Skip to content

api/Mail/Sieve: tokenise filter rule values to align with EGroupware … #241

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
ralfbecker merged 1 commit into from
May 15, 2026
+204 −35
Merged

api/Mail/Sieve: tokenise filter rule values to align with EGroupware … #241

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
226 changes: 191 additions & 35 deletions api/src/Mail/Sieve/Script.php
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -88,6 +88,7 @@ function retrieveRules (Connection $connection)
$anyofbit = 4;
$keepbit = 8;
$regexbit = 128;
$tokenizedbit = 256;

if (!isset($this->name)){
$this->errstr = 'retrieveRules: no script name specified';
Expand Down Expand Up @@ -176,6 +177,7 @@ function retrieveRules (Connection $connection)
$rule['anyof'] = ($bits[8] & $anyofbit);
$rule['keep'] = ($bits[8] & $keepbit);
$rule['regexp'] = ($bits[8] & $regexbit);
$rule['tokenized'] = ($bits[8] & $tokenizedbit);
$rule['bodytransform'] = ($bits[12]??null);
$rule['field_bodytransform'] = ($bits[13]??null);
$rule['ctype'] = ($bits[14]??null);
Expand Down Expand Up @@ -322,11 +324,24 @@ function updateScript (Connection $connection, $utf7imap_fileinto=false, &$vac_r
$newruletext .= 'not ';
$rule['from'] = preg_replace("/^\s*!/","",$rule['from']);
}
$match = ':contains';
if (preg_match("/\*|\?/", $rule['from'])) $match = ':matches';
if ($rule['regexp']) $match = ':regex';
$newruletext .= "address " . $match . " [\"From\"]";
$newruletext .= " \"" . addslashes($rule['from']) . "\"";
// Tokenized mode: only when the explicit 'tokenized' flag is set
// AND regex is off AND value has no wildcards. Existing rules without
// the flag keep their historical behaviour verbatim.
if (!empty($rule['tokenized']) && !$rule['regexp']
&& !preg_match("/\*|\?/", $rule['from']))
{
$newruletext .= self::buildTokenizedSieveTest($rule['from'],
function($term, $not) {
$prefix = $not ? 'not ' : '';
return $prefix . 'address :contains ["From"] "' . addslashes($term) . '"';
});
} else {
$match = ':contains';
if (preg_match("/\*|\?/", $rule['from'])) $match = ':matches';
if ($rule['regexp']) $match = ':regex';
$newruletext .= "address " . $match . " [\"From\"]";
$newruletext .= " \"" . addslashes($rule['from']) . "\"";
}
$started = 1;
}
if ($rule['to']) {
Expand All @@ -335,11 +350,21 @@ function updateScript (Connection $connection, $utf7imap_fileinto=false, &$vac_r
$newruletext .= 'not ';
$rule['to'] = preg_replace("/^\s*!/","",$rule['to']);
}
$match = ':contains';
if (preg_match("/\*|\?/", $rule['to'])) $match = ':matches';
if ($rule['regexp']) $match = ':regex';
$newruletext .= "address " . $match . " [\"To\",\"TO\",\"Cc\",\"CC\"]";
$newruletext .= " \"" . addslashes($rule['to']) . "\"";
if (!empty($rule['tokenized']) && !$rule['regexp']
&& !preg_match("/\*|\?/", $rule['to']))
{
$newruletext .= self::buildTokenizedSieveTest($rule['to'],
function($term, $not) {
$prefix = $not ? 'not ' : '';
return $prefix . 'address :contains ["To","TO","Cc","CC"] "' . addslashes($term) . '"';
});
} else {
$match = ':contains';
if (preg_match("/\*|\?/", $rule['to'])) $match = ':matches';
if ($rule['regexp']) $match = ':regex';
$newruletext .= "address " . $match . " [\"To\",\"TO\",\"Cc\",\"CC\"]";
$newruletext .= " \"" . addslashes($rule['to']) . "\"";
}
$started = 1;
}
if ($rule['subject']) {
Expand All @@ -348,11 +373,21 @@ function updateScript (Connection $connection, $utf7imap_fileinto=false, &$vac_r
$newruletext .= 'not ';
$rule['subject'] = preg_replace("/^\s*!/","",$rule['subject']);
}
$match = ':contains';
if (preg_match("/\*|\?/", $rule['subject'])) $match = ':matches';
if ($rule['regexp']) $match = ':regex';
$newruletext .= "header " . $match . " \"subject\"";
$newruletext .= " \"" . addslashes($rule['subject']) . "\"";
if (!empty($rule['tokenized']) && !$rule['regexp']
&& !preg_match("/\*|\?/", $rule['subject']))
{
$newruletext .= self::buildTokenizedSieveTest($rule['subject'],
function($term, $not) {
$prefix = $not ? 'not ' : '';
return $prefix . 'header :contains "subject" "' . addslashes($term) . '"';
});
} else {
$match = ':contains';
if (preg_match("/\*|\?/", $rule['subject'])) $match = ':matches';
if ($rule['regexp']) $match = ':regex';
$newruletext .= "header " . $match . " \"subject\"";
$newruletext .= " \"" . addslashes($rule['subject']) . "\"";
}
$started = 1;
}
if ($rule['field'] && $rule['field_val']) {
Expand All @@ -362,31 +397,45 @@ function updateScript (Connection $connection, $utf7imap_fileinto=false, &$vac_r
$rule['field_val'] = preg_replace("/^\s*!/","",$rule['field_val']);
}
$end = '';
if (empty($rule['comparator']) || $rule['comparator'] === 'contains')
$isContainsMode = empty($rule['comparator']) || $rule['comparator'] === 'contains';
if ($isContainsMode && !empty($rule['tokenized']) && !$rule['regexp']
&& !preg_match("/\*|\?/", $rule['field_val']))
{
$match = ':contains';
if (preg_match("/\*|\?/", $rule['field_val'])) $match = ':matches';
if ($rule['regexp']) $match = ':regex';
$headerName = $rule['field'];
$newruletext .= self::buildTokenizedSieveTest($rule['field_val'],
function($term, $not) use ($headerName) {
$prefix = $not ? 'not ' : '';
return $prefix . 'header :contains "' . addslashes($headerName) . '" "' . addslashes($term) . '"';
});
}
else
{
$match = ':value "'.$rule['comparator'].'" :comparator "i;ascii-numeric"';
// i;ascii-numeric only accounts for unsigned integers,
// negative numbers are always evaluated as positiv infinity,
// --> we have to check for the value starting with a minus
if (in_array($rule['comparator'], ['gt', 'ge']))
if ($isContainsMode)
{
$newruletext .= 'allof (not header :matches "'.addslashes($rule['field']).'" "-*", ';
$end = ')';
$match = ':contains';
if (preg_match("/\*|\?/", $rule['field_val'])) $match = ':matches';
if ($rule['regexp']) $match = ':regex';
}
elseif (in_array($rule['comparator'], ['lt', 'le']))
else
{
$newruletext .= 'anyof (header :matches "'.addslashes($rule['field']).'" "-*", ';
$end = ')';
$match = ':value "'.$rule['comparator'].'" :comparator "i;ascii-numeric"';
// i;ascii-numeric only accounts for unsigned integers,
// negative numbers are always evaluated as positiv infinity,
// --> we have to check for the value starting with a minus
if (in_array($rule['comparator'], ['gt', 'ge']))
{
$newruletext .= 'allof (not header :matches "'.addslashes($rule['field']).'" "-*", ';
$end = ')';
}
elseif (in_array($rule['comparator'], ['lt', 'le']))
{
$newruletext .= 'anyof (header :matches "'.addslashes($rule['field']).'" "-*", ';
$end = ')';
}
}
$newruletext .= "header " . $match . " \"" . addslashes($rule['field']) . "\"";
$newruletext .= " \"" . addslashes($rule['field_val']) . "\"".$end;
}
$newruletext .= "header " . $match . " \"" . addslashes($rule['field']) . "\"";
$newruletext .= " \"" . addslashes($rule['field_val']) . "\"".$end;
$started = 1;
}
if ($rule['size']) {
Expand All @@ -400,11 +449,22 @@ function updateScript (Connection $connection, $utf7imap_fileinto=false, &$vac_r
if (!empty($rule['field_bodytransform'])){
if ($started) $newruletext .= ", ";
$btransform = " :raw ";
$match = ' :contains';
if ($rule['bodytransform']) $btransform = " :text ";
if (preg_match("/\*|\?/", $rule['field_bodytransform'])) $match = ':matches';
if ($rule['regexp']) $match = ':regex';
$newruletext .= "body " . $btransform . $match . " \"" . $rule['field_bodytransform'] . "\"";
if (!empty($rule['tokenized']) && !$rule['regexp']
&& !preg_match("/\*|\?/", $rule['field_bodytransform']))
{
$bt = $btransform;
$newruletext .= self::buildTokenizedSieveTest($rule['field_bodytransform'],
function($term, $not) use ($bt) {
$prefix = $not ? 'not ' : '';
return $prefix . 'body' . $bt . ':contains "' . addslashes($term) . '"';
});
} else {
$match = ' :contains';
if (preg_match("/\*|\?/", $rule['field_bodytransform'])) $match = ':matches';
if ($rule['regexp']) $match = ':regex';
$newruletext .= "body " . $btransform . $match . " \"" . $rule['field_bodytransform'] . "\"";
}
$started = 1;

}
Expand Down Expand Up @@ -800,4 +860,100 @@ function updateScript (Connection $connection, $utf7imap_fileinto=false, &$vac_r

return true;
}

/**
* Tokenize a free-text filter value honouring the EGroupware search syntax,
* and emit a Sieve test (single or composite allof/anyof/not group).
*
* This helper is only called for rules where the explicit 'tokenized' flag
* is set (third search mode, alongside wildcards and regex). Existing rules
* without the flag use the historical contiguous-substring :contains test.
*
* Supported syntax (matches Addressbook/Calendar/InfoLog conventions, and
* the tokenised IMAP search builder in api/src/Mail.php):
* foo bar -> anyof (test(foo), test(bar)) [OR, default for whitespace]
* foo or bar -> anyof (test(foo), test(bar))
* foo and bar -> allof (test(foo), test(bar))
* foo +bar -> allof (test(foo), test(bar)) [required]
* foo -bar -> allof (test(foo), not test(bar)) [forbidden]
* "foo bar" -> single quoted phrase as one token (preserved verbatim)
*
* @param string $value raw user input from the filter field
* @param callable $factory function(string $term, bool $not): string
* emits ONE complete Sieve test for $term
* @return string the Sieve test (single or composite), suitable for joining
* with ", " inside an outer allof/anyof. Empty when $value empty.
*/
static function buildTokenizedSieveTest($value, callable $factory)
{
$value = trim((string)$value);
if ($value === '') return '';

$tokens = self::parseSieveTokens($value);
if (empty($tokens)) return '';

$items = array();
$nextOp = 'anyof'; // EGW default for whitespace-separated tokens
foreach ($tokens as $tok)
{
$lower = strtolower($tok);
if ($lower === 'and') { $nextOp = 'allof'; continue; }
if ($lower === 'or') { $nextOp = 'anyof'; continue; }

$negate = false;
$term = $tok;
if (strlen($tok) > 1)
{
if ($tok[0] === '+') { $term = substr($tok, 1); $nextOp = 'allof'; }
elseif ($tok[0] === '-') { $term = substr($tok, 1); $nextOp = 'allof'; $negate = true; }
}
if ($term === '') continue;
$items[] = array('op' => $nextOp, 'term' => $term, 'not' => $negate);
$nextOp = 'anyof';
}
if (empty($items)) return '';

if (count($items) === 1)
{
return $factory($items[0]['term'], $items[0]['not']);
}

$hasAnd = false; $hasOr = false;
foreach ($items as $it) { if ($it['op'] === 'allof') $hasAnd = true; else $hasOr = true; }

$subtests = array();
foreach ($items as $it) { $subtests[] = $factory($it['term'], $it['not']); }

if ($hasAnd && !$hasOr) return 'allof (' . implode(', ', $subtests) . ')';
if ($hasOr && !$hasAnd) return 'anyof (' . implode(', ', $subtests) . ')';

$combined = $subtests[0];
for ($i = 1, $n = count($items); $i < $n; $i++)
{
$op = $items[$i]['op'];
$combined = $op . ' (' . $combined . ', ' . $subtests[$i] . ')';
}
return $combined;
}

/**
* Split a filter value into tokens, preserving quoted phrases as single tokens.
*
* @param string $string
* @return array list of token strings (without their surrounding quotes)
*/
static function parseSieveTokens($string)
{
$tokens = array();
if (preg_match_all('/"([^"]*)"|\'([^\']*)\'|(\S+)/u', $string, $m, PREG_SET_ORDER))
{
foreach ($m as $match)
{
if (isset($match[1]) && $match[1] !== '') $tokens[] = $match[1];
elseif (isset($match[2]) && $match[2] !== '') $tokens[] = $match[2];
elseif (isset($match[3]) && $match[3] !== '') $tokens[] = $match[3];
}
}
return $tokens;
}
}
10 changes: 10 additions & 0 deletions mail/inc/class.mail_sieve.inc.php
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -270,13 +270,16 @@ function edit ($content=null)
break;
}
$content['anyof'] = $rules['anyof'] != 0?1:0;
$content['tokenized'] = !empty($rules['tokenized']) ? 1 : 0;
}
else // Adding new rule
{
$this->getRules(null);
$newRulePriority = count($this->rules)*2+1;
$newRules ['priority'] = $newRulePriority;
$newRules ['status'] = 'ENABLED';
// Default 'tokenized' to the user's last choice (persistent preference)
$newRules ['tokenized'] = (int)($GLOBALS['egw_info']['user']['preferences']['mail']['sieve_last_tokenized'] ?? 0);
$readonlys = array(
'button[delete]' => 'true',
);
Expand Down Expand Up @@ -333,6 +336,13 @@ function edit ($content=null)
if( $newRule['anyof'] ) { $newRule['flg'] += 4; }
if( $newRule['keep'] ) { $newRule['flg'] += 8; }
if( $newRule['regexp'] ) { $newRule['flg'] += 128; }
if( !empty($newRule['tokenized']) ) { $newRule['flg'] += 256; }

// Persist the user's last 'tokenized' choice so the next new rule
// inherits it as default (per-user preference, survives logout).
$GLOBALS['egw']->preferences->add('mail', 'sieve_last_tokenized',
!empty($newRule['tokenized']) ? 1 : 0, 'user');
$GLOBALS['egw']->preferences->save_repository(true, 'user');

if (!empty($newRule['field']) && $newRule['comparator'] !== 'contains' && !preg_match('/^[0-9]+/', $newRule['field_val']))
{
Expand Down
3 changes: 3 additions & 0 deletions mail/templates/default/sieve.edit.xet
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -139,6 +139,9 @@
<row>
<et2-checkbox label="Use regular expressions (see wikipedia for information on POSIX regular expressions)" id="regexp"></et2-checkbox>
</row>
<row>
<et2-checkbox label="Use tokenised search syntax (+word required, -word forbidden, &quot;...&quot; literal phrase; same as the EGroupware search box)" id="tokenized"></et2-checkbox>
</row>
<row class="dialogFooterToolbar">
<et2-hbox width="100%">
<et2-button statustext="Saves this rule" label="Save" id="button[save]"></et2-button>
Expand Down
Loading