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.%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[PST_TYPE_MAX];
00022     char *dname;
00023     FILE * output[PST_TYPE_MAX];
00024     int32_t stored_count;
00025     int32_t item_count;
00026     int32_t skip_count;
00027 };
00028 
00029 int       grim_reaper();
00030 pid_t     try_fork(char* folder);
00031 void      process(pst_item *outeritem, pst_desc_tree *d_ptr);
00032 void      write_email_body(FILE *f, char *body);
00033 void      removeCR(char *c);
00034 void      usage();
00035 void      version();
00036 void      mk_kmail_dir(char* fname);
00037 int       close_kmail_dir();
00038 void      mk_recurse_dir(char* dir);
00039 int       close_recurse_dir();
00040 void      mk_separate_dir(char *dir);
00041 int       close_separate_dir();
00042 void      mk_separate_file(struct file_ll *f, int32_t t, char *extension, int openit);
00043 void      close_separate_file(struct file_ll *f);
00044 char*     my_stristr(char *haystack, char *needle);
00045 void      check_filename(char *fname);
00046 int       acceptable_ext(pst_item_attach* attach);
00047 void      write_separate_attachment(char f_name[], pst_item_attach* attach, int attach_num, pst_file* pst);
00048 void      write_embedded_message(FILE* f_output, pst_item_attach* attach, char *boundary, pst_file* pf, int save_rtf, char** extra_mime_headers);
00049 void      write_inline_attachment(FILE* f_output, pst_item_attach* attach, char *boundary, pst_file* pst);
00050 int       valid_headers(char *header);
00051 void      header_has_field(char *header, char *field, int *flag);
00052 void      header_get_subfield(char *field, const char *subfield, char *body_subfield, size_t size_subfield);
00053 char*     header_get_field(char *header, char *field);
00054 char*     header_end_field(char *field);
00055 void      header_strip_field(char *header, char *field);
00056 int       test_base64(char *body, size_t len);
00057 void      find_html_charset(char *html, char *charset, size_t charsetlen);
00058 void      find_rfc822_headers(char** extra_mime_headers);
00059 void      write_body_part(FILE* f_output, pst_string *body, char *mime, char *charset, char *boundary, pst_file* pst);
00060 void      write_schedule_part_data(FILE* f_output, pst_item* item, const char* sender, const char* method);
00061 void      write_schedule_part(FILE* f_output, pst_item* item, const char* sender, const char* boundary);
00062 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);
00063 void      write_vcard(FILE* f_output, pst_item *item, pst_item_contact* contact, char comment[]);
00064 int       write_extra_categories(FILE* f_output, pst_item* item);
00065 void      write_journal(FILE* f_output, pst_item* item);
00066 void      write_appointment(FILE* f_output, pst_item *item);
00067 void      create_enter_dir(struct file_ll* f, pst_item *item);
00068 void      close_enter_dir(struct file_ll *f);
00069 
00070 const char*  prog_name;
00071 char*  output_dir = ".";
00072 
00073 // Normal mode just creates mbox format files in the current directory. Each file is named
00074 // the same as the folder's name that it represents
00075 #define MODE_NORMAL 0
00076 
00077 // KMail mode creates a directory structure suitable for being used directly
00078 // by the KMail application
00079 #define MODE_KMAIL 1
00080 
00081 // recurse mode creates a directory structure like the PST file. Each directory
00082 // contains only one file which stores the emails in mboxrd format.
00083 #define MODE_RECURSE 2
00084 
00085 // separate mode creates the same directory structure as recurse. The emails are stored in
00086 // separate files, numbering from 1 upward. Attachments belonging to the emails are
00087 // saved as email_no-filename (e.g. 1-samplefile.doc or 1-Attachment2.zip)
00088 #define MODE_SEPARATE 3
00089 
00090 
00091 // Output Normal just prints the standard information about what is going on
00092 #define OUTPUT_NORMAL 0
00093 
00094 // Output Quiet is provided so that only errors are printed
00095 #define OUTPUT_QUIET 1
00096 
00097 // default mime-type for attachments that have a null mime-type
00098 #define MIME_TYPE_DEFAULT "application/octet-stream"
00099 #define RFC822            "message/rfc822"
00100 
00101 // output mode for contacts
00102 #define CMODE_VCARD 0
00103 #define CMODE_LIST  1
00104 
00105 // output mode for deleted items
00106 #define DMODE_EXCLUDE 0
00107 #define DMODE_INCLUDE 1
00108 
00109 // Output type mode flags
00110 #define OTMODE_EMAIL        1
00111 #define OTMODE_APPOINTMENT  2
00112 #define OTMODE_JOURNAL      4
00113 #define OTMODE_CONTACT      8
00114 
00115 // output settings for RTF bodies
00116 // filename for the attachment
00117 #define RTF_ATTACH_NAME "rtf-body.rtf"
00118 // mime type for the attachment
00119 #define RTF_ATTACH_TYPE "application/rtf"
00120 
00121 // global settings
00122 int         mode         = MODE_NORMAL;
00123 int         mode_MH      = 0;   // a submode of MODE_SEPARATE
00124 int         mode_EX      = 0;   // a submode of MODE_SEPARATE
00125 int         mode_MSG     = 0;   // a submode of MODE_SEPARATE
00126 int         mode_thunder = 0;   // a submode of MODE_RECURSE
00127 int         output_mode  = OUTPUT_NORMAL;
00128 int         contact_mode = CMODE_VCARD;
00129 int         deleted_mode = DMODE_EXCLUDE;
00130 int         output_type_mode = 0xff;    // Default to all.
00131 int         contact_mode_specified = 0;
00132 int         overwrite = 0;
00133 int         prefer_utf8 = 0;
00134 int         save_rtf_body = 1;
00135 int         file_name_len = 10;     // enough room for MODE_SPEARATE file name
00136 pst_file    pstfile;
00137 regex_t     meta_charset_pattern;
00138 char*       default_charset = NULL;
00139 char*       acceptable_extensions = NULL;
00140 
00141 int         number_processors = 1;  // number of cpus we have
00142 int         max_children  = 0;      // based on number of cpus and command line args
00143 int         max_child_specified = 0;// have command line arg -j
00144 int         active_children;        // number of children of this process, cannot be larger than max_children
00145 pid_t*      child_processes;        // setup by main(), and at the start of new child process
00146 
00147 #ifdef HAVE_SEMAPHORE_H
00148 int         shared_memory_id;
00149 sem_t*      global_children = NULL;
00150 sem_t*      output_mutex    = NULL;
00151 #endif
00152 
00153 
00154 int grim_reaper(int waitall)
00155 {
00156     int available = 0;
00157 #ifdef HAVE_FORK
00158 #ifdef HAVE_SEMAPHORE_H
00159     if (global_children) {
00160         //sem_getvalue(global_children, &available);
00161         //printf("grim reaper %s for pid %d (parent %d) with %d children, %d available\n", (waitall) ? "all" : "", getpid(), getppid(), active_children, available);
00162         //fflush(stdout);
00163         int i,j;
00164         for (i=0; i<active_children; i++) {
00165             int status;
00166             pid_t child = child_processes[i];
00167             pid_t ch = waitpid(child, &status, ((waitall) ? 0 : WNOHANG));
00168             if (ch == child) {
00169                 // check termination status
00170                 //if (WIFEXITED(status)) {
00171                 //    int ext = WEXITSTATUS(status);
00172                 //    printf("Process %d exited with status  %d\n", child, ext);
00173                 //    fflush(stdout);
00174                 //}
00175                 if (WIFSIGNALED(status)) {
00176                     int sig = WTERMSIG(status);
00177                     DEBUG_INFO(("Process %d terminated with signal %d\n", child, sig));
00178                     //printf("Process %d terminated with signal %d\n", child, sig);
00179                     //fflush(stdout);
00180                 }
00181                 if (status != 0) {
00182                     exit(status);
00183                 }
00184                 // this has terminated, remove it from the list
00185                 for (j=i; j<active_children-1; j++) {
00186                     child_processes[j] = child_processes[j+1];
00187                 }
00188                 active_children--;
00189                 i--;
00190             }
00191         }
00192         sem_getvalue(global_children, &available);
00193         //printf("grim reaper %s for pid %d with %d children, %d available\n", (waitall) ? "all" : "", getpid(), active_children, available);
00194         //fflush(stdout);
00195     }
00196 #endif
00197 #endif
00198     return available;
00199 }
00200 
00201 
00202 pid_t try_fork(char *folder)
00203 {
00204 #ifdef HAVE_FORK
00205 #ifdef HAVE_SEMAPHORE_H
00206     int available = grim_reaper(0);
00207     // If children have called sem_post but not exited yet, we could have available > 0 but active_children == max_children
00208     if (available && active_children < max_children) {
00209         sem_wait(global_children);
00210         pid_t child = fork();
00211         if (child < 0) {
00212             // fork failed, pretend it worked and we are the child
00213             return 0;
00214         }
00215         else if (child == 0) {
00216             // fork worked, and we are the child, reinitialize *our* list of children
00217             active_children = 0;
00218             memset(child_processes, 0, sizeof(pid_t) * max_children);
00219             pst_reopen(&pstfile);   // close and reopen the pst file to get an independent file position pointer
00220         }
00221         else {
00222             // fork worked, and we are the parent, record this child that we need to wait for
00223             //pid_t me = getpid();
00224             //printf("parent %d forked child pid %d to process folder %s\n", me, child, folder);
00225             //fflush(stdout);
00226             child_processes[active_children++] = child;
00227         }
00228         return child;
00229     }
00230     else {
00231         return 0;   // pretend to have forked and we are the child
00232     }
00233 #endif
00234 #endif
00235     return 0;
00236 }
00237 
00238 
00239 void process(pst_item *outeritem, pst_desc_tree *d_ptr)
00240 {
00241     struct file_ll ff;
00242     pst_item *item = NULL;
00243 
00244     DEBUG_ENT("process");
00245     create_enter_dir(&ff, outeritem);
00246 
00247     for (; d_ptr; d_ptr = d_ptr->next) {
00248         DEBUG_INFO(("New item record\n"));
00249         if (!d_ptr->desc) {
00250             ff.skip_count++;
00251             DEBUG_WARN(("ERROR item's desc record is NULL\n"));
00252             continue;
00253         }
00254         DEBUG_INFO(("Desc Email ID %#"PRIx64" [d_ptr->d_id = %#"PRIx64"]\n", d_ptr->desc->i_id, d_ptr->d_id));
00255 
00256         item = pst_parse_item(&pstfile, d_ptr, NULL);
00257         DEBUG_INFO(("About to process item\n"));
00258 
00259         if (!item) {
00260             ff.skip_count++;
00261             DEBUG_INFO(("A NULL item was seen\n"));
00262             continue;
00263         }
00264 
00265         if (item->subject.str) {
00266             DEBUG_INFO(("item->subject = %s\n", item->subject.str));
00267         }
00268 
00269         if (item->folder && item->file_as.str) {
00270             DEBUG_INFO(("Processing Folder \"%s\"\n", item->file_as.str));
00271             if (output_mode != OUTPUT_QUIET) {
00272                 pst_debug_lock();
00273                     printf("Processing Folder \"%s\"\n", item->file_as.str);
00274                     fflush(stdout);
00275                 pst_debug_unlock();
00276             }
00277             ff.item_count++;
00278             if (d_ptr->child && (deleted_mode == DMODE_INCLUDE || strcasecmp(item->file_as.str, "Deleted Items"))) {
00279                 //if this is a non-empty folder other than deleted items, we want to recurse into it
00280                 pid_t parent = getpid();
00281                 pid_t child = try_fork(item->file_as.str);
00282                 if (child == 0) {
00283                     // we are the child process, or the original parent if no children were available
00284                     pid_t me = getpid();
00285                     process(item, d_ptr->child);
00286 #ifdef HAVE_FORK
00287 #ifdef HAVE_SEMAPHORE_H
00288                     if (me != parent) {
00289                         // we really were a child, forked for the sole purpose of processing this folder
00290                         // free my child count slot before really exiting, since
00291                         // all I am doing here is waiting for my children to exit
00292                         sem_post(global_children);
00293                         grim_reaper(1); // wait for all my child processes to exit
00294                         exit(0);        // really exit
00295                     }
00296 #endif
00297 #endif
00298                 }
00299             }
00300 
00301         } else if (item->contact && (item->type == PST_TYPE_CONTACT)) {
00302             DEBUG_INFO(("Processing Contact\n"));
00303             if (!(output_type_mode & OTMODE_CONTACT)) {
00304                 ff.skip_count++;
00305                 DEBUG_INFO(("skipping contact: not in output type list\n"));
00306             }
00307             else {
00308                 ff.item_count++;
00309                 if (mode == MODE_SEPARATE) mk_separate_file(&ff, PST_TYPE_CONTACT, (mode_EX) ? ".vcf" : "", 1);
00310                 if (contact_mode == CMODE_VCARD) {
00311                     pst_convert_utf8_null(item, &item->comment);
00312                     write_vcard(ff.output[PST_TYPE_CONTACT], item, item->contact, item->comment.str);
00313                 }
00314                 else {
00315                     pst_convert_utf8(item, &item->contact->fullname);
00316                     pst_convert_utf8(item, &item->contact->address1);
00317                     fprintf(ff.output[PST_TYPE_CONTACT], "%s <%s>\n", item->contact->fullname.str, item->contact->address1.str);
00318                 }
00319                 if (mode == MODE_SEPARATE) close_separate_file(&ff);
00320             }
00321 
00322         } else if (item->email && ((item->type == PST_TYPE_NOTE) || (item->type == PST_TYPE_SCHEDULE) || (item->type == PST_TYPE_REPORT))) {
00323             DEBUG_INFO(("Processing Email\n"));
00324             if (!(output_type_mode & OTMODE_EMAIL)) {
00325                 ff.skip_count++;
00326                 DEBUG_INFO(("skipping email: not in output type list\n"));
00327             }
00328             else {
00329                 char *extra_mime_headers = NULL;
00330                 ff.item_count++;
00331                 if (mode == MODE_SEPARATE) {
00332                     // process this single email message, possibly forking
00333                     pid_t parent = getpid();
00334                     pid_t child = try_fork(item->file_as.str);
00335                     if (child == 0) {
00336                         // we are the child process, or the original parent if no children were available
00337                         pid_t me = getpid();
00338                         mk_separate_file(&ff, PST_TYPE_NOTE, (mode_EX) ? ".eml" : "", 1);
00339                         write_normal_email(ff.output[PST_TYPE_NOTE], ff.name[PST_TYPE_NOTE], item, mode, mode_MH, &pstfile, save_rtf_body, PST_TYPE_NOTE, &extra_mime_headers);
00340                         close_separate_file(&ff);
00341                         if (mode_MSG) {
00342                             mk_separate_file(&ff, PST_TYPE_NOTE, ".msg", 0);
00343                             write_msg_email(ff.name[PST_TYPE_NOTE], item, &pstfile);
00344                         }
00345 #ifdef HAVE_FORK
00346 #ifdef HAVE_SEMAPHORE_H
00347                         if (me != parent) {
00348                             // we really were a child, forked for the sole purpose of processing this message
00349                             // free my child count slot before really exiting, since
00350                             // all I am doing here is waiting for my children to exit
00351                             sem_post(global_children);
00352                             grim_reaper(1); // wait for all my child processes to exit - there should not be any
00353                             exit(0);        // really exit
00354                         }
00355 #endif
00356 #endif
00357                     }
00358                 }
00359                 else {
00360                     // process this single email message, cannot fork since not separate mode
00361                     write_normal_email(ff.output[PST_TYPE_NOTE], ff.name[PST_TYPE_NOTE], item, mode, mode_MH, &pstfile, save_rtf_body, 0, &extra_mime_headers);
00362                 }
00363             }
00364 
00365         } else if (item->journal && (item->type == PST_TYPE_JOURNAL)) {
00366             DEBUG_INFO(("Processing Journal Entry\n"));
00367             if (!(output_type_mode & OTMODE_JOURNAL)) {
00368                 ff.skip_count++;
00369                 DEBUG_INFO(("skipping journal entry: not in output type list\n"));
00370             }
00371             else {
00372                 ff.item_count++;
00373                 if (mode == MODE_SEPARATE) mk_separate_file(&ff, PST_TYPE_JOURNAL, (mode_EX) ? ".ics" : "", 1);
00374                 write_journal(ff.output[PST_TYPE_JOURNAL], item);
00375                 fprintf(ff.output[PST_TYPE_JOURNAL], "\n");
00376                 if (mode == MODE_SEPARATE) close_separate_file(&ff);
00377             }
00378 
00379         } else if (item->appointment && (item->type == PST_TYPE_APPOINTMENT)) {
00380             DEBUG_INFO(("Processing Appointment Entry\n"));
00381             if (!(output_type_mode & OTMODE_APPOINTMENT)) {
00382                 ff.skip_count++;
00383                 DEBUG_INFO(("skipping appointment: not in output type list\n"));
00384             }
00385             else {
00386                 ff.item_count++;
00387                 if (mode == MODE_SEPARATE) mk_separate_file(&ff, PST_TYPE_APPOINTMENT, (mode_EX) ? ".ics" : "", 1);
00388                 write_schedule_part_data(ff.output[PST_TYPE_APPOINTMENT], item, NULL, NULL);
00389                 fprintf(ff.output[PST_TYPE_APPOINTMENT], "\n");
00390                 if (mode == MODE_SEPARATE) close_separate_file(&ff);
00391             }
00392 
00393         } else if (item->message_store) {
00394             // there should only be one message_store, and we have already done it
00395             ff.skip_count++;
00396             DEBUG_WARN(("item with message store content, type %i %s, skipping it\n", item->type, item->ascii_type));
00397 
00398         } else {
00399             ff.skip_count++;
00400             DEBUG_WARN(("Unknown item type %i (%s) name (%s)\n",
00401                         item->type, item->ascii_type, item->file_as.str));
00402         }
00403         pst_freeItem(item);
00404     }
00405     close_enter_dir(&ff);
00406     DEBUG_RET();
00407 }
00408 
00409 
00410 
00411 int main(int argc, char* const* argv) {
00412     pst_item *item = NULL;
00413     pst_desc_tree *d_ptr;
00414     char * fname = NULL;
00415     char *d_log  = NULL;
00416     int c,x;
00417     char *temp = NULL;               //temporary char pointer
00418     prog_name = argv[0];
00419 
00420     time_t now = time(NULL);
00421     srand((unsigned)now);
00422 
00423     if (regcomp(&meta_charset_pattern, "<meta[^>]*content=\"[^>]*charset=([^>\";]*)[\";]", REG_ICASE | REG_EXTENDED)) {
00424         printf("cannot compile regex pattern to find content charset in html bodies\n");
00425         exit(3);
00426     }
00427 
00428     // command-line option handling
00429     while ((c = getopt(argc, argv, "a:bC:c:Dd:emhj:kMo:qrSt:uVwL:8"))!= -1) {
00430         switch (c) {
00431         case 'a':
00432             if (optarg) {
00433                 int n = strlen(optarg);
00434                 acceptable_extensions = (char*)pst_malloc(n+2);
00435                 strcpy(acceptable_extensions, optarg);
00436                 acceptable_extensions[n+1] = '\0';  // double null terminates array of non-empty null terminated strings.
00437                 char *p = acceptable_extensions;
00438                 while (*p) {
00439                     if (*p == ',') *p = '\0';
00440                     p++;
00441                 }
00442             }
00443             break;
00444         case 'b':
00445             save_rtf_body = 0;
00446             break;
00447         case 'C':
00448             if (optarg) {
00449                 default_charset = optarg;
00450             }
00451             else {
00452                 usage();
00453                 exit(0);
00454             }
00455             break;
00456         case 'c':
00457             if (optarg && optarg[0]=='v') {
00458                 contact_mode=CMODE_VCARD;
00459                 contact_mode_specified = 1;
00460             }
00461             else if (optarg && optarg[0]=='l') {
00462                 contact_mode=CMODE_LIST;
00463                 contact_mode_specified = 1;
00464             }
00465             else {
00466                 usage();
00467                 exit(0);
00468             }
00469             break;
00470         case 'D':
00471             deleted_mode = DMODE_INCLUDE;
00472             break;
00473         case 'd':
00474             d_log = optarg;
00475             break;
00476         case 'h':
00477             usage();
00478             exit(0);
00479             break;
00480         case 'j':
00481             max_children = atoi(optarg);
00482             max_child_specified = 1;
00483             break;
00484         case 'k':
00485             mode = MODE_KMAIL;
00486             break;
00487         case 'M':
00488             mode = MODE_SEPARATE;
00489             mode_MH  = 1;
00490             mode_EX  = 0;
00491             mode_MSG = 0;
00492             break;
00493         case 'e':
00494             mode = MODE_SEPARATE;
00495             mode_MH  = 1;
00496             mode_EX  = 1;
00497             mode_MSG = 0;
00498             file_name_len = 14;
00499             break;
00500         case 'L':
00501             pst_debug_setlevel(atoi(optarg));
00502             break;
00503         case 'm':
00504             mode = MODE_SEPARATE;
00505             mode_MH  = 1;
00506             mode_EX  = 1;
00507             mode_MSG = 1;
00508             file_name_len = 14;
00509             break;
00510         case 'o':
00511             output_dir = optarg;
00512             break;
00513         case 'q':
00514             output_mode = OUTPUT_QUIET;
00515             break;
00516         case 'r':
00517             mode = MODE_RECURSE;
00518             mode_thunder = 0;
00519             break;
00520         case 'S':
00521             mode = MODE_SEPARATE;
00522             mode_MH  = 0;
00523             mode_EX  = 0;
00524             mode_MSG = 0;
00525             break;
00526         case 't':
00527             // email, appointment, contact, other
00528             if (!optarg) {
00529                 usage();
00530                 exit(0);
00531             }
00532             temp = optarg;
00533             output_type_mode = 0;
00534             while (*temp > 0) {
00535               switch (temp[0]) {
00536                 case 'e':
00537                     output_type_mode |= OTMODE_EMAIL;
00538                     break;
00539                 case 'a':
00540                     output_type_mode |= OTMODE_APPOINTMENT;
00541                     break;
00542                 case 'j':
00543                     output_type_mode |= OTMODE_JOURNAL;
00544                     break;
00545                 case 'c':
00546                     output_type_mode |= OTMODE_CONTACT;
00547                     break;
00548                 default:
00549                     usage();
00550                     exit(0);
00551                     break;
00552               }
00553               temp++;
00554             }
00555             break;
00556         case 'u':
00557             mode = MODE_RECURSE;
00558             mode_thunder = 1;
00559             break;
00560         case 'V':
00561             version();
00562             exit(0);
00563             break;
00564         case 'w':
00565             overwrite = 1;
00566             break;
00567         case '8':
00568             prefer_utf8 = 1;
00569             break;
00570         default:
00571             usage();
00572             exit(1);
00573             break;
00574         }
00575     }
00576 
00577     if (argc > optind) {
00578         fname = argv[optind];
00579     } else {
00580         usage();
00581         exit(2);
00582     }
00583 
00584 #ifdef _SC_NPROCESSORS_ONLN
00585     number_processors =  sysconf(_SC_NPROCESSORS_ONLN);
00586 #endif
00587     max_children    = (max_child_specified) ? max_children : number_processors * 4;
00588     active_children = 0;
00589     child_processes = (pid_t *)pst_malloc(sizeof(pid_t) * max_children);
00590     memset(child_processes, 0, sizeof(pid_t) * max_children);
00591 
00592 #ifdef HAVE_SEMAPHORE_H
00593     if (max_children) {
00594         shared_memory_id = shmget(IPC_PRIVATE, sizeof(sem_t)*2, 0777);
00595         if (shared_memory_id >= 0) {
00596             global_children = (sem_t *)shmat(shared_memory_id, NULL, 0);
00597             if (global_children == (sem_t *)-1) global_children = NULL;
00598             if (global_children) {
00599                 output_mutex = &(global_children[1]);
00600                 sem_init(global_children, 1, max_children);
00601                 sem_init(output_mutex, 1, 1);
00602             }
00603             shmctl(shared_memory_id, IPC_RMID, NULL);
00604         }
00605     }
00606 #endif
00607 
00608     #ifdef DEBUG_ALL
00609         // force a log file
00610         if (!d_log) d_log = "readpst.log";
00611     #endif // defined DEBUG_ALL
00612     #ifdef HAVE_SEMAPHORE_H
00613         DEBUG_INIT(d_log, output_mutex);
00614     #else
00615         DEBUG_INIT(d_log, NULL);
00616     #endif
00617     DEBUG_ENT("main");
00618 
00619     if (output_mode != OUTPUT_QUIET) printf("Opening PST file and indexes...\n");
00620     RET_DERROR(pst_open(&pstfile, fname, default_charset), 1, ("Error opening File\n"));
00621     RET_DERROR(pst_load_index(&pstfile), 2, ("Index Error\n"));
00622 
00623     pst_load_extended_attributes(&pstfile);
00624 
00625     if (chdir(output_dir)) {
00626         x = errno;
00627         pst_close(&pstfile);
00628         DEBUG_RET();
00629         DIE(("Cannot change to output dir %s: %s\n", output_dir, strerror(x)));
00630     }
00631 
00632     d_ptr = pstfile.d_head; // first record is main record
00633     item  = pst_parse_item(&pstfile, d_ptr, NULL);
00634     if (!item || !item->message_store) {
00635         DEBUG_RET();
00636         DIE(("Could not get root record\n"));
00637     }
00638 
00639     // default the file_as to the same as the main filename if it doesn't exist
00640     if (!item->file_as.str) {
00641         if (!(temp = strrchr(fname, '/')))
00642             if (!(temp = strrchr(fname, '\\')))
00643                 temp = fname;
00644             else
00645                 temp++; // get past the "\\"
00646         else
00647             temp++; // get past the "/"
00648         item->file_as.str = (char*)pst_malloc(strlen(temp)+1);
00649         strcpy(item->file_as.str, temp);
00650         item->file_as.is_utf8 = 1;
00651         DEBUG_INFO(("file_as was blank, so am using %s\n", item->file_as.str));
00652     }
00653     DEBUG_INFO(("Root Folder Name: %s\n", item->file_as.str));
00654 
00655     d_ptr = pst_getTopOfFolders(&pstfile, item);
00656     if (!d_ptr) {
00657         DEBUG_RET();
00658         DIE(("Top of folders record not found. Cannot continue\n"));
00659     }
00660 
00661     process(item, d_ptr->child);    // do the children of TOPF
00662     grim_reaper(1); // wait for all child processes
00663 
00664     pst_freeItem(item);
00665     pst_close(&pstfile);
00666     DEBUG_RET();
00667 
00668 #ifdef HAVE_SEMAPHORE_H
00669     if (global_children) {
00670         sem_destroy(global_children);
00671         sem_destroy(output_mutex);
00672         shmdt(global_children);
00673     }
00674 #endif
00675 
00676     regfree(&meta_charset_pattern);
00677     return 0;
00678 }
00679 
00680 
00681 void write_email_body(FILE *f, char *body) {
00682     char *n = body;
00683     DEBUG_ENT("write_email_body");
00684     if (mode != MODE_SEPARATE) {
00685         while (n) {
00686             char *p = body;
00687             while (*p == '>') p++;
00688             if (strncmp(p, "From ", 5) == 0) fprintf(f, ">");
00689             if ((n = strchr(body, '\n'))) {
00690                 n++;
00691                 pst_fwrite(body, n-body, 1, f); //write just a line
00692                 body = n;
00693             }
00694         }
00695     }
00696     pst_fwrite(body, strlen(body), 1, f);
00697     DEBUG_RET();
00698 }
00699 
00700 
00701 void removeCR (char *c) {
00702     // converts \r\n to \n
00703     char *a, *b;
00704     DEBUG_ENT("removeCR");
00705     a = b = c;
00706     while (*a != '\0') {
00707         *b = *a;
00708         if (*a != '\r') b++;
00709         a++;
00710     }
00711     *b = '\0';
00712     DEBUG_RET();
00713 }
00714 
00715 
00716 void usage() {
00717     DEBUG_ENT("usage");
00718     version();
00719     printf("Usage: %s [OPTIONS] {PST FILENAME}\n", prog_name);
00720     printf("OPTIONS:\n");
00721     printf("\t-V\t- Version. Display program version\n");
00722     printf("\t-C charset\t- character set for items with an unspecified character set\n");
00723     printf("\t-D\t- Include deleted items in output\n");
00724     printf("\t-L <level> \t- Set debug level; 1=debug,2=info,3=warn.\n");
00725     printf("\t-M\t- Write emails in the MH (rfc822) format\n");
00726     printf("\t-S\t- Separate. Write emails in the separate format\n");
00727     printf("\t-a <attachment-extension-list>\t- Discard any attachment without an extension on the list\n");
00728     printf("\t-b\t- Don't save RTF-Body attachments\n");
00729     printf("\t-c[v|l]\t- Set the Contact output mode. -cv = VCard, -cl = EMail list\n");
00730     printf("\t-d <filename> \t- Debug to file.\n");
00731     printf("\t-e\t- As with -M, but include extensions on output files\n");
00732     printf("\t-h\t- Help. This screen\n");
00733     printf("\t-j <integer>\t- Number of parallel jobs to run\n");
00734     printf("\t-k\t- KMail. Output in kmail format\n");
00735     printf("\t-m\t- As with -e, but write .msg files also\n");
00736     printf("\t-o <dirname>\t- Output directory to write files to. CWD is changed *after* opening pst file\n");
00737     printf("\t-q\t- Quiet. Only print error messages\n");
00738     printf("\t-r\t- Recursive. Output in a recursive format\n");
00739     printf("\t-t[eajc]\t- Set the output type list. e = email, a = attachment, j = journal, c = contact\n");
00740     printf("\t-u\t- Thunderbird mode. Write two extra .size and .type files\n");
00741     printf("\t-w\t- Overwrite any output mbox files\n");
00742     printf("\t-8\t- Output bodies in UTF-8, rather than original encoding, if UTF-8 version is available\n");
00743     printf("\n");
00744     printf("Only one of -M -S -e -k -m -r should be specified\n");
00745     DEBUG_RET();
00746 }
00747 
00748 
00749 void version() {
00750     DEBUG_ENT("version");
00751     printf("ReadPST / LibPST v%s\n", VERSION);
00752 #if BYTE_ORDER == BIG_ENDIAN
00753     printf("Big Endian implementation being used.\n");
00754 #elif BYTE_ORDER == LITTLE_ENDIAN
00755     printf("Little Endian implementation being used.\n");
00756 #else
00757 #  error "Byte order not supported by this library"
00758 #endif
00759     DEBUG_RET();
00760 }
00761 
00762 
00763 void mk_kmail_dir(char *fname) {
00764     //make a directory based on OUTPUT_KMAIL_DIR_TEMPLATE
00765     //change to that directory
00766     char *dir, *index;
00767     int x;
00768     DEBUG_ENT("mk_kmail_dir");
00769     dir = pst_malloc(strlen(fname)+strlen(OUTPUT_KMAIL_DIR_TEMPLATE)+1);
00770     sprintf(dir, OUTPUT_KMAIL_DIR_TEMPLATE, fname);
00771     check_filename(dir);
00772     if (D_MKDIR(dir)) {
00773         if (errno != EEXIST) {  // not an error because it exists
00774             x = errno;
00775             DIE(("mk_kmail_dir: Cannot create directory %s: %s\n", dir, strerror(x)));
00776         }
00777     }
00778     if (chdir(dir)) {
00779         x = errno;
00780         DIE(("mk_kmail_dir: Cannot change to directory %s: %s\n", dir, strerror(x)));
00781     }
00782     free (dir);
00783 
00784     //we should remove any existing indexes created by KMail, cause they might be different now
00785     index = pst_malloc(strlen(fname)+strlen(KMAIL_INDEX)+1);
00786     sprintf(index, KMAIL_INDEX, fname);
00787     unlink(index);
00788     free(index);
00789 
00790     DEBUG_RET();
00791 }
00792 
00793 
00794 int close_kmail_dir() {
00795     int x;
00796     DEBUG_ENT("close_kmail_dir");
00797     if (chdir("..")) {
00798         x = errno;
00799         DIE(("close_kmail_dir: Cannot move up dir (..): %s\n", strerror(x)));
00800     }
00801     DEBUG_RET();
00802     return 0;
00803 }
00804 
00805 
00806 char *item_type_to_name(int32_t item_type) {
00807     char *name;
00808     switch (item_type) {
00809         case PST_TYPE_APPOINTMENT:
00810             name = "calendar";
00811             break;
00812         case PST_TYPE_CONTACT:
00813             name = "contacts";
00814             break;
00815         case PST_TYPE_JOURNAL:
00816             name = "journal";
00817             break;
00818         case PST_TYPE_STICKYNOTE:
00819         case PST_TYPE_TASK:
00820         case PST_TYPE_NOTE:
00821         case PST_TYPE_OTHER:
00822         case PST_TYPE_REPORT:
00823         default:
00824             name = "mbox";
00825             break;
00826     }
00827     return name;
00828 }
00829 
00830 
00831 int32_t reduced_item_type(int32_t item_type) {
00832     int32_t reduced;
00833     switch (item_type) {
00834         case PST_TYPE_APPOINTMENT:
00835         case PST_TYPE_CONTACT:
00836         case PST_TYPE_JOURNAL:
00837             reduced = item_type;
00838             break;
00839         case PST_TYPE_STICKYNOTE:
00840         case PST_TYPE_TASK:
00841         case PST_TYPE_NOTE:
00842         case PST_TYPE_OTHER:
00843         case PST_TYPE_REPORT:
00844         default:
00845             reduced = PST_TYPE_NOTE;
00846             break;
00847     }
00848     return reduced;
00849 }
00850 
00851 
00852 // this will create a directory by that name
00853 void mk_recurse_dir(char *dir) {
00854     int x;
00855     DEBUG_ENT("mk_recurse_dir");
00856     check_filename(dir);
00857     if (D_MKDIR (dir)) {
00858         if (errno != EEXIST) {  // not an error because it exists
00859             x = errno;
00860             DIE(("mk_recurse_dir: Cannot create directory %s: %s\n", dir, strerror(x)));
00861         }
00862     }
00863     if (chdir(dir)) {
00864         x = errno;
00865         DIE(("mk_recurse_dir: Cannot change to directory %s: %s\n", dir, strerror(x)));
00866     }
00867     DEBUG_RET();
00868 }
00869 
00870 
00871 int close_recurse_dir() {
00872     int x;
00873     DEBUG_ENT("close_recurse_dir");
00874     if (chdir("..")) {
00875         x = errno;
00876         DIE(("close_recurse_dir: Cannot go up dir (..): %s\n", strerror(x)));
00877     }
00878     DEBUG_RET();
00879     return 0;
00880 }
00881 
00882 
00883 void mk_separate_dir(char *dir) {
00884     size_t dirsize = strlen(dir) + 10;
00885     char dir_name[dirsize];
00886     int x = 0, y = 0;
00887 
00888     DEBUG_ENT("mk_separate_dir");
00889     do {
00890         if (y == 0)
00891             snprintf(dir_name, dirsize, "%s", dir);
00892         else
00893             snprintf(dir_name, dirsize, "%s" SEP_MAIL_FILE_TEMPLATE, dir, y, ""); // enough for 9 digits allocated above
00894 
00895         check_filename(dir_name);
00896         DEBUG_INFO(("about to try creating %s\n", dir_name));
00897         if (D_MKDIR(dir_name)) {
00898             if (errno != EEXIST) { // if there is an error, and it doesn't already exist
00899                 x = errno;
00900                 DIE(("mk_separate_dir: Cannot create directory %s: %s\n", dir, strerror(x)));
00901             }
00902         } else {
00903             break;
00904         }
00905         y++;
00906     } while (overwrite == 0);
00907 
00908     if (chdir(dir_name)) {
00909         x = errno;
00910         DIE(("mk_separate_dir: Cannot change to directory %s: %s\n", dir, strerror(x)));
00911     }
00912 
00913     if (overwrite) {
00914         // we should probably delete all files from this directory
00915 #if !defined(WIN32) && !defined(__CYGWIN__)
00916         DIR * sdir = NULL;
00917         struct dirent *dirent = NULL;
00918         struct stat filestat;
00919         if (!(sdir = opendir("./"))) {
00920             DEBUG_WARN(("mk_separate_dir: Cannot open dir \"%s\" for deletion of old contents\n", "./"));
00921         } else {
00922             while ((dirent = readdir(sdir))) {
00923                 if (lstat(dirent->d_name, &filestat) != -1)
00924                     if (S_ISREG(filestat.st_mode)) {
00925                         if (unlink(dirent->d_name)) {
00926                             y = errno;
00927                             DIE(("mk_separate_dir: unlink returned error on file %s: %s\n", dirent->d_name, strerror(y)));
00928                         }
00929                     }
00930             }
00931             closedir(sdir);     // cppcheck detected leak
00932         }
00933 #endif
00934     }
00935 
00936     DEBUG_RET();
00937 }
00938 
00939 
00940 int close_separate_dir() {
00941     int x;
00942     DEBUG_ENT("close_separate_dir");
00943     if (chdir("..")) {
00944         x = errno;
00945         DIE(("close_separate_dir: Cannot go up dir (..): %s\n", strerror(x)));
00946     }
00947     DEBUG_RET();
00948     return 0;
00949 }
00950 
00951 
00952 void mk_separate_file(struct file_ll *f, int32_t t, char *extension, int openit) {
00953     DEBUG_ENT("mk_separate_file");
00954     DEBUG_INFO(("opening next file to save email type %s\n", item_type_to_name(t)));
00955     if (f->item_count > 999999999) { // bigger than nine 9's
00956         DIE(("mk_separate_file: The number of emails in this folder has become too high to handle\n"));
00957     }
00958     sprintf(f->name[t], SEP_MAIL_FILE_TEMPLATE, f->item_count, extension);
00959     check_filename(f->name[t]);
00960     if (openit) {
00961         if (!(f->output[t] = fopen(f->name[t], "w"))) {
00962             DIE(("mk_separate_file: Cannot open file to save email \"%s\"\n", f->name[t]));
00963         }
00964     }
00965     DEBUG_RET();
00966 }
00967 
00968 
00969 void close_separate_file(struct file_ll *f) {
00970     int32_t t;
00971     DEBUG_ENT("close_separate_file");
00972     for (t=0; t<PST_TYPE_MAX; t++) {
00973         if (f->output[t]) {
00974             struct stat st;
00975             fclose(f->output[t]);
00976             stat(f->name[t], &st);
00977             if (!st.st_size) {
00978                 DEBUG_WARN(("removing empty output file %s\n", f->name[t]));
00979                 remove(f->name[t]);
00980             }
00981             f->output[t] = NULL;
00982         }
00983     }
00984     DEBUG_RET();
00985 }
00986 
00987 
00988 char *my_stristr(char *haystack, char *needle) {
00989     // my_stristr varies from strstr in that its searches are case-insensitive
00990     char *x=haystack, *y=needle, *z = NULL;
00991     if (!haystack || !needle) {
00992         return NULL;
00993     }
00994     while (*y != '\0' && *x != '\0') {
00995         if (tolower(*y) == tolower(*x)) {
00996             // move y on one
00997             y++;
00998             if (!z) {
00999                 z = x; // store first position in haystack where a match is made
01000             }
01001         } else {
01002             y = needle; // reset y to the beginning of the needle
01003             z = NULL; // reset the haystack storage point
01004         }
01005         x++; // advance the search in the haystack
01006     }
01007     // If the haystack ended before our search finished, it's not a match.
01008     if (*y != '\0') return NULL;
01009     return z;
01010 }
01011 
01012 
01013 void check_filename(char *fname) {
01014     char *t = fname;
01015     DEBUG_ENT("check_filename");
01016     if (!t) {
01017         DEBUG_RET();
01018         return;
01019     }
01020     while ((t = strpbrk(t, "/\\:"))) {
01021         // while there are characters in the second string that we don't want
01022         *t = '_'; //replace them with an underscore
01023     }
01024     DEBUG_RET();
01025 }
01026 
01027 
01034 int  acceptable_ext(pst_item_attach* attach)
01035 {
01036     if (!acceptable_extensions || *acceptable_extensions == '\0') return 1;     // acceptable list missing or empty
01037     char *attach_filename = (attach->filename2.str) ? attach->filename2.str
01038                                                     : attach->filename1.str;
01039     if (!attach_filename) return 1; // attachment with no name is always acceptable
01040     char *e = strrchr(attach_filename, '.');
01041     if (!e) return 1;               // attachment with no extension is always acceptable.
01042     DEBUG_ENT("acceptable_ext");
01043     DEBUG_INFO(("attachment extension %s\n", e));
01044     int rc = 0;
01045     char *a = acceptable_extensions;
01046     while (*a) {
01047         if (pst_stricmp(a, e) == 0) {
01048             rc = 1;
01049             break;
01050         }
01051         a += strlen(a) + 1;
01052     }
01053     DEBUG_INFO(("attachment acceptable returns %d\n", rc));
01054     DEBUG_RET();
01055     return rc;
01056 }
01057 
01058 
01059 void write_separate_attachment(char f_name[], pst_item_attach* attach, int attach_num, pst_file* pst)
01060 {
01061     FILE *fp = NULL;
01062     int x = 0;
01063     char *temp = NULL;
01064 
01065     // If there is a long filename (filename2) use that, otherwise
01066     // use the 8.3 filename (filename1)
01067     char *attach_filename = (attach->filename2.str) ? attach->filename2.str
01068                                                     : attach->filename1.str;
01069     DEBUG_ENT("write_separate_attachment");
01070     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));
01071 
01072     if (!attach->data.data) {
01073         // make sure we can fetch data from the id
01074         pst_index_ll *ptr = pst_getID(pst, attach->i_id);
01075         if (!ptr) {
01076             DEBUG_WARN(("Couldn't find i_id %#"PRIx64". Cannot save attachment to file\n", attach->i_id));
01077             DEBUG_RET();
01078             return;
01079         }
01080     }
01081 
01082     check_filename(f_name);
01083     if (!attach_filename) {
01084         // generate our own (dummy) filename for the attachement
01085         temp = pst_malloc(strlen(f_name)+15);
01086         sprintf(temp, "%s-attach%i", f_name, attach_num);
01087     } else {
01088         // have an attachment name, make sure it's unique
01089         temp = pst_malloc(strlen(f_name)+strlen(attach_filename)+15);
01090         do {
01091             if (fp) fclose(fp);
01092             if (x == 0)
01093                 sprintf(temp, "%s-%s", f_name, attach_filename);
01094             else
01095                 sprintf(temp, "%s-%s-%i", f_name, attach_filename, x);
01096         } while ((fp = fopen(temp, "r")) && ++x < 99999999);
01097         if (x > 99999999) {
01098             DIE(("error finding attachment name. exhausted possibilities to %s\n", temp));
01099         }
01100     }
01101     DEBUG_INFO(("Saving attachment to %s\n", temp));
01102     if (!(fp = fopen(temp, "w"))) {
01103         DEBUG_WARN(("write_separate_attachment: Cannot open attachment save file \"%s\"\n", temp));
01104     } else {
01105         (void)pst_attach_to_file(pst, attach, fp);
01106         fclose(fp);
01107     }
01108     if (temp) free(temp);
01109     DEBUG_RET();
01110 }
01111 
01112 
01113 void write_embedded_message(FILE* f_output, pst_item_attach* attach, char *boundary, pst_file* pf, int save_rtf, char** extra_mime_headers)
01114 {
01115     pst_index_ll *ptr;
01116     DEBUG_ENT("write_embedded_message");
01117     ptr = pst_getID(pf, attach->i_id);
01118 
01119     pst_desc_tree d_ptr;
01120     d_ptr.d_id        = 0;
01121     d_ptr.parent_d_id = 0;
01122     d_ptr.assoc_tree  = NULL;
01123     d_ptr.desc        = ptr;
01124     d_ptr.no_child    = 0;
01125     d_ptr.prev        = NULL;
01126     d_ptr.next        = NULL;
01127     d_ptr.parent      = NULL;
01128     d_ptr.child       = NULL;
01129     d_ptr.child_tail  = NULL;
01130 
01131     pst_item *item = pst_parse_item(pf, &d_ptr, attach->id2_head);
01132     // It appears that if the embedded message contains an appointment/
01133     // calendar item, pst_parse_item returns NULL due to the presence of
01134     // an unexpected reference type of 0x1048, which seems to represent
01135     // an array of GUIDs representing a CLSID. It's likely that this is
01136     // a reference to an internal Outlook COM class.
01137     //      Log the skipped item and continue on.
01138     if (!item) {
01139         DEBUG_WARN(("write_embedded_message: pst_parse_item was unable to parse the embedded message in attachment ID %llu", attach->i_id));
01140     } else {
01141         if (!item->email) {
01142             DEBUG_WARN(("write_embedded_message: pst_parse_item returned type %d, not an email message", item->type));
01143         } else {
01144             fprintf(f_output, "\n--%s\n", boundary);
01145             fprintf(f_output, "Content-Type: %s\n\n", attach->mimetype.str);
01146             write_normal_email(f_output, "", item, MODE_NORMAL, 0, pf, save_rtf, 1, extra_mime_headers);
01147         }
01148         pst_freeItem(item);
01149     }
01150 
01151     DEBUG_RET();
01152 }
01153 
01154 
01155 void write_inline_attachment(FILE* f_output, pst_item_attach* attach, char *boundary, pst_file* pst)
01156 {
01157     DEBUG_ENT("write_inline_attachment");
01158     DEBUG_INFO(("Attachment Size is %#"PRIx64", data = %#"PRIxPTR", id %#"PRIx64"\n", (uint64_t)attach->data.size, attach->data.data, attach->i_id));
01159 
01160     if (!attach->data.data) {
01161         // make sure we can fetch data from the id
01162         pst_index_ll *ptr = pst_getID(pst, attach->i_id);
01163         if (!ptr) {
01164             DEBUG_WARN(("Couldn't find ID pointer. Cannot save attachment to file\n"));
01165             DEBUG_RET();
01166             return;
01167         }
01168     }
01169 
01170     fprintf(f_output, "\n--%s\n", boundary);
01171     if (!attach->mimetype.str) {
01172         fprintf(f_output, "Content-Type: %s\n", MIME_TYPE_DEFAULT);
01173     } else {
01174         fprintf(f_output, "Content-Type: %s\n", attach->mimetype.str);
01175     }
01176     fprintf(f_output, "Content-Transfer-Encoding: base64\n");
01177 
01178     if (attach->content_id.str) {
01179         fprintf(f_output, "Content-ID: <%s>\n", attach->content_id.str);
01180     }
01181 
01182     if (attach->filename2.str) {
01183         // use the long filename, converted to proper encoding if needed.
01184         // it is already utf8
01185         pst_rfc2231(&attach->filename2);
01186         fprintf(f_output, "Content-Disposition: attachment; \n        filename*=%s\n\n", attach->filename2.str);
01187     }
01188     else if (attach->filename1.str) {
01189         // short filename never needs encoding
01190         fprintf(f_output, "Content-Disposition: attachment; filename=\"%s\"\n\n", attach->filename1.str);
01191     }
01192     else {
01193         // no filename is inline
01194         fprintf(f_output, "Content-Disposition: inline\n\n");
01195     }
01196 
01197     (void)pst_attach_to_file_base64(pst, attach, f_output);
01198     fprintf(f_output, "\n\n");
01199     DEBUG_RET();
01200 }
01201 
01202 
01203 int  header_match(char *header, char*field) {
01204     int n = strlen(field);
01205     if (strncasecmp(header, field, n) == 0) return 1;   // tag:{space}
01206     if ((field[n-1] == ' ') && (strncasecmp(header, field, n-1) == 0)) {
01207         char *crlftab = "\r\n\t";
01208         DEBUG_INFO(("Possible wrapped header = %s\n", header));
01209         if (strncasecmp(header+n-1, crlftab, 3) == 0) return 1; // tag:{cr}{lf}{tab}
01210     }
01211     return 0;
01212 }
01213 
01214 int  valid_headers(char *header)
01215 {
01216     // headers are sometimes really bogus - they seem to be fragments of the
01217     // message body, so we only use them if they seem to be real rfc822 headers.
01218     // this list is composed of ones that we have seen in real pst files.
01219     // there are surely others. the problem is - given an arbitrary character
01220     // string, is it a valid (or even reasonable) set of rfc822 headers?
01221     if (header) {
01222         if (header_match(header, "Content-Type: "                 )) return 1;
01223         if (header_match(header, "Date: "                         )) return 1;
01224         if (header_match(header, "From: "                         )) return 1;
01225         if (header_match(header, "MIME-Version: "                 )) return 1;
01226         if (header_match(header, "Microsoft Mail Internet Headers")) return 1;
01227         if (header_match(header, "Received: "                     )) return 1;
01228         if (header_match(header, "Return-Path: "                  )) return 1;
01229         if (header_match(header, "Subject: "                      )) return 1;
01230         if (header_match(header, "To: "                           )) return 1;
01231         if (header_match(header, "X-ASG-Debug-ID: "               )) return 1;
01232         if (header_match(header, "X-Barracuda-URL: "              )) return 1;
01233         if (header_match(header, "X-x: "                          )) return 1;
01234         if (strlen(header) > 2) {
01235             DEBUG_INFO(("Ignore bogus headers = %s\n", header));
01236         }
01237         return 0;
01238     }
01239     else return 0;
01240 }
01241 
01242 
01243 void header_has_field(char *header, char *field, int *flag)
01244 {
01245     DEBUG_ENT("header_has_field");
01246     if (my_stristr(header, field) || (strncasecmp(header, field+1, strlen(field)-1) == 0)) {
01247         DEBUG_INFO(("header block has %s header\n", field+1));
01248         *flag = 1;
01249     }
01250     DEBUG_RET();
01251 }
01252 
01253 
01254 void header_get_subfield(char *field, const char *subfield, char *body_subfield, size_t size_subfield)
01255 {
01256     if (!field) return;
01257     DEBUG_ENT("header_get_subfield");
01258     char search[60];
01259     snprintf(search, sizeof(search), " %s=", subfield);
01260     field++;
01261     char *n = header_end_field(field);
01262     char *s = my_stristr(field, search);
01263     if (n && s && (s < n)) {
01264         char *e, *f, save;
01265         s += strlen(search);    // skip over subfield=
01266         if (*s == '"') {
01267             s++;
01268             e = strchr(s, '"');
01269         }
01270         else {
01271             e = strchr(s, ';');
01272             f = strchr(s, '\n');
01273             if (e && f && (f < e)) e = f;
01274         }
01275         if (!e || (e > n)) e = n;   // use the trailing lf as terminator if nothing better
01276         save = *e;
01277         *e = '\0';
01278         snprintf(body_subfield, size_subfield, "%s", s);  // copy the subfield to our buffer
01279         *e = save;
01280         DEBUG_INFO(("body %s %s from headers\n", subfield, body_subfield));
01281     }
01282     DEBUG_RET();
01283 }
01284 
01285 char* header_get_field(char *header, char *field)
01286 {
01287     char *t = my_stristr(header, field);
01288     if (!t && (strncasecmp(header, field+1, strlen(field)-1) == 0)) t = header;
01289     return t;
01290 }
01291 
01292 
01293 // return pointer to \n at the end of this header field,
01294 // or NULL if this field goes to the end of the string.
01295 char *header_end_field(char *field)
01296 {
01297     char *e = strchr(field+1, '\n');
01298     while (e && ((e[1] == ' ') || (e[1] == '\t'))) {
01299         e = strchr(e+1, '\n');
01300     }
01301     return e;
01302 }
01303 
01304 
01305 void header_strip_field(char *header, char *field)
01306 {
01307     char *t;
01308     while ((t = header_get_field(header, field))) {
01309         char *e = header_end_field(t);
01310         if (e) {
01311             if (t == header) e++;   // if *t is not \n, we don't want to keep the \n at *e either.
01312             while (*e != '\0') {
01313                 *t = *e;
01314                 t++;
01315                 e++;
01316             }
01317             *t = '\0';
01318         }
01319         else {
01320             // this was the last header field, truncate the headers
01321             *t = '\0';
01322         }
01323     }
01324 }
01325 
01326 
01327 int  test_base64(char *body, size_t len)
01328 {
01329     int b64 = 0;
01330     uint8_t *b = (uint8_t *)body;
01331     DEBUG_ENT("test_base64");
01332     while (len--) {
01333         if ((*b < 32) && (*b != 9) && (*b != 10)) {
01334             DEBUG_INFO(("found base64 byte %d\n", (int)*b));
01335             DEBUG_HEXDUMPC(body, strlen(body), 0x10);
01336             b64 = 1;
01337             break;
01338         }
01339         b++;
01340     }
01341     DEBUG_RET();
01342     return b64;
01343 }
01344 
01345 
01346 void find_html_charset(char *html, char *charset, size_t charsetlen)
01347 {
01348     const int  index = 1;
01349     const int nmatch = index+1;
01350     regmatch_t match[nmatch];
01351     DEBUG_ENT("find_html_charset");
01352     int rc = regexec(&meta_charset_pattern, html, nmatch, match, 0);
01353     if (rc == 0) {
01354         int s = match[index].rm_so;
01355         int e = match[index].rm_eo;
01356         if (s != -1) {
01357             char save = html[e];
01358             html[e] = '\0';
01359                 snprintf(charset, charsetlen, "%s", html+s);    // copy the html charset
01360             html[e] = save;
01361             DEBUG_INFO(("charset %s from html text\n", charset));
01362         }
01363         else {
01364             DEBUG_INFO(("matching %d %d %d %d\n", match[0].rm_so, match[0].rm_eo, match[1].rm_so, match[1].rm_eo));
01365             DEBUG_HEXDUMPC(html, strlen(html), 0x10);
01366         }
01367     }
01368     else {
01369         DEBUG_INFO(("regexec returns %d\n", rc));
01370     }
01371     DEBUG_RET();
01372 }
01373 
01374 
01375 void find_rfc822_headers(char** extra_mime_headers)
01376 {
01377     DEBUG_ENT("find_rfc822_headers");
01378     char *headers = *extra_mime_headers;
01379     if (headers) {
01380         char *temp, *t;
01381         while ((temp = strstr(headers, "\n\n"))) {
01382             temp[1] = '\0';
01383             t = header_get_field(headers, "\nContent-Type:");
01384             if (t) {
01385                 t++;
01386                 DEBUG_INFO(("found content type header\n"));
01387                 char *n = strchr(t, '\n');
01388                 char *s = strstr(t, ": ");
01389                 char *e = strchr(t, ';');
01390                 if (!e || (e > n)) e = n;
01391                 if (s && (s < e)) {
01392                     s += 2;
01393                     if (!strncasecmp(s, RFC822, e-s)) {
01394                         headers = temp+2;   // found rfc822 header
01395                         DEBUG_INFO(("found 822 headers\n%s\n", headers));
01396                         break;
01397                     }
01398                 }
01399             }
01400             //DEBUG_INFO(("skipping to next block after\n%s\n", headers));
01401             headers = temp+2;   // skip to next chunk of headers
01402         }
01403         *extra_mime_headers = headers;
01404     }
01405     DEBUG_RET();
01406 }
01407 
01408 
01409 void write_body_part(FILE* f_output, pst_string *body, char *mime, char *charset, char *boundary, pst_file* pst)
01410 {
01411     DEBUG_ENT("write_body_part");
01412     removeCR(body->str);
01413     size_t body_len = strlen(body->str);
01414 
01415     if (body->is_utf8 && (strcasecmp("utf-8", charset))) {
01416         if (prefer_utf8) {
01417             charset = "utf-8";
01418         } else {
01419             // try to convert to the specified charset since the target
01420             // is not utf-8, and the data came from a unicode (utf16) field
01421             // and is now in utf-8.
01422             size_t rc;
01423             DEBUG_INFO(("Convert %s utf-8 to %s\n", mime, charset));
01424             pst_vbuf *newer = pst_vballoc(2);
01425             rc = pst_vb_utf8to8bit(newer, body->str, body_len, charset);
01426             if (rc == (size_t)-1) {
01427                 // unable to convert, change the charset to utf8
01428                 free(newer->b);
01429                 DEBUG_INFO(("Failed to convert %s utf-8 to %s\n", mime, charset));
01430                 charset = "utf-8";
01431             } else {
01432                 // null terminate the output string
01433                 pst_vbgrow(newer, 1);
01434                 newer->b[newer->dlen] = '\0';
01435                 free(body->str);
01436                 body->str = newer->b;
01437                 body_len = newer->dlen;
01438             }
01439             free(newer);
01440         }
01441     }
01442     int base64 = test_base64(body->str, body_len);
01443     fprintf(f_output, "\n--%s\n", boundary);
01444     fprintf(f_output, "Content-Type: %s; charset=\"%s\"\n", mime, charset);
01445     if (base64) fprintf(f_output, "Content-Transfer-Encoding: base64\n");
01446     fprintf(f_output, "\n");
01447     // Any body that uses an encoding with NULLs, e.g. UTF16, will be base64-encoded here.
01448     if (base64) {
01449         char *enc = pst_base64_encode(body->str, body_len);
01450         if (enc) {
01451             write_email_body(f_output, enc);
01452             fprintf(f_output, "\n");
01453             free(enc);
01454         }
01455     }
01456     else {
01457         write_email_body(f_output, body->str);
01458     }
01459     DEBUG_RET();
01460 }
01461 
01462 
01463 void write_schedule_part_data(FILE* f_output, pst_item* item, const char* sender, const char* method)
01464 {
01465     fprintf(f_output, "BEGIN:VCALENDAR\n");
01466     fprintf(f_output, "VERSION:2.0\n");
01467     fprintf(f_output, "PRODID:LibPST v%s\n", VERSION);
01468     if (method) fprintf(f_output, "METHOD:%s\n", method);
01469     fprintf(f_output, "BEGIN:VEVENT\n");
01470     if (sender) {
01471         if (item->email->outlook_sender_name.str) {
01472             fprintf(f_output, "ORGANIZER;CN=\"%s\":MAILTO:%s\n", item->email->outlook_sender_name.str, sender);
01473         } else {
01474             fprintf(f_output, "ORGANIZER;CN=\"\":MAILTO:%s\n", sender);
01475         }
01476     }
01477     write_appointment(f_output, item);
01478     fprintf(f_output, "END:VCALENDAR\n");
01479 }
01480 
01481 
01482 void write_schedule_part(FILE* f_output, pst_item* item, const char* sender, const char* boundary)
01483 {
01484     const char* method  = "REQUEST";
01485     const char* charset = "utf-8";
01486     char fname[30];
01487     if (!item->appointment) return;
01488 
01489     // inline appointment request
01490     fprintf(f_output, "\n--%s\n", boundary);
01491     fprintf(f_output, "Content-Type: %s; method=\"%s\"; charset=\"%s\"\n\n", "text/calendar", method, charset);
01492     write_schedule_part_data(f_output, item, sender, method);
01493     fprintf(f_output, "\n");
01494 
01495     // attachment appointment request
01496     snprintf(fname, sizeof(fname), "i%i.ics", rand());
01497     fprintf(f_output, "\n--%s\n", boundary);
01498     fprintf(f_output, "Content-Type: %s; charset=\"%s\"; name=\"%s\"\n", "text/calendar", "utf-8", fname);
01499     fprintf(f_output, "Content-Disposition: attachment; filename=\"%s\"\n\n", fname);
01500     write_schedule_part_data(f_output, item, sender, method);
01501     fprintf(f_output, "\n");
01502 }
01503 
01504 
01505 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)
01506 {
01507     char boundary[60];
01508     char altboundary[66];
01509     char *altboundaryp = NULL;
01510     char body_charset[30];
01511     char buffer_charset[30];
01512     char body_report[60];
01513     char sender[60];
01514     int  sender_known = 0;
01515     char *temp = NULL;
01516     time_t em_time;
01517     char *c_time;
01518     char *headers = NULL;
01519     int has_from, has_subject, has_to, has_cc, has_date, has_msgid;
01520     has_from = has_subject = has_to = has_cc = has_date = has_msgid = 0;
01521     DEBUG_ENT("write_normal_email");
01522 
01523     pst_convert_utf8_null(item, &item->email->header);
01524     headers = valid_headers(item->email->header.str) ? item->email->header.str :
01525               valid_headers(*extra_mime_headers)     ? *extra_mime_headers     :
01526               NULL;
01527 
01528     // setup default body character set and report type
01529     strncpy(body_charset, pst_default_charset(item, sizeof(buffer_charset), buffer_charset), sizeof(body_charset));
01530     body_charset[sizeof(body_charset)-1] = '\0';
01531     strncpy(body_report, "delivery-status", sizeof(body_report));
01532     body_report[sizeof(body_report)-1] = '\0';
01533 
01534     // setup default sender
01535     pst_convert_utf8(item, &item->email->sender_address);
01536     if (item->email->sender_address.str && strchr(item->email->sender_address.str, '@')) {
01537         temp = item->email->sender_address.str;
01538         sender_known = 1;
01539     }
01540     else {
01541         temp = "MAILER-DAEMON";
01542     }
01543     strncpy(sender, temp, sizeof(sender));
01544     sender[sizeof(sender)-1] = '\0';
01545 
01546     // convert the sent date if it exists, or set it to a fixed date
01547     if (item->email->sent_date) {
01548         em_time = pst_fileTimeToUnixTime(item->email->sent_date);
01549         c_time = ctime(&em_time);
01550         if (c_time)
01551             c_time[strlen(c_time)-1] = '\0'; //remove end \n
01552         else
01553             c_time = "Thu Jan 1 00:00:00 1970";
01554     } else
01555         c_time = "Thu Jan 1 00:00:00 1970";
01556 
01557     // create our MIME boundaries here.
01558     snprintf(boundary, sizeof(boundary), "--boundary-LibPST-iamunique-%i_-_-", rand());
01559     snprintf(altboundary, sizeof(altboundary), "alt-%s", boundary);
01560 
01561     // we will always look at the headers to discover some stuff
01562     if (headers ) {
01563         char *t;
01564         removeCR(headers);
01565 
01566         temp = strstr(headers, "\n\n");
01567         if (temp) {
01568             // cut off our real rfc822 headers here
01569             temp[1] = '\0';
01570             // pointer to all the embedded MIME headers.
01571             // we use these to find the actual rfc822 headers for embedded message/rfc822 mime parts
01572             // but only for the outermost message
01573             if (!*extra_mime_headers) *extra_mime_headers = temp+2;
01574             DEBUG_INFO(("Found extra mime headers\n%s\n", temp+2));
01575         }
01576 
01577         // Check if the headers have all the necessary fields
01578         header_has_field(headers, "\nFrom:",        &has_from);
01579         header_has_field(headers, "\nTo:",          &has_to);
01580         header_has_field(headers, "\nSubject:",     &has_subject);
01581         header_has_field(headers, "\nDate:",        &has_date);
01582         header_has_field(headers, "\nCC:",          &has_cc);
01583         header_has_field(headers, "\nMessage-Id:",  &has_msgid);
01584 
01585         // look for charset and report-type in Content-Type header
01586         t = header_get_field(headers, "\nContent-Type:");
01587         header_get_subfield(t, "charset", body_charset, sizeof(body_charset));
01588         header_get_subfield(t, "report-type", body_report, sizeof(body_report));
01589 
01590         // derive a proper sender email address
01591         if (!sender_known) {
01592             t = header_get_field(headers, "\nFrom:");
01593             if (t) {
01594                 // assume address is on the first line, rather than on a continuation line
01595                 t++;
01596                 char *n = strchr(t, '\n');
01597                 char *s = strchr(t, '<');
01598                 char *e = strchr(t, '>');
01599                 if (s && e && n && (s < e) && (e < n)) {
01600                 char save = *e;
01601                 *e = '\0';
01602                     snprintf(sender, sizeof(sender), "%s", s+1);
01603                 *e = save;
01604                 }
01605             }
01606         }
01607 
01608         // Strip out the mime headers and some others that we don't want to emit
01609         header_strip_field(headers, "\nMicrosoft Mail Internet Headers");
01610         header_strip_field(headers, "\nMIME-Version:");
01611         header_strip_field(headers, "\nContent-Type:");
01612         header_strip_field(headers, "\nContent-Transfer-Encoding:");
01613         header_strip_field(headers, "\nContent-class:");
01614         header_strip_field(headers, "\nX-MimeOLE:");
01615         header_strip_field(headers, "\nX-From_:");
01616     }
01617 
01618     DEBUG_INFO(("About to print Header\n"));
01619 
01620     if (item && item->subject.str) {
01621         pst_convert_utf8(item, &item->subject);
01622         DEBUG_INFO(("item->subject = %s\n", item->subject.str));
01623     }
01624 
01625     if (mode != MODE_SEPARATE) {
01626         // most modes need this separator line.
01627         // procmail produces this separator without the quotes around the
01628         // sender email address, but apparently some Mac email client needs
01629         // those quotes, and they don't seem to cause problems for anyone else.
01630         char *quo = (embedding) ? ">" : "";
01631         fprintf(f_output, "%sFrom \"%s\" %s\n", quo, sender, c_time);
01632     }
01633 
01634     // print the supplied email headers
01635     if (headers) {
01636         int len = strlen(headers);
01637         if (len > 0) {
01638             fprintf(f_output, "%s", headers);
01639             // make sure the headers end with a \n
01640             if (headers[len-1] != '\n') fprintf(f_output, "\n");
01641             //char *h = headers;
01642             //while (*h) {
01643             //    char *e = strchr(h, '\n');
01644             //    int   d = 1;    // normally e points to trailing \n
01645             //    if (!e) {
01646             //        e = h + strlen(h);  // e points to trailing null
01647             //        d = 0;
01648             //    }
01649             //    // we could do rfc2047 encoding here if needed
01650             //    fprintf(f_output, "%.*s\n", (int)(e-h), h);
01651             //    h = e + d;
01652             //}
01653         }
01654     }
01655 
01656     // record read status
01657     if ((item->flags & PST_FLAG_READ) == PST_FLAG_READ) {
01658         fprintf(f_output, "Status: RO\n");
01659     }
01660 
01661     // create required header fields that are not already written
01662 
01663     if (!has_from) {
01664         if (item->email->outlook_sender_name.str){
01665             pst_rfc2047(item, &item->email->outlook_sender_name, 1);
01666             fprintf(f_output, "From: %s <%s>\n", item->email->outlook_sender_name.str, sender);
01667         } else {
01668             fprintf(f_output, "From: <%s>\n", sender);
01669         }
01670     }
01671 
01672     if (!has_subject) {
01673         if (item->subject.str) {
01674             pst_rfc2047(item, &item->subject, 0);
01675             fprintf(f_output, "Subject: %s\n", item->subject.str);
01676         } else {
01677             fprintf(f_output, "Subject: \n");
01678         }
01679     }
01680 
01681     if (!has_to && item->email->sentto_address.str) {
01682         pst_rfc2047(item, &item->email->sentto_address, 0);
01683         fprintf(f_output, "To: %s\n", item->email->sentto_address.str);
01684     }
01685 
01686     if (!has_cc && item->email->cc_address.str) {
01687         pst_rfc2047(item, &item->email->cc_address, 0);
01688         fprintf(f_output, "Cc: %s\n", item->email->cc_address.str);
01689     }
01690 
01691     if (!has_date && item->email->sent_date) {
01692         char c_time[C_TIME_SIZE];
01693         struct tm stm;
01694         gmtime_r(&em_time, &stm);
01695         strftime(c_time, C_TIME_SIZE, "%a, %d %b %Y %H:%M:%S %z", &stm);
01696         fprintf(f_output, "Date: %s\n", c_time);
01697     }
01698 
01699     if (!has_msgid && item->email->messageid.str) {
01700         pst_convert_utf8(item, &item->email->messageid);
01701         fprintf(f_output, "Message-Id: %s\n", item->email->messageid.str);
01702     }
01703 
01704     // add forensic headers to capture some .pst stuff that is not really
01705     // needed or used by mail clients
01706     pst_convert_utf8_null(item, &item->email->sender_address);
01707     if (item->email->sender_address.str && !strchr(item->email->sender_address.str, '@')
01708                                         && strcmp(item->email->sender_address.str, ".")
01709                                         && (strlen(item->email->sender_address.str) > 0)) {
01710         fprintf(f_output, "X-libpst-forensic-sender: %s\n", item->email->sender_address.str);
01711     }
01712 
01713     if (item->email->bcc_address.str) {
01714         pst_convert_utf8(item, &item->email->bcc_address);
01715         fprintf(f_output, "X-libpst-forensic-bcc: %s\n", item->email->bcc_address.str);
01716     }
01717 
01718     // add our own mime headers
01719     fprintf(f_output, "MIME-Version: 1.0\n");
01720     if (item->type == PST_TYPE_REPORT) {
01721         // multipart/report for DSN/MDN reports
01722         fprintf(f_output, "Content-Type: multipart/report; report-type=%s;\n\tboundary=\"%s\"\n", body_report, boundary);
01723     }
01724     else {
01725         fprintf(f_output, "Content-Type: multipart/mixed;\n\tboundary=\"%s\"\n", boundary);
01726     }
01727     fprintf(f_output, "\n");    // end of headers, start of body
01728 
01729     // now dump the body parts
01730     if ((item->type == PST_TYPE_REPORT) && (item->email->report_text.str)) {
01731         write_body_part(f_output, &item->email->report_text, "text/plain", body_charset, boundary, pst);
01732         fprintf(f_output, "\n");
01733     }
01734 
01735     if (item->body.str && item->email->htmlbody.str) {
01736         // start the nested alternative part
01737         fprintf(f_output, "\n--%s\n", boundary);
01738         fprintf(f_output, "Content-Type: multipart/alternative;\n\tboundary=\"%s\"\n", altboundary);
01739         altboundaryp = altboundary;
01740     }
01741     else {
01742         altboundaryp = boundary;
01743     }
01744 
01745     if (item->body.str) {
01746         write_body_part(f_output, &item->body, "text/plain", body_charset, altboundaryp, pst);
01747     }
01748 
01749     if (item->email->htmlbody.str) {
01750         find_html_charset(item->email->htmlbody.str, body_charset, sizeof(body_charset));
01751         write_body_part(f_output, &item->email->htmlbody, "text/html", body_charset, altboundaryp, pst);
01752     }
01753 
01754     if (item->body.str && item->email->htmlbody.str) {
01755         // end the nested alternative part
01756         fprintf(f_output, "\n--%s--\n", altboundary);
01757     }
01758 
01759     if (item->email->rtf_compressed.data && save_rtf) {
01760         pst_item_attach* attach = (pst_item_attach*)pst_malloc(sizeof(pst_item_attach));
01761         DEBUG_INFO(("Adding RTF body as attachment\n"));
01762         memset(attach, 0, sizeof(pst_item_attach));
01763         attach->next = item->attach;
01764         item->attach = attach;
01765         attach->data.data         = pst_lzfu_decompress(item->email->rtf_compressed.data, item->email->rtf_compressed.size, &attach->data.size);
01766         attach->filename2.str     = strdup(RTF_ATTACH_NAME);
01767         attach->filename2.is_utf8 = 1;
01768         attach->mimetype.str      = strdup(RTF_ATTACH_TYPE);
01769         attach->mimetype.is_utf8  = 1;
01770     }
01771 
01772     if (item->email->encrypted_body.data) {
01773         pst_item_attach* attach = (pst_item_attach*)pst_malloc(sizeof(pst_item_attach));
01774         DEBUG_INFO(("Adding encrypted text body as attachment\n"));
01775         memset(attach, 0, sizeof(pst_item_attach));
01776         attach->next = item->attach;
01777         item->attach = attach;
01778         attach->data.data = item->email->encrypted_body.data;
01779         attach->data.size = item->email->encrypted_body.size;
01780         item->email->encrypted_body.data = NULL;
01781     }
01782 
01783     if (item->email->encrypted_htmlbody.data) {
01784         pst_item_attach* attach = (pst_item_attach*)pst_malloc(sizeof(pst_item_attach));
01785         DEBUG_INFO(("Adding encrypted HTML body as attachment\n"));
01786         memset(attach, 0, sizeof(pst_item_attach));
01787         attach->next = item->attach;
01788         item->attach = attach;
01789         attach->data.data = item->email->encrypted_htmlbody.data;
01790         attach->data.size = item->email->encrypted_htmlbody.size;
01791         item->email->encrypted_htmlbody.data = NULL;
01792     }
01793 
01794     if (item->type == PST_TYPE_SCHEDULE) {
01795         write_schedule_part(f_output, item, sender, boundary);
01796     }
01797 
01798     // other attachments
01799     {
01800         pst_item_attach* attach;
01801         int attach_num = 0;
01802         for (attach = item->attach; attach; attach = attach->next) {
01803             pst_convert_utf8_null(item, &attach->filename1);
01804             pst_convert_utf8_null(item, &attach->filename2);
01805             pst_convert_utf8_null(item, &attach->mimetype);
01806             DEBUG_INFO(("Attempting Attachment encoding\n"));
01807             if (attach->method == PST_ATTACH_EMBEDDED) {
01808                 DEBUG_INFO(("have an embedded rfc822 message attachment\n"));
01809                 if (attach->mimetype.str) {
01810                     DEBUG_INFO(("which already has a mime-type of %s\n", attach->mimetype.str));
01811                     free(attach->mimetype.str);
01812                 }
01813                 attach->mimetype.str = strdup(RFC822);
01814                 attach->mimetype.is_utf8 = 1;
01815                 find_rfc822_headers(extra_mime_headers);
01816                 write_embedded_message(f_output, attach, boundary, pst, save_rtf, extra_mime_headers);
01817             }
01818             else if (attach->data.data || attach->i_id) {
01819                 if (acceptable_ext(attach)) {
01820                     if (mode == MODE_SEPARATE && !mode_MH)
01821                         write_separate_attachment(f_name, attach, ++attach_num, pst);
01822                     else
01823                         write_inline_attachment(f_output, attach, boundary, pst);
01824                 }
01825             }
01826         }
01827     }
01828 
01829     fprintf(f_output, "\n--%s--\n\n", boundary);
01830     DEBUG_RET();
01831 }
01832 
01833 
01834 void write_vcard(FILE* f_output, pst_item* item, pst_item_contact* contact, char comment[])
01835 {
01836     char*  result = NULL;
01837     size_t resultlen = 0;
01838     char   time_buffer[30];
01839     // We can only call rfc escape once per printf, since the second call
01840     // may free the buffer returned by the first call.
01841     // I had tried to place those into a single printf - Carl.
01842 
01843     DEBUG_ENT("write_vcard");
01844 
01845     // make everything utf8
01846     pst_convert_utf8_null(item, &contact->fullname);
01847     pst_convert_utf8_null(item, &contact->surname);
01848     pst_convert_utf8_null(item, &contact->first_name);
01849     pst_convert_utf8_null(item, &contact->middle_name);
01850     pst_convert_utf8_null(item, &contact->display_name_prefix);
01851     pst_convert_utf8_null(item, &contact->suffix);
01852     pst_convert_utf8_null(item, &contact->nickname);
01853     pst_convert_utf8_null(item, &contact->address1);
01854     pst_convert_utf8_null(item, &contact->address2);
01855     pst_convert_utf8_null(item, &contact->address3);
01856     pst_convert_utf8_null(item, &contact->home_po_box);
01857     pst_convert_utf8_null(item, &contact->home_street);
01858     pst_convert_utf8_null(item, &contact->home_city);
01859     pst_convert_utf8_null(item, &contact->home_state);
01860     pst_convert_utf8_null(item, &contact->home_postal_code);
01861     pst_convert_utf8_null(item, &contact->home_country);
01862     pst_convert_utf8_null(item, &contact->home_address);
01863     pst_convert_utf8_null(item, &contact->business_po_box);
01864     pst_convert_utf8_null(item, &contact->business_street);
01865     pst_convert_utf8_null(item, &contact->business_city);
01866     pst_convert_utf8_null(item, &contact->business_state);
01867     pst_convert_utf8_null(item, &contact->business_postal_code);
01868     pst_convert_utf8_null(item, &contact->business_country);
01869     pst_convert_utf8_null(item, &contact->business_address);
01870     pst_convert_utf8_null(item, &contact->other_po_box);
01871     pst_convert_utf8_null(item, &contact->other_street);
01872     pst_convert_utf8_null(item, &contact->other_city);
01873     pst_convert_utf8_null(item, &contact->other_state);
01874     pst_convert_utf8_null(item, &contact->other_postal_code);
01875     pst_convert_utf8_null(item, &contact->other_country);
01876     pst_convert_utf8_null(item, &contact->other_address);
01877     pst_convert_utf8_null(item, &contact->business_fax);
01878     pst_convert_utf8_null(item, &contact->business_phone);
01879     pst_convert_utf8_null(item, &contact->business_phone2);
01880     pst_convert_utf8_null(item, &contact->car_phone);
01881     pst_convert_utf8_null(item, &contact->home_fax);
01882     pst_convert_utf8_null(item, &contact->home_phone);
01883     pst_convert_utf8_null(item, &contact->home_phone2);
01884     pst_convert_utf8_null(item, &contact->isdn_phone);
01885     pst_convert_utf8_null(item, &contact->mobile_phone);
01886     pst_convert_utf8_null(item, &contact->other_phone);
01887     pst_convert_utf8_null(item, &contact->pager_phone);
01888     pst_convert_utf8_null(item, &contact->primary_fax);
01889     pst_convert_utf8_null(item, &contact->primary_phone);
01890     pst_convert_utf8_null(item, &contact->radio_phone);
01891     pst_convert_utf8_null(item, &contact->telex);
01892     pst_convert_utf8_null(item, &contact->job_title);
01893     pst_convert_utf8_null(item, &contact->profession);
01894     pst_convert_utf8_null(item, &contact->assistant_name);
01895     pst_convert_utf8_null(item, &contact->assistant_phone);
01896     pst_convert_utf8_null(item, &contact->company_name);
01897     pst_convert_utf8_null(item, &item->body);
01898 
01899     // the specification I am following is (hopefully) RFC2426 vCard Mime Directory Profile
01900     fprintf(f_output, "BEGIN:VCARD\n");
01901     fprintf(f_output, "FN:%s\n", pst_rfc2426_escape(contact->fullname.str, &result, &resultlen));
01902 
01903     //fprintf(f_output, "N:%s;%s;%s;%s;%s\n",
01904     fprintf(f_output, "N:%s;", (!contact->surname.str)             ? "" : pst_rfc2426_escape(contact->surname.str, &result, &resultlen));
01905     fprintf(f_output, "%s;",   (!contact->first_name.str)          ? "" : pst_rfc2426_escape(contact->first_name.str, &result, &resultlen));
01906     fprintf(f_output, "%s;",   (!contact->middle_name.str)         ? "" : pst_rfc2426_escape(contact->middle_name.str, &result, &resultlen));
01907     fprintf(f_output, "%s;",   (!contact->display_name_prefix.str) ? "" : pst_rfc2426_escape(contact->display_name_prefix.str, &result, &resultlen));
01908     fprintf(f_output, "%s\n",  (!contact->suffix.str)              ? "" : pst_rfc2426_escape(contact->suffix.str, &result, &resultlen));
01909 
01910     if (contact->nickname.str)
01911         fprintf(f_output, "NICKNAME:%s\n", pst_rfc2426_escape(contact->nickname.str, &result, &resultlen));
01912     if (contact->address1.str)
01913         fprintf(f_output, "EMAIL:%s\n", pst_rfc2426_escape(contact->address1.str, &result, &resultlen));
01914     if (contact->address2.str)
01915         fprintf(f_output, "EMAIL:%s\n", pst_rfc2426_escape(contact->address2.str, &result, &resultlen));
01916     if (contact->address3.str)
01917         fprintf(f_output, "EMAIL:%s\n", pst_rfc2426_escape(contact->address3.str, &result, &resultlen));
01918     if (contact->birthday)
01919         fprintf(f_output, "BDAY:%s\n", pst_rfc2425_datetime_format(contact->birthday, sizeof(time_buffer), time_buffer));
01920 
01921     if (contact->home_address.str) {
01922         //fprintf(f_output, "ADR;TYPE=home:%s;%s;%s;%s;%s;%s;%s\n",
01923         fprintf(f_output, "ADR;TYPE=home:%s;",  (!contact->home_po_box.str)      ? "" : pst_rfc2426_escape(contact->home_po_box.str, &result, &resultlen));
01924         fprintf(f_output, "%s;",                ""); // extended Address
01925         fprintf(f_output, "%s;",                (!contact->home_street.str)      ? "" : pst_rfc2426_escape(contact->home_street.str, &result, &resultlen));
01926         fprintf(f_output, "%s;",                (!contact->home_city.str)        ? "" : pst_rfc2426_escape(contact->home_city.str, &result, &resultlen));
01927         fprintf(f_output, "%s;",                (!contact->home_state.str)       ? "" : pst_rfc2426_escape(contact->home_state.str, &result, &resultlen));
01928         fprintf(f_output, "%s;",                (!contact->home_postal_code.str) ? "" : pst_rfc2426_escape(contact->home_postal_code.str, &result, &resultlen));
01929         fprintf(f_output, "%s\n",               (!contact->home_country.str)     ? "" : pst_rfc2426_escape(contact->home_country.str, &result, &resultlen));
01930         fprintf(f_output, "LABEL;TYPE=home:%s\n", pst_rfc2426_escape(contact->home_address.str, &result, &resultlen));
01931     }
01932 
01933     if (contact->business_address.str) {
01934         //fprintf(f_output, "ADR;TYPE=work:%s;%s;%s;%s;%s;%s;%s\n",
01935         fprintf(f_output, "ADR;TYPE=work:%s;",  (!contact->business_po_box.str)      ? "" : pst_rfc2426_escape(contact->business_po_box.str, &result, &resultlen));
01936         fprintf(f_output, "%s;",                ""); // extended Address
01937         fprintf(f_output, "%s;",                (!contact->business_street.str)      ? "" : pst_rfc2426_escape(contact->business_street.str, &result, &resultlen));
01938         fprintf(f_output, "%s;",                (!contact->business_city.str)        ? "" : pst_rfc2426_escape(contact->business_city.str, &result, &resultlen));
01939         fprintf(f_output, "%s;",                (!contact->business_state.str)       ? "" : pst_rfc2426_escape(contact->business_state.str, &result, &resultlen));
01940         fprintf(f_output, "%s;",                (!contact->business_postal_code.str) ? "" : pst_rfc2426_escape(contact->business_postal_code.str, &result, &resultlen));
01941         fprintf(f_output, "%s\n",               (!contact->business_country.str)     ? "" : pst_rfc2426_escape(contact->business_country.str, &result, &resultlen));
01942         fprintf(f_output, "LABEL;TYPE=work:%s\n", pst_rfc2426_escape(contact->business_address.str, &result, &resultlen));
01943     }
01944 
01945     if (contact->other_address.str) {
01946         //fprintf(f_output, "ADR;TYPE=postal:%s;%s;%s;%s;%s;%s;%s\n",
01947         fprintf(f_output, "ADR;TYPE=postal:%s;",(!contact->other_po_box.str)       ? "" : pst_rfc2426_escape(contact->other_po_box.str, &result, &resultlen));
01948         fprintf(f_output, "%s;",                ""); // extended Address
01949         fprintf(f_output, "%s;",                (!contact->other_street.str)       ? "" : pst_rfc2426_escape(contact->other_street.str, &result, &resultlen));
01950         fprintf(f_output, "%s;",                (!contact->other_city.str)         ? "" : pst_rfc2426_escape(contact->other_city.str, &result, &resultlen));
01951         fprintf(f_output, "%s;",                (!contact->other_state.str)        ? "" : pst_rfc2426_escape(contact->other_state.str, &result, &resultlen));
01952         fprintf(f_output, "%s;",                (!contact->other_postal_code.str)  ? "" : pst_rfc2426_escape(contact->other_postal_code.str, &result, &resultlen));
01953         fprintf(f_output, "%s\n",               (!contact->other_country.str)      ? "" : pst_rfc2426_escape(contact->other_country.str, &result, &resultlen));
01954         fprintf(f_output, "LABEL;TYPE=postal:%s\n", pst_rfc2426_escape(contact->other_address.str, &result, &resultlen));
01955     }
01956 
01957     if (contact->business_fax.str)      fprintf(f_output, "TEL;TYPE=work,fax:%s\n",         pst_rfc2426_escape(contact->business_fax.str, &result, &resultlen));
01958     if (contact->business_phone.str)    fprintf(f_output, "TEL;TYPE=work,voice:%s\n",       pst_rfc2426_escape(contact->business_phone.str, &result, &resultlen));
01959     if (contact->business_phone2.str)   fprintf(f_output, "TEL;TYPE=work,voice:%s\n",       pst_rfc2426_escape(contact->business_phone2.str, &result, &resultlen));
01960     if (contact->car_phone.str)         fprintf(f_output, "TEL;TYPE=car,voice:%s\n",        pst_rfc2426_escape(contact->car_phone.str, &result, &resultlen));
01961     if (contact->home_fax.str)          fprintf(f_output, "TEL;TYPE=home,fax:%s\n",         pst_rfc2426_escape(contact->home_fax.str, &result, &resultlen));
01962     if (contact->home_phone.str)        fprintf(f_output, "TEL;TYPE=home,voice:%s\n",       pst_rfc2426_escape(contact->home_phone.str, &result, &resultlen));
01963     if (contact->home_phone2.str)       fprintf(f_output, "TEL;TYPE=home,voice:%s\n",       pst_rfc2426_escape(contact->home_phone2.str, &result, &resultlen));
01964     if (contact->isdn_phone.str)        fprintf(f_output, "TEL;TYPE=isdn:%s\n",             pst_rfc2426_escape(contact->isdn_phone.str, &result, &resultlen));
01965     if (contact->mobile_phone.str)      fprintf(f_output, "TEL;TYPE=cell,voice:%s\n",       pst_rfc2426_escape(contact->mobile_phone.str, &result, &resultlen));
01966     if (contact->other_phone.str)       fprintf(f_output, "TEL;TYPE=msg:%s\n",              pst_rfc2426_escape(contact->other_phone.str, &result, &resultlen));
01967     if (contact->pager_phone.str)       fprintf(f_output, "TEL;TYPE=pager:%s\n",            pst_rfc2426_escape(contact->pager_phone.str, &result, &resultlen));
01968     if (contact->primary_fax.str)       fprintf(f_output, "TEL;TYPE=fax,pref:%s\n",         pst_rfc2426_escape(contact->primary_fax.str, &result, &resultlen));
01969     if (contact->primary_phone.str)     fprintf(f_output, "TEL;TYPE=phone,pref:%s\n",       pst_rfc2426_escape(contact->primary_phone.str, &result, &resultlen));
01970     if (contact->radio_phone.str)       fprintf(f_output, "TEL;TYPE=pcs:%s\n",              pst_rfc2426_escape(contact->radio_phone.str, &result, &resultlen));
01971     if (contact->telex.str)             fprintf(f_output, "TEL;TYPE=bbs:%s\n",              pst_rfc2426_escape(contact->telex.str, &result, &resultlen));
01972     if (contact->job_title.str)         fprintf(f_output, "TITLE:%s\n",                     pst_rfc2426_escape(contact->job_title.str, &result, &resultlen));
01973     if (contact->profession.str)        fprintf(f_output, "ROLE:%s\n",                      pst_rfc2426_escape(contact->profession.str, &result, &resultlen));
01974     if (contact->assistant_name.str || contact->assistant_phone.str) {
01975         fprintf(f_output, "AGENT:BEGIN:VCARD\n");
01976         if (contact->assistant_name.str)    fprintf(f_output, "FN:%s\n",                    pst_rfc2426_escape(contact->assistant_name.str, &result, &resultlen));
01977         if (contact->assistant_phone.str)   fprintf(f_output, "TEL:%s\n",                   pst_rfc2426_escape(contact->assistant_phone.str, &result, &resultlen));
01978     }
01979     if (contact->company_name.str)      fprintf(f_output, "ORG:%s\n",                       pst_rfc2426_escape(contact->company_name.str, &result, &resultlen));
01980     if (comment)                        fprintf(f_output, "NOTE:%s\n",                      pst_rfc2426_escape(comment, &result, &resultlen));
01981     if (item->body.str)                 fprintf(f_output, "NOTE:%s\n",                      pst_rfc2426_escape(item->body.str, &result, &resultlen));
01982 
01983     write_extra_categories(f_output, item);
01984 
01985     fprintf(f_output, "VERSION: 3.0\n");
01986     fprintf(f_output, "END:VCARD\n\n");
01987     if (result) free(result);
01988     DEBUG_RET();
01989 }
01990 
01991 
01999 int write_extra_categories(FILE* f_output, pst_item* item)
02000 {
02001     char*  result = NULL;
02002     size_t resultlen = 0;
02003     pst_item_extra_field *ef = item->extra_fields;
02004     const char *fmt = "CATEGORIES:%s";
02005     int category_started = 0;
02006     while (ef) {
02007         if (strcmp(ef->field_name, "Keywords") == 0) {
02008             fprintf(f_output, fmt, pst_rfc2426_escape(ef->value, &result, &resultlen));
02009             fmt = ", %s";
02010             category_started = 1;
02011         }
02012         ef = ef->next;
02013     }
02014     if (category_started) fprintf(f_output, "\n");
02015     if (result) free(result);
02016     return category_started;
02017 }
02018 
02019 
02020 void write_journal(FILE* f_output, pst_item* item)
02021 {
02022     char*  result = NULL;
02023     size_t resultlen = 0;
02024     char   time_buffer[30];
02025     pst_item_journal* journal = item->journal;
02026 
02027     // make everything utf8
02028     pst_convert_utf8_null(item, &item->subject);
02029     pst_convert_utf8_null(item, &item->body);
02030 
02031     fprintf(f_output, "BEGIN:VJOURNAL\n");
02032     fprintf(f_output, "DTSTAMP:%s\n",                     pst_rfc2445_datetime_format_now(sizeof(time_buffer), time_buffer));
02033     if (item->create_date)
02034         fprintf(f_output, "CREATED:%s\n",                 pst_rfc2445_datetime_format(item->create_date, sizeof(time_buffer), time_buffer));
02035     if (item->modify_date)
02036         fprintf(f_output, "LAST-MOD:%s\n",                pst_rfc2445_datetime_format(item->modify_date, sizeof(time_buffer), time_buffer));
02037     if (item->subject.str)
02038         fprintf(f_output, "SUMMARY:%s\n",                 pst_rfc2426_escape(item->subject.str, &result, &resultlen));
02039     if (item->body.str)
02040         fprintf(f_output, "DESCRIPTION:%s\n",             pst_rfc2426_escape(item->body.str, &result, &resultlen));
02041     if (journal && journal->start)
02042         fprintf(f_output, "DTSTART;VALUE=DATE-TIME:%s\n", pst_rfc2445_datetime_format(journal->start, sizeof(time_buffer), time_buffer));
02043     fprintf(f_output, "END:VJOURNAL\n");
02044     if (result) free(result);
02045 }
02046 
02047 
02048 void write_appointment(FILE* f_output, pst_item* item)
02049 {
02050     char*  result = NULL;
02051     size_t resultlen = 0;
02052     char   time_buffer[30];
02053     pst_item_appointment* appointment = item->appointment;
02054 
02055     // make everything utf8
02056     pst_convert_utf8_null(item, &item->subject);
02057     pst_convert_utf8_null(item, &item->body);
02058     pst_convert_utf8_null(item, &appointment->location);
02059 
02060     fprintf(f_output, "UID:%#"PRIx64"\n", item->block_id);
02061     fprintf(f_output, "DTSTAMP:%s\n",                     pst_rfc2445_datetime_format_now(sizeof(time_buffer), time_buffer));
02062     if (item->create_date)
02063         fprintf(f_output, "CREATED:%s\n",                 pst_rfc2445_datetime_format(item->create_date, sizeof(time_buffer), time_buffer));
02064     if (item->modify_date)
02065         fprintf(f_output, "LAST-MOD:%s\n",                pst_rfc2445_datetime_format(item->modify_date, sizeof(time_buffer), time_buffer));
02066     if (item->subject.str)
02067         fprintf(f_output, "SUMMARY:%s\n",                 pst_rfc2426_escape(item->subject.str, &result, &resultlen));
02068     if (item->body.str)
02069         fprintf(f_output, "DESCRIPTION:%s\n",             pst_rfc2426_escape(item->body.str, &result, &resultlen));
02070     if (appointment && appointment->start)
02071         fprintf(f_output, "DTSTART;VALUE=DATE-TIME:%s\n", pst_rfc2445_datetime_format(appointment->start, sizeof(time_buffer), time_buffer));
02072     if (appointment && appointment->end)
02073         fprintf(f_output, "DTEND;VALUE=DATE-TIME:%s\n",   pst_rfc2445_datetime_format(appointment->end, sizeof(time_buffer), time_buffer));
02074     if (appointment && appointment->location.str)
02075         fprintf(f_output, "LOCATION:%s\n",                pst_rfc2426_escape(appointment->location.str, &result, &resultlen));
02076     if (appointment) {
02077         switch (appointment->showas) {
02078             case PST_FREEBUSY_TENTATIVE:
02079                 fprintf(f_output, "STATUS:TENTATIVE\n");
02080                 break;
02081             case PST_FREEBUSY_FREE:
02082                 // mark as transparent and as confirmed
02083                 fprintf(f_output, "TRANSP:TRANSPARENT\n");
02084             case PST_FREEBUSY_BUSY:
02085             case PST_FREEBUSY_OUT_OF_OFFICE:
02086                 fprintf(f_output, "STATUS:CONFIRMED\n");
02087                 break;
02088         }
02089         if (appointment->is_recurring) {
02090             const char* rules[] = {"DAILY", "WEEKLY", "MONTHLY", "YEARLY"};
02091             const char* days[]  = {"SU", "MO", "TU", "WE", "TH", "FR", "SA"};
02092             pst_recurrence *rdata = pst_convert_recurrence(appointment);
02093             fprintf(f_output, "RRULE:FREQ=%s", rules[rdata->type]);
02094             if (rdata->count)       fprintf(f_output, ";COUNT=%u",      rdata->count);
02095             if ((rdata->interval != 1) &&
02096                 (rdata->interval))  fprintf(f_output, ";INTERVAL=%u",   rdata->interval);
02097             if (rdata->dayofmonth)  fprintf(f_output, ";BYMONTHDAY=%d", rdata->dayofmonth);
02098             if (rdata->monthofyear) fprintf(f_output, ";BYMONTH=%d",    rdata->monthofyear);
02099             if (rdata->position)    fprintf(f_output, ";BYSETPOS=%d",   rdata->position);
02100             if (rdata->bydaymask) {
02101                 char byday[40];
02102                 int  empty = 1;
02103                 int i=0;
02104                 memset(byday, 0, sizeof(byday));
02105                 for (i=0; i<6; i++) {
02106                     int bit = 1 << i;
02107                     if (bit & rdata->bydaymask) {
02108                         char temp[40];
02109                         snprintf(temp, sizeof(temp), "%s%s%s", byday, (empty) ? ";BYDAY=" : ";", days[i]);
02110                         strcpy(byday, temp);
02111                         empty = 0;
02112                     }
02113                 }
02114                 fprintf(f_output, "%s", byday);
02115             }
02116             fprintf(f_output, "\n");
02117             pst_free_recurrence(rdata);
02118         }
02119         switch (appointment->label) {
02120             case PST_APP_LABEL_NONE:
02121                 if (!write_extra_categories(f_output, item)) fprintf(f_output, "CATEGORIES:NONE\n");
02122                 break;
02123             case PST_APP_LABEL_IMPORTANT:
02124                 fprintf(f_output, "CATEGORIES:IMPORTANT\n");
02125                 break;
02126             case PST_APP_LABEL_BUSINESS:
02127                 fprintf(f_output, "CATEGORIES:BUSINESS\n");
02128                 break;
02129             case PST_APP_LABEL_PERSONAL:
02130                 fprintf(f_output, "CATEGORIES:PERSONAL\n");
02131                 break;
02132             case PST_APP_LABEL_VACATION:
02133                 fprintf(f_output, "CATEGORIES:VACATION\n");
02134                 break;
02135             case PST_APP_LABEL_MUST_ATTEND:
02136                 fprintf(f_output, "CATEGORIES:MUST-ATTEND\n");
02137                 break;
02138             case PST_APP_LABEL_TRAVEL_REQ:
02139                 fprintf(f_output, "CATEGORIES:TRAVEL-REQUIRED\n");
02140                 break;
02141             case PST_APP_LABEL_NEEDS_PREP:
02142                 fprintf(f_output, "CATEGORIES:NEEDS-PREPARATION\n");
02143                 break;
02144             case PST_APP_LABEL_BIRTHDAY:
02145                 fprintf(f_output, "CATEGORIES:BIRTHDAY\n");
02146                 break;
02147             case PST_APP_LABEL_ANNIVERSARY:
02148                 fprintf(f_output, "CATEGORIES:ANNIVERSARY\n");
02149                 break;
02150             case PST_APP_LABEL_PHONE_CALL:
02151                 fprintf(f_output, "CATEGORIES:PHONE-CALL\n");
02152                 break;
02153         }
02154         // ignore bogus alarms
02155         if (appointment->alarm && (appointment->alarm_minutes >= 0) && (appointment->alarm_minutes < 1440)) {
02156             fprintf(f_output, "BEGIN:VALARM\n");
02157             fprintf(f_output, "TRIGGER:-PT%dM\n", appointment->alarm_minutes);
02158             fprintf(f_output, "ACTION:DISPLAY\n");
02159             fprintf(f_output, "DESCRIPTION:Reminder\n");
02160             fprintf(f_output, "END:VALARM\n");
02161         }
02162     }
02163     fprintf(f_output, "END:VEVENT\n");
02164     if (result) free(result);
02165 }
02166 
02167 
02168 void create_enter_dir(struct file_ll* f, pst_item *item)
02169 {
02170     memset(f, 0, sizeof(*f));
02171     f->stored_count = (item->folder) ? item->folder->item_count : 0;
02172     pst_convert_utf8(item, &item->file_as);
02173     f->dname = (char*) pst_malloc(strlen(item->file_as.str)+1);
02174     strcpy(f->dname, item->file_as.str);
02175 
02176     DEBUG_ENT("create_enter_dir");
02177     if (mode == MODE_KMAIL) {
02178         int32_t t;
02179         mk_kmail_dir(item->file_as.str);
02180         for (t=0; t<PST_TYPE_MAX; t++) {
02181             if (t == reduced_item_type(t)) {
02182                 f->name[t] = (char*) pst_malloc(strlen(item->file_as.str)+strlen(OUTPUT_TEMPLATE)+30);
02183                 sprintf(f->name[t], OUTPUT_TEMPLATE, item->file_as.str, item_type_to_name(t));
02184             }
02185         }
02186     } else if (mode == MODE_RECURSE) {
02187         int32_t t;
02188         mk_recurse_dir(item->file_as.str);
02189         for (t=0; t<PST_TYPE_MAX; t++) {
02190             if (t == reduced_item_type(t)) {
02191                 f->name[t] = strdup(item_type_to_name(t));
02192             }
02193         }
02194         if (mode_thunder) {
02195             FILE *type_file = fopen(".type", "w");
02196             fprintf(type_file, "%d\n", item->type);
02197             fclose(type_file);
02198         }
02199     } else if (mode == MODE_SEPARATE) {
02200         // do similar stuff to recurse here.
02201         int32_t t;
02202         mk_separate_dir(item->file_as.str);
02203         for (t=0; t<PST_TYPE_MAX; t++) {
02204             if (t == reduced_item_type(t)) {
02205                 f->name[t] = (char*) pst_malloc(file_name_len);
02206                 memset(f->name[t], 0, file_name_len);
02207             }
02208         }
02209     } else {
02210         // MODE_NORMAL
02211         int32_t t;
02212         for (t=0; t<PST_TYPE_MAX; t++) {
02213             if (t == reduced_item_type(t)) {
02214                 f->name[t] = (char*) pst_malloc(strlen(item->file_as.str)+strlen(OUTPUT_TEMPLATE)+30);
02215                 sprintf(f->name[t], OUTPUT_TEMPLATE, item->file_as.str, item_type_to_name(t));
02216             }
02217         }
02218     }
02219 
02220     if (mode != MODE_SEPARATE) {
02221         int32_t t;
02222         for (t=0; t<PST_TYPE_MAX; t++) {
02223             if (f->name[t]) {
02224                 if (!overwrite) {
02225                     int x = 0;
02226                     char *temp = (char*) pst_malloc (strlen(f->name[t])+10); //enough room for 10 digits
02227 
02228                     sprintf(temp, "%s", f->name[t]);
02229                     check_filename(temp);
02230                     while ((f->output[t] = fopen(temp, "r"))) {
02231                         DEBUG_INFO(("need to increase filename because one already exists with that name\n"));
02232                         x++;
02233                         sprintf(temp, "%s%08d", f->name[t], x);
02234                         DEBUG_INFO(("- bump file name and try \"%s\"\n", temp));
02235                         if (x == 99999999) {
02236                             DIE(("create_enter_dir: Why can I not create a folder %s? I have tried %i extensions...\n", f->name[t], x));
02237                         }
02238                         fclose(f->output[t]);
02239                     }
02240                     if (x > 0) { //then the f->name should change
02241                         free (f->name[t]);
02242                         f->name[t] = temp;
02243                     } else {
02244                         free(temp);
02245                     }
02246                 }
02247                 check_filename(f->name[t]);
02248                 if (!(f->output[t] = fopen(f->name[t], "w"))) {
02249                     DIE(("create_enter_dir: Could not open file \"%s\" for write\n", f->name[t]));
02250                 }
02251                 DEBUG_INFO(("f->name = %s\nitem->folder_name = %s\n", f->name[t], item->file_as.str));
02252             }
02253         }
02254     }
02255     DEBUG_RET();
02256 }
02257 
02258 
02259 void close_enter_dir(struct file_ll *f)
02260 {
02261     int32_t t;
02262     DEBUG_INFO(("processed item count for folder %s is %i, skipped %i, total %i \n",
02263                 f->dname, f->item_count, f->skip_count, f->stored_count));
02264     if (output_mode != OUTPUT_QUIET) {
02265         pst_debug_lock();
02266             printf("\t\"%s\" - %i items done, %i items skipped.\n", f->dname, f->item_count, f->skip_count);
02267             fflush(stdout);
02268         pst_debug_unlock();
02269     }
02270     for (t=0; t<PST_TYPE_MAX; t++) {
02271         if (f->output[t]) {
02272             if (mode == MODE_SEPARATE) DEBUG_WARN(("close_enter_dir finds open separate file\n"));
02273             fclose(f->output[t]);
02274             f->output[t] = NULL;
02275         }
02276         if (f->name[t]) {
02277             struct stat st;
02278             stat(f->name[t], &st);
02279             if (!st.st_size) {
02280                 DEBUG_WARN(("removing empty output file %s\n", f->name[t]));
02281                 remove(f->name[t]);
02282             }
02283             free(f->name[t]);
02284             f->name[t] = NULL;
02285         }
02286     }
02287     free(f->dname);
02288 
02289     if (mode == MODE_KMAIL)
02290         close_kmail_dir();
02291     else if (mode == MODE_RECURSE) {
02292         if (mode_thunder) {
02293             FILE *type_file = fopen(".size", "w");
02294             fprintf(type_file, "%i %i\n", f->item_count, f->stored_count);
02295             fclose(type_file);
02296         }
02297         close_recurse_dir();
02298     } else if (mode == MODE_SEPARATE)
02299         close_separate_dir();
02300 }
02301 

Generated on 30 Oct 2016 for 'LibPst' by  doxygen 1.6.1