|
a/src/internfile/mh_mbox.cpp |
|
b/src/internfile/mh_mbox.cpp |
| 1 |
#ifndef lint
|
1 |
#ifndef lint
|
| 2 |
static char rcsid[] = "@(#$Id: mh_mbox.cpp,v 1.3 2007-12-13 06:58:21 dockes Exp $ (C) 2005 J.F.Dockes";
|
2 |
static char rcsid[] = "@(#$Id: mh_mbox.cpp,v 1.4 2008-08-29 13:05:12 dockes Exp $ (C) 2005 J.F.Dockes";
|
| 3 |
#endif
|
3 |
#endif
|
| 4 |
/*
|
4 |
/*
|
| 5 |
* This program is free software; you can redistribute it and/or modify
|
5 |
* This program is free software; you can redistribute it and/or modify
|
| 6 |
* it under the terms of the GNU General Public License as published by
|
6 |
* it under the terms of the GNU General Public License as published by
|
| 7 |
* the Free Software Foundation; either version 2 of the License, or
|
7 |
* the Free Software Foundation; either version 2 of the License, or
|
|
... |
|
... |
| 113 |
// Emacs VM botch ? Adds tz between hour and year
|
113 |
// Emacs VM botch ? Adds tz between hour and year
|
| 114 |
// From dockes Wed Feb 23 10:31:20 +0100 2005
|
114 |
// From dockes Wed Feb 23 10:31:20 +0100 2005
|
| 115 |
// From dockes Fri Dec 1 20:36:39 +0100 2006
|
115 |
// From dockes Fri Dec 1 20:36:39 +0100 2006
|
| 116 |
// The modified regexp gives the exact same results on the ietf mail archive
|
116 |
// The modified regexp gives the exact same results on the ietf mail archive
|
| 117 |
// and my own's.
|
117 |
// and my own's.
|
|
|
118 |
// Update, 2008-08-29: some old? Thunderbird versions apparently use a date
|
|
|
119 |
// in "Date: " header format, like: From - Mon, 8 May 2006 10:57:32
|
|
|
120 |
// This was added as an alternative format. By the way it also fools "mail" and
|
|
|
121 |
// emacs-vm, Recoll is not alone
|
|
|
122 |
//
|
| 118 |
static const char *frompat =
|
123 |
static const char *frompat =
|
| 119 |
#if 0 //1.9.0
|
124 |
#if 0 //1.9.0
|
| 120 |
"^From .* [1-2][0-9][0-9][0-9]$";
|
125 |
"^From .* [1-2][0-9][0-9][0-9]$";
|
| 121 |
#endif
|
126 |
#endif
|
| 122 |
#if 1
|
127 |
#if 1
|
| 123 |
"^From[ ]+[^ ]+[ ]+" // From toto@tutu
|
128 |
"^From[ ]+[^ ]+[ ]+" // From whatever
|
| 124 |
"[[:alpha:]]{3}[ ]+[[:alpha:]]{3}[ ]+[0-3 ][0-9][ ]+" // Date
|
129 |
"[[:alpha:]]{3}[ ]+[[:alpha:]]{3}[ ]+[0-3 ][0-9][ ]+" // Date
|
| 125 |
"[0-2][0-9]:[0-5][0-9](:[0-5][0-9])?[ ]+" // Time, seconds optional
|
130 |
"[0-2][0-9]:[0-5][0-9](:[0-5][0-9])?[ ]+" // Time, seconds optional
|
| 126 |
"([^ ]+[ ]+)?" // Optional tz
|
131 |
"([^ ]+[ ]+)?" // Optional tz
|
| 127 |
"[12][0-9][0-9][0-9]" // Year, unanchored, more data may follow
|
132 |
"[12][0-9][0-9][0-9]" // Year, unanchored, more data may follow
|
|
|
133 |
"|" // Or standard mail Date: header format
|
|
|
134 |
"^From[ ]+[^ ]+[ ]+" // From toto@tutu
|
|
|
135 |
"[[:alpha:]]{3},[ ]+[0-3]?[0-9][ ]+[[:alpha:]]{3}[ ]+" // Date Mon, 8 May
|
|
|
136 |
"[12][0-9][0-9][0-9][ ]+" // Year
|
|
|
137 |
"[0-2][0-9]:[0-5][0-9](:[0-5][0-9])?" // Time, secs optional: 10:57(:32)?
|
| 128 |
;
|
138 |
;
|
| 129 |
#endif
|
139 |
#endif
|
| 130 |
// "([ ]+[-+][0-9]{4})?$"
|
140 |
// "([ ]+[-+][0-9]{4})?$"
|
| 131 |
static regex_t fromregex;
|
141 |
static regex_t fromregex;
|
| 132 |
static bool regcompiled;
|
142 |
static bool regcompiled;
|