64(char *body)
01319 {
01320     int b64 = 0;
01321     uint8_t *b = (uint8_t *)body;
01322     DEBUG_ENT("test_base64");
01323     while (*b) {
01324         if ((*b < 32) && (*b != 9) && (*b != 10)) {
01325             DEBUG_INFO(("found base64 byte %d\n", (int)*b));
01326             DEBUG_HEXDUMPC(body, strlen(body), 0x10);
01327             b64 = 1;
01328             break;
01329         }
01330         b++;
01331     }
01332     DEBUG_RET();
01333     return b64;
01334 }
01335 
01336 
01337 void find_html_charset(char *html, char *charset, size_t charsetlen)
01338 {
01339     const int  index = 1;
01340     const int nmatch = index+1;
01341     regmatch_t match[nmatch];
01342     DEBUG_ENT("find_html_charset");
01343     int rc = regexec(&meta_charset_pattern, html, nmatch, match, 0);
01344     if (rc == 0) {
01345         int s = match[index].rm_so;
01346         int e = match[index].rm_eo;
01347         if (s != -1) {
01348             char save = html[e];
01349             html[e] = '\0';
01350                 snprintf(charset, charsetlen, "%s", html+s);    // copy the html charset
01351             html[e] = save;
01352             DEBUG_INFO(("charset %s from html text\n", charset));
01353         }
01354         else {
01355             DEBUG_INFO(("matching %d %d %d %d\n", match[0].rm_so, match[0].rm_eo, match[1].rm_so, match[1].rm_eo));
01356             DEBUG_HEXDUMPC(html, strlen(html), 0x10);
01357         }
01358     }
01359     else {
01360         DEBUG_INFO(("regexec returns %d\n", rc));
01361     }
01362     DEBUG_RET();
01363 }
01364 
01365 
01366 void find_rfc822_headers(char** extra_mime_headers)
01367 {
01368     DEBUG_ENT("find_rfc822_headers");
01369     char *headers = *extra_mime_headers;
01370     if (headers) {
01371         char *temp, *t;
01372         while ((temp = strstr(headers, "\n\n"))) {
01373             temp[1] = '\0';
01374             t = header_get_field(headers, "\nContent-Type:");
01375             if (t) {
01376                 t++;
01377                 DEBUG_INFO(("found content type header\n"));
01378                 char *n = strchr(t, '\n');
01379                 char *s = strstr(t, ": ");
01380                 char *e = strchr(t, ';');
01381                 if (!e || (e > n)) e = n;
01382                 if (s && (s < e)) {
01383                     s += 2;
01384                     if (!strncasecmp(s, RFC822, e-s)) {
01385                         headers = temp+2;   // found rfc822 header
01386                         DEBUG_INFO(("found 822 headers\n%s\n", headers));
01387                         break;
01388                     }
01389                 }
01390             }
01391             //DEBUG_INFO(("skipping to next block after\n%s\n", headers));
01392             headers = temp+2;   // skip to next chunk of headers
01393         }
01394         *extra_mime_headers = headers;
01395     }
01396     DEBUG_RET();
01397 }
01398 
01399 
01400 void write_body_part(FILE* f_output, pst_string *body, char *mime, char *charset, char *boundary, pst_file* pst)
01401 {
01402     DEBUG_ENT("write_body_part");
01403     if (body->is_utf8 && (strcasecmp("utf-8", charset))) {
01404         // try to convert to the specified charset since the target
01405         // is not utf-8, and the data came from a unicode (utf16) field
01406         // and is now in utf-8.
01407         size_t rc;
01408         DEBUG_INFO(("Convert %s utf-8 to %s\n", mime, charset));
01409         pst_vbuf *newer = pst_vballoc(2);
01410         rc = pst_vb_utf8to8bit(newer, body->str, strlen(body->str), charset);
01411         if (rc == (size_t)-1) {
01412             // unable to convert, change the charset to utf8
01413             free(newer->b);
01414             DEBUG_INFO(("Failed to convert %s utf-8 to %s\n", mime, charset));
01415             charset = "utf-8";
01416         }
01417         else {
01418             // null terminate the output string
01419             pst_vbgrow(newer, 1);
01420             newer->b[newer->dlen] = '\0';
01421             free(body->str);
01422             body->str = newer->b;
01423         }
01424         free(newer);
01425     }
01426     removeCR(body->str);
01427     int base64 = test_base64(body->str);
01428     fprintf(f_output, "\n--%s\n", boundary);
01429     fprintf(f_output, "Content-Type: %s; charset=\"%s\"\n", mime, charset);
01430     if (base64) fprintf(f_output, "Content-Transfer-Encoding: base64\n");
01431     fprintf(f_output, "\n");
01432     if (base64) {
01433         char *enc = pst_base64_encode(body->str, strlen(body->str));
01434         if (enc) {
01435             write_email_body(f_output, enc);
01436             fprintf(f_output, "\n");
01437             free(enc);
01438         }
01439     }
01440     else {
01441         write_email_body(f_output, body->str);
01442     }
01443     DEBUG_RET();
01444 }
01445 
01446 
01447 void write_schedule_part_data(FILE* f_output, pst_item* item, const char* sender, const char* method)
01448 {
01449     fprintf(f_output, "BEGIN:VCALENDAR\n");
01450     fprintf(f_output, "VERSION:2.0\n");
01451     fprintf(f_output, "PRODID:LibPST v%s\n", VERSION);
01452     if (method) fprintf(f_output, "METHOD:%s\n", method);
01453     fprintf(f_output, "BEGIN:VEVENT\n");
01454     if (sender) {
01455         if (item->email->outlook_sender_name.str) {
01456             fprintf(f_output, "ORGANIZER;CN=\"%s\":MAILTO:%s\n", item->email->outlook_sender_name.str, sender);
01457         } else {
01458             fprintf(f_output, "ORGANIZER;CN=\"\":MAILTO:%s\n", sender);
01459         }
01460     }
01461     write_appointment(f_output, item);
01462     fprintf(f_output, "END:VCALENDAR\n");
01463 }
01464 
01465 
01466 void write_schedule_part(FILE* f_output, pst_item* item, const char* sender, const char* boundary)
01467 {
01468     const char* method  = "REQUEST";
01469     const char* charset = "utf-8";
01470     char fname[30];
01471     if (!item->appointment) return;
01472 
01473     // inline appointment request
01474     fprintf(f_output, "\n--%s\n", boundary);
01475     fprintf(f_output, "Content-Type: %s; method=\"%s\"; charset=\"%s\"\n\n", "text/calendar", method, charset);
01476     write_schedule_part_data(f_output, item, sender, method);
01477     fprintf(f_output, "\n");
01478 
01479     // attachment appointment request
01480     snprintf(fname, sizeof(fname), "i%i.ics", rand());
01481     fprintf(f_output, "\n--%s\n", boundary);
01482     fprintf(f_output, "Content-Type: %s; charset=\"%s\"; name=\"%s\"\n", "text/calendar", "utf-8", fname);
01483     fprintf(f_output, "Content-Disposition: attachment; filename=\"%s\"\n\n", fname);
01484     write_schedule_part_data(f_output, item, sender, method);
01485     fprintf(f_output, "\n");
01486 }
01487 
01488 
01489 void write_normal_email(FILE* f_output, char f_name[], pst_item* item, int mode, int mode_MH, pst_file* pst, int save_rtf, int embedding, char** extra_mime_headers)
01490 {
01491     char boundary[60];
01492     char altboundary[66];
01493     char *altboundaryp = NULL;
01494     char body_charset[30];
01495     char buffer_charset[30];
01496     char body_report[60];
01497     char sender[60];
01498     int  sender_known = 0;
01499     char *temp = NULL;
01500     time_t em_time;
01501     char *c_time;
01502     char *headers = NULL;
01503     int has_from, has_subject, has_to, has_cc, has_date, has_msgid;
01504     has_from = has_subject = has_to = has_cc = has_date = has_msgid = 0;
01505     DEBUG_ENT("write_normal_email");
01506 
01507     pst_convert_utf8_null(item, &item->email->header);
01508     headers = valid_headers(item->email->header.str) ? item->email->header.str :
01509               valid_headers(*extra_mime_headers)     ? *extra_mime_headers     :
01510               NULL;
01511 
01512     // setup default body character set and report type
01513     strncpy(body_charset, pst_default_charset(item, sizeof(buffer_charset), buffer_charset), sizeof(body_charset));
01514     body_charset[sizeof(body_charset)-1] = '\0';
01515     strncpy(body_report, "delivery-status", sizeof(body_report));
01516     body_report[sizeof(body_report)-1] = '\0';
01517 
01518     // setup default sender
01519     pst_convert_utf8(item, &item->email->sender_address);
01520     if (item->email->sender_address.str && strchr(item->email->sender_address.str, '@')) {
01521         temp = item->email->sender_address.str;
01522         sender_known = 1;
01523     }
01524     else {
01525         temp = "MAILER-DAEMON";
01526     }
01527     strncpy(sender, temp, sizeof(sender));
01528     sender[sizeof(sender)-1] = '\0';
01529 
01530     // convert the sent date if it exists, or set it to a fixed date
01531     if (item->email->sent_date) {
01532         em_time = pst_fileTimeToUnixTime(item->email->sent_date);
01533         c_time = ctime(&em_time);
01534         if (c_time)
01535             c_time[strlen(c_time)-1] = '\0'; //remove end \n
01536         else
01537             c_time = "Fri Dec 28 12:06:21 2001";
01538     } else
01539         c_time = "Fri Dec 28 12:06:21 2001";
01540 
01541     // create our MIME boundaries here.
01542     snprintf(boundary, sizeof(boundary), "--boundary-LibPST-iamunique-%i_-_-", rand());
01543     snprintf(altboundary, sizeof(altboundary), "alt-%s", boundary);
01544 
01545     // we will always look at the headers to discover some stuff
01546     if (headers ) {
01547         char *t;
01548         removeCR(headers);
01549 
01550         temp = strstr(headers, "\n\n");
01551         if (temp) {
01552             // cut off our real rfc822 headers here
01553             temp[1] = '\0';
01554             // pointer to all the embedded MIME headers.
01555             // we use these to find the actual rfc822 headers for embedded message/rfc822 mime parts
01556             // but only for the outermost message
01557             if (!*extra_mime_headers) *extra_mime_headers = temp+2;
01558             DEBUG_INFO(("Found extra mime headers\n%s\n", temp+2));
01559         }
01560 
01561         // Check if the headers have all the necessary fields
01562         header_has_field(headers, "\nFrom:",        &has_from);
01563         header_has_field(headers, "\nTo:",          &has_to);
01564         header_has_field(headers, "\nSubject:",     &has_subject);
01565         header_has_field(headers, "\nDate:",        &has_date);
01566         header_has_field(headers, "\nCC:",          &has_cc);
01567         header_has_field(headers, "\nMessage-Id:",  &has_msgid);
01568 
01569         // look for charset and report-type in Content-Type header
01570         t = header_get_field(headers, "\nContent-Type:");
01571         header_get_subfield(t, "charset", body_charset, sizeof(body_charset));
01572         header_get_subfield(t, "report-type", body_report, sizeof(body_report));
01573 
01574         // derive a proper sender email address
01575         if (!sender_known) {
01576             t = header_get_field(headers, "\nFrom:");
01577             if (t) {
01578                 // assume address is on the first line, rather than on a continuation line
01579                 t++;
01580                 char *n = strchr(t, '\n');
01581                 char *s = strchr(t, '<');
01582                 char *e = strchr(t, '>');
01583                 if (s && e && n && (s < e) && (e < n)) {
01584                 char save = *e;
01585                 *e = '\0';
01586                     snprintf(sender, sizeof(sender), "%s", s+1);
01587                 *e = save;
01588                 }
01589             }
01590         }
01591 
01592         // Strip out the mime headers and some others that we don't want to emit
01593         header_strip_field(headers, "\nMicrosoft Mail Internet Headers");
01594         header_strip_field(headers, "\nMIME-Version:");
01595         header_strip_field(headers, "\nContent-Type:");
01596         header_strip_field(headers, "\nContent-Transfer-Encoding:");
01597         header_strip_field(headers, "\nContent-class:");
01598         header_strip_field(headers, "\nX-MimeOLE:");
01599         header_strip_field(headers, "\nX-From_:");
01600     }
01601 
01602     DEBUG_INFO(("About to print Header\n"));
01603 
01604     if (item && item->subject.str) {
01605         pst_convert_utf8(item, &item->subject);
01606         DEBUG_INFO(("item->subject = %s\n", item->subject.str));
01607     }
01608 
01609     if (mode != MODE_SEPARATE) {
01610         // most modes need this separator line.
01611         // procmail produces this separator without the quotes around the
01612         // sender email address, but apparently some Mac email client needs
01613         // those quotes, and they don't seem to cause problems for anyone else.
01614         char *quo = (embedding) ? ">" : "";
01615         fprintf(f_output, "%sFrom \"%s\" %s\n", quo, sender, c_time);
01616     }
01617 
01618     // print the supplied email headers
01619     if (headers) {
01620         int len = strlen(headers);
01621         if (len > 0) {
01622             fprintf(f_output, "%s", headers);
01623             // make sure the headers end with a \n
01624             if (headers[len-1] != '\n') fprintf(f_output, "\n");
01625             //char *h = headers;
01626             //while (*h) {
01627             //    char *e = strchr(h, '\n');
01628             //    int   d = 1;    // normally e points to trailing \n
01629             //    if (!e) {
01630             //        e = h + strlen(h);  // e points to trailing null
01631             //        d = 0;
01632             //    }
01633             //    // we could do rfc2047 encoding here if needed
01634             //    fprintf(f_output, "%.*s\n", (int)(e-h), h);
01635             //    h = e + d;
01636             //}
01637         }
01638     }
01639 
01640     // record read status
01641     if ((item->flags & PST_FLAG_READ) == PST_FLAG_READ) {
01642         fprintf(f_output, "Status: RO\n");
01643     }
01644 
01645     // create required header fields that are not already written
01646 
01647     if (!has_from) {
01648         if (item->email->outlook_sender_name.str){
01649             pst_rfc2047(item, &item->email->outlook_sender_name, 1);
01650             fprintf(f_output, "From: %s <%s>\n", item->email->outlook_sender_name.str, sender);
01651         } else {
01652             fprintf(f_output, "From: <%s>\n", sender);
01653         }
01654     }
01655 
01656     if (!has_subject) {
01657         if (item->subject.str) {
01658             pst_rfc2047(item, &item->subject, 0);
01659             fprintf(f_output, "Subject: %s\n", item->subject.str);
01660         } else {
01661             fprintf(f_output, "Subject: \n");
01662         }
01663     }
01664 
01665     if (!has_to && item->email->sentto_address.str) {
01666         pst_rfc2047(item, &item->email->sentto_address, 0);
01667         fprintf(f_output, "To: %s\n", item->email->sentto_address.str);
01668     }
01669 
01670     if (!has_cc && item->email->cc_address.str) {
01671         pst_rfc2047(item, &item->email->cc_address, 0);
01672         fprintf(f_output, "Cc: %s\n", item->email->cc_address.str);
01673     }
01674 
01675     if (!has_date && item->email->sent_date) {
01676         char c_time[C_TIME_SIZE];
01677         struct tm stm;
01678         gmtime_r(&em_time, &stm);
01679         strftime(c_time, C_TIME_SIZE, "%a, %d %b %Y %H:%M:%S %z", &stm);
01680         fprintf(f_output, "Date: %s\n", c_time);
01681     }
01682 
01683     if (!has_msgid && item->email->messageid.str) {
01684         pst_convert_utf8(item, &item->email->messageid);
01685         fprintf(f_output, "Message-Id: %s\n", item->email->messageid.str);
01686     }
01687 
01688     // add forensic headers to capture some .pst stuff that is not really
01689     // needed or used by mail clients
01690     pst_convert_utf8_null(item, &item->email->sender_address);
01691     if (item->email->sender_address.str && !strchr(item->email->sender_address.str, '@')
01692                                         && strcmp(item->email->sender_address.str, ".")
01693                                         && (strlen(item->email->sender_address.str) > 0)) {
01694         fprintf(f_output, "X-libpst-forensic-sender: %s\n", item->email->sender_address.str);
01695     }
01696 
01697     if (item->email->bcc_address.str) {
01698         pst_convert_utf8(item, &item->email->bcc_address);
01699         fprintf(f_output, "X-libpst-forensic-bcc: %s\n", item->email->bcc_address.str);
01700     }
01701 
01702     // add our own mime headers
01703     fprintf(f_output, "MIME-Version: 1.0\n");
01704     if (item->type == PST_TYPE_REPORT) {
01705         // multipart/report for DSN/MDN reports
01706         fprintf(f_output, "Content-Type: multipart/report; report-type=%s;\n\tboundary=\"%s\"\n", body_report, boundary);
01707     }
01708     else {
01709         fprintf(f_output, "Content-Type: multipart/mixed;\n\tboundary=\"%s\"\n", boundary);
01710     }
01711     fprintf(f_output, "\n");    // end of headers, start of body
01712 
01713     // now dump the body parts
01714     if ((item->type == PST_TYPE_REPORT) && (item->email->report_text.str)) {
01715         write_body_part(f_output, &item->email->report_text, "text/plain", body_charset, boundary, pst);
01716         fprintf(f_output, "\n");
01717     }
01718 
01719     if (item->body.str && item->email->htmlbody.str) {
01720         // start the nested alternative part
01721         fprintf(f_output, "\n--%s\n", boundary);
01722         fprintf(f_output, "Content-Type: multipart/alternative;\n\tboundary=\"%s\"\n", altboundary);
01723         altboundaryp = altboundary;
01724     }
01725     else {
01726         altboundaryp = boundary;
01727     }
01728 
01729     if (item->body.str) {
01730         write_body_part(f_output, &item->body, "text/plain", body_charset, altboundaryp, pst);
01731     }
01732 
01733     if (item->email->htmlbody.str) {
01734         find_html_charset(item->email->htmlbody.str, body_charset, sizeof(body_charset));
01735         write_body_part(f_output, &item->email->htmlbody, "text/html", body_charset, altboundaryp, pst);
01736     }
01737 
01738     if (item->body.str && item->email->htmlbody.str) {
01739         // end the nested alternative part
01740         fprintf(f_output, "\n--%s--\n", altboundary);
01741     }
01742 
01743     if (item->email->rtf_compressed.data && save_rtf) {
01744         pst_item_attach* attach = (pst_item_attach*)pst_malloc(sizeof(pst_item_attach));
01745         DEBUG_INFO(("Adding RTF body as attachment\n"));
01746         memset(attach, 0, sizeof(pst_item_attach));
01747         attach->next = item->attach;
01748         item->attach = attach;
01749         attach->data.data         = pst_lzfu_decompress(item->email->rtf_compressed.data, item->email->rtf_compressed.size, &attach->data.size);
01750         attach->filename2.str     = strdup(RTF_ATTACH_NAME);
01751         attach->filename2.is_utf8 = 1;
01752         attach->mimetype.str      = strdup(RTF_ATTACH_TYPE);
01753         attach->mimetype.is_utf8  = 1;
01754     }
01755 
01756     if (item->email->encrypted_body.data) {
01757         pst_item_attach* attach = (pst_item_attach*)pst_malloc(sizeof(pst_item_attach));
01758         DEBUG_INFO(("Adding encrypted text body as attachment\n"));
01759         attach = (pst_item_attach*) pst_malloc(sizeof(pst_item_attach));
01760         memset(attach, 0, sizeof(pst_item_attach));
01761         attach->next = item->attach;
01762         item->attach = attach;
01763         attach->data.data = item->email->encrypted_body.data;
01764         attach->data.size = item->email->encrypted_body.size;
01765         item->email->encrypted_body.data = NULL;
01766     }
01767 
01768     if (item->email->encrypted_htmlbody.data) {
01769         pst_item_attach* attach = (pst_item_attach*)pst_malloc(sizeof(pst_item_attach));
01770         DEBUG_INFO(("Adding encrypted HTML body as attachment\n"));
01771         attach = (pst_item_attach*) pst_malloc(sizeof(pst_item_attach));
01772         memset(attach, 0, sizeof(pst_item_attach));
01773         attach->next = item->attach;
01774         item->attach = attach;
01775         attach->data.data = item->email->encrypted_htmlbody.data;
01776         attach->data.size = item->email->encrypted_htmlbody.size;
01777         item->email->encrypted_htmlbody.data = NULL;
01778     }
01779 
01780     if (item->type == PST_TYPE_SCHEDULE) {
01781         write_schedule_part(f_output, item, sender, boundary);
01782     }
01783 
01784     // other attachments
01785     {
01786         pst_item_attach* attach;
01787         int attach_num = 0;
01788         for (attach = item->attach; attach; attach = attach->next) {
01789             pst_convert_utf8_null(item, &attach->filename1);
01790             pst_convert_utf8_null(item, &attach->filename2);
01791             pst_convert_utf8_null(item, &attach->mimetype);
01792             DEBUG_INFO(("Attempting Attachment encoding\n"));
01793             if (attach->method == PST_ATTACH_EMBEDDED) {
01794                 DEBUG_INFO(("have an embedded rfc822 message attachment\n"));
01795                 if (attach->mimetype.str) {
01796                     DEBUG_INFO(("which already has a mime-type of %s\n", attach->mimetype.str));
01797                     free(attach->mimetype.str);
01798                 }
01799                 attach->mimetype.str = strdup(RFC822);
01800                 attach->mimetype.is_utf8 = 1;
01801                 find_rfc822_headers(extra_mime_headers);
01802                 write_embedded_message(f_output, attach, boundary, pst, save_rtf, extra_mime_headers);
01803             }
01804             else if (attach->data.data || attach->i_id) {
01805                 if (acceptable_ext(attach)) {
01806                     if (mode == MODE_SEPARATE && !mode_MH)
01807                         write_separate_attachment(f_name, attach, ++attach_num, pst);
01808                     else
01809                         write_inline_attachment(f_output, attach, boundary, pst);
01810                 }
01811             }
01812         }
01813     }
01814 
01815     fprintf(f_output, "\n--%s--\n\n", boundary);
01816     DEBUG_RET();
01817 }
01818 
01819 
01820 void write_vcard(FILE* f_output, pst_item* item, pst_item_contact* contact, char comment[])
01821 {
01822     char*  result = NULL;
01823     size_t resultlen = 0;
01824     char   time_buffer[30];
01825     // We can only call rfc escape once per printf, since the second call
01826     // may free the buffer returned by the first call.
01827     // I had tried to place those into a single printf - Carl.
01828 
01829     DEBUG_ENT("write_vcard");
01830 
01831     // make everything utf8
01832     pst_convert_utf8_null(item, &contact->fullname);
01833     pst_convert_utf8_null(item, &contact->surname);
01834     pst_convert_utf8_null(item, &contact->first_name);
01835     pst_convert_utf8_null(item, &contact->middle_name);
01836     pst_convert_utf8_null(item, &contact->display_name_prefix);
01837     pst_convert_utf8_null(item, &contact->suffix);
01838     pst_convert_utf8_null(item, &contact->nickname);
01839     pst_convert_utf8_null(item, &contact->address1);
01840     pst_convert_utf8_null(item, &contact->address2);
01841     pst_convert_utf8_null(item, &contact->address3);
01842     pst_convert_utf8_null(item, &contact->home_po_box);
01843     pst_convert_utf8_null(item, &contact->home_street);
01844     pst_convert_utf8_null(item, &contact->home_city);
01845     pst_convert_utf8_null(item, &contact->home_state);
01846     pst_convert_utf8_null(item, &contact->home_postal_code);
01847     pst_convert_utf8_null(item, &contact->home_country);
01848     pst_convert_utf8_null(item, &contact->home_address);
01849     pst_convert_utf8_null(item, &contact->business_po_box);
01850     pst_convert_utf8_null(item, &contact->business_street);
01851     pst_convert_utf8_null(item, &contact->business_city);
01852     pst_convert_utf8_null(item, &contact->business_state);
01853     pst_convert_utf8_null(item, &contact->business_postal_code);
01854     pst_convert_utf8_null(item, &contact->business_country);
01855     pst_convert_utf8_null(item, &contact->business_address);
01856     pst_convert_utf8_null(item, &contact->other_po_box);
01857     pst_convert_utf8_null(item, &contact->other_street);
01858     pst_convert_utf8_null(item, &contact->other_city);
01859     pst_convert_utf8_null(item, &contact->other_state);
01860     pst_convert_utf8_null(item, &contact->other_postal_code);
01861     pst_convert_utf8_null(item, &contact->other_country);
01862     pst_convert_utf8_null(item, &contact->other_address);
01863     pst_convert_utf8_null(item, &contact->business_fax);
01864     pst_convert_utf8_null(item, &contact->business_phone);
01865     pst_convert_utf8_null(item, &contact->business_phone2);
01866     pst_convert_utf8_null(item, &contact->car_phone);
01867     pst_convert_utf8_null(item, &contact->home_fax);
01868     pst_convert_utf8_null(item, &contact->home_phone);
01869     pst_convert_utf8_null(item, &contact->home_phone2);
01870     pst_convert_utf8_null(item, &contact->isdn_phone);
01871     pst_convert_utf8_null(item, &contact->mobile_phone);
01872     pst_convert_utf8_null(item, &contact->other_phone);
01873     pst_convert_utf8_null(item, &contact->pager_phone);
01874     pst_convert_utf8_null(item, &contact->primary_fax);
01875     pst_convert_utf8_null(item, &contact->primary_phone);
01876     pst_convert_utf8_null(item, &contact->radio_phone);
01877     pst_convert_utf8_null(item, &contact->telex);
01878     pst_convert_utf8_null(item, &contact->job_title);
01879     pst_convert_utf8_null(item, &contact->profession);
01880     pst_convert_utf8_null(item, &contact->assistant_name);
01881     pst_convert_utf8_null(item, &contact->assistant_phone);
01882     pst_convert_utf8_null(item, &contact->company_name);
01883     pst_convert_utf8_null(item, &item->body);
01884 
01885     // the specification I am following is (hopefully) RFC2426 vCard Mime Directory Profile
01886     fprintf(f_output, "BEGIN:VCARD\n");
01887     fprintf(f_output, "FN:%s\n", pst_rfc2426_escape(contact->fullname.str, &result, &resultlen));
01888 
01889     //fprintf(f_output, "N:%s;%s;%s;%s;%s\n",
01890     fprintf(f_output, "N:%s;", (!contact->surname.str)             ? "" : pst_rfc2426_escape(contact->surname.str, &result, &resultlen));
01891     fprintf(f_output, "%s;",   (!contact->first_name.str)          ? "" : pst_rfc2426_escape(contact->first_name.str, &result, &resultlen));
01892     fprintf(f_output, "%s;",   (!contact->middle_name.str)         ? "" : pst_rfc2426_escape(contact->middle_name.str, &result, &resultlen));
01893     fprintf(f_output, "%s;",   (!contact->display_name_prefix.str) ? "" : pst_rfc2426_escape(contact->display_name_prefix.str, &result, &resultlen));
01894     fprintf(f_output, "%s\n",  (!contact->suffix.str)              ? "" : pst_rfc2426_escape(contact->suffix.str, &result, &resultlen));
01895 
01896     if (contact->nickname.str)
01897         fprintf(f_output, "NICKNAME:%s\n", pst_rfc2426_escape(contact->nickname.str, &result, &resultlen));
01898     if (contact->address1.str)
01899         fprintf(f_output, "EMAIL:%s\n", pst_rfc2426_escape(contact->address1.str, &result, &resultlen));
01900     if (contact->address2.str)
01901         fprintf(f_output, "EMAIL:%s\n", pst_rfc2426_escape(contact->address2.str, &result, &resultlen));
01902     if (contact->address3.str)
01903         fprintf(f_output, "EMAIL:%s\n", pst_rfc2426_escape(contact->address3.str, &result, &resultlen));
01904     if (contact->birthday)
01905         fprintf(f_output, "BDAY:%s\n", pst_rfc2425_datetime_format(contact->birthday, sizeof(time_buffer), time_buffer));
01906 
01907     if (contact->home_address.str) {
01908         //fprintf(f_output, "ADR;TYPE=home:%s;%s;%s;%s;%s;%s;%s\n",
01909         fprintf(f_output, "ADR;TYPE=home:%s;",  (!contact->home_po_box.str)      ? "" : pst_rfc2426_escape(contact->home_po_box.str, &result, &resultlen));
01910         fprintf(f_output, "%s;",                ""); // extended Address
01911         fprintf(f_output, "%s;",                (!contact->home_street.str)      ? "" : pst_rfc2426_escape(contact->home_street.str, &result, &resultlen));
01912         fprintf(f_output, "%s;",                (!contact->home_city.str)        ? "" : pst_rfc2426_escape(contact->home_city.str, &result, &resultlen));
01913         fprintf(f_output, "%s;",                (!contact->home_state.str)       ? "" : pst_rfc2426_escape(contact->home_state.str, &result, &resultlen));
01914         fprintf(f_output, "%s;",                (!contact->home_postal_code.str) ? "" : pst_rfc2426_escape(contact->home_postal_code.str, &result, &resultlen));
01915         fprintf(f_output, "%s\n",               (!contact->home_country.str)     ? "" : pst_rfc2426_escape(contact->home_country.str, &result, &resultlen));
01916         fprintf(f_output, "LABEL;TYPE=home:%s\n", pst_rfc2426_escape(contact->home_address.str, &result, &resultlen));
01917     }
01918 
01919     if (contact->business_address.str) {
01920         //fprintf(f_output, "ADR;TYPE=work:%s;%s;%s;%s;%s;%s;%s\n",
01921         fprintf(f_output, "ADR;TYPE=work:%s;",  (!contact->business_po_box.str)      ? "" : pst_rfc2426_escape(contact->business_po_box.str, &result, &resultlen));
01922         fprintf(f_output, "%s;",                ""); // extended Address
01923         fprintf(f_output, "%s;",                (!contact->business_street.str)      ? "" : pst_rfc2426_escape(contact->business_street.str, &result, &resultlen));
01924         fprintf(f_output, "%s;",                (!contact->business_city.str)        ? "" : pst_rfc2426_escape(contact->business_city.str, &result, &resultlen));
01925         fprintf(f_output, "%s;",                (!contact->business_state.str)       ? "" : pst_rfc2426_escape(contact->business_state.str, &result, &resultlen));
01926         fprintf(f_output, "%s;",                (!contact->business_postal_code.str) ? "" : pst_rfc2426_escape(contact->business_postal_code.str, &result, &resultlen));
01927         fprintf(f_output, "%s\n",               (!contact->business_country.str)     ? "" : pst_rfc2426_escape(contact->business_country.str, &result, &resultlen));
01928         fprintf(f_output, "LABEL;TYPE=work:%s\n", pst_rfc2426_escape(contact->business_address.str, &result, &resultlen));
01929     }
01930 
01931     if (contact->other_address.str) {
01932         //fprintf(f_output, "ADR;TYPE=postal:%s;%s;%s;%s;%s;%s;%s\n",
01933         fprintf(f_output, "ADR;TYPE=postal:%s;",(!contact->other_po_box.str)       ? "" : pst_rfc2426_escape(contact->other_po_box.str, &result, &resultlen));
01934         fprintf(f_output, "%s;",                ""); // extended Address
01935         fprintf(f_output, "%s;",                (!contact->other_street.str)       ? "" : pst_rfc2426_escape(contact->other_street.str, &result, &resultlen));
01936         fprintf(f_output, "%s;",                (!contact->other_city.str)         ? "" : pst_rfc2426_escape(contact->other_city.str, &result, &resultlen));
01937         fprintf(f_output, "%s;",                (!contact->other_state.str)        ? "" : pst_rfc2426_escape(contact->other_state.str, &result, &resultlen));
01938         fprintf(f_output, "%s;",                (!contact->other_postal_code.str)  ? "" : pst_rfc2426_escape(contact->other_postal_code.str, &result, &resultlen));
01939         fprintf(f_output, "%s\n",               (!contact->other_country.str)      ? "" : pst_rfc2426_escape(contact->other_country.str, &result, &resultlen));
01940         fprintf(f_output, "LABEL;TYPE=postal:%s\n", pst_rfc2426_escape(contact->other_address.str, &result, &resultlen));
01941     }
01942 
01943     if (contact->business_fax.str)      fprintf(f_output, "TEL;TYPE=work,fax:%s\n",         pst_rfc2426_escape(contact->business_fax.str, &result, &resultlen));
01944     if (contact->business_phone.str)    fprintf(f_output, "TEL;TYPE=work,voice:%s\n",       pst_rfc2426_escape(contact->business_phone.str, &result, &resultlen));
01945     if (contact->business_phone2.str)   fprintf(f_output, "TEL;TYPE=work,voice:%s\n",       pst_rfc2426_escape(contact->business_phone2.str, &result, &resultlen));
01946     if (contact->car_phone.str)         fprintf(f_output, "TEL;TYPE=car,voice:%s\n",        pst_rfc2426_escape(contact->car_phone.str, &result, &resultlen));
01947     if (contact->home_fax.str)          fprintf(f_output, "TEL;TYPE=home,fax:%s\n",         pst_rfc2426_escape(contact->home_fax.str, &result, &resultlen));
01948     if (contact->home_phone.str)        fprintf(f_output, "TEL;TYPE=home,voice:%s\n",       pst_rfc2426_escape(contact->home_phone.str, &result, &resultlen));
01949     if (contact->home_phone2.str)       fprintf(f_output, "TEL;TYPE=home,voice:%s\n",       pst_rfc2426_escape(contact->home_phone2.str, &result, &resultlen));
01950     if (contact->isdn_phone.str)        fprintf(f_output, "TEL;TYPE=isdn:%s\n",             pst_rfc2426_escape(contact->isdn_phone.str, &result, &resultlen));
01951     if (contact->mobile_phone.str)      fprintf(f_output, "TEL;TYPE=cell,voice:%s\n",       pst_rfc2426_escape(contact->mobile_phone.str, &result, &resultlen));
01952     if (contact->other_phone.str)       fprintf(f_output, "TEL;TYPE=msg:%s\n",              pst_rfc2426_escape(contact->other_phone.str, &result, &resultlen));
01953     if (contact->pager_phone.str)       fprintf(f_output, "TEL;TYPE=pager:%s\n",            pst_rfc2426_escape(contact->pager_phone.str, &result, &resultlen));
01954     if (contact->primary_fax.str)       fprintf(f_output, "TEL;TYPE=fax,pref:%s\n",         pst_rfc2426_escape(contact->primary_fax.str, &result, &resultlen));
01955     if (contact->primary_phone.str)     fprintf(f_output, "TEL;TYPE=phone,pref:%s\n",       pst_rfc2426_escape(contact->primary_phone.str, &result, &resultlen));
01956     if (contact->radio_phone.str)       fprintf(f_output, "TEL;TYPE=pcs:%s\n",              pst_rfc2426_escape(contact->radio_phone.str, &result, &resultlen));
01957     if (contact->telex.str)             fprintf(f_output, "TEL;TYPE=bbs:%s\n",              pst_rfc2426_escape(contact->telex.str, &result, &resultlen));
01958     if (contact->job_title.str)         fprintf(f_output, "TITLE:%s\n",                     pst_rfc2426_escape(contact->job_title.str, &result, &resultlen));
01959     if (contact->profession.str)        fprintf(f_output, "ROLE:%s\n",                      pst_rfc2426_escape(contact->profession.str, &result, &resultlen));
01960     if (contact->assistant_name.str || contact->assistant_phone.str) {
01961         fprintf(f_output, "AGENT:BEGIN:VCARD\n");
01962         if (contact->assistant_name.str)    fprintf(f_output, "FN:%s\n",                    pst_rfc2426_escape(contact->assistant_name.str, &result, &resultlen));
01963         if (contact->assistant_phone.str)   fprintf(f_output, "TEL:%s\n",                   pst_rfc2426_escape(contact->assistant_phone.str, &result, &resultlen));
01964     }
01965     if (contact->company_name.str)      fprintf(f_output, "ORG:%s\n",                       pst_rfc2426_escape(contact->company_name.str, &result, &resultlen));
01966     if (comment)                        fprintf(f_output, "NOTE:%s\n",                      pst_rfc2426_escape(comment, &result, &resultlen));
01967     if (item->body.str)                 fprintf(f_output, "NOTE:%s\n",                      pst_rfc2426_escape(item->body.str, &result, &resultlen));
01968 
01969     write_extra_categories(f_output, item);
01970 
01971     fprintf(f_output, "VERSION: 3.0\n");
01972     fprintf(f_output, "END:VCARD\n\n");
01973     if (result) free(result);
01974     DEBUG_RET();
01975 }
01976 
01977 
01985 int write_extra_categories(FILE* f_output, pst_item* item)
01986 {
01987     char*  result = NULL;
01988     size_t resultlen = 0;
01989     pst_item_extra_field *ef = item->extra_fields;
01990     const char *fmt = "CATEGORIES:%s";
01991     int category_started = 0;
01992     while (ef) {
01993         if (strcmp(ef->field_name, "Keywords") == 0) {
01994             fprintf(f_output, fmt, pst_rfc2426_escape(ef->value, &result, &resultlen));
01995             fmt = ", %s";
01996             category_started = 1;
01997         }
01998         ef = ef->next;
01999     }
02000     if (category_started) fprintf(f_output, "\n");
02001     if (result) free(result);
02002     return category_started;
02003 }
02004 
02005 
02006 void write_journal(FILE* f_output, pst_item* item)
02007 {
02008     char*  result = NULL;
02009     size_t resultlen = 0;
02010     char   time_buffer[30];
02011     pst_item_journal* journal = item->journal;
02012 
02013     // make everything utf8
02014     pst_convert_utf8_null(item, &item->subject);
02015     pst_convert_utf8_null(item, &item->body);
02016 
02017     fprintf(f_output, "BEGIN:VJOURNAL\n");
02018     fprintf(f_output, "DTSTAMP:%s\n",                     pst_rfc2445_datetime_format_now(sizeof(time_buffer), time_buffer));
02019     if (item->create_date)
02020         fprintf(f_output, "CREATED:%s\n",                 pst_rfc2445_datetime_format(item->create_date, sizeof(time_buffer), time_buffer));
02021     if (item->modify_date)
02022         fprintf(f_output, "LAST-MOD:%s\n",                pst_rfc2445_datetime_format(item->modify_date, sizeof(time_buffer), time_buffer));
02023     if (item->subject.str)
02024         fprintf(f_output, "SUMMARY:%s\n",                 pst_rfc2426_escape(item->subject.str, &result, &resultlen));
02025     if (item->body.str)
02026         fprintf(f_output, "DESCRIPTION:%s\n",             pst_rfc2426_escape(item->body.str, &result, &resultlen));
02027     if (journal && journal->start)
02028         fprintf(f_output, "DTSTART;VALUE=DATE-TIME:%s\n", pst_rfc2445_datetime_format(journal->start, sizeof(time_buffer), time_buffer));
02029     fprintf(f_output, "END:VJOURNAL\n");
02030     if (result) free(result);
02031 }
02032 
02033 
02034 void write_appointment(FILE* f_output, pst_item* item)
02035 {
02036     char*  result = NULL;
02037     size_t resultlen = 0;
02038     char   time_buffer[30];
02039     pst_item_appointment* appointment = item->appointment;
02040 
02041     // make everything utf8
02042     pst_convert_utf8_null(item, &item->subject);
02043     pst_convert_utf8_null(item, &item->body);
02044     pst_convert_utf8_null(item, &appointment->location);
02045 
02046     fprintf(f_output, "UID:%#"PRIx64"\n", item->block_id);
02047     fprintf(f_output, "DTSTAMP:%s\n",                     pst_rfc2445_datetime_format_now(sizeof(time_buffer), time_buffer));
02048     if (item->create_date)
02049         fprintf(f_output, "CREATED:%s\n",                 pst_rfc2445_datetime_format(item->create_date, sizeof(time_buffer), time_buffer));
02050     if (item->modify_date)
02051         fprintf(f_output, "LAST-MOD:%s\n",                pst_rfc2445_datetime_format(item->modify_date, sizeof(time_buffer), time_buffer));
02052     if (item->subject.str)
02053         fprintf(f_output, "SUMMARY:%s\n",                 pst_rfc2426_escape(item->subject.str, &result, &resultlen));
02054     if (item->body.str)
02055         fprintf(f_output, "DESCRIPTION:%s\n",             pst_rfc2426_escape(item->body.str, &result, &resultlen));
02056     if (appointment && appointment->start)
02057         fprintf(f_output, "DTSTART;VALUE=DATE-TIME:%s\n", pst_rfc2445_datetime_format(appointment->start, sizeof(time_buffer), time_buffer));
02058     if (appointment && appointment->end)
02059         fprintf(f_output, "DTEND;VALUE=DATE-TIME:%s\n",   pst_rfc2445_datetime_format(appointment->end, sizeof(time_buffer), time_buffer));
02060     if (appointment && appointment->location.str)
02061         fprintf(f_output, "LOCATION:%s\n",                pst_rfc2426_escape(appointment->location.str, &result, &resultlen));
02062     if (appointment) {
02063         switch (appointment->showas) {
02064             case PST_FREEBUSY_TENTATIVE:
02065                 fprintf(f_output, "STATUS:TENTATIVE\n");
02066                 break;
02067             case PST_FREEBUSY_FREE:
02068                 // mark as transparent and as confirmed
02069                 fprintf(f_output, "TRANSP:TRANSPARENT\n");
02070             case PST_FREEBUSY_BUSY:
02071             case PST_FREEBUSY_OUT_OF_OFFICE:
02072                 fprintf(f_output, "STATUS:CONFIRMED\n");
02073                 break;
02074         }
02075         if (appointment->is_recurring) {
02076             const char* rules[] = {"DAILY", "WEEKLY", "MONTHLY", "YEARLY"};
02077             const char* days[]  = {"SU", "MO", "TU", "WE", "TH", "FR", "SA"};
02078             pst_recurrence *rdata = pst_convert_recurrence(appointment);
02079             fprintf(f_output, "RRULE:FREQ=%s", rules[rdata->type]);
02080             if (rdata->count)       fprintf(f_output, ";COUNT=%u",      rdata->count);
02081             if ((rdata->interval != 1) &&
02082                 (rdata->interval))  fprintf(f_output, ";INTERVAL=%u",   rdata->interval);
02083             if (rdata->dayofmonth)  fprintf(f_output, ";BYMONTHDAY=%d", rdata->dayofmonth);
02084             if (rdata->monthofyear) fprintf(f_output, ";BYMONTH=%d",    rdata->monthofyear);
02085             if (rdata->position)    fprintf(f_output, ";BYSETPOS=%d",   rdata->position);
02086             if (rdata->bydaymask) {
02087                 char byday[40];
02088                 int  empty = 1;
02089                 int i=0;
02090                 memset(byday, 0, sizeof(byday));
02091                 for (i=0; i<6; i++) {
02092                     int bit = 1 << i;
02093                     if (bit & rdata->bydaymask) {
02094                         char temp[40];
02095                         snprintf(temp, sizeof(temp), "%s%s%s", byday, (empty) ? ";BYDAY=" : ";", days[i]);
02096                         strcpy(byday, temp);
02097                         empty = 0;
02098                     }
02099                 }
02100                 fprintf(f_output, "%s", byday);
02101             }
02102             fprintf(f_output, "\n");
02103             pst_free_recurrence(rdata);
02104         }
02105         switch (appointment->label) {
02106             case PST_APP_LABEL_NONE:
02107                 if (!write_extra_categories(f_output, item)) fprintf(f_output, "CATEGORIES:NONE\n");
02108                 break;
02109             case PST_APP_LABEL_IMPORTANT:
02110                 fprintf(f_output, "CATEGORIES:IMPORTANT\n");
02111                 break;
02112             case PST_APP_LABEL_BUSINESS:
02113                 fprintf(f_output, "CATEGORIES:BUSINESS\n");
02114                 break;
02115             case PST_APP_LABEL_PERSONAL:
02116                 fprintf(f_output, "CATEGORIES:PERSONAL\n");
02117                 break;
02118             case PST_APP_LABEL_VACATION:
02119                 fprintf(f_output, "CATEGORIES:VACATION\n");
02120                 break;
02121             case PST_APP_LABEL_MUST_ATTEND:
02122                 fprintf(f_output, "CATEGORIES:MUST-ATTEND\n");
02123                 break;
02124             case PST_APP_LABEL_TRAVEL_REQ:
02125                 fprintf(f_output, "CATEGORIES:TRAVEL-REQUIRED\n");
02126                 break;
02127             case PST_APP_LABEL_NEEDS_PREP:
02128                 fprintf(f_output, "CATEGORIES:NEEDS-PREPARATION\n");
02129                 break;
02130             case PST_APP_LABEL_BIRTHDAY:
02131                 fprintf(f_output, "CATEGORIES:BIRTHDAY\n");
02132                 break;
02133             case PST_APP_LABEL_ANNIVERSARY:
02134                 fprintf(f_output, "CATEGORIES:ANNIVERSARY\n");
02135                 break;
02136             case PST_APP_LABEL_PHONE_CALL:
02137                 fprintf(f_output, "CATEGORIES:PHONE-CALL\n");
02138                 break;
02139         }
02140         // ignore bogus alarms
02141         if (appointment->alarm && (appointment->alarm_minutes >= 0) && (appointment->alarm_minutes < 1440)) {
02142             fprintf(f_output, "BEGIN:VALARM\n");
02143             fprintf(f_output, "TRIGGER:-PT%dM\n", appointment->alarm_minutes);
02144             fprintf(f_output, "ACTION:DISPLAY\n");
02145             fprintf(f_output, "DESCRIPTION:Reminder\n");
02146             fprintf(f_output, "END:VALARM\n");
02147         }
02148     }
02149     fprintf(f_output, "END:VEVENT\n");
02150     if (result) free(result);
02151 }
02152 
02153 
02154 void create_enter_dir(struct file_ll* f, pst_item *item)
02155 {
02156     pst_convert_utf8(item, &item->file_as);
02157     f->type         = item->type;
02158     f->stored_count = (item->folder) ? item->folder->item_count : 0;
02159 
02160     DEBUG_ENT("create_enter_dir");
02161     if (mode == MODE_KMAIL)
02162         f->name = mk_kmail_dir(item->file_as.str);
02163     else if (mode == MODE_RECURSE) {
02164         f->name = mk_recurse_dir(item->file_as.str, f->type);
02165         if (mode_thunder) {
02166             FILE *type_file = fopen(".type", "w");
02167             fprintf(type_file, "%d\n", item->type);
02168             fclose(type_file);
02169         }
02170     } else if (mode == MODE_SEPARATE) {
02171         // do similar stuff to recurse here.
02172         mk_separate_dir(item->file_as.str);
02173         f->name = (char*) pst_malloc(file_name_len);
02174         memset(f->name, 0, file_name_len);
02175     } else {
02176         f->name = (char*) pst_malloc(strlen(item->file_as.str)+strlen(OUTPUT_TEMPLATE)+1);
02177         sprintf(f->name, OUTPUT_TEMPLATE, item->file_as.str);
02178     }
02179 
02180     f->dname = (char*) pst_malloc(strlen(item->file_as.str)+1);
02181     strcpy(f->dname, item->file_as.str);
02182 
02183     if (overwrite != 1) {
02184         int x = 0;
02185         char *temp = (char*) pst_malloc (strlen(f->name)+10); //enough room for 10 digits
02186 
02187         sprintf(temp, "%s", f->name);
02188         check_filename(temp);
02189         while ((f->output = fopen(temp, "r"))) {
02190             DEBUG_INFO(("need to increase filename because one already exists with that name\n"));
02191             DEBUG_INFO(("- increasing it to %s%d\n", f->name, x));
02192             x++;
02193             sprintf(temp, "%s%08d", f->name, x);
02194             DEBUG_INFO(("- trying \"%s\"\n", f->name));
02195             if (x == 99999999) {
02196                 DIE(("create_enter_dir: Why can I not create a folder %s? I have tried %i extensions...\n", f->name, x));
02197             }
02198             fclose(f->output);
02199         }
02200         if (x > 0) { //then the f->name should change
02201             free (f->name);
02202             f->name = temp;
02203         } else {
02204             free(temp);
02205         }
02206     }
02207 
02208     DEBUG_INFO(("f->name = %s\nitem->folder_name = %s\n", f->name, item->file_as.str));
02209     if (mode != MODE_SEPARATE) {
02210         check_filename(f->name);
02211         if (!(f->output = fopen(f->name, "w"))) {
02212             DIE(("create_enter_dir: Could not open file \"%s\" for write\n", f->name));
02213         }
02214     }
02215     DEBUG_RET();
02216 }
02217 
02218 
02219 void close_enter_dir(struct file_ll *f)
02220 {
02221     DEBUG_INFO(("processed item count for folder %s is %i, skipped %i, total %i \n",
02222                 f->dname, f->item_count, f->skip_count, f->stored_count));
02223     if (output_mode != OUTPUT_QUIET) {
02224         pst_debug_lock();
02225             printf("\t\"%s\" - %i items done, %i items skipped.\n", f->dname, f->item_count, f->skip_count);
02226             fflush(stdout);
02227         pst_debug_unlock();
02228     }
02229     if (f->output) {
02230         if (mode == MODE_SEPARATE) DEBUG_WARN(("close_enter_dir finds open separate file\n"));
02231         struct stat st;
02232         fclose(f->output);
02233         stat(f->name, &st);
02234         if (!st.st_size) {
02235             DEBUG_WARN(("removing empty output file %s\n", f->name));
02236             remove(f->name);
02237         }
02238         f->output = NULL;
02239     }
02240     free(f->name);
02241     free(f->dname);
02242 
02243     if (mode == MODE_KMAIL)
02244         close_kmail_dir();
02245     else if (mode == MODE_RECURSE) {
02246         if (mode_thunder) {
02247             FILE *type_file = fopen(".size", "w");
02248             fprintf(type_file, "%i %i\n", f->item_count, f->stored_count);
02249             fclose(type_file);
02250         }
02251         close_recurse_dir();
02252     } else if (mode == MODE_SEPARATE)
02253         close_separate_dir();
02254 }
02255 

Generated on 27 Dec 2013 for 'LibPst' by  doxygen 1.6.1